CHAPTER 1

INTRODUCTION

1.1 Background of the Study

There are numerous types of investments in capital market such as bond, stock, certificate
, commercial papers, contracts, warrants, EFTs and mutual fund. Mutual funds are financi
al institutions which pool resources from investors and make diversified investment in sto
ck, bonds and other securities of capital market. Mutual funds have become one of the lar
gest financial intermediaries, across the globe in the corporate world. Mutual Fund is a m
ode of investment in which numerous investors pool their savings to reap the benefits wit
h a determined strategy of returns. This type of investment area is helpful both for institut
ions as well as the individuals ( Afza and Rauf, 2009).

Mutual Funds are designed to invest in a large number of specific securities and market in
struments. These are tailored for small investors, who participate with small amounts of m
oney and secure high return with the help of professional managers. The fund managers m
anage the money of the savers and invest in multiple securities, available in financial mar
kets, such as stocks, bonds, commodities, precious metals and other remunerative busines
ses. This group of persons is called as Fund management and the Fund managers are resp
onsible for investment decisions. The Fund managers organize their highly diversified por
tfolios, in which they select such securities that fetch maximum returns against very low r
isks. Profit can also be earned from mutual Funds by appreciation in the original price of t
he investment units held by investors (Sipra, 2006).

Capital market helps in promoting investment and gearing up economic development. Mu

tual fund is a key instrument of the capital market. It has a very long history and a signific
ant impact on the country’s financial market. From its inception the growth of mutual fun

1
ds is not very fast and it took a really long time to evaluate to the present and the most mo
dern mutual funds Industry. Mutual Funds emergence occurred for the first time in Nether
lands in the18th century after it was introduced into Switzerland, Scotland then to United
States of America in the 19th century. Mutual funds were first introduced in 1962, in Paki
stan, due to the establishment of National Investment Trust (NIT) by the government of P
akistan. NIT Limited launched first of all open end equity funds. Just after four years in 1
966 another type of fund named Investment Corporation of Pakistan was established, whi
ch offered number of funds in its portfolio of closed-ended mutual funds? Investment Cor
poration of Pakistan (ICP) mutual funds has been floated and renounced by the ICP in ear
ly nineties 1994-95 a lot of mutual funds companies were launched in the private sector o
f Pakistan and mutual funds of Pakistan became an organization with a lot of roots (Shah
& Hijazi, 2005).

Pakistan’s financial sector started liberalizing in 1990 due to the fact, when private playe
rs were allowed to operate in the market. But year 2000 was a year of change which expe
rienced huge dramatic changes, when investors started interest in investment through prof
essional managers in different mutual funds. Due to the interest of investors in mutual fun
d, it is considered to be the dominant nonbanking finance institutions (NBFIs) in the coun
try ( Afza and Rauf, 2009). According to Research survey (MUFAP, 2006) conducted on
Mutual fund Industry, the size of Mutual fund Industry of Pakistan in 2006 was Rs 171 bi
llion US $ 2.38 Billion. Pakistan mutual fund industry now offering a variety of mutual fu
nds appealing to attract different types of investors. The MUFAP currently deals in mone
y market fund, fixed income funds, equity funds, balanced funds, Islamic funds, tracker fu
nds and fund of fund. In 2008 the total value of mutual fund industry was reported as 385.
5 billion rupees. This was due to the rise in value of open ended mutual funds, which sho
wed Rs.331.6 billion assets. The value of close ended mutual funds in the same year was r
eported as Rs.54 billion (Nazir & Nawaz, 2010 ). Pakistan mutual funds has shown treme
ndous raise in the last five years or so. The numbers of mutual funds has increased, as tota

2
l mutual funds operating in MUFAP were 135 in 2010, 158 in 2013 and 170 in 2014. In 2
013 the fund’s net asset were reported as Rs.356867 in millions in which Rs.332702 milli
on are the share of open ended funds and Rs.24165 in million representing close ended fu
nds. Whereas pension fund has shown Rs.4822 million. The value of all types mutual fun
d net assets increased to Rs.413973 million, which represents Rs.386568 million for open
ended mutual fund, Rs.21417 million for close ended mutual funds and Rs.5988 represent
ing voluntary pension funds (MUFAP, 2015). These figures shows that in Pakistan this in
dustry is growing at a rapid growth and investors are having trust and confidence regardin
g their investment. This consistent growth in this industry has motivated the researchers to
conduct the researches in this particular industry and to highlight its importance to the gen
eral public and investors.

Mutual fund pools the saving of investors and invests in profitable avenues or units trust.
Mutual funds Management Companies increase in Pakistan at a steady rate. Almost 20 A
MC are operating in MUFAP. These AMCS offer a verity of funds to the investors, i.e. eq
uity, income, balanced, asset protected, Islamic etc. The Pakistan mutual funds associatio
n currently deals in open ended, close ended and pension funds.

The mutual fund data is mostly available in developed countries, that is why most of the s
tudies in mutual fund performance have been conducted in developed world and very few
studies have been carried out in developing and emerging world. More ever the prevailing
studies of mutual fund performance in emerging markets using very traditional approache
s of measuring funds’ performance, which shows that most of the fund’s managers do not
capture the important market issues associated with mutual funds return. In Pakistan very
few studies have investigated the impact of fund characteristics or determinants on fund p
erformance (Sipra,2006, Afza and Rauf, 2009, Nazir and Nawaz, 2010). These researcher
s have focused on traditional ratios and single factor model.

This study attempts to analyze the mutual fund performance in Pakistan and attempts to u

3
nderstand the investors investing behavior in mutual funds. There are many studies in the
performance analysis field in respect to Pakistan; however this study is a worthy one as it
applies more advance data analysis tools and comprehensive data sets.

This study as mentioned above covers two aspects of mutual fund industry in Pakistan i.e.
the mutual fund performance evaluation and mutual fund investors investing behavior in
mutual fund. Mutual Fund performance can be affected by variety of factors (characteristi
cs) i.e. expense ratio, management fee, fund cash flow, fund family, fund size, fund liqui
dity and fund age. These factors have a key impact on the fund return over the period of ti
me. These funds characteristics are investigated in this current study to know their relatio
nship with the funds’ adjusted return.

Behavioral Finance is considered the important area in the field of finance. As investors m
ore informed, aware and have through knowledge of market dynamics and volatility. The
y respond immediately to new products in the financial market and try to gather informati
on. Mutual fund industry has provided a chance to many investors to easily invest in stoc
ks, which seemingly impossible for them. The mutual fund schemes provide a variety of o
ptions to different individuals across the globe irrespective of their age, financial position,
return expectations and tolerance of risk. In the past few years, we have seen a dramatic g
rowth of the Pakistan Mutual Fund industry with many private players bringing global ex
pertise to the Pakistan Mutual Fund industry. The behavior of the Investor can be affected
by many internal as well as external factors. Brand name of AMC, advertisement, risk and
return profile of the investment, quality of goods or services etc. All these factors plays a
crucial and significant role in determining the decision making process of investors. In Pa
kistan, so for none of a comprehensive study has been conducted to explore the mutual fu
nd investors investing behavior towards mutual fund. As there are a lot of investors in mu
tual fund industry across Pakistan, which need to be enquired for the reason of investment
in mutual fund. Investor behavior in mutual fund has been rarely touched by the researche

4
r in Pakistan. However in india some of the studies have analyzed the mutual fund investo
r behavior ( Chander ,2004, Miglani,2007, Kumar ,2008).

1.2 Statement of Problem

This research study aimed to investigate the performance of mutual Fund industry in Paki
stan with a view to Ascertain the operational effectiveness of various kinds of funds, cond
uct cost and risk analysis and identify factors affecting its competitiveness. The study also
seek to investigate the skills of the managers in capturing the market variation, as it is alw
ays a dilemma for the fund managers whether they capture the market variations in Pakist
an. More ever the study investigate the impact of fund characteristics as it has an impact o
n funds return and do effect the mutual fund investors behavior Rather than the market fac
tors there are some of the funds characteristics which also affect its performance. The opti
mal level of maintenance in term of these fund characteristics is always been a challenge t
o the fund managers, in order to outperform industry and market. The positive or negative
impact of these characteristics has been a concern for the fund managers and investors. Th
erefore a comprehensive study needs to be conducted to understand that how the fund’s m
anagers capture the market variations and how its characteristics, which are internally, des
igned by the managers, affect their returns. Rather than the fund performance evaluation, t
he individual investors are faced with considerable ambiguity when it comes to their fund
investment decision. Investors face a dizzying array of choices. Investors are also subject
ed to a barrage of performance statistics on the funds. While these performance data prov
ide signals of the fund manager skill, the investors clearly face a great deal of uncertaint
y about the quality of the signals.

1.3 Research Objectives

5
The following objectives are being followed in this research study

a. To describe the prevailing Pakistan’s Mutual Fund industry and conduct its
performance analysis for the period 2010 to 2014.
b. To identify the different factors (characteristics) affecting the mutual fund
performance and operational effectiveness of various funds with a view to find
strength and weaknesses.
c. To compare the performance of different mutual fund categories in the shape of
portfolios. .
d. To identify the best model among the three assets pricing models in explaining
mutual fund performance in Pakistan.
e. To understand the mutual fund investors investing behavior in mutual fund.

f. To Suggest measures to improve the performance of mutual funds industry in line w

ith regulating the investors investing behavior in a more rational way and positive
direction.

1.4 Research Questions

In order to resolve the research issue, the research sought answers of the following questi
ons.

a. Which model better predicts the mutual fund performance in Pakistan?

b. What kinds of effects these factors have on mutual fund performance?

c. Is there any difference in operational effectiveness of these factors towards fund output
(return) considering the sample funds of this study?

d .How risk adjusted return of this sample mutual funds behave with market return?

6
e. What are the factors affecting the mutual fund performance in Pakistan?

f. How investors of mutual fund behave towards mutual fund in Pakistan?

g. What kind of perception, attitude and behavior fund investors have towards the differen
t funds and their characteristics in Pakistan?

1.5 Significance and Contributions of the Study

This research ensures vitally important contributions in both literature and practical persp
ectives. First, this research has been conducted in emerging market having different chara
cteristics from that of the developed market behavior. As most of similar studies have bee
n conducted in developed world.

Second, this research will help in filling the gap of nonexistence of the findings from deve
loping and emerging markets. As the findings will be carried to the emerging markets. Th
e literature in emerging market is too small and thin, it will add value in the literature fro
m emerging market perspectives.

Third, this study uses a more extensive dataset than has been used in any previous mutual
fund studies of emerging markets. This research has constructed a data sets which has foc
used on the different determinants of mutual funds over a longer and recent period and in
a comprehensive way than the previous researchers in Pakistani mutual funds did. This is
the first time a mutual fund study has used data comprehensively investigating the mutual
fund through advanced models like CAPM, Fama French 3- factor and Carhart 4- factor
model and most important GRS test has been applied on mutual fund data, finding which
model better predict mutual fund performance in Pakistan. Furthermore, this is vitally imp
ortant study as it covers almost all kinds of open ended mutual funds except money marke
t funds of all AMCs.

7
Fourth, this study covers new area like knowing about the mutual fund investor’s behavio
r towards Mutual Fund, which has not been touched in Pakistan and none of an extensive
study regarding the investor behavior in mutual fund has been conducted. This research w
ill add quality literature in the field of behavioral finance particularly about the investor b
ehavior toward Mutual fund.

Fifth, this research has explored the issue of liquidity, fund family size, fund age, cash flo
w and fund size of almost all types of funds about which there are contradictory views an
d opinions of the researchers. This research will add contribution regarding this issue and
will add true results from Pakistan mutual fund perspective.

Sixth, this study uses the monthly and annual data of the annual reports which exactly pic
tures the fund performance in true manner.

Seventh, this research has multinomial logistic regression for the investor’s behavior data,
which have not been used before in the context of Pakistan mutual fund studies. So this w
ill add a valuable addition to the Pakistan mutual fund studies index in term of empirical s
tatistical methods.

Finally, it is hoped, this study will help and will be useful for individuals, asset managem
ent companies and institutional investors in selecting mutual funds. This study can help th
e managers to put inputs in their strategies regarding positioning and maximizing funds re
turn.

1.6 Justification or Rationale of the Study

In Pakistan very few studies have investigated the impact of fund characteristics or determ
inants on fund performance (Sipra,2006, Afza and Rauf, 2009, Nazir and Nawaz, 2010). A
ll these researchers have used the more traditional measures i.e sharp ratio, Treynor ratio a
nd Jenesan alpha for calculating the funds, risk adjusted return and then found the impact

8
of the fund characteristics on these adjusted return. However, these traditional ratios have
certain assumptions and give weak results and less valuable in the modeling perspective f
or the funds analysis. This research is unique in the sense, that it focuses on advanced mo
deling which has not been applied by researchers in the context of Pakistan mutual fund e
valuation. This research applies CAPM, Fama French Three factor model and Carhart – F
our factor models. GRS test has been applied to find which model better explain Pakistani
mutual fund performance.
This research is of more value as this study not only investigates the mutual funds’ perfor
mance of Pakistan mutual fund industry through more advanced models used for investig
ating funds performances but also intends to know about the mutual fund investor’s perce
ption and behavior towards mutual fund. Mutual fund investor behavior is very new area o
f research in Pakistan. This research uses more advanced statistical tool for the analysis of
investors’ behavior, i.e Multinomial Logistic Regression has been applied to find the relat
ionship between non matric dependent variables and matric independent variables and sim
ple tables, bars, pie and curve have been used. All factors which are related to mutual fun
d performance and attract or demotivate mutual fund investors have been used in the inve
stor behavior analysis of this research study.

1.7 Salient Features of Literature Review

A large number of studies showing the fund persistent in its performance provide ample
mixed evidences. Some earlier studies indicated that superior performance persist
ence does not persist over time (Sharpe, 1966; Jensen, 1968, McDonald, 1974; Shawk
y, 1982; Chang & Lewllen, 1984; Henriksson, 1984; Grinblatt & Titman, 1989b; Ipp
olito, 1989; Kahn & Rudd, 1995). Few studies found some performance persistence (Carl
son, 1970; Kon & Jen, 1979; Lehmann & Modest,(1987) and some found mixed evidence
- performance persistence in early 1970s which did not go beyond 1980s (Malkiel, 1995).
Similarly Heffernan (2001) demonstrated ambiguous evidence of performance persisten
ce, particularly for short horizons. No evidence of persistence in performance has been fo

9
und among close ended funds in UK fund industry (Dimson and Minio- Kozerski, 2001);
but stronger performance persistence has been found in UK pension funds (Tonks, 2005).
Different studies i.e Molson (2003) argued that fund size has significant impact on the fun
ds expense ratio the same was documented by, Tang Cheong (2007) and Zera et. al ( 2007
). Many researchers found that fund size has positive relationship with fund performance (
Gorman,1991, Grinblatt Titman,1994; Peterson et al., 2001, Nazir and Nawaz,2010). But
some of the researchers found negative relationship between fund size and fund return, as
evidenced from the studies of (Jang & hung, 2003, karlson and persson,2005; Haslem et
al., 2008 ; Bablos et al., 2009). Chen et al. (2004) argued that small size funds better perfo
rm. Expense ratio is also a key determinant of fund performance. There are contradictory
views of researcher about the impact of expense ratio on fund performance. Ippolito (198
9) argued that as the expense ratio increases it assure the better performance for high expe
nses and has positive impact on the fund return. Some other researcher found the same fin
dings (Droms and Walker, 1996, Dowen and Mann, 2007; Afza and Rauf, 2009, Nazir an
d Nawaz, 2010). Turnover is also the vital factor of fund performance. Turn over effects t
he fund performance. Carhart (1997) found a negative relationship between turn over and
fund performance. The same negative coefficient was found by Afza & Rauf (2010). Wh
ile some other researchers found a positive relationship between turn over and fund perfor
mance (Soderlind et al., 2000; wermers et al., 2000). Liquidity is also a very essential dete
rminant which affects the fund performance. There is diverse kind of feelings researchers
have about the impact of liquidity on fund performance. Glenn (2004) argued that liquidit
y has significant negative impact on fund performance. Similar results have been reported
by many researchers (Dukes & Davis, 2006; Afza & Rauf, 2010; Nazir & Nawaz, 2010).
Management fee is another variable affecting the fund performance. George (2001) argue
d that management efficiency has positive impact on fund performance. Other researchers
also found similar findings (Pushner et al., 1999; Gallaghar, 2003; Joseph, 2004). The net
cash flow is the indication of the growth in the funds’ assets. As the movement of cash wo
uld have an impact on the funds beta and hence, the performance will be affected. It is du

10
e to the fact the funds receiving huge amount cash inflow could not capitalize the investm
ent in profitable business avenues. This makes a larger cash position in the portfolio and s
ubsequently lowers the mutual fund’s beta. Conversely, large cash outflows force the man
gers to liquidate its assets. These help funds managers from trading on information and c
ause him to show negative market timing ability. Warther (1995) suggested that net flow
negatively effects the portfolio beta (Ferson and Schadt, 1996). Edelen (1999) argued tha
t there is negative association between the funds abnormal return and investors’ flows. Fu
rthermore, Benson and Faff (2006) studied the association of money flows with equity fun
d performance. They found evidence supporting Warther’s (1995) and Edelen’s (1999) w
ho found that money flow has negative impact on funds’ performance. Gruber (1996) arg
ued that funds with higher net flow generate higher return as compared to lower net flows
funds. Zheng (1999) documented that funds having positive cash flows outperform those h
aving negative cash flows. Sapp and Tiwari (2004) argued that smart money is vital as it a
ffects the stock momentum phenomenon. They predicted that having control stock return
momentum, the smart money effect would likely to be disappeared. Dahquist et al. (2000
) found that lagged flow has an impact on the funds’ performance. However, Gharghori et
al. (2007) argued that the smart money effect would not be explained by the stock return
momentum. Ferreira et al. (2009) documented the above findings. The relationship betwe
en fund age and fund performance has not been properly explained in previous studies. It i
s presumed that Funds normally experience higher costs at the introduction stage as a hug
e amount is allocated for advertising and promotion activities. They argued that old fund d
o perform better in comparison of the young incepted funds.. Nevertheless, one could arg
ue that managers of young funds are more active. This is confirmed by Blake and Timmer
mann (1998), who revealed that funds are likely to perform best during their first year of e
xistence. Similarly, Otten & Bams (2002) found a negative correlation between performa
nce and fund age in some European countries during the period 1991-1998. Nevertheless,
Prather et al. (2004) and Ferreira et al.. (2009) found no relationship between age and fun
d performance. Afza & Rauf (2009) investigated the fund age impact on fund performanc

11
e and found insignificant positive impact on fund performance in Pakistani context.

The fund family is another characteristic of the fund which has vital impact on funds’ perf
ormance and the decision making of the fund’s managers. Ensuring more funds in a famil
y helps to get economies of scale. (Khorana and Servaes, 1999, 2005). A larger fund fami
ly helps in better research quality. Furthermore, it helps in reducing the search cost especi
ally for investors due to easy recognition (Gaspar et al., 2006).

However, Chen et al. (2002) found that fund family does not significantly impact the perf
ormance of fund. In contrast, Guedj & Papastakaikoudi (2004) revealed that persistence o
f performance among large fund families. In addition, Ferreira et al. (2009) revealed that f
und age has significant role in the fund performance of both US and international funds.

This research also focuses on understanding investors investing behavior in mutual. Little
efforts have been made by researchers to study investors’ behavior towards mutual funds
particularly in the context of Pakistan. Madhusudhan & Bodekar (1996) argued that mos
t of the investors are inclined towards mutual funds due to the more liquidity, safety and c
apital appreciation. Sunder (1998) documented that the curve of awareness in small cities
is poor as compared to big cities and found that some factors i.e. brand image and return a
re vital for investment. Tripathy (2002) argued that almost all investors were unsatisfied e
xcept UTI investors. Chander (2004) concluded that some inefficiencies in management,
de- regulations and poor control are the main causes of non- investment in mutual fund.

Miglani (2007) conducted a study to know the investor behavior toward mutual funds’ pe
rformance and return. He made his analysis on the perception views of investor in decisio
n making. Kumar (2008) conducted a study on investor behavior towards mutual funds in
which he assessed the investor’s performance, awareness, and investor’s preference in dif
ferent mutual funds categories, schemes and other factors which effect the investor selecti
on of mutual funds. Sultana (2010) conducted a research in which she found on the demo
graphical factors age and gender on the risk tolerance level of the investors. Bodekar (19
96) conducted study on investor behavior in mutual funds and focus on fund schemes, inf

12
ormation sources, risk element and NAV of different funds. He found that investor like po
sitive NAV low risk and more safety, due to which they invest in mutual fund. Jan and Ja
in (2012) the study was argued that each demographical factor have effect on buying beha
vior pattern of investor, the research conduct that every demographical factor (gender, age
, qualification, occupation, income and education) had significant bearing on investor buy
ing behavior process. Barber & zheng (2000) conducted study in investor behavior towa
rds mutual funds and the research argued three types of results. 1). Investor buys the funds
on the basis of past performance.

2) Approximately 40 percent investors are also selling the fund on past performance.

3) Investor do not like to buy the fund with highly commission fee or other highly transact
ion cost. Geete ,& Desai (2010)analyzed investor behavior and the study concluded that
main factor which effect investor decision are money growth, return & security factor of
money reliability of the company and also seek the company profile, they also follow com
pany market brand value. Barber (2003) conducted research and found that investors are
mostly sensitive to pay front-end load and less sensitive for other operating expense. The r
esearcher also provided through managers that investor do not like in highly front-end loa
d against of the operating expenses.

Arathy et al. ( 2015) conducted study in the investor behavior and stated that better past e
xperience, tax benefit, return potential, liquidity, low cost and transparency are the major f
actors that motivate the investors to invest in mutual funds. Ansari & Moid (2013) stated
that investors, investing decision in mutual funds are due to risk factor, growth and additi
onal income. Rajeswari & Ramamurthy (2001) conducted study in investor behavior in m
utual funds. The study determines that the factor influencing investment decision is produ
ct quality (performance of the fund) and brand name. These two factors were the main fac
tor which affects mostly the decision making in mutual fund. Kumar & Goel (2012) foun
d that majority of investor, invest in growth scheme, the study also concluded that growth
is the aim of the investor in mutual fund. Sharma (2011) in his study revealed that invest

13
ors switch from one mutual fund to other , the reason behind to switch from one to anothe
r mutual fund are same family fund, investor mind set, current market condition, poor fun
d management and poor selection of portfolio.

1.8 Theoretical Frameworks

The literature of both mutual fund performance and investors investing behavior in Mutua
l Fund was reviewed and during the phase of reviewing the literature, the following theore
tical frameworks have been formulated.

14
1.8.1 Performance Analysis

15
1.8.2 Investors Investing Behavior in Mutual Funds

Performance of Record

Fund reputation and Brand Name

Scheme portfolio of Investment

Reputation of fund manager

Types of mutual fund
Withdraw facilities investors

Favorable rating by agency

Innovativeness of scheme

Product with tax benefit

Entry and exit load

Minimum initial investment

Disclosure of derivation from the

Original pattern

Figure A.2 Investor behaviors in Mutual funds analysis

The above frame work represents the fund factors on the biases of which the mutual fund
investor behavior has been investigated. These factors are the dependent variables upon
which the investor behavior has been investigated with matric or categorical independent

16
variables.

1.9 Brief Overview of Research Methodology

This study seeks to highlight mutual fund performance as first part and knowing mutual f
und investors investing behavior as second part, so this is a quantitative research using bo
th primary data and secondary data for mutual fund investor behavior and fund performan
ce. This research can also be called an applied research. The investors investing behavior i
n mutual fund have been analyzed through convenience sampling and mutual fund perfor
mance has been analyzed through random sampling technique. The monthly data of 153 o
pen ended mutual fund collected to find the adjusted return on the bases of CAPM, Fama
French 3- factor model and Carhart four factor models. This monthly data has been collec
ted from 2009 to find previous year alpha of all three models. The monthly data has been
collected from the fund NAV on MUFAP site. This study applies GRS test to understand
which model better predict the mutual fund performance in Pakistan. Most of the studies r
egarding mutual fund performance have used the traditional measure for the evaluation of
fund performance in Pakistan. That is why this study is unique as applying more sophistic
ated and advanced techniques for fund performance evaluation. This study investigates th
e impact of fund characteristics i.e. expense ratio, fund size, fund age, fund family size, ca
sh flow and liquidity on the risk adjusted return of the fund through regression over the s
tudy period of 2010 to 2014. The annual data of fund determinants (characteristics) have
been collected from funds annual reports. These characteristics of funds are very vital for
the fund return and its overall performance.

This research finds out the key impact of these fund characteristics which is an area of int
erest and area of concern for the asset managers and AMCs. This research also intends to
find the investor behavior of mutual fund investors for which bars, tables, pie and multino
mial logistic regression have been applied to find the relationship between non matric dep
endent variables, representing mutual fund investor behavior and matric or categorical var

17
iables representing the independent variables.

18
1.10 Research Scheme

This research has few chapters, chapter one consists of the introduction, problem statemen
t, objectives, research questions, significance, need and newness, salient of literature, brie
f methodology, theoretical frameworks and research scheme. Chapter two literatures of th
is research, theoretical framework, and hypothesis. Chapter three contains the methodolog
y of this research, which represents the study design, data collection methods, population
& sampling and the statistical tools. Chapter four consists of an over view of Pakistan mu
tual fund industry. Chapter five contains the comprehensive data analysis of both perform
ance evaluation and investor behavior parts of this research and also contains findings. Ch
apter six consists of recommendations, future research directions and conclusion.

19
 CHAPTER 2
REVIEW OF LITERATURE
2.1 Introduction
This chapter illustrates and explains the history of mutual fund and its growth, the different
performance measures and models used in the analysis of stocks and funds evaluation,
different mutual funds’ performance studies conducted in Pakistan and across the globe,
studies illustrate mutual fund selection and timing abilities, mutual fund performance
persistence studies, studies related to investors decision making in mutual funds, fund
flows studies and studies focusing on mutual funds investors investing behavior.

2.2 History and Growth of Mutual Fund

Mutual Fund is an ancient vehicle of investment. Mutual funds collect the savings of small
investors and invest collectively in stocks, bonds, or money market instruments (Shah and
Hijazi, 2005). The investment company concepts date to Europe in the late 1700s, Dutch
Merchant Abraham van Katwitch Invited subscriptions from investors with limited
means.” The materialization of “investment Pooling“ in England in the 1800s brought the
concept closer to U.S. shores. The enactment of two British Laws, the Joint Stock
Companies Acts of 1862 and 1867, permitted investors to share in the profits of an
investment enterprise, and limited investor liability to the amount of investment capital
devoted to the enterprise. May be more outstandingly, the British fund model established
a direct link with U.S. Securities markets, serving finance the development of the post –
Civil War U.S. economy. The Scottish American Investment Trust, Formed on February,
1873 by fund pioneer Robert Fleming, invested in the economic potential of the United
States, Chiefly through American railroad bonds. It was introduced in Europe, in
particular, Great Britain. Robert Fleming set up in 1868 the first investment trust which is
well known as Foreign and colonial investment trust, which could manage the money of
investors in Scotland by investing in variety of stocks. This investment trust along other

20
same kind of investment trusts started operations in Britain and the U.S. Later on these
trusts, resembled today’s close ended mutual funds. The first mutual fund in the U.S,
Massachusetts investor’s trust, was set up in March 1924. This was the open ended mutual
fund. By Nov, 1925 all these funds were open ended, having redemption feature. Similarly,
they had almost all the features of a good modern Mutual Funds like sound investment
policies and restrictions, open endness, self– liquidating features, a publicized portfolio,
simple capital structure, excellent and professional fund management and
diversification etc. Hence they are the honored grandparents of today’s funds. Prior to these
funds all the initial investment companies were closed – ended companies. Therefore, it
can be said that although the basic concept of diversification and professional fund
management, were picked by U.S.A. From England Investment Companies “The Mutual
Fund is an American Creation.”
Because of their exclusive feature, open – ended Mutual Funds rapidly became very
popular. By 1929, there were 19 open – ended Mutual Funds in USA with total assets of
$ 140 million. But the 1929 Stock Market crash followed by great depression of 1930
ravaged the U.S. Financial Market as well as the Mutual Fund Industry. This necessitated
tighter regulations for mutual funds and for Financial Sectors. Hence, to protect the
interest of the common investors, U.S. Government passed various Acts, such a Securities
Act 1933, Securities Exchange Act 1934 and the Investment Companies Act 1940. A
committee called the National Committee of Investment Company (Now, Investment
Company Institute), was also formed to Co-operate with the Federal Regulatory Agency
and to keep informed of trends in Mutual Fund Legislation (Shah and Hijazi, 2005).
The stock market crashed in 1929 which created great depression, and the outbreak of the
Second World War slackened the pace of growth of the mutual fund industry across
the world. However the Innovations in the types and products of mutual funds increased
in 1950s and 1960s. This boom era of developing mutual funds give birth to the first
international stock mutual fund in US in 1940.
In the year 1976, the first tax exempt municipal bond funds emerged and in 1979, the first

21
money market mutual fund was introduced in the capital market.

The latest additions are the international bond funds during the period 1986 to1990. The
mutual fund industry witnessed phenomenal growth during eighties and nineties, when
the mutual fund industry showed outstanding increase in the size, number, categories,
assets, schemes and shareholders. In the early eighties US mutual fund industry registered
a ten-fold growth. Since 1996, due to dramatic increase in mutual fund volume across the
world, its assets have exceeded bank deposits. The mutual fund industry is so called
virtually rival to banking industry.
As a result of these measures, the Mutual Fund Industry began to develop speedily and the
total net assets of the Mutual Funds Industry increased from $ 448 million in 1940 to $ 2.5
billion in 1950. The number of shareholder’s accounts increased from 296000, to more than
one Million during 1940 – 1951. “As a result of renewed interest in Mutual Fund Industry
they grew at 18% annual compound rate reaching peak of their rapid growth curve in the
late 1960s.” Mutual funds trend started in 1962, in Pakistan, when Government of Pakistan
established National Investment Trust (NIT). NIT Limited launched first of all open end
equity funds. Just after four years in 1966 another type of fund named Investment
Corporation of Pakistan was established which offered a series of closed-ended mutual
funds. Investment Corporation of Pakistan (ICP) mutual funds has been floated and
renounced by the Investment Corporation of Pakistan, in early nineties 1994-95 a lot of
mutual funds companies were launched in the private sector of Pakistan and mutual funds
of Pakistan became an organization with a lot of roots. The size of Mutual fund Industry of
Pakistan in 2006 was Rs.171 billion or US $ 2.38 Billion. Pakistan mutual fund industry
now offering a variety of mutual funds appealing to attract different types of investors. The
MUFAP currently deals in money market fund, fixed income funds, equity funds, balanced
funds, Islamic funds, tracker funds and fund of fund. The total assets size of the mutual fund
industry increased to 385.5 billion rupees in 2008. This rise in assets was primarily due to
Open-Ended Funds having value of Rs. 331.6 billion. The value of the Closed-Ended at the

22
end of 2008 was Rs.54 billion, (Shah and Hijazi, 2008). Pakistan mutual has shown a rise
in the last five years or so. The numbers of mutual funds has increased as in total mutual
funds operating in MUFAP were 135 in 2010, 158 in 2013 and 170 in 2014. In 2013 the
fund’s net asset were reported as Rs.356867 in millions in which Rs.332702 million are the
share of open ended funds and Rs.24165 in million representing close ended funds.
Whereas, pension fund has shown Rs.4822 million. In 2014 the sum of net assets of open,
close ended and pension fund are Rs.413973 million, which represent Rs.386568 million
for open ended mutual fund, Rs.21417 million for close ended mutual funds and Rs.5988
representing voluntary pension funds (Mufap, 2015). These figures shows that in Pakistan
this industry is growing at a rapid growth and investors are having trust and confidence
regarding their investment.

2.3 Mutual Fund Performance Measures

Prior to 1965, comparison of the fund returns was the only method available to the
portfolio managers to evaluate mutual fund return. Perhaps in one of the earliest studies
on mutual fund return and performance evaluation was conducted by Close (1952) who
compared the performance of open ended and close ended funds and found that the mean
returns of close ended funds to be higher, in spite of three times more sales and popularity
of open ended funds. Need for performance criteria arose in the study of Brown and
Vickers (1963) who observed that different types of fund require different performance
criteria. Subsequently various performance measures have been developed, primarily
taking inputs from the pioneering work on Modern Portfolio Theory (Markowitz, 1952;
1959) and Capital asset pricing theory also known as CAPM (e.g. Sharpe, 1964; Lintner,
1965; Treynor, 1965; Mossin, 1966).
It was the year of 1965 which proved a turnaround year in the subject of mutual fund return
performance evaluation. Treynor (1965) presented a new way of viewing and presenting
performance results. Treynor (1965) discussed the impact of market on portfolio returns
and risk aversion by investors. He graphically developed measure for management

23
performance. But the real quantification of mutual fund performance evaluation came
from the work done by Sharpe (1966). He developed a linear relationship between
expected return on portfolio and its associated risk. His biggest contribution to the
financial literature is that of concept of risk premium (known as beta). Sharpe (1966)
introduced a new method of performance measurement in the form of Sharpe ratio (reward
to variability ratio) and reported that higher Sharpe ratios are desirable. The mutual fund
performance measures received a further fillip from the study done by Jensen (1968)
who developed a model for mutual fund performance (practical adaptation of CAPM),
that statistically relates fund’s performance to a benchmark. Jensen introduced the term
alpha (Jensen’s Alpha) in regression equation used in the model of performance
measurement. Superior security price forecasting is indicated by the positive alpha and
negative alpha indicates either a poor security selection or the existence of high expenses
(as a result of frequent trading or other factors). Carlson (1970) reported the regression of
fund returns on broader market index have a higher unexplained variance, which can be
significantly reduced if a more related benchmark (like a specific type mutual fund index)
is used as a market proxy. Similar view was later observed by Lehman and Modest
(1987), who empirically stated that Jensen measures have been sensitive to the
choice of benchmarks. Although alphas and betas of mutual fund performance attained
wide publicity and were adopted on wider scale as standard tools for measuring
mutual fund performance, Miller and Gressis (1980) explained that the estimates of fund’s
alpha and beta may provide misleading information if non stationary is present in the risk
return relationship which is ignored. Authors found non stationary betas and reported that
there was no statistically significant relationship between fund betas and market rate of
return. Grinblatt and Titman (1989) analyzed both actual and gross portfolio returns, and
reported that superior performance may exist in certain type of funds (for example
growth funds, aggressive growth funds and smaller funds), but due to higher expenses,
abnormal profit returns were eliminated. Grinblatt and Titman (1993) criticized Jensen
measures on account of its sensitivity to the choice of benchmark portfolio. The authors

24
introduced new measures of portfolio performance named as “portfolio change measure”
that did not require the use of benchmark portfolios but employs portfolio holdings. Fama
French (1993) introduced 3-factor model for measuring the portfolios return, which got
importance in the field of measuring portfolio performance over additional two factors
bases i.e. size factor and value factor rather just using the market factor. This was a step
towards muilti-factor model in measuring the portfolios performances. The new measure
used insights from event study methodology and was not subjected to survivorship bias
and had some statistical computation advantages. Later some more new measures of
mutual fund return performance have been developed such as measures depending
on Bayesian performance evaluation, where investors choose to invest in an active fund
when the prior point estimate of alpha is positive or the measures based on event study
methodology utilizing mutual fund holdings (Kothari and Warner, 2001). The DEA
technique has been widely used by many researchers in analyzing. Murthi et al. (1997)
first used this technique in the analysis of mutual funds. After that so many researchers
used this technique to understand the efficiency of funds. DEA is very familiar as it covers
the results of both traditional measures like Jensen and Sharp Index alpha also compare
the results generated through these traditional measures with the more sophisticated
measures. The traditional measures do not properly find out the efficiency and inefficiency
but this method do well in doing so.
Worthington (1998) used the DEA technique to find the comparison. His research found
that only single input is not a good mechanism to find the results, he suggested that
multiple units should be used to compare the efficiency of two units. Galagedera &
Silvapulle (2002) also used DEA to measure the efficiency of 257 Australian investment
funds. DEA is vital and able to use many factors rather than output and risk which indicate
the funds operation well and have not pattern like Sharp and Jensen. The observations
suggested that while selecting input and output are effective to figure out the structure and
size of the unit.
Although mutual fund is considered to have stocks which outperform the market, however

25
the funds return are found to be lower. The difference between net fund’s return and stock
funds return are due to the costs like expense ratio, and transaction cost (Wermers, 2000).
Net sales and redemptions plays a vital role in the mutual funds’ performance and have
been found that funds with positive net sales do well in comparison to funds having more
redemption (Dickson et al., 2000). Pastor and Stambaugh (2002) further modified the
existing performance methodology and measure and included non-benchmark assets.
Researchers argued that the estimation of alpha or sharp ratio can be improved with the
use of value, size and momentum factors. In addition difference between Fama French
and CAPM alphas, importance was reduced, when non-benchmark assets measures
were included. Cohen et al. (2005) developed new measure of performance measurement
depending on the similarity of the portfolio holdings of the given fund manager with the
successful manager (in terms of covariance between the weights of the portfolio holdings).
These new measures were likely to be powerful in measuring performance as compared
to the traditional measure been used for performance analysis. Regarding analyzing
performance of mutual funds, a number of studies showed different results. Few studies
among them have found negative alpha or showed underperformance with respect to the
benchmarks used (Sharpe, 1966; Treynor and Mazury, 1966; Jensen, 1968; McDonald,
1974; Elton et al., 1993; Gruber, 1996; Daniel et al., 1997). Some studies i.e. Malkiel
(1995) concluded that negative alphas are due to net returns and positive alphas mainly
due to the gross returns, but neither alpha was found statistically significant different from
zero. Trying to explain the cause of the underperformance, Wermers (2000) found that
underperformance of equity funds is mainly due to non-stock holding portfolios while
comparing to the market average. But some studies showed inconclusive evidence while
using conditional performance model. Ferson and Schadt (1996) found that alphas had a
mean value of zero (as compared to negative alphas found in most of the studies). The
same findings were obtained by Ferson and Warther (1996) who applied conditional
measures on 63 funds and argued that most often the funds outperform the market return
after having been adjusted for the risk.

26
Performance analysis has been done analyzing both open ended and close ended and
underperformance was found in both close ended and open ended (Anderson et al., 1996);
in bond funds (Blake et al., 1993). There lies an inconclusive evidence regarding out
performance of international funds (Cumby & Glen,1990) but significant evidence
though can be found of underperformance in the context of international bond funds
(Detzler,1999).
Some of the studies however found different results on the bases of risk adjusted return
and found that these funds outperform the market (for example, Ippolito, 1989;
Grinblatt ;Fama French ). Cahart (1997) showed that to earn abnormal return, funds
managers’ use momentum strategy. The same momentum anomaly was pointed out by
Jegadeesh and Titman (1993).They found that funds performing smoothly for about 12
months keep on going the same momentum and perform well in future too. Therefore, if
fund managers like to employ this phenomenon to have abnormal returns, seemingly not
good practices of value added. Subsequently, Cahart (1997) proposed a new four factor
model which do have the factors of the Fama French 3- factor model(1993) in excess of
his own momentum. His momentum variable is the equally weighted portfolio of the
stock’s highest 30% eleven-month returns, lagged one month, minus the equally weighted
portfolio of stocks with the lowest 30% eleven-month returns, lagged one month. Cahart’s
(1997) four-factor model has got more importance in measuring mutual fund performance.

There are critics on muilti-factor models, especially on Fama French 3-factor and carhart
4-factor model that these models have no theoretical background. Hover many researchers
argued that these model do well in measuring the portfolios return (Fletcher and
Forbes, 2002; Otten and Bams, 2004; Hubner, 2007). But some of the researchers i.e
Ferson and Schadt (1996) argued that the performance measures mentioned above are
based on unconditional expected returns and risk, meaning that the portfolio’s betas are
fixed for the whole observation period. As a result, Ferson and Schadt (1996) mitigate the
drawback of traditional measures by incorporating interaction variables in order to capture

27
time-varying expectations. These interaction variables are a vector of the predetermined
economic variables. They assume that the conditional beta is a linear function of a vector
of the predetermined variables. In addition to the conditional betas, these predetermined
variables can also be incorporated with the portfolio alpha in order to allow time-varying
in abnormal performance. Carhart (1997) made an extension to the Fama three factor
model and it is a new model being used for the performance of funds.
This conditional beta is vital to be used instead of unconditional beta to deal well with
capturing the dynamic strategy to help the managers in capturing the market variations.
The researchers argued that by using conditional betas the portfolio performance can be
improved, as it well capture the market variations.(Ferson and Schadt, 1996; Ferson and
Warther, 1996; Sawicki 2001; Roy and Deb, 2003).

Timing abilities vital for the managers simply by decomposing the performance into
selecting and timing abilities. The well-known approaches in this regard are, the quadratic
regression method (Treynor and Mazuy ; 1966 ) This model suggest that the characteristic
line is not straight as volatility offer variations. Keeping in view the stock variations the
managers’ move from high volatile portfolios to less volatile portfolios. Titman (1989)
Found that although there are numerous studies signifying underperformance, some latest
studies are indicating the trend in the opposite direction. Kosowski et al. (2006) re-
examined the performance evaluation results of US mutual fund industry from the period
1962 to 1994 and using bootstrap methods authors found that some of the managers
achieved superior genuine performance. Avramov and Wermers (2006) also pointed in the
same direction. The whole debate on out-performance and under-performance by mutual
funds signifies that mutual funds under-performed in the earlier years but in the later years
mixed performance was evident. That is there are large number of insignificant alphas (not
significantly different from zero) and presence of very few mutual funds having
significantly positive alphas (Barras et al., 2009)
2.3 Mutual Fund Performance Persistence

28
Performance persistence (or whether historical performance persists into the future) is one
of the most widely discussed and cited issue in the mutual fund literature. A large number
of studies have been conducted on the subject that provides ample mixed evidence.
While some earlier studies indicated that superior performance persistence does
not persist over time (Sharpe, 1966; Jensen, 1968, McDonald, 1974; Shawky, 1982;
Chang and Lewllen, 1984; Henriksson, 1984; Grinblatt & Titman, 1989; Ippolito, 1989;
Kahn and Rudd, 1995). Few studies found some performance persistence (Carlson, 1970;
Konand Jen, 1979; Lehmann and Modest,(1987) and some found mixed evidence -
performance persistence in early 1970s which did not go beyond 1980s (Malkiel, 1995).
Studies in early 1990s suggested that some mutual funds have persistent superior mutual
fund performance (Grinblatt and Titman, 1992; Hendricks et al., 1993; Goetzmann and
Ibboston, 1994; Elton et al., 1996; Gruber, 1996; Blake et al., 1999 etc). Some persistence
performance had weak support (Grinblatt and Titman, 1992); some evidence has very
strong support (Hendricks et al., 1993). Persistence in performance had been even found
by using conditional or time varying alphas or betas as compared to unconditional or
average ones (Christopherson et al., 1998). Some studies have found performance
persistence only in fixed income mutual funds and not in equity (Kahn and Rudd, 1995).

Hendricks et al. (1993) found persistence for four quarters and reversal thereof (they called
their findings as “hot hand” phenomenon). The “hot hands” phenomenon of Hendricks et
al. (1993) was criticized by some researchers as they found persistence of losing
performance (Brown and Goetzmann, 1995; Carhart, 1997), which they called “icy
hands”. Brown & Goetzmann (1995) highlighted a very pertinent point in practical
terms “that past patterns yield a clue about which funds to avoidbut do not provide strong
indications about which funds will outperform their benchmark in the future”.
Later studies raised some doubt as they suggested that superior performance persistence
found in earlier studies may be due to survivorship bias and attrition bias of mutual fund
samples in earlier studies (Brown et al., 1992; Brown & Goetzmann,1995; Malkiel, 1995)

29
or due to some naïve momentum investment strategy (Carhart,1997; Daniel et al., 1997;
Wermers, 1997). Survivorship bias is present because some funds disappear during the
period under study. Using data with surviving funds will bias upwards the true
performance of the funds, because high performance funds will tend to be over represented
in the sample. Attrition effect refers to the fact that the attrition of poor performers alters
persistence measures because it changes the composition of the sample. Carhart et al.
(2002) discussed the impact of survivorship bias on performance persistence and
demonstrated that survivorship bias, if removed, weakens performance persistence.
Further, the authors found that survivorship bias increases with sample length, but at the
decreasing rate (Carhartet al., 2002). Lehmann and Modest (1987), Hendricks et al. (1993)
and Wermers (1997), among others, have found evidence of persistence in fund
performance over short horizons of one to three years. Grinblatt and Titman (1992) and
Elton et al. (1996) concluded that past risk-adjusted performance is predictive of future
performance over periods as long as three years. Yet, Brown and Goetzmann (1995)
concluded that even if there has been some predictability it is quite difficult to detect.
Carhart (1997), Daniel et al. (1997), Wermers (2000) and Pastor and Stambaugh (2002b)
argued that most of this persistence has been due to factors other than managerial ability
Outside the US, several studies on performance persistence have been conducted in
United Kingdom also. The evidence for performance persistence in UK has been found
to be nil (Fletcher 1999a; 1999b); limited (Brown et al, 1997); persistence of poor
performance (Quigley &Sinquefield, 2000); and persistence of performance (Lundeet al.,
1999; Blake & Timmerman, 1998).

Allen & Tan (1999) supported evidence of persistence in performance in the long run but
not in the short run, and with both raw and risk adjusted returns. Similarly Heffernan
(2001) demonstrated ambiguous evidence of performance persistence, particularly for
short horizons. No evidence of persistence in performance has been found among close
ended funds in UK fund industry (Dimson&Minio- Kozerski, 2001); but stronger

30
performance persistence has been found in UK pension funds (Tonks, 2005).
Using Monte Carlo simulations, Horst and Verbeek (2000) found persistence in mutual
fund performance to be more pronounced than previous studies. Horst et al. (2001)
discussed about “look ahead bias” that may result in spurious performance persistence.
“Look ahead bias” occurs because persistence studies typically utilize a ranking period
and an evaluation period. Funds tend to disappear in non-random way in any of these
periods, causing “look ahead bias” thereby generating spurious performance
persistence.
Berk and Green (2004) discussed the cause of absence of performance persistence
from the perspective of investor fund flows. Investors direct their money to successful
managers, thereby decreasing the expected marginal rate of returns for the fund and hence
absence of performance persistence. Similar findings were observed in the study
conducted by Bollen and Busse (2005).

2.4 The Effects of Cost and Expense on Mutual Fund Performance

Brad M. Barber et al. (2003) argued that while purchasing the mutual fund the investor
look at the cost associated with mutual funds like the front load, back load fee and other
services fee of the managers. As the mutual funds guarantee performance on the bases of
advertising, marketing and information’s cost. They argued that there is negative relations
between fund flows and front-end load fees. They also documented that commission
charged by the funds brockers will negatively impact the funds flows. In contrast, they
found no relation between operating expenses and fund flows. Additional they argued that
mutual fund marketing and advertising represent fund’s operating expenses which,
account for this surprising result.
Many researchers documented that most of funds in US funds have dropped the ratio of
charging expenses like most of US equity funds have dropped the front load fee from 91%
to 35% during the period 1992 to 1999. These practices changed the overall volume of
assets been managed by the AMCs (EGB, 2004). They argued that the front load fee can

31
affect the volume of assets.

It was evidenced that investors like to pay low fee for managing their mutual funds units
and most of the researchers found that the investors do not like the front load fee and
commission (Brad M. Barber et al., 2003).

Many researchers found that marketing expenses help to improve the fund performance
as marketing expenses helps to motivate investors, which increase the asset volume of a
particular mutual fund. Marketing expenses represent the portion of overall operating
expenses of the mutual fund.
Several studies have shown a negative relationship between funds’ operating expense ratio
and fund performance (Gruber, 1996 and Carhart,1997). Thus, this reflects that investors
avoid to purchase high operating funds. Generally, investors pay fees to mutual funds
through operating expense ratios applied to assets under management or through load fees
charged when investors purchase (or less commonly sell) a mutual fund. When purchasing
funds through a broker, investors pay a commission to the broker for some mutual funds,
but not for others, which are designated as non-transaction fee (NTF) funds.
Survey and other experimental evidences supported that mutual fund investors are
indifferent and cannot create a tradeoff regarding the fee charged by mutual funds. Wilcox
(1998) investigated 50 consumers, who currently invest in mutual funds having different
expense ratio and load combinations. He documented that 46 of the 50 investors did not
like load fee in the expense ratios. Alexander, Jones, and Nigro (1998) documented that
all surveyed 2000 investors, 84 percent of respondents were agreed with the fact that the
higher the expenses of funds the more the return.

There is surprisingly little empirical research on how investors consider expenses when
investing in mutual funds. The only empirical work that we are aware of is Sirri and
Tufano(1998) who found a negative correlation between funds flow and funds expenses.

32
The most important study in determining the factors affecting Expense Ratio was done in
Malaysia by Soo-Wah Low (2008) who observed that both funds size and fund family has
an impact on the expense ratio and found that these negatively affect the expense ratio.
The same kind of observations was documented by Babalosy at el. al (2009) who
investigated Greek equity funds. These findings remained consistent with the studies of
Geranio and Zanotti (2005) who determined and intensively explored the expense ratio
of Italian funds and its impact on the funds’ performance.
Rompotis (2008) investigated the different categories of Greek Mutual funds industry i.e.
money market, bond, equity and balanced fund. He found that equity funds have more
expense ratio as compared to money market and other categories of funds. He also found
that funds, managed by banks, charge more than other categories of funds.

Many other academic studies like Molson (2003), Tang Cheong (2007) and Zeraet. Al
( 2007)argued that the size of fund has impact on the funds expense ratio.Their researches
revealed that larger the fund size, the lower the expense ratio and it is all due to economies
of scale and reduction in marginal cost. Zera et al. (2007) investigated the impact of
corporate governance on the expense ratio and found it as a key factor determining the
expense ratio.
Some other studies were also conducted covering qualitative aspects of the funds like the
investment strategy, the management style, investor preferences and distribution network
carry an impact on expense ratio. Laplante (2001) explored the impact of these variables.
He also found that age, size and nature of fund have also association with expense ratio.
The research concluded that the fund expenses are important attribute for selecting a fund
for investment. The investors need to have proper information about the expenses
associated with different fund categories, so that he can make decision with rational
judgment basis.

33
2.5 Mutual Fund Performance Studies in Pakistan
Mehreen and Nawazesh (2011) analyzed the performance of Pakistan’s mutual fund
industry for the period 2006–10, a period which is characterized both by bullish and
bearish markets. The analysis revealed that Islamic funds have better growth as compared
to the conventional funds. Income funds were found suffered due to underdeveloped bond
market, and hight-bill rates resulted in negative excess returns in this period. For stock
funds, market indices and size are significant factors that indicate a preference for large-
cap stocks of managers. This results in consistently negative or insignificant alphas
meaning that no fund outperforms the market. Their study eventually found, all funds did
not outperform well and did not capture the market variations. The same kind of studies
were conducted by researchers like, Afza and Rauf (2009) and Nazir and Nawaz (2010)
both these studies focused and concentrated on finding the determinants of mutual
fund performance. Afza and Rauf explored the performance of open- ended mutual fund
performance for the period 1999–2006 by using quarterly Sharpe ratio. They found that
most of the funds determinants are insignificant; found that past returns predicting future
returns and found other factors such as expense ratio and fund’s asset size were impacting
funds’ performance insignificantly.
Nazir and Nawaz, on contrary analyzed very small sample of 13 mutual funds and found
that asset size does matter and it has positive impact on the on performance. They also
found that management fee has positive impact on the performance of mutual fund.
Sipra (2006) and Shah &Hijazi (2005) conducted researches in evaluating mutual fund
performance. Sipra concluded that only 30 percent of funds analyzed could outperform the
market. Shah and Hijazi found that the funds outperforming the market are mainly due to
the restricted sample of funds employed in their study as compared to Sipra. Several
different funds have evolved in Pakistan and, given their number; it is now possible to
conduct an evaluation according to fund type. Further, other studies have so far been
restricted to traditional Sharpe and Treynor measures for analysis.

34
Nafees et al. (2013) investigated the level of risk in the returns of mutual funds in Pakistan.
This study exhibited the use of measures such as standard deviation, beta and alpha along
with ARCH and GARCH models to find risk. This study analyzed the close and open
ended mutual funds index. The results revealed the close ended mutual funds are more
risky than open ended funds. It was recommended that managers need to be more active
to cover and capture the market volatility.Saeed (2004) found that Pakistan mutual fund
industry has got more growth in which the regulators are vital as the more they will ensure
the code of corporate governance the better the performance of mutual fund industry.
Cheema and Shah (2006) argued that mutual funds are now the vitally important capital
market instrument liked by the investors and general public. They found that investment
in mutual funds has grown up, this phenomenal growth shows the popularity of mutual
funds in Pakistan and hence the world.
Arshad (2013) argued that Mutual funds are the most studied areas of research in
developed countries. These funds are effectively managed and endure to perform better
and yield higher returns due to the professional asset managers. Mutual funds help the
small investors in yielding them better returns and reducing their unsystematic risks in
their financial decisions. The study found and revealedthat age, size, nature and sponsors
of the fund have significant impact on the expense ratio of fund. The research found that
size of fund and age is positively related with expense ratio. The research found that the
expense ratio is a bit high as compared to the local market. The regulators are advised to
cap the expense ratio at certain level to ensure the confidence of the investors, and to
provide the investors a chance to earn more and better. Ali (2015) investigated the
performance of open ended mutual fund in term of their risk and return. The performance
was analyzed through sharp and Jensen ratio and. the study found that out of total mutual
funds schemes, analyzed only 3 mutual funds scheme outperformed the benchmark.
Zulfiqar et al. (2011) investigated the performance of 22 close ended mutual funds listed
on kirachi stock exchange for the period 1999 to2009. The funds were analyzed at their
different states of stock market like Normal, Boom & Recession by using performance

35
measure ratios such as Jensen, Sharpe, Treynor and Sortino. In addition funds were ranked
according to their performance. The results found that all these funds have underperformed
in different states of stock market. The results suggested that the low performance than the
bench mark is due to the inefficient management. Among all the investigated funds for
their performance only Asian Stock Fund shows the best performance but not up to the
benchmark, while performance of Meezan balanced fund was found the worst. Iqbal
Khalid et al. (2009) analyzed the performance of close ended mutual funds categorizing
them for the purpose of investor’s guidance. The funds’ performance was analyzed
through traditional measure like Sharpe Performance Index, Treynor Performance
Index, Jensen Alpha, Sortino Ratio And Informational/ Appraisal Ratio. The results
found very weak results for the selected close ended mutual funds for the period 2001 to
2008. The analyses were made to compare the performance of these selected categories
and found none of the category outperforming the market in true sense, meaning that
Pakistani market is very dynamic and volatile and the managers need proper attention to
capture the variations.
Mansoor and Bhatti (2011) conducted study to find the comparison in the performance of
conventional and Islamic portfolios of mutual funds. The results found that both portfolios
have performed better than the market portfolio within the period. However, the result has
shown on average the Islamic portfolio provides slightly less returns relative to the
Conventional counterparts. The result revealed a statistically significant difference between
the standard deviation of the portfolios, indicating that the Islamic portfolio is riskier than
the Conventional portfolio. The results also revealed that both Islamic and Conventional
portfolios were depended on the market portfolio of which the former portfolio was closely
mirrored to the market movement in relation to the latter portfolio.
Mumtaz et al. (2014) conducted a research study to investigate risk and returns profile of
the sample funds relative to the respective market Islamic and Conventional benchmarks
using panel data analysis from July 2007 to June 2012. Consistent with previous studies
conducted to analyze the performance of Islamic funds; on average the results indicated

36
that there is statistically insignificant difference in return performance of the Islamic
funds relative to the benchmarks. The result also revealed that there is a superior fund
selectivity skill but inferior market timing expertise among the Islamic fund managers
within the period of the study. On the contrary, results signified that the Islamic funds
have been able minimize the risks as compared to the benchmarks and their returns
performance is comparable to the market benchmark. Hence, the outcome of this study
would benefit potential investor and market players towards participating in mutual fund
industry. The same kind of study to evaluate mutual fund performance was conducted by
Shah et al. (2012) investigated the performance evaluation of conventional and Islamic
funds in the context of Pakistan. The performance evaluation was made on the bases of
risk, return, selectivity and diversification of funds. The research was conducted using
the data of 94 conventional and 31 Islamic funds. The results found that Islamic funds are
well diversified and managed as the results obtained better sharp ratio and R-square for
the Islamic funds as compared to the conventional funds.
Razzaq et al. (2012) conducted a research study in the area of mutual fund and investigated
9 Islamic mutual funds that are managed by different funds managers of Pakistan. The
researchers examined that returns of funds are according to their level of risk. The study
also investigated the performance and ability of funds managers by the help of different
models, such as sharp ratio, Turnover ratio, Jensen’s Alpha and information ratio. First
returns were calculated than ratios were estimated, where the results represented that
Islamic funds have significant growth in previous years which indicate that in Pakistan
Islamic funds are growing and investor are attracted by these funds. To check the
performance of Islamic mutual funds, first the average monthly returns were calculate, net
asset value and market index value were specified. The findings indicated that the
performance of mutual fund is speedily growing in Pakistani Islamic Mutual Funds
industry. The similar study relating the performance of Islamic fund was Sheikh (2012).
The researcher stated that Islamic finance works and is based on Shari’ah principles. They
found growth in the Islamic funds but suggested some strategic direction and suggestion

37
which need to be taken for the uplift of Islamic funds. They found several theoretical
problems which need to be addressed and well sort out. The paper emphasized on the
regulatory frame work need to be aligned. Ali and Qudos (2012) investigated the
performance of Pakistani mutual funds. The research investigated the rate of return
associated with different mutual funds and tried to know that how funds managers can
minimize the risk of different portfolios. The study was conducted using 15 mutual funds
for the period 2005 to 2009. The most traditional measures like Sharp ratio and Jensen
ratio were applied for the study. The study found that most of the funds were not
performing upto the mark or bench mark set by the managers of the funds. The study to
investigate the return measures about the mutual fund in Pakistan was conducted by
Khan (2008). He argued that Fund unless adjusted for the risk would not account for the
true measures. He further argued that one year performance does not reflect true
performance picture of fund’s performance returns and investors like to invest in those
funds which provide consistence performance over the period of time. He argued that
money market funds performed below par the market. The similar study to investigate risk
and return of mutual fund was conducted by Jabeen and Dar (2012 ). They argued that
Profitable employments of spare funds not only generate income tosaver but also create
positive impact on country’s economy, thus having two fold benefits. Despite the vast
research on mutual funds, little is known is about mutual funds in Pakistan. The overall
image that emerges from literature is that Islamic mutual funds are not flourishing in
Pakistan. This comparative study was aimed to study risk and return of mutual funds in
Pakistan for the period of 2007to 2012. This research investigated the performance of
mutual funds on the basis of net asset values of conventional and Islamic funds using
Sharpe ratio, Jensen alpha and Treynor ratio. Results showed that in year 2008, 2010, 2011
and 2012 Conventional funds delivered better in return than Islamic funds.

2.6 Empirical Studies in Mutual Fund Performance across the World

Due to the phenomenal growth in mutual fund industry during the past few decades, there

38
have been extensive researches conducted in this area with a variety of different data sets,
periods and methodology used.
In the early stages of mutual fund performance was analyzed through direct comparison
between its returns and other portfolios with similar risk were normally used. For example,
Friend et al.. (1970) and Elton et al. (2002) analyzed the performance of mutual funds
during 1960-1968 by dividing mutual funds into three risk categories (high, medium and
low risk). After categorizations they were compared to the random portfolios with the
same risk categories. The results indicated that the mutual funds did worse than the
randomly selected portfolios.

Jensen (1968), used a single-index model following the capital asset pricing modeland
investigated the performance of 115 U.S open-end mutual funds with various
investment objectives (growth, income, balance, etc.) for the period 1945 to 1964 on the
bases of S&P500 index as a market benchmark. He found that out of the sample funds
taken for the investigation, only three mutual funds out performed on the bases of T-
statistics and the rest underperformed the bench mark. He suggested that the managers do
not have a clear strategy of capturing the market variation and should reduce its expenses
and other loads.
In contrast to Jensen (1968), Ippolito (1989) employed a similar model but used more
updated data; he investigated 143 U.S mutual funds for the period 1965-1984 to investigate
the efficiency of the capital markets and mutual fund performance to see whether active
funds yield better return to compensate for the higher fee charged. He found that only
twelve funds significantly outperformed and only four funds significantly underperformed
the market. The rest funds were in between. He concluded that, on the net of fees and
expenses apart from the load fees basis, mutual funds, overall, have outperformed passive
funds and are large enough to compensate for the load charges, a result which is consistent
with the concept of costly information in an efficient market (Grossman and Stiglitz,
1980).

39
Grinblatt and Titmann (1989) investigated the mutual fund performance for the period
1975 to 1984 on the bases of funds both actual returns and gross returns. They applied the
same Jensen’s single-index measure with four sets of benchmarks. They found better
significant results in term of performance in growth funds when gross returns data were
employed but found different weak results for funds actual returns. Hence, they concluded
that growth funds do well in outperforming the market but the evidence disappeared
because of its high expenses.
Cumby and Glen (1990) analyzed the performance of 15 U.S.-based international funds
for the period 1982-1988. They also employed Jensen’s measure of single index and
documented the same kind of findings as by Grinblatt andTitmann (1989). They found
positive alphas in only three funds, however the alphas were statistically significant. They
also checked the market timing ability for the sample mutual funds’ performance. They
used Treynor and Mazuy’s timing model, and found a negative market timing ability.
Similarly, Malkiel (1995) also investigated fund performance of U.S market during
period 1972-1991 on the bases of Jensen’s single-factor model. He found that the sample
funds underperformed, as the average alpha generated is equal to -0.6% having very low
t-statistic value. He eventually found that on average, mutual funds have underperformed
benchmark, both before and after fees and expenses have been deducted. Gruber (1996)
investigated the performance of equity fund performance during period 1985 to 1994 on
the bases of relative return to the market, Jensen’s measure and other multifactor models
used for the analysis of funds. He analyzed 270 funds argued that both single index model
and muilti factor model underperform the market by 1.56% and 0.65% per year using
respectively. Similar to the studies above, in the U.K based, Blake and Timmermann
(1998) investigated a large dataset of funds for the period 1972-1995, using muilti-factor
three index model, which includes T-bills and bond and dividend yield. The results found
an underperformance by 1.8% per annum of all funds, documenting the weak performance
by managers in capturing the market variations. Quigley and Sinquefield (2000)
documented the findings similar to Blake and Timmermann (1998) for the period 1978-

40
1997 from UK based funds.
Some other studies investigated and evaluated funds’ performance in more detail by
decomposing performance into selectivity and timing ability. For instance studies related
to the market timing ability include Treynor and Mazuy (1966). Who studied the market
timing ability of 57 U.S mutual funds for the period 1953 and 1962 on the bases of
quadratic regression approach. They found that only one fund out of total fifty seven funds
exhibited significant market timing.
Kon (1983) argued that fund manager if has timing ability will adjusts risk level ahead of
the market variations. Hence, he tested the timing ability by investigating the stationary of
the fund’s systematic risk. Heanalyzed thirtyseven fundsfor the period 1960 to 1976 and
concluded that not all some funds have significant timing ability. In contrast, Bollen and
Busse (2001) argued that more often active fund managers do well in adjusting their
portfolios at high frequency. They analyzed the funds daily data for the period 1984 to
1995. They found that, using daily data, 40% of funds shown significance and positive
market timing ability.
Matallin-Saez (2006) analyzed the performance of mutual funds in the context of Spain
and investigated the effect of omitting a relevant benchmark. He applied a variety of
measures like Jensen’s measure, several multi-index models and Treynor and Mazuy’s
market timing, and found that the performance of Spanish mutual funds is not up to the
mark, set by the fund’s managers and funds showed negative market timing but that these
results are not statistically significant. He found that the effect of omitting the benchmark
will leads tomore perverse market timing.
While there was much evidence of inferiority in mutual fund performance from earlier
times, studies in mutual fund performance have become much more comprehensive in the
last decade. The recent studies apply more completed data and sophisticated. Ferson and
Schadt(1996) argued that the all the single- and multifactor measures are biased, since
portfolio risk and returns are fixed through time (known as the unconditional measure).
Due to this fact, they proposed in their model a conditional measure which allows time-

41
varying. They applied both conditional and unconditional measures to investigate the
performance of sixty seven mutual funds in the U.S market during for the period 1968-
1990. They used five predetermined variables for their conditional measure and
incorporate it with Jensen’s single factor measure. Their results showed negative Jensen’s
alphas in overall fund performance. However, the alphas shifted to positive pattern when
predetermined variables were included. They also applied their conditional method to
Treynor and Mazuy’s (1966) and Henriksson and Merton’s (1981) market timing
measures and used 3 self- constructed buy-and-hold portfolios to test the market timing
models, as well as data of sixty seven mutual funds. They concluded that the unconditional
market timing models are miss specified, since the results showed negative market timing
performance even if they are in the buy-and- hold strategy portfolios. They found that
when conditional measures are replaced with unconditional than most of the negative
coefficients are removed. Therefore, they confirm that using their conditional model brings
both statistical and economic significance and makes the performance of the funds look
better.
In a similar other study, Sawicki and Ong (2000), applied both unconditional and
conditional Jensen’s measures, as well as Treynor and Mazuy’s market timing model to
find the performance of Australian funds for 1983 to 1995. They found weak evidence of
positive performance and found negative market timing performance of most of the funds.
In consistent to Ferson and Schadt (1996), they documented the statistical significance of
incorporating lagged information variables, specially, about dividend yield. They also
confirmed that by using the conditional model will shift the alphas to positive signs.
Dahlquist (2000) investigated Swedish funds’ performance during period 1993 to 1997,
applying conditional measure. He found that equity funds performed well during the
period.
Daniel et al. (1997) applied and used their characteristic-based performance measure to
investigate the performance of 2500 US funds for the period 1975-1994. Rather than
conventional belief that fund managers are unable to outperform the market, they

42
revealed that fund managers have selectivity ability but no timing ability. They found
that growth funds performed well. Their results have been confirmed by Chen et al..
(2000), who investigated the mutual funds during 1975 to 1995. They concluded that fund
managers have selectivity ability. They found that fund managers are not successful in
outperforming their stocks around 2% per year. However, selectivity ability vanished
because fund managers usually hold stocks for longer than a year.
Wermers (2000) analyzed mutual fund performance during 1975 to 1994. He showed that
fund managers do well as for as their selectivity skills are concerned but style and high
expenses has vanished the chance of abnormal return. He found that on average, the funds
outperform the market by 1.3%, however their net fund return is 1% lower comparing to
the market. These are due to high trading costs and expenses.
Kosowski et al. (2006) analyzed the mutual fund performance for the period 1975-2002.
The results generated positive alpha but was low as compared to market. The results found
that some of the managers outperformed the market. The results found that only few funds
managers could capitalize the abnormal return.
Using similar measures to Kosowski et al.,Cuthbertson et al.. (2008) investigated the
performance of mutual funds and found different results than US. They found that in UK
the returns are due to luck rather than skills.
While most of the studies are carried on performanceof equity portfolios, Comer (2006)
in contrast investigated the hybrid fund market timing performance. He applied multi-
index model which was based on quadratic variables for stock and bond returns. He found
little evidence of timing ability for the period 1981-1991. He found the significance ability
of the fund’s managers as for as timing ability is concerned.
As a variety of performance measures have been proposed in the literature and evidence
of fund performance is still mixed, it is still not certain which of them provides a valid
performance measure and several studies have been conducted to answer this question.
For example, Grinblatt and Titman (1994) investigated the mutual fund performance by
comparing the funds’ performance to the benchmark and finding the relationship between

43
funds’ performance and funds attributes. They applied three measures and four
benchmarks. They analyzed 109 portfolios and 279 mutual funds for the purpose during
the period 1974 and 1984. They found that mutual fund performance is more sensitive to
the choice of benchmarks than measurement methods are. The CRSP value-weighted
index is the most inefficient benchmark because it admits size-related bias. The study
found P8 index appears to be the most reliable index in this study. In contrast, the choices
of performance measure are not sensitive to fund performance since few funds in this study
exhibit positive market timing.
Kothari and Warner (2001) studied the properties of performance measures for mutual
funds. They employ simulation procedures and check the performance using five different
performance measures, both regression-based and characteristic-based measures. They
found that performance measures are unreliable and have little ability to detect any large
magnitude of abnormal performance, in particular funds whose styles are different from
the value-weighted benchmark market portfolio.
Fletcher and Forbes (2002) analyze mutual performance by validating the different
benchmarks used to evaluate performance in the context of UK. They used a similar
approach to that of Grinblatt and Titmann (1994), employing passive portfolios, formed
on the basis of industry and stock characteristics. They used the data of 724 UK unit trusts,
to test the specification of two performance measures across five different benchmarks.
They found that all five benchmarks have some biases.
Cahart’s (1997) analyzed mutual fund performance through using the additional
momentum factor to the fama French 3- factor model. He found that using the conditional
model yields similar inferences to the unconditional model.
Otten and Bams (2004) analyzed mutual fund performance using different approaches.
They applied the step-wise process to understand the added value of additional variables,
such as size factor, book-to-market factor, momentum factor, bond index and a vector of
information variables. They used a more comprehensive and richer dataset than the
previous studies did in the analyses of the data on both aggregate and style levels. They

44
found the superiority of the conditional factor models over unconditional models and that
Cahart’s four-factor model is the best model to explain mutual fund returns. They conclude
that conditional models add value in the performance measurement of mutual fund.

Baruaet al. (1991) probably made the first attempt to compare the performance of UTI
mutual fund scheme with the overall market. They found out performance by UTI
schemes. Further in the study of data on 11 mutual funds from 1990 to 1994, findings
revealed that almost all mutual funds under performed the over allmarket (Shah&
Thomas, 1994). Jayadev (1996) found out performance to a limited extent. Later on Gupta
and Sehgal (1997) conducted study on mutual fund performance from 1992 to 1996, and
found inconsistency in mutual fund performance and systematic risk characteristics.
Murthiet al. (1997) advocated a new index to measure portfolio performance by the name
of Data envelopment portfolio efficiency index (DPEI), as a relative measure that does not
require the specification of a benchmark. Authors found that managers in aggressive
growth, asset al.location, and income and equity income funds were more efficient in
utilizing resources. Authors also found that efficiency was not related to size and costs of
the fund.

Number of studies demonstrated that either most of the mutual fund schemesare unable
to outperform the broader market or results have been found to be inconclusive (for
example Sethu, 1999; Chakrabarti&Rungta, 2000; Singla& Singh; 2000; Gupta, 2000).
Narasimhan and Vijayalakshmi (2001) found that in almost all the mutual funds under
study, risk level was high as compared to the returns. On the contrary Turanet al. (2001)
found considerably low level of risk in the selected mutual fund schemes. Studies
conducted in later years also observed the inconclusive evidence of mutual fund out
performance over all (Anand&Murugaiah, 2006) or exclusively for growth schemes
(Gupta & Gupta, 2004); equity linked saving schemes (Tripathy,2004a) and balanced
schemes (Chandel&Verma, 2005).

45
In a study comparing both private and public sector funds on various accounts, Panwar
and Madhumati (2006) studied 6 public sector sponsored and 12 private sector sponsored
mutual funds from 2002 to 2005 and examined the differences in the characteristics of the
assets; portfolio diversification and effect of diversification on the investment
performance. Three measures of performance were used in the study namely – Jensen
alpha; Sharpe information ratio and eSDAR or excess standard deviation adjusted return
(Statman,2000). Authors concluded that there was no statistical significant difference
in the mean returnsof public sponsored and private sponsored mutual funds; In terms of
average standard deviation, average variance and average coefficient of variance, there
was a statistical difference between public sponsored, private sponsored domestic and
private sponsored foreign mutual funds. Study also found that there was statistical
significant difference in terms of ESDAR (used as a measure of investment
performance) and Residual variance (used as measure of portfolio diversification
measure) between public sponsored and private sponsored mutual funds. Roy and Deb
(2004) carried out an extensive study on performance and its persistence with respect to
mutual funds. Authors studied 133 open ended mutual fund schemes (from equity, debt
and balanced universe of schemes) from the period 1999 to 2003. Authors used conditional
framework of performance measurement (Ferson & Schadt, 1996; Christophersonet
al.,1998) and found that when fund’s beta was conditioned to lagged economic
information variable (various lagged economic information variables used were interest
rates; dividend yields; term structure yield spread; and a dummy for April effect), the fund
performance did not change. On the other hand when alpha was also controlled for lagged
economic information variables, the fund performance on an average became
significantly negative. Authors concluded that on an average the Indian mutual fund
managers only capture the opportunities from the available economic information; they
do not contribute anything beyond it. Authors also studied the performance persistence
of the Indian mutual fund schemes. To measure the past performance both conditional
(conditional alpha, conditional time varying alpha) and unconditional (unconditional

46
alpha) measures of performance were used. Authors observed that the conditional measure
especially time varying conditional alpha predicted the future fund returns
significantly.
Robert (1988) investigated the effect of size of the mutual fund, using the casual
relationship of fund’s net assets and return. He found that small funds in US do well in
performance as compared to big funds. He argued that small funds better manage the risk.
Moreover, some other researchers have also shown the smaller size of the fund, the more
its operating effectiveness.
Gorman (1991) also found similar results as documented by Robert that small mutual funds
(mostly measured by their net assets) perform in most cases faintly better than large mutual
funds. Another key factor mentioned by Gorman was that most often funds speedily wear
out the economies of scale, which lead to decline in returns. (Becker and Vaughan, 2001;
Chen et al.., 2004).Brown (1995) also found out that funds return shrinks over a period of
time. He also pointed out that past performance of mutual funds is good predictor of
performance in future of mutual funds. McLoed and Mathotra (1994) analyzed the impact
of mutual fund expense as vital and has managers consider it as returns. To complement
the research of McLoed and Mathotra, Korkeamaki and Smythe ( 2004 ) examined in detail
this relationship and provided a different school of thought that investors were not
compensated for paying higher expenses for higher risk adjusted returns. Carhart (1997)
investigated in detail the relationship between fund turnover and funds returns and found
out a negative relationship between them. Similarly, Livingston and O’Neal (1998) have
shown that there is a negative relationship between returns and fund expenses.
Soderlind et al. (2000), Korkeamaki and Smythe, (2004) after the dawn of mutual funds
both in developed and under developed countries, there are number of researchers who
have experimentally studied between the connection of open ended funds’ performance
with its characteristics in different point in time especially in developed countries.
Soderlind et al. (2000) also found that better performance is achieved by smaller mutual
funds having less equity.

47
Aggarwal and Gupta (2007) carried out a performance study on quarterly returns of all
equity oriented mutual funds from 2002 to 2006. Using CAPM and Fama French Model
of investment performance, the authors found contrasting views on mutual funds’
performance. With respect to CAPM, the value added by mutual funds was evident but
it was not the same case with the Fama French Model. Moreover in Fama French
Model, factor portfolios behaved differently on account of the mutual fund performance.
Authors finally concluded by recommending for more systematic research on the issues
involved.
Deb (2008) studied 96 mutual funds from the period 2000 to 2005 for examining
the return based style analysis of equity mutual funds in India. Authors used quadratic
optimization of an asset class factor model (Sharpe, 1988, 1992) to determine style
benchmarks. In total 11 mutually exclusive asset classes were developed for style
analysis. Authors in their findings concluded that on the average, Indian equity mutual
fund managers have not been able tobeat their style benchmarks.
Some of the funds in india outperform the market (Sondhi& Jain, 2005). Especially out
performance has been observed in case of sectoral funds like FMCG, IT and in equity
linked saving schemes (Chandel &Verma, 2005) although the study suffered from small
sample selection. Sondhi and Jain (2005, 2006) also found that open ended equity mutual
funds, small size funds, and private equity mutual funds performed better as compared to
their counterparts. Further in some studies especially during the bear equity phase of 1998
to 2002, mutual fund performance was found to be satisfactory as compared to the broader
market (Sapar & Madava, 2003).
Regarding performance persistence, no evidence of short term performance persistence
has been found in Indian mutual funds (Chander, 2005). Sehgal and Jhanwar (2006)
found no short term persistence using monthly mutual fund data, but were able to find
some evidence of short term persistence using daily data.
Most of the studies on market timing ability and stock selection ability in Indian mutual
fund industry found negative or neutral timing ability (for example Gupta, 2001). In a

48
study conducted by Mishra (2002), author examined timing and selectivity skill of mutual
funds and concluded that selected mutual funds had no timing ability, even though at
individual level some of the schemes had timing skill. On the other hand most of the
studies conducted to assess the stock selection ability found positive stock selection ability
(for example Shah & Thomas, 1994; Jayadev,1998; Rao&Venkateshwaralu, 1998; Gupta,
2001; Chander, 2005; Sondhi& Jain,2006; Deb et al., 2007 etc)
Summarizing the debate on performance persistence and fund’s manager ability with
respect to stock selection and timing, Sehgal and Jhanwar (2008) argued that multi factor
benchmarks provided better selectivity and timing measures, which, in turn, may be
further improved by using daily data instead of other frequencies. Authors also observed
that mutual fund timing ability should be examined in a multi-dimensional framework
with additional measures for timing of style characteristics.

2.7 Stock Selection and Timing Ability of Mutual Fund Managers

Performance measurement and performance persistence are of great interest to both
investors and academicians. If performance persists, it indicates superior skills and
abilities of mutual fund managers. Mutual fund literature identifies two prominent
abilities, that if present, identifies the superior performance of the fund manager. These
two abilities are namely, Market timing ability means correctly assessing the direction of
the market and adjusting the portfolio positions accordingly. Stock selection ability
means correctly forecasting the price movements of the individual stocks and correctly
identifying whether the individual stocks are undervalued or overvalued, and then taking
portfolio positions accordingly. There are two types of models (unconditional and
conditional models) for assessing the presence of market timing and stock selection
ability. Traditional or conditional models (Treynor& Mazuy, 1966; Henriksson &
Merton, 1981) regarded “any information” with the future market return as superior
information, while Conditional models (Ferson & Schadt, 1996) assumes semi strong
market efficiency.

49
The mutual fund literature concerning market timing and stock selection ability is vast and
varied. Some studies found market timing ability to be neutral (for example Treynor &
Mazuy, 1966; Kon 1983; Henricksson; 1984; Chang & Lewellen, 1984) or inferior (Veit&
Cheney, 1982; Chen et al., 1992; Kao et al., 1998). Some studies also provided the
contrasting results of positive market timing ability (Chen & Jang, 1994; Bello &
Janjigian, 1997) especially using daily data in contrast to monthly data employed in
earlier studies (Bollen & Busse, 2001). In a study using conditional models, Becker et al.
(1999) used conditional market timing model based on linking of manager’s payoff
function to the excess returns (over a benchmark) of a fund and timing based on public
information. Authors concluded that the conditional market timing model removes the
negative sign associated with the timing ability as reported in traditional unconditional
models, but conditional models do not yield significant evidence of conditional timing
ability.
The stock selection ability has been found to be less (Chang &Lewllen, 1985) or positive
(Lee &Rahman, 1990; Cuthberstonet al., 2008; Baker et al., 2007). In addition,Volkman
(1999) concluded that there is a negative correlation between fund’s timing and
selectivity performance (perverse timing performance). From the cited review it can be
summarized that most of the studies reveal neutral or inferior market timing ability and
most to positive stock selection ability of the fund manager.

2.8 Factors affecting Mutual Fund Performance

A large number of factors affect mutual fund performance. Factors affecting mutual fund
performance can be segmented as fund related factors (like fund size, expense ratio, fund
style, fund age, fund fees and loads, fund flows etc); factors related to fund family (fund
family size, management structure of the fund etc) or related to fund manager (like
managerial tenure or experience, managerial education); factors related to country
(country’s economic and financial development, country’s border or geography in
case of international funds etc) and factors affecting environment in which mutual fund

50
operates (like economic and legal environment). The factors, which affect mutual fund
performance, are also laterally the determinants of mutual fund flows, since performance
is the major determinant of flow.
Relation between mutual fund size and performance has been widely studied in mutual fu
nd literature. Some researchers argued that managers have a strong incentive to incre
ase fund size (because in the mutual fund industry compensation is typically a fixed perce
ntage of assets under management), at the expense of investor returns (Becker & Vaugh
an, 2001). Molson (2003) argued that fund size has significant impact on the funds expen
se ratio the same was documented by, Tang Cheong (2007) and Zera et. al ( 2007). Many
researchers found that fund size has positive relationship with fund performance (Gorman
,1991, Grinblatt Titman,1994; Peterson et al., 2001, Nazir & Nawaz,2010). But some of t
he researchers found negative relationship between fund size and fund return, as evidence
d from the studies of (Jang & hung, 2003, karlson &persson,2005; Haslem et al.,2008 and
Bablos et al.,2009). Chen et al. (2004) argued that small size funds better perform. Expens
e ratio is also a key determinant of fund performance. There are contradictory views of re
searcher about the impact of expense ratio on fund performance. Ippolito(1989) argued th
at as the expense ratio increases it assure the better performance for high expenses and ha
s positive impact on the fund return. Some other researcher found the same findings (Dro
ms & Walker, 1996, Dowen & Mann, 2007; Afza & Rauf, 2009, Nazir & Nawaz,2010). T
urnover is also the vital factor of fund performance. Turn over effects the fund performan
ce. Carhart(1997) found a negative relationship between turn over and fund performance.
The same negative coefficient was found by Afza & rauf (2010). While some other resea
rchers found a positive relationship between turn over and fund performance (Soderlind e
t al.,2000, wermers et al., 2000). Liquidity is also a very essential determinant which affec
ts the fund performance. There is diverse kind of feelings researchers have about the impa
ct of liquidity on fund performance. Glenn (2004) argued that liquidity has significant neg
ative impact on fund performance. Similar results have been reported by many researcher
s (Dukes &Davis, 2006, Afza & Rauf, 2010; Nazir & Nawaz, 2010). Management fee is a

51
nother variable affecting the fund performance. George (2001) argued that management e
fficiency has positive impact on fund performance. Other researchers also found similar fi
ndings (Pushner et al.., 1999; Gallaghar, 2003; Joseph, 2004). The net cash flow is the in
dication of the growth in the funds’ assets. As the movement of cash would have an imp
act on the funds beta and hence, the performance will be affected. It is due to the fact th
e funds receiving huge amount cash inflow could not capitalize the investment in profita
ble business avenues. This makes a larger cash position in the portfolio and subsequent
ly lowers the mutual fund’s beta. Conversely, large cash outflows force the mangers to l
iquidate its assets. These help funds managers from trading on information and c
ause him to show negative market timing ability. Warther (1995) suggested that net flo
w negatively effects the portfolio beta (Ferson and Schadt, 1996). Edelen (1999) argued
that there is negative association between the funds abnormal return and invest
ors’ flows. Furthermore, Benson and Faff (2006) studied the association of money flows
with equity fund performance. They found evidence supporting Warther’s (1995) and E
delen’s (1999) who found that money flow has negative impact on funds’ performan
ce. Gruber (1996) argued that funds with higher net flow generate higher return as compa
red to lower net flows funds. Zheng (1999) documented that funds having positive cash f
lows outperform those having negative cash flows. Sapp and Tiwari (2004) argued that s
mart money is vital as it affects the stock momentum phenomenon. They predicted that h
aving control stock return momentum, the smart money effect would likely to be disappea
red. Dahquist et al.. (2000) found that lagged flow has an impact on the funds’ performan
ce. However, Gharghori et al.. (2007) argued that t h e smart money effect would not be
explained by the stock return momentum. Ferreira et al.. (2009) documented the abov
e findings. The relationship between fund age and fund performance has not been p
roperly explained in previous studies. It is presumed that Funds normally experience hig
her costs at the introduction and found that older funds do well than younger funds.. Nev
ertheless, one could argue that managers of young funds are more active. This is confir
med by Blake and Timmermann (1998), who revealed that funds are likely to perform b

52
est during their first year of existence. Similarly, Otten and Bams (2002) found a negati
ve correlation between performance and fund age in some European countries during th
e period 1991-1998. Nevertheless, Prather et al.. (2004) and Ferreira et al.. (2009) found
no relationship between age and fund performance. Afza & Rauf (2009) investigated the f
und age impact on fund performance and found insignificant positive impact on fund perf
ormance in Pakistani context.

The fund family is another characteristic of the fund which has vital impact on funds’ perf
ormance and the decision making of the fund’s managers. Ensuring more funds in a famil
y helps to get economies of scale. (Khorana & Servaes, 1999, 2005). A larger fund famil
y helps in better research quality. Furthermore, it helps in reducing the search cost espec
ially f o r investors due to easy recognition (Gaspar et al.., 2006).
However, Chen et al. (2002) found that fund family does not significantly impact
the performance of fund. In contrast, Guedj and Papastakaikoudi (2004) revealed that
persistence of performance among large fund families. In addition, Ferreira et al.. (2009)
revealed that fund age has significant role in the fund performance of both US and
international funds. Large sized mutual funds present several advantages like
spreading fixed expenses over a large asset base; more research resources; more
investment opportunities; ability to negotiate better spreads and decline in brokerage
commissions (Brennan & Hughes, 1991). Despite all these advantages large funds
found problems in performance persistence (Gruber, 1996; Berk& Green, 2004). Also due
to their large size, large funds often trade large block of shares, attracting attention of
market participants (signaling impact) and therefore incurring significantly higher impact
costs (Lowenstein, 1997; Chen et al., 2004); other reasons can be administrative
complexities and change in optimal behavior of fund manager (Indro et al., 1999). On the
other hand, small funds have been found to be more active and agile (Cremers and
Petajisto, 2009).

53
Researchers therefore have found mixed evidence in relationship between fund
performance and fund size (Grinblatt& Titman, 1989b; 1994). Some have found negative
relationship or diseconomies of scale (Indroet al., 1999; Berk& Green, 2004). As fund size
increases, performance erodes due to organizational diseconomies (Stein, 2002); inability
to apply investment strategy on a large scale (Pollet and Wilson, 2008) or trading costs
(Perold& Solomon, 1991; Edelenet al., 2007). Many types of organizational diseconomies
explain why small organizations outperform large organizations. Such diseconomies can
be bureaucracy and related coordination costs (Williamson, 1975; 1988); or hierarchy
costs (Aghion & Tirole, 1997; Stein, 2002).
In a significant study discussing the relationship between fund performance and size
conducted by Chen et al. (2004), author found that fund returns decline with lagged fund
size and this relationship has been more pronounced among funds that have to invest in
small and illiquid stocks, suggesting that adverse effects of scale are related to liquidity.
Authors finally concluded that scale erodes fund performance because of interaction
between liquidity and organizational diseconomies. Indroet al. (1999) found that as the
fund size increased three years return increased but at the same time there have been
diminishing marginal rate of return from increasing total assets under management. Some
research pieces have also found evidence of positive relationship between fund size and
performance (Ottens&Bams, 2002), but more recent work points out towards
diseconomies of scale.
Much has been written about the mutual fund performance as affected by the investment
style. One of the earliest researches that indicated that style plays significant role in
portfolio performance is study conducted by McDonald (1974) in which he found that risk
adjusted performance of more aggressive funds dominated that of similar conservative
funds. Such significant relationship was also evident in other studies (like Malkiel, 1995).
In a similar vein it was found that value stock portfolios outperform growth stock portfolio
on risk adjusted basis (Capaulet al., 1993; Lakonishok et al., 1992; Fama and French,
1998; Chan and Lakonishok, 2004). Brown and Harlow (2002) observed that more style

54
consistent funds significantly outperformed less style consistent funds on risk adjusted
basis. Authors also found a positive relationship between measures of fund style
consistency and persistence of its future performance.
Some authors have also criticized traditional methods of classifying investment
styles due to their subjective nature (Bartolomeo & Witkowski, 1997) and new
classification systems based on style factors have been developed (Brown & Goetzmann,
1997). The subjective nature of styling and misclassification can be due to the result of
‘tournament hypothesis’ as advocated by Brown et al. (1996) according to which
managers of different funds in the same objective class have different incentives to
adjust portfolio risk depending on relative performance. Evidence also exists that
investment style and performance persistence are connected (Teo & Woo, 2004) in which
authors demonstrated that portfolio of past winners and losers continue to copy their
previous behavior, and persistence effect declines slowly in style adjusted returns.
More style consistent funds exhibit less portfolio turnover (therefore less transaction costs)
also managers are less likely to make errors in asset al.location and security selection
(Barberis & Shleifer, 2003). Style drift for opportunity may lead to alteration of portfolio
risk and thereby generating suboptimal performance (Huang et al., 2009) and finally for
style consistent fund manager, evaluation becomes easier and much easier for them to
signal their superior skills to the investors.

There is some contrasting evidence also, under performance can result from being in close
to style (Asnesset al., 2000). More style consistent fund managers are unlikely to capture
the benefits of timing as available from style rotation (Swinkles and Tjong-A-Tjoe, 2007)
and during certain business cycles less consistent style managers may out perform more
consistent style managers.
The effect of fund’s age on performance is mixed. While younger funds may be considered
as agile and active, they may lack experience. Some researchers have found that younger
funds performed better than the older ones (Otten & Bams, 2002; Cremers & Petajisto,

55
2009). Some have found no relationship between age and fund’s performance (Chen et al.,
2004).
The relation between mutual fund returns and fees provides a test of the value of active
management. Empirical evidence on relation between mutual fund performance and
fees is mixed. Some researchers have found positive relationship between fees and better
net performance (Ippolito, 1989), some have found insignificant relationship between
fees and performance (Chen et al., 2004), especially between fees and net performance
(Carhart, 1997; Dahlquistet al., 2000), or negative relationship between fees and
performance (Gil-Bazo& Ruiz-Verdu, 2009). Gruber (1996) pointed out that on average
actively managed funds under performed index funds, but index funds charge lower fees.
Apart from expenses, funds also charge loads in the form of front end (at the time of
purchase), or back end (at the time of sale). Loads have been found to discourage
redemptions (Chordia, 1996) and are effective tool in controlling fund’s volatility (Green
et al., 2007). Some researchers have found no relationship between loads and
performance (Chen et al, 2004) and some have even found negative relationship
between loads and mutual fund performance (Carhart, 1997; Polet and Wilson, 2008).
Further Morey (2003) found that size of load has little predictive ability in determining
future performance, as no significant difference was found between high load and low load
funds after adjusting for loads.
Chalmers et al. (1999) estimated the annual trading costs of fund managers’ and found
strong negative association with the return performance. In fact authors found that trading
costs have more explanatory power for fund returns as compared to turnover. Researchers
tried to explain this finding in different contexts. One of the contexts they cited was
that fund managers do not follow the rule that informed trader initiates the trade only when
the expected value of information is more as compared to the trading costs (Grossman
&Stiglitz, 1980). Authors finally concluded that low cost proxy for trading costs might be
valuable information, as trading costs are important in explaining fund returns. Dellva and
Olson (1998) studied 568 mutual funds over a period of 1987-1992 for determining the

56
fundamental differences in various types of fees with respect of expenses and fund
performance. Authors found that superior performance funds usually have lower
expense ratios; front-end load (entry load) funds usually have lower risk adjusted
performance; expenses increased when turnover increased but with no corresponding
increase in performance due to an increase in turnover. Also, the superior performance
was not indicated when there was no fee.
Relationship between fund flows and fund performance came to highlight when Gruber
(1996) proposed ‘Smart money’ hypothesis according to which investors are able to
detect skilled managers and direct their money to them. Therefore there can be a
positive relationship between flow and future fund returns. Empirical evidence exists
that funds having net flows in previous three months outperform the funds that
experience outflows in the same time period (Gruber, 1996; Zheng, 1999). But some
researchers have found no significant relationship between fund flows and fund
performance (Dahlquist et al., 2000).
A rich literature exists that examine a relationship between job tenure (proxy for
managerial experience) and job performance. Most of the earlier studies show that
experience with a task improves job performance (Christensen-Szalanskiet al., 1983)
Research has also demonstrated that experience has led to increase in knowledge but not
the skills (Nass, 1994). More significant result with respect to mutual fund has been found
to be in the study conducted by Golec (1996) in which tenure (experience) of the fund
manager has been found to be most significant and positive prediction of the fund
performance. Some other studies have also found significant and positive relationships
between managerial tenure and mutual fund performance (like Mikhail et al., 1997);
betweenspecific experience and performance (Clement, 1999); while some researchers
have found general experience and not specific experience as related to the performance
(Fortin et al., 2008). In some studies the relationship between managerial experience and
performance has been found to be negative or with inconclusive evidence (Jacob et al.,
1999; Costa et al., 2006).

57
In the closely related studies regarding the impact of manager’s age on mutual fund
performance, some have found no significant relationship (like Chevalier & Ellison,
1999a; Fortin et al.., 1999), negative relationship due to impact of aging (for example
Peterson et al., 2001) and positive relationships because of learning and experience
(Filbeck and Tompkins, 2004; Ferreira et al., 2006) but such positive influence decreases
with the longer term horizon (Ejara & Nag, 2009). Bliss and Potter (2002) found that in
contrary to the popular belief, female fund managers were found to be more risk seekers
and tend to outperform as compared to their male counterparts, but after controlling for
various parameters, authors found no significant difference between male and female fund
managers performance in cross sectional regression.
Literature has cited relationship between managerial education and fund performance.
Importance of education in performance has been highlighted and professional
education qualifications like MBA or CFA designations have been found to be associated
with better performance (Topel, 1991; Golec, 1996). Some studies pointed out more
consistent performance by CFA charter holders (Switzer & Huang, 2007). Researches have
found that investment managers attending more selective undergraduate universities
exhibited higher raw excess returns (Chevalier & Ellison, 1997) but not higher risk
adjusted excess returns (Chevalier & Ellison, 1999b).
Fund families do affect mutual fund performance as large fund families have incentives
and means to push some of their funds (Guedz & Papastaikoudi, 2003). Size has altogether
a different impact, when is related to the fund family. There can be more economies of
scale, synergies and wider sharing of costs if fund family is large. Also there are economies
of scale in trading commissions (Chen et al., 2004) thereby reducing cost of opening new
funds (Khorana & Serveas, 1999). Therefore impact of fund family size on mutual fund
performance has been found to be positive (Chen et al., 2004). Yet some researchers
reported an evidence of cross subsidization in large fund families (Scharfstein& Stein,
2000), which states that larger fund families transfer fund performance from low fee fund
to higher fee funds and create ‘stars’ in the fund family that generate flows not only to

58
itself but to the entire family (Nanda, 2004; Gasper et al., 2006). Because of this spillover
effect, mutual fund families strategically transfer performance across member funds to
favor those more likely to increase overall family profits (Gaspar et al., 2006). Huij and
Verbeek (2007) found that after positive returns, low marketing expense funds in high
marketing expense fund families have substantially larger inflows as compared to the
similar funds that are operating in low marketing expense families, whereas both have
similar outflows after negative returns. Authors explained this phenomenon as higher
visibility effects of lower marketing expense funds, therefore lowering of search
costs and subsidization effect of low marketing expense funds by other family members.
Fund families also seriously consider resource allocation to different funds and have a
favorable attitude towards the well performing funds. One of the main resources fund
family has that of the fund manager. Since literature has shown that personal
characteristics of the fund manager can help predict superior stock selection ability
(Chevalier & Ellison 1999a; 1999b). Therefore mutual fund families place due importance
on this resource and their performance both in terms of asset growth rate and return is
linked to top management turnover (Khorana, 1996; 2001). There also exists empirical
evidence that relates lagged fund return with promotions and demotions of the
managers (Hu et al., 2000). On other hand, Ivkovic (2002) argued that being part of
large family might improve performance because of other spillovers. Massa (2003)
found a statistically strong and negative relationship between performance and degree of
product differentiation, which exists at both the family and category level.
The classical decision making model has been widely used in financial research
(Fiegenbaum, 1990) which argues that decision makers are knowledgeable and rational
with access to all the information necessary to make valid decisions. With this perspective,
it is argued that different alternatives to the same problem should lead to same optimal
performance outcome, whether decision is made by individual, group or organization.
Behavioral decision making model (for example Barber and Odean, 2000) offers a
contrasting perspective for the study of human choice behaviour. Behavioral decision

59
making model prompted some studies in individual versus group decision making
(Sniezek & Henry, 1990; Vollrath et al., 1989) which concluded that recall and recognition
of relevant information is better in groups. Behavioral decision making theory argues that
group members tend to pool and integrate their resources when the task is complex and is
completed under high levels of uncertainty (Hinszet al., 1997). But still some studies
observed that individuals operating in group decision making may be subject to
polarization ( Burnstein & Vinolur, 1977).

The question of whether management structure affects performance of fund is relevant in

the above cited context. Much of the literature on individual versus group decision making
and its relationship to performance has been cited in the context of laboratory settings
(Hogarth & Reder, 1987). When individual fund manager manages the fund, there is less
uncertainty in decision making and also small funds can only be managed by individual
fund managers. On the other hand, funds, which are managed by team members, have
more resources more investment alternatives which can help reduce uncertainty (Tindaleet
al.., 1993), thereby improving the performance. But in team managed funds there can be
‘hierarchy costs’ (Chen et al., 2004) and larger funds may experience organizational
diseconomies. Also in team managed funds, there is more competition to convince an idea
(Aghion & Tirole, 1997; Stein, 2002). The impact of team management on funds’
performance has been found insignificant (Prather & Middleton, 2002; Bliss et al., 2008)
or poor (Chen et al., 2004; Massa et al.., 2009).
There are hardly any studies that relate macro-economic factors of the country with the
mutual fund performance. Yet some researchers have found positive relationship between
country’s economic development and mutual fund performance (Christofferson &
Sarkissian, 2009). Financial development has some advantage for mutual fund
performance because of higher liquidity and lower trading costs. Mutual fund performance
has also been linked with business cycle as Lynch and Musto (2003) found that conditional
mutual fund performance moves with the business cycles.

60
Literature has also cited that effect of geographic distance and country’s boundaries
on investment performance. The underlying concept is whether local or foreign investors
are at information advantage. Some studies have demonstrated that local investors
outperform foreign investors (Choeet al., 2005; Dvorak, 2005; Teo, 2009). Some have
found evidence that foreign investors are better informed (Grinblatt & Keloharju, 2001;
Froot & Ramadorai, 2008) and some have found no difference (Kang & Stulz, 1997).
Ahmed and Nanda (2005) examined the performance of 191 emerging market open ended
stock and bond mutual funds over the period 1980-2000, conditional on US monetary
policy. The authors found that that while the emerging market stock mutual funds under
performed (lower returns, higher risk and lower Sharpe ratios) their corresponding country
indices, the performance of the bond funds was in line with their benchmarks. Further
during restrictive policy regimes, the emerging market stock mutual funds have higher
Sharpe ratios (the largest difference was found for Asia Pacific funds). Similar results were
found for the emerging market bond mutual funds.
In a very comprehensive and landmark study on determinants of mutual fund performance
(Ferreira et al, 2006) studied funds from 27 countries. The major findings of the
study were: 1) Domestic funds outperformed international funds and US domiciled
funds outperformed any other funds elsewhere in the world. 2) Negative relationship
was found between fund size and performance but that was not evident for non-US and
international funds. 3) Fund age and fees were negatively related to mutual fund
performance. 4) Funds belonging to large fund families, individually managed funds and
funds widely distributed in several countries had better performance. 5) Domestic funds
located in developed countries, those with liquid stock markets and strong legal
institutions, had better performance.

2.9 Comprehensive and Detailed Studies on Mutual Fund Selection Behavior

In one of the studies dealing with quantitative approach to mutual fund selection,
multi-criterion approach to mutual fund selection was proposed by Cook and Hebner

61
(1993). The study found that the most popular approach for evaluating mutual funds
employs only single criteria, the fund’s mean, risk-adjusted rate of return. Multi-criteria
methodology involves numerous factors like standard deviation of fund’s alpha, front and
back end load fees, the level of diversification, quality of service and so on. It was also
recognized that individual investors possess dissimilar attributes and preferences and
hence allow investors to formulate different ratings (and consequently rankings) of the set
of competing mutual funds, which ultimately gets reflected in their purchase / selection
decision.
In a landmark study of the mutual fund purchase decision by investors, Capon et al. (1996)
explored the extent to which investors make purchase decisions consistent with modern
finance theory. The study investigated the manner in which investors make investment
decisions for mutual funds. Study reported that investors’ consider many non-performance
related variables. Study results demonstrated that the mutual fund investment decision was
better considered in a multi-attribute framework, where return and risk were merely two
aspects in a set of attribute whose importance varies across investors. The authors also
argued that as mutual fund purchase value increases, investors would behave in a more
rational manner, simply because of the magnitude of potential gains and losses
involved. Authors found published performance rankings, advertising and commission
based financial advisors as major sources of information. In the study, although past
mutual fund returns and level of risk were rated the two most important factors in
aggregate, several additional factors were also found to be relevant, namely - amount of
sale charges, management fees, reputation of fund manager, fund family, clarity of the
fund's accounting statement, recommendation from a financial magazine or newsletter,
availability of telephone switching, the fact that funds are already owned in that
family, and a friend's recommendation.
Alexander et al. (1998) studied the characteristics, investor knowledge and sources of
information for mutual fund investors. Authors surveyed 2000 randomly selected mutual
fund investors. Data on demographics, financial, fund ownership attributes, familiarity

62
with costs, investment risk and sources of information was collected from the surveyed
investors. Authors reported the major findings as:
1) most widely used source of information was the fund prospectus followed by
employer provided material, newspapers and magazines, friends and work
presentations in that order.
2) Investors were aware of risk in investing in mutual funds.
3) Slight positive relationship was found between the current and future performance. 4)
Financial literacy has been found to be higher for prospectus users as compared to
investors using other sources of information.
Saraoglu & Detzler (2002) based on analytic hierarchy process presented a framework for
selection of mutual funds that took into account unique preferences and constraints of
individual investors. Model helps the advisor and investor to list the individual selection
criteria (both qualitative and quantitative) and the relative importance of each
criterion. They concluded that the model is flexible, user friendly and consistent
throughout the portfolio decision making process.
Jones et al. (2005) discussed the fund selection ability of mutual fund advisors, while
recommending them to their clients. Authors surveyed 500 financial advisors in order to
assess their decision process in buying mutual funds. The authors concluded that financial
advisors put great importance on objective information sources such as independent
performance rankings and much less emphasis on fund advertising and popular press
publications. Financial advisors also emphasized performance relative to other funds with
similar objective, style and risk, while selecting mutual funds for recommendations. They
didn’t much emphasized on sales load and other fees. Authors highlighted the value
that financial advisors contribute to their clients as they emphasize important information
in the mutual fund selection which sometimes, individual investors cannot or are unable
to do so.
Bollen (2007) examined mutual fund attributes and the investor behavior especially with
respect to investments in socially responsible funds. Author found significantly lower

63
monthly fund flow volatility in socially responsible funds as compared to conventional
funds. Author also tried to establish relationship between annual funds flows and lagged
fund returns (socially responsible funds in comparison to conventional funds) and the
finding was that investors in socially responsible funds exhibited a significantly larger
response to lagged positive returns and smaller response to lagged negative returns.
Author reported that evidence suggests that preference of socially responsible investor can
be represented by conditional multi attribute utility function (deriving utility from
socially responsible attribute) especially when these funds deliver positive returns.
Bergstresser et al. (2009) attempted to quantify the benefits that investors get in exchange
for what they pay to brokers or financial advisors for fund selection advice. Authors
compared broker sold and direct sold funds from 1996 to 2004 and found that there were
no sound tangible benefits to the investors from broker sold funds. The brokers sold
channel did not show any evidence of superior aggregate market timing ability. As
compared to direct sold funds, broker sold funds delivered lower risk adjusted returns even
before subtracting distribution costs. They attributed the findings explaining that
investors buy through brokers either were getting substantial non tangible benefits or
there was a conflict of interest between brokers and their clients.

2.10 Behavioral Decision Framework in Mutual Fund Selection Process

One of the three theoretical frameworks under which mutual fund investors makes
decision about the fund selection is the behavioral decision framework, which has gained
importance since the pioneering efforts of behavioral researchers. Modern finance
theory advocates the concept of optimal choice in financial decision making, but is too
narrow and limited in its approach in fully describing the actual choice (Debondt & Thaler,
1995). In behavioral finance framework of mutual fund selection, emphasis is on
describing psychological biases that affect investor behavior. Numerous psychological
biases have been discussed and described in detail in behavioral finance literature. Some
of the important biases that have been identified to influence mutual fund investors

64
are namely: Representativeness heuristic (Kahneman & Tversky, 1974) which further can
cause biases like framing (Tversky & Kahneman, 1986) and mental accounting
(Hirshleifer, 2001); prospect theory (Kahneman & Tversky, 1979) and the resulting loss
aversion and disposition effect (Shefrin & Statman, 1985; Odean, 1998); mental
accounting (Kahneman&Lovallo,1993); cognitive dissonance (Goetzman & Peles, 1997);
Status quo bias (Samuelson & Zeckhauser, 1988) and the related endowment bias
(Kahnemanet al., 1992); and over confidence bias. Some of the relevant behavioral studies
related to mutual fund selection decision are presented in this section.
Kahneman and Tversky (1974) examined a heuristic process (which they called
representativeness) which they defined as the “subjective probability of the event, or a
sample, is determined by a degree to which it is similar in essential characteristic to its
parent population, and reflects the salient features of the process by which it is generated”
Representativeness heuristic or (focusing on similarities) can be interpreted in a way, that
past performance of the mutual fund becomes a dominant selection criteria for the
investor, irrespective of the future expectation of returns.

Kahneman & Tversky (1979) provided theoretical explanation for why investors sell
their winning funds. They made use of the prospect theory, which they regard as the
alternative to the utility theory. They argued that people are loss averse; they have an
asymmetric attitude to gains and losses. Investors get less utility from gaining, say Rs 100
than they would lose if they lost Rs 100. If investors use the purchase price of their mutual
funds as a reference point, prospect theory predicts that mutual fund investors would be
more likely to sell their winning mutual funds than their losers. Tversky and Kahneman
(1986) described about rational theory and framing of decisions. Authors observed
that decisions problems if presented as alternate descriptions give rise to different
preferences, which is contrary to the principle of invariance that underlies theory of
rational choice. Authors concluded that violation of the rational theory of choice is due to
the framing of decisions.

65
Samuelson and Zeckhausar (1988) discussed about status quo bias that is doing nothing
or maintaining one’s current decision. Authors studies on health and retirement plans of
individuals and found that status quo bias was substantial in real decision making. Shefrin
and Statman(1985) commented upon the concept of Kahneman and Tversky’s
approach of aversion to loss realization. The authors discussed this phenomenon into
wider theoretical framework, which included elements of mental accounting, regret
aversion, self-control and tax consideration. They concluded that gain and loss realization
explanation is not limited to the concept of tax realization, but also depends on other
elements of this framework.
Goetzmann and Peles (1997) discussed why some investors retain their mutual funds
which consistently under perform. They hypothesized that this may be due to investors
supporting their own past decisions (cognitive dissonance) and believing that something
they own is superior to something they do not own (endowment effect). Authors carried
out a survey on mutual fund investors and found that cognitive processes used by
investors for inaction were based on biased past performance beliefs and this has
been true even for the well informed investor. Authors also found strong evidence
for endowment effect. Odean (1998) demonstrated strong disposition affect among
investors, which was not motivated by desire to rebalance portfolio or to avoid higher
trading costs of low priced stocks. Nor disposition effect was justified by subsequent
portfolio performance. Authors concluded such behavior to be sub optimal and lead to
lower return.

In an extensive study on behavior of mutual fund investors, Barber et al. (2000)

contributed significant findings to the mutual fund literature. Authors studied purchase
and sale decisions of 30000 households from 1990 to 1996 and found three important
findings.

1) Investors bought fund with strong past performance.

66
2) Investors sold funds with strong past performance, but were reluctant to sell their
loosing funds; they were twice as likely to sell a winning fund as compared to sell a losing
fund.

3) Investors were sensitive to the form in which expenses were charged; they show
asymmetric behavior to fees as they were less likely to buy mutual funds with high
transaction fees but they were rather insensitive to fund’s operating expense. Authors
associated the above findings with the behavioral biases of the investors, they argued
that investors displayed behavioral bias of representative heuristic (when they
bought fund with strong past performance); disposition effect (when the investors were
reluctant to sell their losing mutual fund shares) and framing effect (when they reacted
differently to different forms of expenses). Authors further observed that buying past
winning funds may be productive (because of some evidence on performance
persistence) but selling past winning funds and being insensitive to the operating expense
ratios is rather not rational and clearly counterproductive.

Chen et al (2004) discussed the behavior of Chinese investors. Authors gathered

sample from 1998 to 2002. Authors found that individual investors exhibited behavioral
biases and made trading mistakes. Specifically they exhibited overconfidence,
disposition effect and representativeness heuristic bias. Authors found that more
experienced investors were more inclined towards making trading mistakes, specifically
exhibiting the disposition effect and representativeness. Authors further found mixed
evidence for relationship between experience and over confidence.

2.11 Consumer Decision Framework in mutual Fund Selection Process

A large number of factors influence individual purchase decision (Ersamuset al., 2001).
Several researchers have regarded a consumer decision making process as a five stage
process (for example Schiffman & Kanuk, 2000). These five stages are: 1) Factors
influencing the cognitive process leading up to problem recognition (environmental

67
factors), 2) Information search, 3) Evaluation of alternatives, 4) Choice and 5)
Evaluation of outcomes. In addition, factors relating to consumer social environment
like cultural, economic and institutional environment, demographic, social and
personal characteristics have an influence on consumer purchase decision (Chisnall,
1995). These influences in totality lead to need recognition and the consumer starts with
information collection process. As such consumers can get information from external and
internal sources of information. From mutual fund perspective, the internal sources of
information can be (for example, past experience) and external sources of information
can be (for example; agents, families, friends and advertisements). These sources get
transformed into selection criteria (like performance, risk, services etc), which
ultimately form intentions and investors, arrive at final decision making.

Howard (1969) argued that higher is the financial risk involved in the purchase decision,
more extensive is the search for information. Although consumer decision making process
requires the use of objective criteria for evaluation of alternatives (Schiffman & Kanuk,
2000) not all products are evaluated in same fashion (for example Engle et al., 1990). In
financial services industry, purchasing behavior is highly influenced by type of financial
product being purchased (Betts,1994). Same was pointed out by Capon et al. (1996) who
argued that selection criteria employed for mutual fund selection have been determined by
personal factors; brand or product features and purchase context.

Ersamus et al. (2001) further observed that complexity of decision making process rises
with the product importance, and if financial risk involved is high, consumers usually
engages in more extensive search for information. Although selection criteria help in
comparing the alternatives, final decision making is the result of application of specific
decision rules like heuristics or multi attribute decision rule (for example Fishbein &
Azjan, 1975).

68
But there is a limit to information processing ability, and adding to it complex and vast
information (especially in case of financial product like mutual fund) makes investor to
look at alternative decision making rules ( Kivetz & Simonson, 2000). In view of it other
variables like variables related with the visibility of the fund (for example fund size, fund
family size, media coverage) acts as determinant of mutual fund flows. Since the financial
products are complex to understand, investor usually calls for assistance of third party as
a substitute for active information search. As a consequence, investor puts great emphasis
on brand image and past performance (Beckett et al, 2000).

2.12 Mutual Fund Flow Determinants

Mutual fund new flows are usually defined as change in total net asset value minus asset
appreciation. Most of the empirical studies make use of normalized flows to correct for
fund size effect. Flows are usually regarded at the end of the period. Some researchers also
considered aggregate flows (category flows) as different from individual flows (for
example Sirri & Tufano, 1998; Elton et al., 2003) despite the high correlation between
the two. While earlier research indicated a weak relationship between performance and
flow (for example Smith, 1978; Woerheide, 1982), subsequent research presented a robust
relationship. Most of the studies indicated, that performance flow relationship is
significantly positive (for example Ippolito, 1992), different studies (for example
Cashmanet al., 2006) also indicated that performance flow relationship is different for
different type of funds. In contrary to the evidence provided by performance persistence
literature, which states that performance persistence is strong in case of poor performers,
investors direct their investments in top performing funds (Goriaev et al., 2002; Barber et
al., 2005).
Several studies ( Spitz, 1970; Chevalier & Ellison, 1997; Sirri & Tufano, 1998) have
documented that abnormal positive returns generate disproportionately more inflows than
abnormal negative returns would generate outflows. Further because of spillover effects

69
in mutual fund families (Nanda et al., 2003) performance flow relationship is convex in
nature. This means that it is sufficient for the mutual fund family to only have some well
performing funds in order to experience large inflow in its assets under management.
Harless & Peterson (1998) premised on two models for the observation that poor funds
continue to exist. Two models used were – holdings in mutual funds are determined on
the bases of past risk adjusted returns and are subject to partial adjustment. The other
model is behavioral concept of Representativeness heuristic. Researchers studied 100
funds over a period of 1977-1992 and found that investors do not move in accordance with
Jensen’s alpha prescription. Rather judgments about the fund preference have been
heavily biased by recent gross returns.
Further it has been found that performance in more recent years has the strongest impact
on mutual fund flows (Berkowitz and Ketowitz, 2000) but some studies have also
demonstrated the importance of performance achieved in last 6-8 months (Cashmanet al.,
2006). Edwards and Zhang (1998) examined the relationship between aggregate
mutual fund flows into stock and bond mutual funds and the stock and bond mutual fund
returns. Fund flows were examined from 1961 to 1996 (30 year period in equity funds)
and for 1976 to 1996 (20 year period in bond funds). The study found that the relationship
of flows and return runs in the opposite direction as stronger relationship has been found
between returns and flows (as compared to the relationship between flow and return). That
is to say higher return causes large equity and bond fund flows. The results were found to
be robust to both the econometric procedures used namely Granger causality analysis and
Instrumental variables method.
Warther (1995) observed that flows into stock funds and bond funds were strongly
correlated whereas there was a negative correlation between stock funds and money
market funds. The other findings of the study were – flows into stock funds, bond funds
and precious metal funds were correlated with returns from their respective sectors;
no positive relationship was found between flows and lagged returns. Further no strong
association was found between fund flows and parameters like investor sentiment; small

70
stock behavior and close end fund discount behavior. Santini and Aber (1998) further
extensively studied Warther’s (1995) idea and explored the relationship between
aggregate mutual fund flows and several variables like interest rate levels; risk adjusted
and non-risk adjusted performance measures; changes in personal disposable income.
Authors found that new money flows were negatively correlated to real long term
interest rates and positively related to disposable personal income and stock market
performance. However, on the similar lines of Warther (1995) they found no statistical
significance between flows and lagged performance.
Zheng (1999) studied the fund selection ability of mutual fund investors. Using Grinblatt
and Titman’s (1993) performance measure, author found that mutual fund investors were
able to make buying and selling decisions on the basis of good assessment of short term
performance. The study confirmed the ‘Smart Money Effect’ in that the aggregate newly
invested money in equity mutual funds was able to forecast short term future fund
performance. That is to say that those funds that receive more money subsequently
performed better as compared to the funds that lose money. Although no statistical
evidence was found that investors could beat the market by investing in the funds with
positive money flow, yet there was evidence that positive money flow in small funds
tended to outperform the market. The smart money effect was found tobe short lived to
the extent of 30 months. Author further concluded that the smart money effect was not due
to macroeconomic information but was due to fund specific information.

Studies have also demonstrated negative relationship of volatility and net cash flow (for
example Shuet al., 2002). Investors have been found to be more sensitive to risk adjusted
performance measures (Guercio and Tkac, 2002; Shuet al., 2002). But other performance
metrics like raw measures of performance evaluation or ranked measures have been
equally found to be valuable (Berkowitz & Ketowitz, 2000).
Some researchers have further found that the rating or ranking system as more
predominant in explaining fund flows (Guercio & Tkac, 2002) as compared to other

71
measures and more so there was an asymmetric flow when rating was changed, more flow
when upgrade as compared to less outflow when downgrade in rating happened (Guercio
&Tkac, 2002). Evidence on influence of loads on mutual fund flows is mixed. While some
researchers, argued that it has a negative impact (Cashmanet al., 2006) because it involves
cost and therefore discourage flows. Other researchers (for example Berkowitz and
Kotowitz, 2000) argued that load funds are mostly intermediaries promoted, therefore
investors in these kind of funds ingeneral are not financially sophisticated to move in or
out; as a result load has a negative impact on flow.

Media coverage and attention has a positive relationship in mutual fund flow (Sirri &
Tufano, 1998; Reuter & Zitzewitz, 2006). Empirical evidence also suggests that other
marketing efforts and advertisements have a positive effect on subsequent flows (Capon
et al., 1996; Barber et al, 2005). Jain and Wu (2000) examined the signaling hypothesis
of mutual fund advertising – whether it is used to attract more money in mutual funds.
This may be because of belief of investors in performance persistence or lowering of
search costs. All funds studied were found to have superior performance in the pre-
advertisement period. Authors found that in the post advertisement period,
performance of the mutual funds, on the average was significantly inferior as
compared to the benchmarks. Authors further found that advertising attracts money to the
extent of 20% more in comparison to the non-advertised funds of the similar
characteristics. Some studies have also considered mutual fund buy or sell
recommendations in the media. For example, Reuter and Zitzewitz (2006) concluded that
media mentions of the mutual funds have limited ability on the future performance but
they have significant economic impact in terms of investor cash flows.

Fund age and fund size are also related to mutual fund flows. Research has revealed that
older and larger funds attract less relative flows (Barber et al., 2005). Adding to itlagged
performance flow sensitivity has been found to be higher for small and younger funds

72
(Goriaevet al., 2002). Fund expenses have been found to lower fund performance thereby
having a negative impact on fund flows (Elton et al., 2003; 2004), but reactions have not
been similar to funds in various groupings (Barber et al., 2005). Also investors have been
found to be more sensitive to load charges than to expense ratios (Wilcox, 2003; Nanda,
2004). In fact both significant positive and negative effects of loads and expenses on
subsequent flows have been documented and results are at the most ambiguous (Barber
et al, 2005; Cashman et al, 2006). Researchers have also examined the impact of only
marketing costs and have found that marketing efforts account for the positive impact of
operating expenses on flows (Ferris & Chance,1987; Elton et al., 2003). Impact of
managerial incentive fees on fund flows has also been found to be positive (Elton et al.,
2003) because of signaling of superior manager skills. Mutual fund investors behave
differently in context of mutual fund purchase (Capon et al., 1996; Wilcox, 2003) as they
rely on different information sources and attributes to finalize their investment decisions.
Surprisingly, research indicates that more biasness in financial decision making develops
with the increase in financial literacy (Goetzmann & Peles, 1997; Wilcox, 2003).
Fund family size can be positively related to fund flows as it lowers search costs by
increasing visibility. More diversified fund family can provide better services (in form of
scheme switching) to the investors and less diversified fund families signal their focused
objectives. In spite of documented better performance by the less diversified fund families
(Siggelkow, 2003), investors pour more money in highly diversified fund families
(Rockinger, 1996). Evidence on number of funds in the fund family is also mixed, having
both negative relationship (Goriaev, 2002) and positive relationship (Rockinger, 1996)
with the fund flows. Presence of star fund in the fund family also results in more positive
net flow to the other funds in the family, because of spillover effects (Nanda, 2004).

Brokers and financial advisors have been found to exercise substantial influence on
purchase of load funds (Zhao, 2006). Authors also observed that there is more likelihood
that brokers and financial advisors direct investor monies into smaller but high load funds,

73
which may experience better performance due to their smaller size. On the other hand,
investors who have been directly investing their money, go for larger funds due to high
visibility. But some studies have found that there is no substantial and tangible benefit by
investing through brokers (Bergstresseret al., 2009) Another example of other explanatory
variables is evident from the cash flows in socially responsible funds. The volatility of
cash flows in the socially responsible funds wasfound to be low (Bollen, 2007) in spite of
the fact that the returns of the socially responsible funds either matched the conventional
counterparts (Hamilton et al., 1993; Statman, 2000; Geczyet al., 2003) or underperformed
them (Bauer et al.,2005).
Duca (2005) demonstrated the role of lower asset transfer cost (lower mutual fund load),
lower income risk and diversification considerations in increasing mutual fund ownership
in US households. Increasing ownership of stock or mutual fund due to lower asset transfer
cost was reported in several other studies also (Reid & Miller, 1999; Saito, 1995; Heaton
& Lucas, 2000). Other possible reasons in mutual fund investments are retirement
preparedness (Morgan, 1994) and other demographic reasons.
Providing a sort of conclusive evidence on mutual fund flows and its various determinants,
Capon et al (1996) observed that information source and selection criteria are inter
related. Moreover, demographic characteristic and investor behavior varies. Although
performance takes a central role in investment decision process, other criteria also play an
important role in mutual fund selection

2.13 Studies linked with Mutual Fund Selection Behavior

Many studies relating to preference, perception and attitude of investors towards
mutual funds have been performed with respect to mutual fund .. Although studies directly
relating to fund selection criteria are very few in number yet they are important in
discussing behavior of investors while selecting the mutual funds. Important studies are
presented below in chronological order.
Probably starting with Jambodekar (1996) study, that revealed that income schemes and

74
open-ended schemes were more preferred than growth schemes and close-ended schemes
during the then prevalent market conditions. Major factors which investors look in
mutual fund were safety of principal, liquidity and capital appreciation in the order of
importance; newspaper and magazines were the first source of information through which
investors got to know about mutual funds / schemes and investor service was a major
differentiating factor in the selection of mutual fund schemes. Other studies like that of
Shankar (1996) stressed on following of consumer product distribution model by Asset
Management Company to capture the market. Other determinants of mutual fund flows
have been found to be tax concessions, especially in the case of salaried people (Sikidar&
Singh, 1996); role of mutual fund agents, age and income of investor (Sundar,1998);
brand image (Chakrabarti&Rungta, 2000); and also psychological and sociological
factors (Shanmugham, 2000). In addition to finding factors for mutual fund selection,
several studies have been conducted for segmenting the investors on the basis of
their characteristics, investment size etc (Rajan 1997;1998; Kiran & Rao, 2004). In an
extensive study on mutual fund selection criteria, Rajeswari & Moorthy (2002) studied
the financial behavior and factors influencing fund / scheme selection of retail investors.
The authors conducted factor analysis using principal component analysis, to identify
the investor’s underlying fund/scheme selection criteria, so as to group them into
specific market segment for designing of the appropriate market strategy by Asset
Management Company. The study was conducted through a survey of 350 mutual fund
investors in 10 urban and semi urban centers. The survey revealed that the most preferred
investment vehicle was bank deposit and mutual funds ranked 4th in order among 8
choices. Growth schemes were ranked first followed by income schemes and balanced
schemes. The first preference was for open-ended schemes and only 15% investors’
favored close ended schemes. Investors looked for safety first in mutual fund products,
followed by good returns, tax benefit, liquidity and capital appreciation. The survey
revealed that respondents on their own made the scheme selection decision, and the other
sources influencing their selection decisions were newspapers and magazines, brokers

75
and agents, television, friend’s suggestion and direct mail in that order. Further 44% of
the respondents reported that they use internet facility to know more about mutual funds
and 37% respondents preferred to get routine /special information like daily NAV,
dividend, bonus etc through automated response system while 53% preferred personal
communication. Regarding influential fund selection factors - factors that relate to product
were namely intrinsic qualities of the product, portfolio management record, and image.
Two factors related to sponsor related factors namely infrastructure and reputation; three
factors related to service related factors namely subsequent disclosure and fringe
benefits were found to be influential in fund / scheme selection.

Singh and Vanita (2002) found that absolute returns from mutual funds and name of the
promoters were the basic criteria used by the investors in mutual fund scheme selection.
Also investors preferred to invest in the private sector, open ended and balanced schemes
of mutual funds. Some later studies like Kiran and Rao (2004) identified investor group
segments using the demographic and psychographics characteristics of investors
using multinomial logistic regression and factor analysis. In another study by Singh and
Chander (2004) perception of investors towards mutual funds was studied. It was found
that investors favorably perceive the daily disclosures of net asset value (mainly young
investors in the age group of 20-35 years) and tax rebates (mainly salaried and retired
class of investors) provided on investments in mutual funds. At the same time, they
disliked the way the industry has beenregulated, poor role and ineffectiveness of
controlling bodies like SEBI, less returns on an average as compared to the selected
benchmarks and inefficient management by the mutual fund companies.
In ascertaining the factors yielding behavioral biases, Kumar (2006) examined
whether individual investors exhibited stronger behavioral biases when value ambiguity
or information uncertainty was higher. Author used six year data from 1991 to 1996 of
retail holdings and trades and found that at times when stock markets have been difficult
to value, investors exhibit behavioral tendencies of overconfidence, stronger disposition

76
effect, limited attention and Representativeness. In addition when there has been greater
market wide uncertainty, these behavioral biases become stronger.
In the study by Ranganathan (2006) attempt was made to study the aspects of fund
selection behavior of individual mutual fund investors in Mumbai. With the help of
random and judgment sampling techniques author selected 100 mutual fund investors
during 2004. Author found that investors mainly preferred growth scheme; and open ended
structure. Investors gave high importance to published sources of information (for example
newspapers and magazines). Using factor analysis, author found seven principals
components among three groups. In the fund related qualities, the major selection criteria
were intrinsic fund qualities, credibility of image and flexible investment facilities.
Among fund sponsor related qualities, the major selection criteria were reputation of the
fund manager and competent performance. Among investor related services, major
selection criteria were transparent disclosure and fringe benefits. Author also used
multinomial logistic regression and found the factors of competent performance, flexible
investment facilities, reputation of the fund sponsor and fringe benefits as significant
factors explaining the fund selection behavior.
In a study conducted on mutual fund purchase by high net worth individuals (HNWI) in
India (Sharma, 2006) it was found that two channels of mutual fund purchases dominated
for HNWIs investors – direct or through a broker. The important information source for
the mutual fund investor was mutual fund performance ranking, recommendations of
business associates, advertising etc. in that order. The important fund selection criteria
were investment performance record, responsiveness to enquiries, fund managers
reputation, confidentiality and fund management fees e.t.c in that order. Singh (2006)
found that the majority of the investors in the age group of 20 - 35 years and salaried
people preferred close ended growth schemes over other schemes. Authors also
observed that major sources of fund information were brokers or financial advisors,
newspaper, finance journal, banker, friend’s or relative advice, finance journal, TV,
internet etc in that order. Singh and Chander (2006) argued that mutual funds were not

77
among the first three choices to invest. As regards to investment in mutual funds, close
ended funds were preferred over the others. Sectorial funds were not much sought after.
The major reason for investment in mutual funds was the good returns. The important
result that emerged in this study was that occupation (and not age) is significantly
associated with the perception of investors about the returns received from the mutual
funds, but at the same time occupation and age does not play any role in the selection of
particular type of mutual fund scheme. Also it was observed that most of the investors
base their investment decisions on the advice of brokers, professionals and financial
advisors followed by independent review of newspaper advertisements. Sundar et al.
(2007) studied 56 mutual fund schemes from 9 different sectors. Authors concluded that
it is feasible to time the market and take appropriate buy and sell decisions, but the same
is not feasible with all the mutual fund schemes. Das et al. (2008) carried out investor
survey in Cuttack to assess the behavior of retail investor with special reference to
products of insurance and mutual fund. Simple random sampling was used to select 100
respondents from these two cities. Authors found that graduates and post graduates were
investing more in insurance as compared to professionals who were investing more in
mutual funds; government servants were investing to the maximum extent (as a percentage
of their gross savings); males were more likely to be investors as compared to females;
the most important objective of investment was capital growth followed by tax saving and
retirement plan, in that order; investors were more likely to invest in life insurance as
compared to mutual funds and government savings; investors preferred to invest in open
ended funds; there was a positive correlation between mutual fund’s past performance
and brand image; most of the investors have found that newspapers and magazines are the
major source of information as compared to direct communication from the company.

2.14 Investors Behavior and Perception in Mutual Fund

Little efforts have been made by researchers to study investors’ behavior towards mutual

78
funds in particular..Madhusudhan & Jam bodekar (1996) argued that investor expects
better services, safety of principal, liquidity and capital appreciation. SyamaSunder (1998)
argued that the awareness was poor in small cities. Brand image and return were the prime
factors for investment. Panda and Tripathy (2002) found that investors were unsatisfied
except fromUTI. Singhand Chander (2004) concluded that poor regulation and control,
under–performance and inefficient management are the causes of non-investment.
Desiganetal(2006) found that women investors are hesitant for investment in mutual funds.
Rakesh and Sirinivas (2013) conducted a research knowing the investing behavior of
Andrapradesh investor’s behavior towards mutual funds. They inquired investors for their
age, income, education, categories of fund, sources of information, kind of fund, features
of scheme like, Diversification, return, safety, risk, transaction cost and tax benefits. They
found that most of the investors invest in mutual fund due to diversified portfolios, safe
investment, professional management, low transaction cost and the tax benefits associated
with mutual fund investment. Vyas (2013) conducted a research knowing the mutual fund
investor behavior and perceptions. The study used a wide range survey questionnaire which
covered the age, size, education, savings, income and different fund attributes and features
like fund reputation , fund NAV, sources of information, fund age, risk , safety, switching,
tax benefit, Diversification and Protection by government agency. The results found that
most of the investors invest in safe and efficient portfolios.
Pariharetal (2009) found that respondent’s age, gender and income were significantly
associated with their attitude whereas education and occupations were not associated with
the same. Pandey (2011) found that younger people aged below 35, graduate people and
the salaried person were easier to sell the funds and there was a large untapped market
there. Ippolito (1992) and Bogle (1992) reported that fund selection by investors is based
on past performance of the funds and money flows into winning funds more rapidly than
they flow out of losing funds. Goetzman (1993) and Grubber (1996) studied the ability of
investors to select funds and found evidence to support selection ability among active fund
investors. Malhotra and Robert (1997) found that the preoccupation of MF investors with

79
using performance evaluation as selection criteria is misguided because of volatility of
returns. The findings of Ferris and Chance (1987), Trzeinka and Zwing (1990), and Chance
and Ferris (1991) are consistent with the findings of Malhotra and Robert (1997). Zheng
(1998) examined the fund selection ability of MF investors and found that the investor’s
decisions are based on short-term fund performance and investors use fund specific in
formation in their selection decision. Vidyashankar (1990), Agarwal G.D. (1992), Gupta
(1993) Atmaramani(1996), Madhusudan(1996) and Srinivasan (1999) have conducted
extensive researches regarding investor expectations, protection, awareness and fund
selection behavior. Few striking ones among the other studies are given below.
Gupta L.C (1993) conducted a house hold investor survey with the objective to provide to
investor preferences on MFs and other financial assets. Jambodekar (1996) conducted a
study to assess the awareness of Mutual fund among investors, to identify the information
sources influencing the buyer decision and the factors influencing the choice of a particular
fund. The study revealed that income schemes and open-ended schemes are preferred over
growth schemes and close-ended schemes during the prevalent market conditions.
Investors look for Safety of Principal, Liquidity and Capital Appreciation in order of
importance; Newspapers and Magazines are the first source of information through which
investors get to know about MFs/ Schemes and the investor service is the major
differentiating actor in the selection of MFs. Sikidar and Singh (1996) conducted a research
with objective to understand the behavioral aspects of the investors of the North Eastern
region towards equity and MFs investment portfolio. The results indicated that the salaried
and self-employed invest in mutual funds due to tax concessions. UTI and SBI schemes
were popular in that part of the country then and other funds had not proved to be a big hit
during the time when the survey was done. Rajan (1997) divided the investors on the basis
of their characteristics, investment size, and the relationship between stage in life cycle of
the investors and their investment pattern. Sunder (1998) found found that the awareness
about mutual fund was poor in small cities i.e Vishakapatnam. Agents play a key role to
familiarize the MF culture. The research found open-end schemes as preferred. The

80
research found that ages and income of investors are the two important determinants in the
selection of fund/scheme. The results indicated that brand image and return are investors,
prime considerations. Shanmugham (2001) conducted a survey of 201 individual investors
to study the information sourcing by investors, their perception of various investment
strategy dimensions and the factors motivating share investment decisions, and reported
that, psychological and sociological factors dominated economic factors in share
investment decisions. Rajeshwari & Moorthy (2002) studied the financial behavior and
factors influencing fund/scheme selection of retail investors by conducting Factor Analysis
using Principal Component Analysis, to identify the investor’s underlying fund/scheme
selection criteria, so as to group them in to specific market segment for designing of the
appropriate marketing strategy. Kiran and Rao (2004) identified investor group segments
using the demographic and psychographic characteristics of investors using two statistical
techniques, namely: Multinomial Logistic Regression (MLR) and Factor Analysis.
Miglani (2007) analyzed the perception views of investor in decision making. Kumar
(2008) conducted a study on investor behavior towards mutual funds in which he assessed
the investor’s performance, awareness, and investor’s preference in different mutual funds
categories, schemes and other factors which effect the investor selection of mutual funds.
Sultana (2010) conducted a research in which she found on the demographical factors age
and gender on the risk tolerance level of the investors. Bodekar (1996) conducted study to
assess investor behavior on fund schemes, information sources, risk element and NAV of
different funds. He found that investor like positive NAV low risk and more safety, due to
which they invest in mutual fund. Jan and Jain (2012) the study was argued that each
demographical factor have effect on buying behavior pattern of investor, the research
conduct that every demographical factor (gender, age, qualification, occupation, income
and education) had significant bearing on investor buying behavior process. Barber &
zheng (2000) conducted study in investor behavior towards mutual funds and the research
argued three types of results. 1stinvestor buys the funds on the basis of past performance.
2ndapproxiamently 40 percent investors are also sell the fund on past performance. 3rd

81
investor do not like to buy the fund with highly commission fee or other highly transaction
cost. Desai (2010) analyzed investor behavior and the study concluded that main factor
which effect investor decision are money growth, return & security factor of money
reliability of the company and also seek the company profile, they also follow company
market brand value. Barber (2003) conducted research and found that investors are mostly
sensitive to pay front-end load and less sensitive for other operating expense. The
researcher also provided through managers that investor do not like in highly front-end
load against of the operating expenses. Arathy et al. ( 2015) conducted study in the investor
behavior and stated that better post experience, tax benefit, return potential, liquidity, low
cost and transparency are the major factors that motivate the investors to invest in mutual
funds. Ansari & Moid (2013) stated that investors, investing decision in mutual funds are
due to risk factor, growth and additional income. Rajeswari and Ramamurthy (2001)
conducted study in investor behavior in mutual funds. The study determines that the factor
influencing investment decision is product quality (performance of the fund) and brand
name. These two factors were the main factor which affects mostly the decision making in
mutual fund. Kumar & Goel (2012) their study found that majority of investor, invest in
growth scheme, the study also concluded that growth is the aim of the investor in mutual
fund. Sharma (2011) the study revealed that investors switch from one mutual fund to
other , the reason behind to switch from one to another mutual fund are same family fund,
investor mind set, current market condition, poor fund management and poor selection of
portfolio.

2.15 Theoretical Frameworks

The following theoretical framework has been framed for the performance analysis of this
study based on the literature review of this research. This study finds the open ended funds,
adjusted return through CAPM, Fama French Three Factor and Carhart four Factor model.
After finding the adjusted return of these funds the GRS test is applied to understand that
which among the three models better explain the mutual fund performance in Pakistan.

82
After the GRS the fund adjusted returns are regressed on the selected funds characteristics.
The investor behavior towards mutual fund is analyzed through questions asked from the
investors during the data collection of this study. Their perception was judged through
different funds attributes and characteristics.

83
2.15.1 Performance Analysis

84
2.15.2 Investors Investing Behavior in Mutual Funds

Attitude towards instruments

Performance of Record

Fund reputation and Brand Name

Scheme portfolio of Investment

Reputation of fund manager

Investors Behavior
Withdraw facilities

Favourable rating by agency

Innovativeness of scheme

Product with tax benefit

Entry and exit load

Minimum initial investment

Disclosure of derivation from the

Original pattern

Figure: B.2 Investor behavior frame work

The above frame work represents the fund factors on the biases of which the mutual fund
investor behavior has been investigated. These factors are the dependent variables upon
which the investor behavior has been investigated with matric or categorical independent
variables.

85
2.16 Research Hypothesis

The following research hypotheses have been developed for the analysis of this research

study.

2.16.1 Performance Analysis Hypothesis

H0: CAPM, Fama French 3-Factor and Carhart 4-Factor model equally well to predict
and explain mutual fund performance in Pakistan.
H1: CAPAM, Fama French 3-Factor and Carhart 4-Factor model do not equally well
to predict and explain mutual fund performance in Pakistan.
H0: Fund size has negative impact on Mutual Fund Performance.
H2: Fund size has Positive impact on Mutual Fund Performance.
H0: Expense Ratio has negative impact on Mutual Fund Performance.
H3: Expense Ratio has positive impact on Mutual Fund Performance.
H0: Liquidity has negative impact on Mutual Fund Performance.
H4: Liquidity has positive impact on Mutual Fund Performance.
H0: Cash flow has negative impact on Mutual Fund Performance.
H5: Cash flow has Positive impact on Mutual Fund Performance.
H0: Fund age has negative impact on Mutual Fund Performance.
H6: Fund age has Positive impact on Mutual Fund Performance.
H0: Fund family has negative impact on Mutual Fund Performance.
H7: Fund family positive impact on Mutual Fund Performance.
2.16.2 Investors Investing Behavior Hypothesis
H1: Most of the investors prefer to invest in mutual fund as the faster investment in all

investment avenues.

86
H0: Most of the investors do not prefer to invest in mutual fund as the faster investment in

all investment avenues

H2: Most of the investors invest with the view to have tax benefit on short term biases rely

upon information provided by agents to invest in Mutual funds

H0: Most of the investors do not invest with the view to have tax benefit on short term

biases rely upon information provided by agents to invest in Mutual funds.

H3: Most of the investors switch to another scheme for more return most often.

H0: Most of the investors do not switch to another scheme for more return most often.

H4: Most of the investors opt for investmentin open-ended funds.

H0: Most of the investors do not opt for investmentin open-ended funds.

H5: Most of the investors invest in mutual fund due to positive NAV factor.

H0: Most of the investors do not invest in mutual fund due to positive NAV factor.

H6: Most of the investors invest due to fund reputation and brand name.

H0: Most of the investors do not invest due to fund reputation and brand name.

H7: Most of the investor invests as it is easily investing opportunity.

H0: Most of the investor invests as it is easily investing opportunity

H8: Most of the investor invests due to disclosure of NAV.

87
H0: Most of the investor do not invests due to disclosure of NAV

H9: Most of the investor comes to know through broker and agents

H0: Most of the investor does not come to know through broker and agents.

H10: Most of the investorinvests in mutual fund due to the protection of their return and

principle by MUFAP.

H0: Most of the investor do not invests in mutual fund due to the protection of their

return and principle by MUFAP

88
 CHAPTER 3
RESEARCH METHODOLOGY
This chapter describes the research methodology followed during the course of this
research study. Herein the different steps taken to gather the data of mutual fund
performance are elucidated and the investor’s behavior in mutual Funds is determined. The
chapter is arranged in the following sequence:
a. Philosophy of the Research.
b. Nature and Type of the Research.
c. The Research design including data collection and sampling plan.
d. Data Analysis techniques and Models.
e. Operational definitions and measurement of the variables.

3.1 The Research Philosophy

The research in hand has two different components, which were investigated
independently. The first part that deals with evaluating Mutual funds’ performance was
predeterminantly based on secondary data, which was collected from Mutual fund
Association of Pakistan (MUFAP). The second part of the research determines the
investors’ investing behavior in mutual fund in Pakistan which is mainly based on the
primary data, which was collected through questionnaire. The investors’ behavior was
measured through a questionnaire, which was administered personally to them; therefore
the research can be placed in positivist research paradigm.
Positivism helps in capturing reliability and relies in the use of research instrument
(Blaxter, Hughes, & Tight, 2006). It is referred to as deductive approach, which deals with
proving and disapproving hypothesis of a research (Green, 2008). However, the positivism
mainly links with quantitative approach (Blaxter, Hughes, & Tight, 2006)
Many researchers developed frameworks for positivism; however the most important
contributor, Robson (2002) who identified 5 stages through which positivist research

89
proceeds.
a) Hypothesis deduction from theory

b) Hypothesis Expression in operational terms

c) Testing the operational hypothesis

d) Examining the specific outcome of the researchers

e) Modification of the theory, if necessary

Considering Ester & smith (2002) characteristics as reference, point the research holds the
similarity to the criteria and fall in positivism.
a) The observer must be independent. Hence the researcher was not the
part of the population of the study.

b) Human interest should be irrelevant. The researcher has interest neither

in the results nor in the responses of the investors investing mutual
funds.

c) Explanation must demonstrate causality. The correlation and regression

of the data demonstrate the desired relationship.

d) Research Hypothesis needs to be taken from the review of literature,

which has been ensured in this research.

e) Concept needs to be operationalized. This research has measured and

operationally defined all the variables.

f) Unit of analysis should be reduced to the simplest term and the same
has been ensured, as all open mutual fund operating on MUFAP site and
investors investing in mutual fund have been considered as the unit of
analysis.

90
 g) Random sampling technique for both performance analysis of mutual
fund and investors investing behavior was dully applied.

h) Statistical method should be applied in such a research and consequently

the data analysis of this research’s both parts was subjected to statistical
analysis.

3.2 Type and Nature of the Research

The research is both descriptive and correlational in nature and falls in the domain of
applied research.
Applied Research is conducted to find out solution of an immediate problem confronted
by the organization or managers (Sekran, 2003).Basic or fundamental research is
conducted for invention of new knowledge.
This research is correlational, explanatory and descriptive at the same time. The correlation
and impact of the independent variables have been investigated in this research study on
the mutual funds return and the investors ‘behavior towards mutual funds. The data was
collected through questionnaires and then descriptive tools like mean, median, standard
deviation and Bars have been generated to reflect the responses of the investors in statistical
terms.

3.3 The Research Design

A research design is the logical plan of the research, which shows that how a research has
to be conducted. Research design identifies and develops the path through which the data
is collected and analyzed to get to the solution (Sekran, 2003).
This research has been concluded on the basis of the following steps shown in table 3.1
Table 3.1 Research design steps
1 Purpose of Research Testing Hypothesis
2 Nature of Research Quantitative research

91
 3 Types of the study or Descriptive, Correlation& causal study
investigation
4 Study setting Contrived
5 Sampling Random sampling
6 Instrument and measurement Pre-used variables & questionnaire
7 Data Collection Secondary for performance analysis and
primary for investor behavior
8 Data Analysis Descriptive, correlation, regression and
Multinomial logistic regression,
Reliability
9 Findings and conclusion Results after testing hypothesis
Sources: Sekaran, (2003)
This study reviewed the literature and during this stage, variables reflecting Mutual Fund
performance and investors’ investing behavior in Mutual fund were identified anda great
emphasize was laid on their concepts, measurements and relationships.
On the basis of literature review, theoretical frameworks were developed and Hypotheses
were formulated. To substantiate hypotheses, data for mutual fund performance was
collected through secondary sources and data of investors’ behavior was collected through
questionnaires. The data obtained was formally analyzed through statistical tools and
interpreted to substantiate the hypotheses of the study.

3.4 Population of the Research

Population means the entire group of units, people, event or things of interest that the
researcher wishes to investigate (Sekran, 2003). For example a researcher wants to know
the Islamic trend in the banking system in KPK then all the banks of KPK are considered
as the population in this research.
The population of mutual fund performance was framed from the total number of mutual
funds traded on the MUFAP site and the population of investors’ investing behavior
consists of all mutual fund investors in Pakistan. However, while investigating the Mutual
fund performance, the research investigated all open ended mutual Funds as its population.

92
As far as, the population of mutual Funds investors’ investing behavior is concerned, this
research investigates the behavior of all mutual funds investors registered with all AMCs
for the purpose of investment in Mutual Fund.

3.5 Sample Size and Technique

Sampling is the process of selecting a sample (units, people, and organization) from the
population of interest and by studying the sample the results are generalized to the whole
population from which the samples were chosen. A sample represents the subset of the
population, which truly represents the entire population (Sekran, 2003).
This research study covers two aspects i.e. the performance analysis of mutual fund and
investors’ behavior in Mutual fund. The sample size of the performance analysis has been
considered as all the mutual funds incepted in January, 2009 or before. However, this study
has used the data of those mutual funds traded on MUFAP from 2009 to 2014. The monthly
data of 153 open ended mutual fund was analyzed through the three different Asset Pricing
Models i.e. CAPM, Fama French 3-Factor Model and Carhart 4-Factor Model to
investigate the adjusted return of the mutual funds’ portfolios, and the penal data of 44 open
ended mutual funds selected through simple random sampling was used to find the impact
of fund characteristics on funds’ performance. These selected funds annual data has been
used for the period 2010 to 2014 for the regression purpose knowing the funds
characteristics’ impact on funds adjusted return.
The investors’ behavior towards mutual fund has been analyzed through simple
convenience sampling techniques. Saunders et al. (2009) pointed that the more the size of
sample representing population the better the results. This research has included foregoing
responses in view ofthe data of 230 mutual fund investors across Pakistan.Roscoe (1975)
argued that at least the sample should consist of 30 respondents and up to 500 respondents
is a true representation in any research study. Tatman (1998) argued that to have statistically
significant results, there must be 100 or above respondents. Keeping in view the arguments
of the above mentioned researchers about sample size, 230 investors investing in mutual

93
fund data has been analyzed in this study.

3.6 Sources of Data and Collection Methods

The data of this research has been collected through two ways. The secondary data for the
analysis of mutual fund performance was collected from the financial reports and annual
reports of the Mutual funds, KSE sources and MUFAP web site. The monthly data has been
collected for the period 2009 to 2014. The penal data of fund characteristics has been used
for the performance analysis of Mutual funds. The annual data of funds characteristics was
collected for the period 2010 to 1014 from the sample funds annual reports. The funds
adjusted returns were calculated through CAPM, Fama French 3- factor and carhart 4-
factor model. Then adjusted returns regressed through funds characteristics, generated
through the above Asset Pricing Models.
This research investigates the investors ‘behavior towards mutual fund as well, so the data
used in behavior analysis has been gathered through the survey questionnaires, which were
administered personally to Peshawar, Quetta, Lahore, Karachi and Islamabad regions based
investors. The names and addresses of the investors who were administered questionnaires
were obtained from Asset Management Companies. Initially a pilot study was conducted
on 20 respondents to check the reliability of the questionnaire. These respondents were
further stratified/ categorized on the basis of their demographic parameters, such as salaried
person, professional and business men.

3.7 Data Management and Treatment Steps of Performance Analysis

The following steps have been used in the data preparing, construction of portfolios, and
management of the data and the application of tools on the data.

1. Collected Data of all open ended mutual funds from 1-1-2009

94
2. Collected T-bills 12-months rates from 2010 converted the rates to monthly and
merged that with Mutual funds data.

3. Collected KSE-100 Index data from 2010 to 2014, and merged that with Mutual
funds data.

4. Collected daily share prices data of KSE listed firms (around 680 firms) for
calculation of momentum, SMB, and HML factors.

5. Removed future contracts data from share prices.

6. Removed those firms which had observations less than 60 (619 firms left)

7. Converted daily share prices data to monthly by keeping only the last day of a
month

8. Generated monthly log stock returns of the available firms

9. Created unique year-month serial number for each stock, with the name `mser' to
be used later when merging with mutual funds data

10. Stocks financial data merged with share prices data (609 firms left) for SMB and
HML factors

11. Created market capitalization and market-to-book ratio

12. Dropped firm-year observations that have negative equity or stock returns equal to
zero (567 firms left; 29635 observations dropped)

95
13. Dropped firms’ year outlier observations, i.e. returns above or below 100%

14. SMB and HML factor calculations: Generated size and BM rankings of firms in
each year, size had two groups, Big and Small, B/M had three groups, Low,
Medium, High. Generated six portfolio returns in each month, Portfolios are BL,
BM, BH, SL, SM, SH, where B is for Big size and S is for Small Size, L, M, and H
are for Low, Medium, High B/M ratios. Generated SMB and HML factors, as SMB
= ( SL + SM + SH)/3 - ( BL + BM + BH)/3, and HML = ( SH + BH)/2 - ( SL +
BL)/2

15. Calculated Momentum Factor in each month: Calculated 11 month’s cumulative

returns for each stock. Ranked all stocks on their 11 months cumulative returns.
Found momentum returns in each month as a difference between average returns
of top 30% minus bottom 30% stock returns. Saved file in v5 SMB and HML
Factors and full monthly data.

16. Created ten portfolios of mutual funds on the basis of their cumulative last one year
returns, where lowest returns mutual funds are placed in P1 and mutual funds with
highest returns are placed in P10. Each portfolio contain on average 15 mutual
funds in each year.

17. Applied CAPM, Fama and French, and Carhart models to the decile portfolios to
select a best model among these competing models. GRS test was applied to select
best model among the competing models. CAPM turned out to be the best model.

18. Finally, risk-adjusted alphas were calculated for each mutual fund using CAPM.
Risk-adjusted alphas were then regressed on firm characteristics such as fund size,
turnover ratio, cash flows, lagged alpha etc.

96
3.8 Data Analysis tools
Given below are the statistical tools which are used in this study both for performance
analysis and investors’ investing behavior.

3.8.1 CAPM

The following CAPM model has been used to find the performance of the 10 decile
portfolios

.(Ri-Rf) = α + β(Rm-Rf) + ε……………………………………1

Where
(Ri-Rf): It represents the actual risk premium on a given stock
β(Rm-Rf): It represents the expected risk premium as suggested by CAPM
α is the intercept.

The same model was tested by previous researchers in their performance analysis studies
( Huij and Verbeek ,2006; EGB ,2004).
3.8.2 Fama French 3-Factor Model
The following Fama French 3-factor model has been used in performance analysis of
mutual fund.

(Ri-Rf) = α+ β1(rm-rf) + β2(SMB) + β3(HML) + ε…………………………..2

(Ri-Rf) : Denotes the actual risk premium on a given stock

SMB: It represents the monthly premium of the size factor. It is the difference in return on
a portfolio that consists of small caps funds and those that contain large caps funds.
HML: It represents the monthly premium of the book to market factor. Or it has been
calculated as the difference in return between a portfolio of high book to market stock and
a portfolio of low book to market stocks.

97
α : It is the intercept
This model has been widely tested by many researchers in the fields of mutual funds’
performance analysis and stock portfolio analysis. After the introduction of this model in
1993 by Fama and French. It was extensively used by researchers in the performance
analysis of stocks and funds. However it has not been explored by researchers in the field
of mutual funds’ performance analysis in Pakistan. EGB (2004) and Huij and Verbeek
(2006) tested this model in the performance evaluation of mutual funds.

3.8.3 Carhart 4-Factor Model

The following carhart model has been used for risk adjusted return of mutual fund.
(Ri-Rp) = a + β1 (rm-rf) + β2(SMB) + β3(HML) + β4(MOM) + ε………………….3
(Ri-Rf) : It denotes the actual risk premium on a given stock
SMB: It represents the monthly premium of the size factor. It is the difference in return on
a portfolio that consists of small caps funds and those that contain large caps funds.
HML: It represents the monthly premium of the book to market factor. Or it has been
calculated as the difference in return between a portfolio of high book to market stock and
a portfolio of low book to market stocks. This factor has been measure as measured by
many researchers ( Fama French, 1993; EGB, 2004).
MOM: It represents the difference in return on portfolio of past winners and a portfolio of
past losers. This is used here in this research so as to know that investors buy funds with
high return in the anticipation of high return in future and they sell low return funds. This
factor has been measured with pattern of previous researchers( Carhart,1997; EGB ,2004).
α : It is the intercept

This model was first used by Carhart (1997) knowing the persistence in mutual funds’
performance. This model has been tested by many researchers in developed world in the
performance evaluation of mutual fund ( e.g. EGB, 2004 ; Edelan ,2007; Huij and
Verbeek ,2006). However very rarely touched in developing and emerging markets

98
specially Pakistan by the researchers in performance evaluation of mutual funds.
3.8.4 Gibbons Ross Shanken Test

GRS = F (N, T – N- 1)

Gibbons et al. (1989) primarily used this model to find the differences in the alphas of the
style portfolios. The same model was tested and justified in their studies, relating to mutual
fund performance (Huij and Verbeek, 2006).

3.8.5 Descriptive statistics

In these statistics the mean, maximum, minimum and standard deviation of variables have
been calculated. It has been used by EGB (2004) in his research in the performance
analysis. Afza and Rauf (2009) also used the same technique for finding the mean and
standard deviation of the funds adjusted return and funds characteristics.

3.8.6 Correlation
Correlation is used to estimate the strength of the relationship between variables. Pearson’s
Product Moment correlation coefficient is used to assess this relationship.

The degree and type of relationship between any two variables in which they vary together
over a period of time. It shows the connection between variables. For example variation in
the level of expenditures or savings with the variation in the income level. Correlation
exists in two ways. The Pearson’s correlation matrix indicates the direction, strength and
significance of the bivariate relationship among all other variables which are measured at
interval ratio. The correlation relationship shows that how one variable
(Dependent variable) varies by the variation of other one or more variables
(Independent variables).

99
The correlation has been used for knowing the correlation of the funds attribute with the
fund return. EGB (2004) used the same tool in finding association among alphas and the
funds characteristics. Afza & Rauf (2009) also used the same technique in finding the
correlation between funds adjusted return and funds characteristics.
3.8.7 Regression Analysis
Regression is a statistical tool that forecasts the change in dependent variable on the basis
of change in one or more independent variables. For example how much change is caused
in sale revenue due to increased population and income level. In this example sale revenue
is dependent variable and income and population are independent variables. The dependent
variable is assumed to be random variable where as independent variables are assumed to
have fixed values that is they are chosen non-randomly. The relation between the expected
values of dependent variable and the independent variable is called regression relationship.
The dependence of a variable on two or more than two independent variable is called
multiple regressions.
In linear regression, the model specification is that the dependent variable is a linear
combination of the parameters.
The following multiple penal regression equation is used
Adjusted Return = β0 + β1Fund size+ β2Expense ratio + β3lequadity + β4 fund age+
β5fund family +β6 fund cash flow+ ε
EGB (2004) used the same tool in finding the impact of funds characteristics on alphas of
funds.

3.8.8 Descriptive Analysis

The table, bars, pies and curves have been used to analyze the responses of the investors
investing in mutual fund. The same approaches were used by researchers in the mutual
fund investors’ behavior analysis i.e.Rajeshwari&Murthy (2002), Kiran&Rao; Singh
&Chander(2004), Ranganat(2006)

100
3.8.9 Multinomial Logistic Regression
Multinomial logistic regression has been used to investigate the relationship of matric
independent or categorical variables with non-matric dependent variables representing
investor behavior. The same model was applied by Galagedera and Silvapulle (2002) and
Ranganat (2006) in their mutual funds investors’ behavior analysis.

3.8.10 Reliability Analysis

Cronbach alpha has been generated to find the reliability of the responses of the investors.
The alpha values of fund characteristics are in the feasible ranges.
3.8.11 Validity of the Questionaire
To ensure validity of the data, the questionnaire developed and validated by Rakesh &
Srinivas (2013), Vyas (2012) Galagedera & Silvapulle(2002) and Ranganat(2006), was
adopted. Further to ensure validity of target respondents only those respondents were
approach/ administered questions whose record was supplied by AMCs. All these AMCs
were kind enough to provide the details about the investors of different mutual funds.
Furthermore Factor analysis was applied to find the interrelations among and sufficiency
of sample size.

101
3.9 Operational Definitions and Measurement of Variables
3.9.1 LCAPM alpha
It is the previous year alpha. EGB (2004) used the same
3.9.2. Funds Size (ITNA)
Fund size is the total amount of investment made by the investors. Natural logarithm has
been taken of the net asset value of assets. The operationalization of fund size in this
research has been in the light of the studies in mutual funds by EGB (2004), Edelan et al.
(2007) who used the natural log of net assets, and the same natural log of the net asset
value was taken by (Afza & Rauf, 2009), who conducted study in Pakistan.
3.9.3. Expense Ratio
Operating mutual fund involve different types of expenses. In this research the expense
ratio has been calculated as the total fund expenses divided by total net asset value of the
fund. This operationalization supports the definition of expense ratio by EGB (2004),
Karlssen & person (2005), Haslem et al. (2008), and Afza&Rauf, (2009) who defined
expense ratio as the amount of total funds expenses divided by total net asset value of fund.
3.9.4. Turnover (liquidity)
In case of Open-Ended Mutual fund, investors can redeem their funds at any time. Due to
the chance of redemption, fund management keeps cash or liquid assets. This study is trying
to find out whether increase in cash balance effect return on investment or not. Liquidity
has been measured as the log of fund’s total cash on annual basis. This measurement has
been supported by (Afza & Rauf, 2009).
3.9.5. Fund Family (Ifam)
In this research the fund family has been calculated as the log of the number of funds in
the funds family. Chen et al. (2002), Khorana & Servaes (2005) and Gaspar et al. (2006)
have also measured the fund family as measured in this research study.
3.9.6. Fund Age (Iage)
It represents the number of years, funds has operated since its inception and has been
calculated as the log of the number of years of fund existence.Otten and Bams (2002),

102
Prather et al. (2004), Gaspar et al. (2006) and Ferreira et al. (2009) also defined it
operationally as in this research.
3.9.7 Cash Flow (C flow)
In this research cash flow has been measured as the net growth in funds asset. It has been
calculated as the (funds actual cash flow – expected cash flow) / TNA at time t+1. The
same formula for the calculation has been used by EGB (2004).

3.9.8 Measurement of the Questionnaire

The questionnaire used for the analysis of investors’ behavior towards mutual fund has
been adopted from the questions used by Rakesh & Srinivas (2013), Vyas (2012)
Galagedera & Silvapulle(2002) and Ranganat(2006) questions have also been adopted.
However some other questions have also been added by researcher, keeping in view the
nature of Pakistan mutual fund industry.

103
 CHAPTER 4

AN OVER VEIW OF MUTUAL FUND INDUSTRY IN PAKISTAN

4.1 Mutual Fund Association of Pakistan

MUFAP is an organization licensed by Government of Pakistan, to deal, operate and
protect Mutual fund in Pakistan. The assets management companies are registered
with the SECP, and can launch Mutual funds under NBFC rules 2008.
4.2 Meaning of Mutual Fund
Mutual funds represent the investments of small investors through professional and skillful
managers, who collected these funds in different advances of stock market like bond,
shares, certificate and other financial instruments. The managers make the diversified
portfolios of equity in debt instrument. Mutual funds are professionally allocated to ensure
safety and returns for the investors.
4.2.1 Types of Mutual Fund Traded in Pakistan
Mutual funds deals in three kind of mutual fund e.g, open ended, close ended and voluntary
pension Mutual fund.
4.2.1.1 Open-Ended

These funds are traded in MUFAP on their NAV. These funds are continuously redeemed
and purchased by investors. New units are also being launched and traded at their daily
NAV. The units of Open ended mutual funds are purchased and redeemed through AMCs,
which offer daily prices of respective Open ended mutual funds.
4.2.1.2 Close-Ended
These funds are like public limited company. It shares are issued through IPO and traded
on stock exchange. After the issuance of this fund, its daily share price is reflected on
secondary market (stock exchange).
4.2.1.3 Voluntary pension Fund

104
These are newly administered funds which can be traded on both MUFAP and stock
exchange or where they want to operate
4.3 Structure of Mutual Fund
a) Asset management companies hold control and operation of funds.

b) AMCs are registered under companies’ ordinance 1984.

c) When AMCs want to launch new funds, they have to established trust Deed, which
will be made between AMCs and trustee.

d) Trust deed to be ensured with the proper approval of SECP under NBFC rules 2008.

e) Trustee work as a custodian of the fund and ensure that funds amount is invested in
profitable avenues and the manager follow the predetermined investment policy.

f) As per the rules of Pakistan, banks and control depositary companies approved by
SECP can work as a trustee.

g) Currently control depositary companies of Pakistan works as a trustee of the fund

of the industry.

h) Overall the SECP is the main regulator of mutual fund industry and have very strict
and straight rules for issuance license to AMCs for launching mutual funds.

4.4 Categories of Mutual Fund

SECP the Regulator has categorized the Schemes of mutual funds as under:-
4.4.1 Equity Scheme
These are called equity funds, which invest in stock more often. The sources of revenues
of these funds include dividends and capital gains, which further document the long term
growth of funds.
4.4.2 Balanced Scheme
These kinds of funds invest in both debts instruments as well as stock. These funds provide
investors growth through diversified investment..

4.4.3 Asset al.location Fund

These funds allocate funds in multiple securities to have maximum diversification and

105
exploit the different investment choices in capital market.
4.4.4 Fund of Fund Scheme
These funds invest in other mutual funds which seem to be better performer in future.
These funds have a diversified portfolio of almost, equity, fixed income, money market ad
balanced funds.
4.4.5 Shariah Compliant (Islamic) Scheme
These funds invest in shariah complaint securities like sukak and ijara sukak.
4.4.6 Capital Protected Scheme
In this type of funds the original amount been invested is guaranteed with some capital
gain at the maturity of contract.
4.4.7 Index Tracker Scheme
This fund invest in securities which mirror the whole market, like investment in KSE
100.these funds work within the specified selected index.
4.4.8 Money Market Scheme
These are called the safer and stable funds. These funds invest in safest short term debts
instruments like treasury bills
4.4.9 Income Scheme
These funds invest in long term debt instruments i.e. TFCS, T bills or preference shares to
provide a continuous steady stream of fixed income
4.4.10 Aggressive Fixed Income Scheme
These funds invest in fixed income securities.
4.4.11 Commodity Scheme
These funds invest in commodities like gold etc. Most often of their these invest 70% of
their investment (assets) in commodity future contracts.
An investor can invest in any of the above categories of funds in accordance with his
requirements and appetite for risk. For example those who want to earn high returns
over a longer period can invest in Equity Funds whereas those who want to invest for
short term with reasonable return can invest in Money Market Fund.

4.5 Advantages of Investment in Mutual Fund

Mutual Fund plays a vital role in growth of capital market. Most of the investors favor

106
investment in Mutual Funds due to the following reasons.

4.5.1 Accessibility
Investment in Mutual Funds is very easy. Just on prescribed form lying with AMCs, the
investor has to write the number of units and to make payment for the units he wants to
buy of any Mutual Fund at the daily prescribed price.
4.5.2 Liquidity
Mutual Fund is very liquid, as the investors can sale it any time at the daily NAV or its
share price both MUFAP and KSE Trading.
4.5.3 Diversification
Investment in Mutual Fund is a kind of investment in portfolio, comprised of the
investment of so many investors, and is allocated by the fund managers across
different market instruments.
4.5.4 Professional Management
The fund of units holders in Mutual Funds are allocated by technically and
professionally skill full managers, who assess the investment advances and ensure
investment at places which are safer and yield healthy returns.

4.6 Tax Credit on investment to Individual

As per income tax ordinance 2001,sextion 62, a tax payer who is a resident and hasn’t the
status of a company, is entitled to tax credit on investment in new shares being offered by
any public limited company listed on stock exchange in Pakistan.
The following the Tax credit criteria for individuals.
a) Maximum tax credit Rs 220,417 on annual taxable income of Rs 6 million
around 22%.
b) Maximum Tax credit Rs 203,571 on taxable income Rs 7 million around 20%
tax rate.

4.7 Capital Gain Tax

107
The following withholding Capital gain tax is required for the Mutual Fund.
a) Within twelve months holding period of security, 12.5% Tax.
b) From 12 Months to 24 Months, 10% Capital gain tax is required.
c) If holding period is above 24 months, than 0% withholding Capital gain tax is
applied.
4.8 Investment in Mutual Fund
Following is the procedure for both individual and firm, if invests in Mutual Fund.
4.8.1 Individuals
Individual investors have to fill the prescribed form, CNIC Copy, Number of units want
to purchase, Affidavit of Zakat and cheque in favor the trustee of Fund.
4.8.2 Corporate
The corporate Pension Fund investment has to fulfill the following criteria.
a) Trust deed/ Memorandum and Article of Associations
b) Trustee approval document
c) Form representing purchase units
d) BOD power of Attorney
e) NTN
f) CNIC of the persons, who are authorized
g) Cheque in favor of the Trustee

4.9 Disinvest in Mutual Fund

The payment to the unit holders after redemption is made in 6 days’ time. The payment is
made through bank transfer or cross cheque by the relevant AMC.

4.10 Dividend Payment

In Mutual Fund industry the dividend is paid after specific interval i-e monthly, quarterly
or on annual basis. The payment dividend very from one category to another. In case an
investor wants to reinvest its dividend, he must inform the AMS, and he will be provided

108
new units.

4.11 Capital Gain Taxes.

Mutual Funds are required to hold Capital gain tax. Mutual Funds hold 12.5% for securities
less than 12 months holding, 10% for above twelve months to 24 months and 0% for above
24 months.

4.12 Risk Disclaimer

Market risk can affect the Mutual Funds Return. As the Mutual Fund NAV goes up and
down over the period due to market variations. All the investors are advised to critically
and carefully check the AMC investment policies and also check the risk disclosure
mechanism of the AMC.

4.13 Performance Picture of Pakistan Mutual Fund Industry in the Period of this Study
4.13.1 Open Ended Mutual Fund Performance
The below table, bar and pie shows the performance of open ended mutual fund over the
period of 2010 to 2014. The below table shows that the number of open ended mutual fund
has been increased from 105 to 153 over the period of 2010 to 2014. And the same can be
found in bar and pie as well.
Table 4.1 Open ended funds

` 2010 2011 2012 2013 2014

Open Ended 105 118 133 138 153

Source: MUFAP

109
 Open Ended Open Ended
200
150 2010

100 2011

50 2012 Ope…

0 2013
2010 2011 2012 2013 2014
2014

Figure D.1Bar of open ended

Figure D.2 Pie of open ended

4.13.2 Close Ended Mutual Fund Performance

The below table, bar and pie shows the performance of close ended mutual fund over the
period of 2010 to 2014. The below table shows that the number of close ended mutual fund
has been decreased from 21 to 14 over the period of 2010 to 2014. And the same can be
found in bar and pie as well.
Table 4.2 close ended Mutual fund

2010 2011 2012 2013 2014

Close- 21 16 15 09 14
Ended
Mutual
Fund

Sources: MUFAP

110
 Close Ended

25

20

15

10 Clos Ended

0
2010 2011 2012 2013 2014

Figure D.3 Bar of Close Ended

Figure D.4 Pie of close ended

4.13.3 Voluntary Pension Scheme

The below table, bar and pie shows the performance of voluntary pension mutual fund over
the period of 2010 to 2014. The below table shows that the number of voluntary pension
mutual fund has been increased from 9 to 13 over the period of 2010 to 2014. And the same

111
can be found in bar and pie as well.

Table 4.3 Voluntary pension Funds

2010 2011 2012 2013 2014

Voluntary Pension Funds 09 09 11 11 13

Figure D.5 Bar of voluntary pension mutual fund

112
 Voluntary pension mutual fund

2010
2011
2012
2013
2014

Figure D.6 Pie of voluntary pension mutual fund

4.13.4 Open ended Mutual Fund Net Asset value
The below bar and pie represent the total net asset value of open ended fund over the period
of 2010 to 2014. The total net asset value showing tremendous increase over the period of
2010 to 2014.
Open ended mutual fund net asset value

113
 Open ended Fund Net Asset value in Millions

400000
350000
300000
250000
200000
Open ended Fund Net
150000 Asset value et asset value
in Millions
100000
50000
0
2010 2011 2012 2013 2014

Figure D.7 Bar of open ended fund net asset value

114
 Figure D.8 pie of open ended fund net asset value
4.13.5 Close Ended Mutual Fund Net Asset value
The below bar and pie represent the total net asset value of close ended fund over the period
of 2010 to 2014. The total net asset value showing decreasing trend over the period of 2010
to 2014.
4.13.6 Asset Management Companies Operating in Pakistan
The below bar and pie represent the number of asset management companies worked over
the period of 2010 to 2014. In 2010 total 26 AMCs were operating and in 2014 the number
dropped to 20 AMCs.

Asset management copanies

30

25

20

15 Asset management
copanies
10

0
2010 2011 2012 2013 2014

Figure D.9 Bar of asset management companies

115
Figure D.10 Pie of asset management companies

116
4.13.7 Open Ended Mutual Fund Return Over the Study Period
The below table shows the return of different categories of open ended mutual funds
operating in MUFAP over the period of 2010 to 2014.
Table 4.4 Open ended funds return
Open- Ended Mutual Fund Return During The Period of this research Study
category of fund 2010 2011 2012 2013 2014

Aggressive income 8.40% -2.12 1.45 8.14 5.14

Asset al.location 17.93 12.19 6.76 23.42 14.78
Balanced 14.25 16.38 13.4 36.65 23.7
Capital Protected 7.22 9.71 3.27 11.38 12.5
Commodity 0 0 0 -17.14 3.56
Equity Fund 18.76 25.04 9.12 56.42 47.34
Fund of fund 13.99 31.7 14.69 35.93 9.93
Income Fund 9.44 11.02 11.08 9.73 9.32
Index Tracker Fund 29.79 22.45 7.33 44.78 35.37
Islamic Money Market Fund 10.1 10.98 10.69 8.06 8.7
Money market Fund 10.63 11.85 11.16 9.05 8.18

Source: MUFAP
4.13.8 The Return performance of Close Ended Funds
Table 4.5 Close Ended Funds Return
Close- ended funds during the period
2010 2011 2012 2013 2014
Aggressive income 0.00% 0 0 0 0
2010 2011 2012 2013 2014
Asset al.location Fund 0 0 0 0 0
2010 2011 2012 2013 2014
Balanced Fund 0 0 10.97 59.74 36.74
2010 2011 2012 2013 2014
Capital Protected fund 8.09 0 0 0
2010 2011 2012 2013 2014
Commodity Fund 0 0 0 0 0

117
 2010 2011 2012 2013 2014
Equity Fund 13.63 18.89 10.97 59.74 36.74
2010 2011 2012 2013 2014
Fund of fund 26.44 42.46 21.34 41.02 45.11
2010 2011 2012 2013 2014
Income Fund 12.02 11.67 11.23 7.84 7.66
2010 2011 2012 2013 2014
Index Tracker Fund 0 0 0 0 0
2010 2011 2012 2013 2014
Islamic Money Market Fund 0 0 0 0 0
2010 2011 2012 2013 2014
Money market Fund 0 0 0 0 0

Source: MUFAP

4.13.9 The Return of Voluntary Pension Scheme

Table 4.6 Voluntary pension scheme
VMS Mutual Fund Return During The Period of this research Study
2010 2011 2012 2013 2014
Equity sub Funds 20.00% 26.45 15.87 60.87 42.44
2010 2011 2012 2013 2014
Debts sub equity Fund 7.64 10.6 10.47 48 7.23
2010 2011 2012 2013 2014
Money market sub fund 3.45 10.94 10.11 8.75 7.32
2010 2011 2012 2013 2014
commodity sub fund 0 0 0 0 4.64

Source: MUFAP

118
 CHAPTER 5
DATA ANALSIS AND FINDINGS
This chapter focuses on mutual fund performance analysis of Pakistan industry through
different models and statistical techniques. The mutual fund performance has been
analyzed through CAPM, Fama French 3-Factor and Carhart 4- factor model. The GRS
test has also been used to understand which model among the three better predicts and
explain mutual fund performance. Descriptive statistics, correlation and regression have
also been used in performance analysis portion of this chapter. Investors investing behavior
in mutual fund has been analyzed through tables, bars, charts and Multinomial logistic
Regression.
5.1 Mutual Fund Performance Analysis
5.1.1 Mutual Fund Performance Analysis through CAPM
Table 5.1 CAPM

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

VARIABLES P1 P2 P3 P4 P5 P6 P7 P8 P9 P10

rm_rf 1.046*** 1.853*** 1.293*** 0.542*** 0.838*** 2.042*** 1.979*** 3.414*** 4.950*** 10.690***

(0.216) (0.443) (0.358) (0.158) (0.191) (0.484) (0.379) (0.663) (0.524) (1.657)

Constant 0.010 0.019 0.027 0.035*** 0.050*** 0.016 0.065*** 0.078** 0.110*** 0.200**

(0.012) (0.022) (0.020) (0.009) (0.010) (0.027) (0.021) (0.037) (0.029) (0.093)

Observations 68 56 68 68 68 68 68 56 56 44

R-squared 0.263 0.245 0.165 0.152 0.226 0.212 0.293 0.329 0.623 0.498

Standard errors in parentheses

*** p<0.01, ** p<0.05, * p<0.1

The table 5.1 shows the results of CAPM model applied on the data of mutual fund. The

119
 results indicate that portfolio returns are significantly related to market factor. All the
portfolios from P1 to P10 have significant coefficients. The portfolios with high return have
significant intercept but high as compared to the low return portfolios. The results indicate
that CAPM perform poorly in explaining portfolios with higher return. As the intercept
zero or closer to zero means that return of that portfolio is better explain by the market
factor. So funds with higher return are not well predicted by the CAPM meaning that the
manager do not cope well capturing the market variation. P1 is the lowest performing
portfolio and P10 is the better perform portfolios in term of return.

5.1.2 Mutual Fund Performance Analysis through Fama French 3-Factor Model
Table 5.2 Fama French 3-Factor Model

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
VARIABLES P1 P2 P3 P4 P5 P6 P7 P8 P9 P10

rm_rf 0.847*** 1.347** 1.074** 0.423*** 0.614*** 2.034*** 1.995*** 3.948*** 5.022*** 11.979***
(0.261) (0.571) (0.450) (0.135) (0.223) (0.612) (0.414) (0.817) (0.664) (2.033)
SMB 0.123 -0.166 0.198 0.290* -0.035 0.580 1.067** 1.889** 0.282 1.531
(0.303) (0.776) (0.523) (0.156) (0.258) (0.711) (0.480) (0.907) (0.737) (2.222)
HML -0.547** -1.283* -0.894* -0.094 -0.313 -1.018 0.031 0.274 -0.204 -1.132
(0.265) (0.659) (0.457) (0.136) (0.226) (0.621) (0.420) (0.812) (0.660) (1.770)
Constant 0.022 0.046 0.051* 0.028*** 0.053*** 0.056 0.065*** 0.103** 0.137*** 0.397***
(0.015) (0.031) (0.026) (0.008) (0.013) (0.036) (0.024) (0.050) (0.040) (0.125)

Observations 56 44 56 56 56 56 56 44 44 32
R-squared 0.299 0.292 0.217 0.213 0.223 0.275 0.335 0.377 0.628 0.608
Standard errors in parentheses
*** p<0.01, ** p<0.05, p<0.1
The table 5.2 shows the results of Fama French Three factor Model. This Model explains
the mutual fund portfolios results over the period of time of this research study. The results
of this model indicate that market factor significantly explain the return of all 10 portfolios
generated for the analysis of this model. The market factor co- efficient of all portfolios i.e.
from P1 to P10 are statistically significant. However the significance of size and value
factor (SMB and HML) do not have sufficient explanatory power, as their co-efficient are
insignificant in majority of the portfolios. The value of alpha increases in size and statistical

120
significance, as moving from P1 to P10 and hence, the R-square too. .Therefore it is
suggested that Fama French model do not offer proper explanation and prediction of mutual
fund data as compared to CAPM and hence do not showing any improvement over CAPM
model. It means that size and value factor of the overall market premium is not captured
by the fund managers.

121
 5.1.3 Mutual Fund Performance Analysis through Carhart 4- Factor Model
Table 5.3 Carhart 4-Factor Model

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

VARIABLES P1 P2 P3 P4 P5 P6 P7 P8 P9 P10

rm_rf 0.978*** 1.417** 1.324*** 0.488*** 0.640*** 2.387*** 2.107*** 4.042*** 4.693*** 11.892***

(0.262) (0.560) (0.447) (0.135) (0.232) (0.606) (0.426) (0.873) (0.692) (2.260)

SMB 0.052 0.045 0.062 0.255 -0.049 0.387 1.006** 1.855* 0.400 1.547

(0.297) (0.770) (0.507) (0.153) (0.262) (0.686) (0.483) (0.923) (0.731) (2.269)

HML -0.811*** -1.766** -1.397*** -0.225 -0.366 -1.730** -0.195 0.104 0.392 -0.996

(0.288) (0.709) (0.493) (0.149) (0.255) (0.667) (0.470) (0.967) (0.766) (2.284)

MOM 0.481** 0.978 0.915** 0.238* 0.097 1.294** 0.411 0.258 -0.906 -0.199

(0.236) (0.592) (0.404) (0.122) (0.209) (0.547) (0.385) (0.778) (0.616) (2.051)

Constant 0.020 0.041 0.048* 0.027*** 0.052*** 0.050 0.063** 0.102** 0.140*** 0.395***

(0.015) (0.031) (0.025) (0.008) (0.013) (0.034) (0.024) (0.050) (0.040) (0.129)

Observations 56 44 56 56 56 56 56 44 44 32

R-squared 0.352 0.338 0.289 0.267 0.226 0.347 0.350 0.379 0.647 0.608

Standard errors in parentheses

The table 5.3 presents the results of Carhart model used for the performance measurements
of mutual funds of this study. The results indicate that market factor significantly explain
the return of all mutual funds sorted in different 10 portfolios. However the significance of
SMB and HML do not have sufficient explanatory power due to the fact, that both SMB
and HML having insignificant co-efficient in majority of the portfolios .The MOM also
showing weak explanatory power as compared to market factor. The value of intercept
increases in size and statistical significance, as we goes from P1 to P10. The results suggest
that both SMB and HML along with MOM showing weakly explaining funds return.

122
 Therefore based on the results, it has been found that Carhart model does not offer
significant improvement over CAPM model, however having a bit improved performance
as compared to Fama and French model in explaining mutual fund return in Pakistan. So
the size, value and momentum premium factors of the overall market premium do not offer
proper explanatory power for the performance measurement of mutual fund return in
Pakistan.
5.1.4 Mutual Fund Performance Analysis through GRS
Table 5.4 GRS Test
GRS tests Mean Absolute Intercepts and Returns Dispersions
GRS_F- P- GRS- P- Mean Abs.
Factors test value Wald value Alpha A|ai|/A|ri| A(a2i)/A(r2i) Avg. Adj R2

10- Returns portfolios

rm_rf 9.04375 0 136.7867 0 0.056154 0.626596 0.365279 0.300686

rm_rf SMB HML 14.41312 0 281.8566 0 0.088005 0.982003 1.203135 0.319066

rm_rf SMB HML
MOM 14.55226 0 301.3173 0 0.086227 0.96217 1.17689 0.345428

GRS tests Mean Absolute Intercepts and Returns Dispersions

The table 5.4 shows the result of GRS test. The test has been applied to find out which
model among the three better explain and predict mutual fund data in Pakistan. GRS tests
whether the combined intercepts of 10 regressions are equal to zero.
If the intercepts are equal to zero, then the model is a good model.
However, the above statistics show that none of the models can sufficiently explain returns.
However, CAPM seems better explaining the mutual fund performance than the other two
models used in the performance analysis of this study.

123
5.1.5 Descriptive Statistic
Table 5.5 Descriptive statistics

Variable Obs Mean Std. Dev. Min Max

capm_alpha 162 -0.002 0.007 -0.019 0.019
ff_alpha 160 0.002 0.012 -0.025 0.060
carhat_alpha 160 0.002 0.013 -0.027 0.038
exp_ratio 166 0.079 0.156 0.000 0.924
Lage 156 2.103 0.380 0.693 2.565

Lfam 164 1.874 0.610 0.000 2.833

Cflow 163 -0.085 0.476 -1.709 0.666
Ltna 166 20.482 1.310 17.888 25.364

Turnover 166 17.923 1.762 10.920 22.415

This table 5.5 shows the descriptive statistics of the variables of this study. The results
show a mean of CAPM alpha as -0.002 with a minimum value -0.019 and having maximum
value of 0.019 and standard deviation value of 0.007. The Fama French alpha shows a
mean value of 0.002, with a minimum mean value -0.025 and maximum mean value 0.060
and standard deviation value of 0.012, the Carhart alpha showing a mean value of 0.002,
the same as that of Fama French alpha value. The result indicates the minimum Carhart
alpha mean value as -0.027 and the maximum mean value as 0.038 with standard deviation
0.013. The results showing 0.079 as the mean value of expense ratio, minimum mean value
0.000 and maximum mean value 0.924 with standard deviation 0.156. The Fund age mean
value is 2.103, with a minimum mean value 0.693 and maximum mean value 2.565 and
having standard deviation of 0.380. The Fund Family showing 1.874 as mean value with
0.000 as minimum mean having standard deviation of 0.610 cash flow showing -0.085 as
mean value with -1.709 as minimum value and 0.666 as the maximum mean value and the
same variable showing 0.476 as the standard deviation Fund size (Itna)showing mean value
20.482 with 17.888 and 25.364 being the minimum and maximum mean value. This

124
 variable shows 1.310 as the standard deviation of the fund size. Liquidity showing 17.923
as the mean value with 10.920 and 22.415 as the minim and maximum mean values.
Liquidity showing 1.762 as the standard deviation.

5.1.6 Correlation Analysis

Table 5.6 Correlation

Table 5.6 shows the results of correlation matrix between the variables used for the mutual

capm_a~a ff_alpha carhat~a exp_ra~o Lage lfam cflow ltna turnover

capm_alpha 1
ff_alpha 0.7583 1
carhat_alpha 0.7749 0.8681 1
exp_ratio 0.0059 -0.0601 -0.0105 1
Lage 0.14 0.1795 0.1427 0.0218 1
Lfam -0.0181 0.1057 0.075 0.0173 0.1413 1
Cflow 0.1642 0.2072 0.2078 -0.0875 -0.0975 0.1561 1
Ltna 0.1256 0.0988 0.0492 -0.2363 0.0094 0.005 0.0264 1
-
Turnover -0.1815 -0.2505 -0.2622 -0.0206 0.0798 0.0246 0.1176 0.1366 1
fund performance analysis of this research study. CAPM Alphas indicates the funds risk
adjusted Return generated through CAPM model. Carhart alpha represents the risk adjusted
Return calculated through carhard-4 Factor model. Expense ratio represents the total fund
expense on the bases of total net asset value of the fund. Lage represent the log of ht number
of years of a fund existence. Lfam represent the fund family size, which is the log of
number of Funds in the Funds family. Cash flow represents the cash flow of the fund over
the period of time. ITna represents the size of the fund, which is the log of Fund’s total net
asset value. Liquidity represents the amount of cash or cash equallant which fund has for
redemption of units’ purposes. The results indicate that CAPM alpha is positively
correlated with expense ratio, Fund age, Fund cash flow, and Itna. Suggesting that as
expense ratio, fund age, Fund cash flow and Fund size increase, than the Fund’s adjusted
Return tends to increase as well, however the CAPM alpha showing negative Correlation

125
with Lfam and Liquidity. Suggesting that as Fund’s Family and Fund’s Liquidity increases
fund adjusted returns tends to decrease. The results show negative Correlation of Fama
French alpha with expense ratio and liquidity, whereas positive correlation with Fund age,
Fund Family, cash flow and Fund size was found. The Carhart alpha show the similar
correlation as that of the Fama French alpha has.

5.1.7 Regression Analysis

Table 5.7 Regression
(1) (2)
VARIABLES CAPM CAPM_lagged alphas

lcapm_alpha 0.567***
(0.069)
exp_ratio 0.225 0.448
(0.359) (0.302)
Lage 0.422** 0.182
(0.178) (0.167)
Lfam -0.079 -0.139
(0.099) (0.090)
Cflow 0.251** 0.253**
(0.122) (0.109)
Ltna 0.087* 0.007
(0.045) (0.039)
Turnover -0.078** -0.015
(0.032) (0.028)
year2 -

year3

year4

Constant -1.356 -0.107

126
 (1.046) (0.911)

R-squared 0.112 0.448

Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1

According to GRS test, CAPM was the preferred model. Other models can be reported in
Appendix for further details. Table 5.7 shows the Regression results of the CAPM alpha
regressed through the fund characteristics used as independent variables in this study.
The two columns show two different regressions. First regression uses current alpha of
CAPM as dependent variable. Whereas expense ratio, Fund age (Iage), fund family (Ifam),
cash flows, fund size (ltna) and Liquidity are independent variables. In the first regression
fund age, cash flow and fund size significantly enhance fund performance. While turnover
(liquidity) significantly negatively impacts the fund performance. Fund family (Ifam)
shows negative insignificant impact on fund performance whereas expense ratio showing
positive insignificant impact on fund performance. The results are consistent with the
findings of many previous researchers. As fund age showing significant impact on the fund
performance which is very similar to the findings of Afza & Rauf (2009).Haslem et al.
(2008), Babolos et al. (2009) and Belgacem (2011) who also found positive results
regarding fund age, which supports the results of this study. These researcher also found
that as the fund get older it performance improve over the period of time. Cash flow
indicates positive significant impact on the fund performance in the results table output of
this research, which is very consistent with the findings of Gruber (1996), Zheng (1999),
Elton et al. (2002) and Tiwari (2004). These researchers also found that funds with
excessive cash flow perform better.
As the results of this study show that fund size (Itna) has positive significant impact on
fund performance. Which support the findings of many researchers (Gorman, 1991,
Grinblatt Titman, 1994; Peterson et al., 2001,Hesle et al.l, 2008, Babolos et al.l, 2009,
Nazir& Nawaz,2010). These researchers also found that as the fund size increases than it

127
tends to increase the fund return. The results of this research indicate liquidity as negatively
significantly impacts the fund return. Many other previous researchers also found and
documented similar results, like Glenn (2004) argued that liquidity has significant negative
impact on fund performance. Similar results have been reported by many researchers
(Dukes & Davis,2006, Afza & Rauf, 2010, Nazir& Nawaz, 2010).
The results indicate that expense ratio has positive insignificant impact on the fund risk
adjusted return, which is very similar to the findings of Afza & Rauf (2009). Greenblatt &
Titman (1994) and Drooms (1996) also found positive results about expense ratio.
The results show a statistically negative insignificant impact of fund family (Ifam) on the
fund adjusted return . This is very consistent with the findings of Otten & Bams (2002) and
Karlsson & Perssson (2005). The second regression showing the lagged alpha to check
whether there exist any persistence in mutual fund return in Pakistani mutual funds. The
lagged alpha has positive co-efficient, which shows that best performer in past continue
better in the present as well. However the inclusion of lagged alpha has affected the
significance of other independent variables, which means that lagged performance absorb
the effect of firm age and firm size. The lagged alpha is highly significant. Majority of other
variables become insignificant, however cash flows remain significant.

5.2 Investors Investing Behavior in Mutual Fund

Investor investing behavior has been analyzed through descriptive like tables, bars, pies
and curves. The 230 mutual fund investors have been analyzed and shown their responses
about the questions asked through these research tools. Besides these tools the mutual fund
investors behavior has been analyzed through the more advanced financial modeling i.e.
Multinomial Logistic Regression to know the relationship between non matric dependent
variables and matric independent or categorical variables.

5.2.1 Investors behavior in Mutual Fund through Descriptive

128
Q1 Gender

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 total 210 were male investors and the remaining 20 were
female investors in mutual Fund.

Table: 5.8 of Q1

Gender Male Female

210 20

Figure E.1Bar of (Q1)

129
 Figure E.2 Pie of (Q1)

Figure E.3 Curve of (Q1)

Q2 Ages of the Respondents of this Study.

The below table, bar, pie and curve shows the ages of the investors, from whom the data
for this research was collected. The data collected represent that 20 investors in mutual
funds were having ages less than 30 years. 50 investors of the total 230 were having ages
between 30 t0 40 and 160 respondents were found having ages above 40 years.

Table 5.9 of Q2

130
Less than 30 Between 30 to 40 Above
40
20 50 160

200

150

100
Series1

50

0
Agelessthen 30 between 30 t0 40 above 40

Figure E.4 Bar of (Q2)

Age
lessthen 30

between 30 t0 40

above 40

Figure E.5 Pie of (Q2)

131
200

150

100 Series2
Series1
50

0
Agelessthen 30 between 30 t0 40 above 40

Figure E.6 Curve of (Q2)

Table 5.10 of Q3

132
Q3 Education Level of the Respondents of this Study

The below table, bar, pie and curve represents the education level of the mutual fund
investors. The data collected from the mutual fund investors shows that 60 investors out of
230 were having education of post graduate level. 80 investors were found having the
graduate level of education. 40 investors out of 230 Shows the education of undergraduate
level. 35 investors were found having diploma and investing in mutual fund industry. 15
out of 230 were found having high school level education. And none of the 230 investors
inquired for the research was found having education less than high school level.

Table 5.10 of Q 3

post grad under Dipl high less than high

graduate uate graduate oma school school
60 80 40 35 15 0

Education

100
80
60
40 Education
20
0
less then…
under…
graduate
post graduate

diploma
high school

Figure E.7 Bar of (Q3)

133
 Education
post graduate

graduate

Figure E.8 Pie of (Q3)

Education

100
80
60 Education
40
20
0
less…
under…

high…
post…
graduate

diploma

Figure E.9 Curve of (Q3)

Q4 Occupation of the Respondents

The below table, bar, pie and curve represent the occupation of the investors in mutual fund
across Pakistan. The results showed that 70 out of 230 investors who were assessed for the
research purpose, were professional, 60 investors were found having different businesses.
55 investors investing in mutual fund were salaried persons and 45 investors in mutual fund
were found been retired from their services.

Table No 5.11 of Q4

134
Professional business Salaried Retired
70 60 55 45

occupation

80
60
40
20
0 occupation

Figure E 10 Bar of (Q4)

occupation

professional

business

salaried

retired

Figure E. 11 Pie of (Q4)

135
 occupation

80
70
60
50
40
30
20
10
0 occupation

Figure E. 12 Curve of (Q4)

Q5 Marital Status of the Respondents

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 investors surveyed, 190 investors were married and 40
investors were unmarried investors, investing in mutual Fund.

Table No 5.12 of Q5
Married Unmarried
190 40

136
 marital status

200

150

100
marital status
50

0
married unmarried

Figure E.13 Bar of (Q5)

marital status

married

unmarried

Figure E.14 Pie of (Q5)

marital status
200

150

100
marital status
50

0
married unmarried

Figure E.15 Curve of (Q5)

137
Q6 Annual Income Levels of the Respondents

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 investors in mutual funds.95 investors were having less
than Rs. 500000, 35 were found having annual income above 5 lac to 10 lac. 50 investors
were found having annual income above 1000000 to 15 lac and 50 investors were found
having annual income above 15 lac.

Table No 5.13 of Q6

Less than Rs 500000 500000 to1000000 1000000 to 1500000 Above 1500000

95 35 50 50

Figure E.16 Bar of (Q6)

138
 Figure E.17 Pie of (Q6)

Figure E.18 Curve of (Q6)

Q7 Location of the Respondents

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total investors investing in mutual fund , total 79 investors
were surveyed from Peshawar, 50 mutual fund investors of Islamabad were surveyed where
as 15 of Karachi, 71 of Lahore, and 15 of Quetta mutual fund investors were enquired for
knowing their behavior and perception.

139
Table No 5.14 of Q7

Peshawar Islamabad Karachi Lahore Quetta

79 50 15 71 15

Figure E.19 Bar of (Q7)

Figure E.20 Pie of (Q7)

140
 Figure E.21 Curve of (Q7)

Q8 Do you Prefer Investment in Mutual Funds to other Saving Avenue in Future?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total 230 mutual fund investors analysed for the purpose of
knowing their behavior and perception towards mutual funds 215 investors said yes
regarding their preference ,10 investors said no and only 5 investors in mutual fund were
not sure regarding their preference . The maximum nuber of mutual fund investors showed
giving preference to mutual fund in future as well.

Table 5.15 of Q8

Yes No Not sure

215 10 5

141
Figure 5.22 Bar of (Q8)

Figure 5.23 Pie of (Q8)

142
 Figure E.24 Curve of (Q8)

Q9 Generally you prefer

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 investors surveyed knowing their behavior and perception
65 of them liked to be invested in equity funds. 35 and 23 investors of mutual funds liked
investment in capital protected and money market funds respectively. 27 investors were of
the veive to invest in asset al.location funds. 24 investors were found preferring investment
in fund of fund, 16 investors of mutual fund were viewing and preferring investment in
income funds. 7 and 23 investors were found giving general preference to balanced and
Islamic funds.

Table 5.16 of Q9

Equity fund 65 Fund of Fund 24

Capital protected 35 Income fund 16
Money market 23 Balanced fund 7
Asset al.location fund 27 Islamic fund 23

143
Figure E.25 Bar of (Q9)

Figure E.26 Curve of (Q9)

Figure E.27 Bar of (Q9)

144
Q10 Which Kind of Mutual Fund You Would You Prefer?
The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total sample investors questioned, 180 investors gave
preference to invest in open ended scheme, 37 investors preferred investment in close
ended scheme and only 13 mutual fund investors liked investment in voluntary pension
funds.

Table 5.17 of Q10

open ended scheme Close ended Scheme voluntary pension Fund

180 37 13

Figure E.28 Bar of (Q10)

145
 Figure E.29 Pie of (Q10)

Figure E.30 Curve of (Q10)

Q11 You Prefer Investment in Mutual Funds due to.

The below table, bar, pie and curve represent the responses of mutual fund investors towards
mutual fund. The data of this research was collected from total 230 mutual Fund investors
in Pakistan. Out of the total 75 investors preferred investment in mutual fund due to its
safety, 5investors were found preferring investment in mutual funds due to tax benefits

146
attached with mutual funds.15 investors preferred investment in mutual fund due to its
capital appreciation, 45 investors preferred investment in mutual fund due to professional
management, 7 investors preferred investment in mutual fund as they believed it as source
of diversification. 31 and 9 mutual funds investors preferred investment in mutual fund due
to its liquidity and flexibility. 7 investors preferred investment in mutual fund due to its
return characteristics. 15 investors of mutual funds preferred investment in mutual fund due
to the benefits attached with mutual fund and 21 investors liked the easy entry and exit of
funds
Table 5.18 of Q11
Safety 75 Liquidity 31
tax benefits 5 Flexibility 9
capital appreciation 15 Return 7
Professional management 45 Benefit 15
Diversification 7 Easy entry and exit 21

Figure E.31 Bar of (Q11)

147
 Figure E.32 Curve of (Q11)

Q12 How Do You Invest in These Options?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 total 91 investors said the agents being the reason of
investment, 73 investors of mutual funds said the reason of investment inspiring and
motivating from family and friends and invest through them. 66 investors believed that due
self-interest we are investing in mutual fund and personally go to AMCs for purchase and
sale of mutual funds units.

Table 5.19 of Q12

Agent Family and friends Self

91 73 66

148
Figure E.33 Bar of (Q12)

Figure E.34 Pie of (Q12)

149
 Figure E.35 Curve of (Q12 )

Q13 You think Fund Performance Record as

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 total 65 mutual fund investors thought the fund
performance record as highly important, 120 investors thought fund performance record
as important for the motivation of the investors. 30 investors considered it as some what
important, 5 investors thought fund record as not very important and 10 mutual fund
investors thought fund performance record as not at all important.

Table 5.20 of Q13

Highly Important Some what Not very Not at all

important important important important
65 120 30 5 10

150
 you think fund performance
record as
140
120
100 you think fund
80 performance record
60
40 as
20
0

Figure E.36 Bar of (Q13)

Figure 5.37 Pie of (Q13)

Figure 5.38 Curve of (Q13)

Q14 Fund Reputation Brand Name

151
The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 total 61 mutual fund investors thought the fund reputation
brand name as highly important, 82 investors thought fund reputation brand name as
important for the motivation of the investors. 57 investors considered it as some what
important, 17 investors thought fund reputation brand name as not very important and 13
mutual fund investors thought fund reputation brand name as not at all important.

Table 5.21 of Q14

Highly Important Some what Not very Not at all

important important important important
61 82 57 17 13

Figure 5.39 of Q14

152
 Figure E.40 Pie of (Q14)

Figure E.41 curve of Q14

Q15 Schemes Portfolio of Investment

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 total mutual fund investors questioned for the purpose of
analysis.52 investors thought scheme portfolio as highly important, 87 thought it as

153
important, 45 investors were considering it as somewhat important, 31 investors of mutual
funds considering scheme portfolio as not very important and only 15 mutual fund
investors considering it as not at all important.

Table 5.22 of Q15

Highly Important Somewhat Not very Not at all

important important important important
52 87 45 31 15

schemes portfolio investment

60
50
40
30
20
10
0

schemes portfolio
investment

Figure E.42 Bar of (Q15)

154
 Figure E.43 pie of (Q15)

Figure E.44 Curve of (Q15)

Q16 Reputation of the Fund Manager

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total 230 investors of mutual fund questioned for their
behavior, 98 investors considered fund manager reputation as highly important, 68

155
investors considering it as important, 30 investors thought fund manager reputation as
somewhat important, 18 investors investing in mutual funds were found considering it not
very important and 16 viewed the fund manager reputation as not at all important.

Table 5.23 of Q16

Highly Important Somewhat not very not at all

important important important important
98 68 30 18 16

Figure E.45 Bar of (Q16)

156
 Figure E.46 Pie of (Q16)

Q17 Withdrawal Facilities

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total 230 ,71 investors of mutual fund considering the
withdrawal facility of mutual fund as highly important , 92 mutual fund investors
considered the same as important, 30 investors considering the fund withdrawal facility

157
as somewhat important,25 were found having view it as not very important and only and
the rest 12 mutual fund investors considering the withdrawal facility as not at all important.
Table 5.24 of Q17
Highly Important Somewhat Not very Not at all
important important important important
71 92 30 25 12

Figure E.48 Bar of (Q17)

Figure E.49 Pie of (Q17)

158
 Figure E.50 curve of Q17

Q18 Favorable Rating by Rating Agencies

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total investors questioned, 25 investors considered that
favorable rating of fund is highly important, 30 investors considered the favorable rating
of fund as important, 91 investors considering the favorable rating of the fund as somewhat
important and 71 investors considered it as not very important and only 13 investors
thought it as not at all important.

Table 5.25 of Q18

Highly Important Somewhat Not very Not at all

important important important important
25 30 91 71 13

159
Figure E.51 Bar of (Q18)

Figure E.52 Pie of (Q18)

160
Figure E.53 Curve of (Q18)

161
Q19 Innovativeness of the scheme

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total 230 investors questioned, 45 investors considered
innovativeness of the scheme as highly important, 61 considered it as important, while 72
investors of mutual fund think the innovativeness as somewhat important, 46investors
thought it as not very important and only 6 investors considered it as not at all important.

Table 5.26 of Q19

Highly Important Somewhat Not very Not at all
important important important important
45 61 72 46 6

Figure E.54 Bar of (Q19)

162
 Figure E.55 Pie of (Q19)

Figure E.56 Curve of (Q19)

Q20 Products with Tax Benefits

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total 230 investors, 73 investors considered mutual fund
investment as investment in product having tax benefit and thought it as highly important,
32 considered it as important, 52 thought the tax benefit with fund as somewhat important,
38 investors considering it as not very important and 35 investors considering it as not at
all important.

163
 Table 5.27 of Q20

Highly Important Somewhat Not very Not at all

important important important important
73 32 52 38 35

Figure E.57 Bar of (Q20)

Figure E.58 Pie of (Q20)

164
 Figure E.59 Curve of (Q20)

Q21 Entry & Exit load

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total 230 mutual fund investors surveyed for the purpose of
the analysis of this study, 105 investors thought the fund entry and exit load as highly
important, 31 mutual fund investors considered it as important, 45 viewed that entry and
exit load is somewhat important, 37 investors of mutual fund thought it as not very
important and 12 mutual fund investors considered the fund entry and exit load as not at
all important. As entry load is charged by the fund at the time of investment in fund and
exit load is charged at the of redemption by the unit holders.

Table 5.28 of Q21

165
Highly Important Somewhat Not very Not at all
important important important important
105 31 45 37 12

166
Figure E.60 Bar of (Q21)

Figure E.61 Pie of (Q21)

167
 Figure E.62 Curve of (Q21)

Q22 Minimum Initial Investments

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total 230 investors ,78 investors were viewing minimum
investment as highly important, 53 said it as important, 62 investors thought it as somewhat
important, 19 investors considered it as not very important, and 18 investors considered
it as not at all important

Table 5.29 of Q22

highly Important somewhat not very not at all

important important important important
78 53 62 19 18

168
Figure E.63 Bar of (Q22)

Figure E.64 Pie of (Q22)

169
 Figure E.65 Curve of (Q22)

Q23 Disclosure of NAV on Every Day

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total 230 investors of mutual fund, 82 investors considered
disclosure of daily NAV as highly important, 72 termed it as important, 35 investors
considered it as somewhat important, 17 viewed it as not very important and 24 considered
it as not at all important.

Table 5.30 of Q23

highly Important somewhat not very not at all

important important important important
82 72 35 17 24

170
 Figure E.66 Bar of (Q23)

Figure E.67 Curve of (Q23)

Q24 Disclosure of Derivation From the Original Pattern

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total investors questioned, 92 investors viewed the disclosure
of derivation from original pattern as highly important, 35 considered it as important, 61
investors saying that it is somewhat important, 7 considered it as not very important and
35 said it is not all important.

171
Table 5.31 Of Q24

highly Important somewhat not very not t all

important important important important
92 35 61 7 35

Figure E.68 Bar of (Q24)

Figure E.69 Pie of (Q24)

172
 Figure E.68Curve of (Q24)

Q25 How Did You Come to Know About Mutual Fund Investment Scheme?
The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 12 mail, 7 radio, 8 tv, 135 broker and 68 financial magazine.
Table 5.32 of Q25
Mail Radio Tv Broker/agent Financial magazine
12 7 8 135 68

Figure E.71 Bar of (Q25)

173
 Figure E.72 Pie of (Q25)

Figure E.73 Curve of (Q25)

Q26 Do You Think Mutual Fund Investing is a Best Alternative to Easily Investing?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 187 mutual fund investors agreed with the ease option and
said yes, 35 investors said no and 8 investors of the sample investors surveyed for the
purpose said we don’t know.

174
Table 5.33 of Q26
Yes No don’t know
187 35 8

Figure E.74 Bar of (Q26)

Figure E.75 Pie of (Q26)

175
 Figure E.76 Curve of (Q26)

Q27 What Kind of Communication You Like to Know About Your Fund
Performance?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of them total 60 liked automated response, 20 personal visit, 40
telephone and 110 no preferences liked as for their communication with mutual Fund.

Table 5.34 of Q17

Automated Personal Visit Telephone No Preferences

response
60 20 40 110

176
 Figure E.77 Bar of (Q27)

Figure E.78 Pie of (Q27)

177
 Figure E.77 Curve of (Q27)

Figure E.79 curve of (Q27)

Q28 Investment in Mutual Fund Helps You Realize the Benefits of Stock Market
Investing?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. When the respondents were questioned most of them i.e. 182 agreed
with the fact that mutual fund Realy give u realization of the benefits in stock markets. 31
of the surveyed investors said not at all realizing and agreed with no option, whereas 17 of
the total 230 mutual fund investors said we do not know about this fact.
Table 5.35 of Q28
Yes No Don’t Know

182 31 17

178
 Figure E.80 Bar of (Q28)

Figure E.81 Pie of (Q28)

Figure E.80 Curve of (Q28)

179
 Figure E.82 curve of (Q28)

Q29 Mutual Fund Investing Gives a Definite Positive Return?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 total 180 believed that mutual fund investment give
definite return whereas 38 and 12 investors considered it with No option and do not know
options respectively
Table 5.36 of Q29
.
Yes No Don’t know

180 38 12 Figure
E.83 Bar of
(Q29)

Figure E.84 Pie of (Q29)

180
 Figure E.83 Curve of (Q29)

Figure E.85 curve of (Q29)

Q30 Mutual Fund Returns and Principals are Fully Protected and Granted by
MUFAP.

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan This question represent the behavior of the investors to know their
investment protection mechanism by any government institution in Pakistan. In Pakistan
mostly the funds are traded in MUFAP and KSE. These organizations provide protection
and establish rules for the investors. Out of total 230 investors questioned, 217 investors
agreed that MUFAP really protect us in term of return and rules whereas, 2 investors said
no and 11 were found not aware and said we do not know.
Table 5.37 of Q30
Yes No don’t know
217 2 11

181
 Figure E.86 Bar of (Q30)

Figure E.87 Pie of (Q30)

182
 Figure E.88 Curve of (Q30)

Q31 Bank Sponsored MF gives more Return than Bank Fixed Deposit?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of 230 investors questioned for this research study ,135 said off
course bank sponsored mutual funds give more return than bank fixed, where as another
65 said no and 30 investors said we do not know about this logic.

Table 5.38 of Q31

Yes No don’t know

135 65 30

160
140
120 bank sponsered MF
give more return
100 than bank fixed
80 deposit
60 bank sponsered MF
give more return
40
than bank fixed
20 deposit
0
yes no don’t know

Figure E.89 Bar of (Q31)

183
 Figure E.90 Pie of (Q31)

Figure E.91 Curve of (Q31)

Q32 Is Entry and Exit out of Mutual Fund Easy?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total number being surveyed in this research study. The huge
number like 161 agreed with ease in entry and exit whereas 15 investors believed that there
is difficulty in entry and exit due to some load fee and other procedures and they went with
no option another 54 investors said we do not know about this.

184
Table 5.39 of Q32

Yes No don’t know

161 15 54

Figure E.92 Bar of (Q32)

Figure E.93 Pie of (Q32)

185
 180
160
140
120 entry and exit
100 out of mutual
80 fund is easyyes
60 entry and exit
40 out of mutual
20 fund is easyyes
0
yes no don’t
know

Figure E.94 Curve of (Q32)

Q33 Due to Professional Investment, a Better Return is expected from Mutual Fund

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. This question Represent the behavior of investors towards the
importance of professional management. 187 investors believed that investment in mutual
fund yield better return as these investments are managed through professional managers.
9 Investor disagreed and said no regarding better return due to better professional
management. 34 investors said we do not know.

Table 5.40 of Q33

Yes No don’t know
187 9 34

186
Figure E.95 Bar of (Q33)

Figure E.96 Pie of (Q33)

187
 Figure E.97 curve of (Q33)

Q34 Ups and down neither of Stock Market will nor affect the Return from Mutual
Fund?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. Out of total 230, only 87 investors agreed that up and down in capital
market would not affect the mutual fund return. Whereas 98 said it will affect the mutual
fund return and 45 investors said we do not know.

Table 5.41 of Q34

Yes No don’t know

87 98 45

188
Figure E.98 Bar of (Q34)

Figure E.99 Pie of (Q34)

189
 Figure E.98 Curve of (Q34)

Figure E.100 curve of (Q34)

Q35 MUFAP helps in protecting the Mutual Fund Industry and the Unit Holders.

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. About this question 180 said yes MUFAP helps in protecting mutual
fund industry, 10 investors said NO and 40 investors said we do not know about this.

Table 5.42 of Q35

Yes No don’t know

180 10 40

190
Figure E.101 Bar of (Q35)

Figure E.102 Pie of (Q35)

191
 Figure E.103 Curve of (Q35)

Q36 How Often You Switch the Schemes of Mutual Fund in a Year?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. The below table shows how often an investor switch the scheme of
mutual fund in a year. out of total 230 investors 135 investors said we never switch the
mutual fund scheme, 35 said we change one or two time, 41 investors said we change the
scheme three or four times and similarly 19 investors said more than four times.

Table 5.43 of Q 36

Never one or two times three or four more than four

times time
135 35 41

192
Figure E.104 Bar of (Q36)

Figure E.105 Pie of (Q36)

193
Figure E.106 Curve of (Q36)

194
Q37 After Investment, how frequently you monitor the Performance of Mutual Fund?

The below table, bar, pie and curve represent the responses of mutual fund investors
towards mutual fund. The data of this research was collected from total 230 mutual Fund
investors in Pakistan. The below chart shows the frequency of the monitoring that how
frequently a mutual fund investment are monitored by investors. 50 investors said we
weekly monitor the performance of mutual fund, 71 investors said we monitor it once a
month, 95 said once a year and 14 said we rarely monitor the performance of the mutual
fund.

Table 5.44 Q37

Weekly Once in Once in year Rarely

month
50 71 95 14

Figure E.107 Bar of (Q37)

5.2.2 Investors Investing Behavior in Mutual Fund through Multinomial Logistic

Regression

195
Q1 Gender
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here gender is the dependent variable and male and female as
independent or categorical variables. 1 representing male mutual fund investors and
These tables show the gender of the mutual fund investors investing in mutual fund. The
relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level.
Table 5.45Case Processing Summary (Q1)

Case Processing Summary

N Marginal Percentage

1 210 91.3%
Gender
2 20 8.7%
Valid 230 100.0%

Missing 0

Total 230

Subpopulation 1

Table 5.46 Model fitting Information (Q1)

Model Fitting Information

Model Model Fitting Criteria Likelihood Ratio Tests

196
 -2 Log Likelihood Chi-Square df Sig.

Intercept Only 4.751

Final 4.751 .000 0 .

Table 5.47 Likelihood Ratio Test(Q1)

Likelihood Ratio Tests

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 187.697 182.946 1 .000

Table 5.48 Parameter Estimates (Q1)

Parameter Estimates
Gendera B Std. Error Wald Df Sig.
1 Intercept 2.351 .234 100.964 1 .000
a. The reference category is: 2.

197
Q2 Ages

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here age is the dependent variable and male and female as independent
or categorical variables. 1 represents investors having ages below 30 years, 2 represent
mutual fund investors having ages above 30 up to 40. And 3represent mutual fund investors
having ages above 40 and invest in mutual fund.
These tables show the ages of the mutual fund investors investing in mutual fund. The
relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level. The results indicate that mutual fund investors above 40 years of ages are more
inclined towards investment in mutual fund.
Table 5.49 Case Processing Summary(Q2)
Case Processing Summary
N Marginal
Percentage
1 20 8.7%
Age 2 50 21.7%
3 160 69.6%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

198
 Table 5.50 Model Fitting Information(Q2)
Model Fitting Information
Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.

Intercept Only 10.233

Final 10.233 .000 0 .

Table 5.51 Likelihood Ratio test(Q2)

Likelihood Ratio Tests
Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 149.165 138.932 2 .000

Table 5.52 Parameter Estimates(Q2)

Parameter Estimates
Agea B Std. Error Wald Df Sig.
1 Intercept -2.079 .237 76.872 1 .000
2 Intercept -1.163 .162 51.540 1 .000
a. The reference category is: 3.

Q3 Education
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here education is the dependent variable levels of education of the mutual
fund investors are the independent variables.. 1 representing investors of mutual fund
having post-graduation and invest in mutual fund. 2 represents investors having graduation
and invest in mutual fund.3 represent investors having under graduation , 4 showing

199
investors of mutual fund with diploma level, 5 represent investors having high school level
education and 6 represent investors investing with less than high school level education.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level. The result shows.
Table 5.53 Case Processing Summary (Q3)

Case Processing Summary

N Marginal
Percentage

1 60 26.1%

2 80 34.8%

Education 3 40 17.4%

4 35 15.2%

5 15 6.5%
Valid 230 100.0%
Table Missing 0 5.54 Model
Fitting Total 230 Information (Q3)
Subpopulation 1

Model Fitting Information

Model Model Fitting Likelihood Ratio Tests

Criteria

-2 Log Likelihood Chi-Square Df Sig.

Intercept Only 20.366

200
 Final 20.366 .000 0 .

Q4 Occupation
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here gender is the dependent variable and male and female as
independent or categorical variables. 1 representing professional, 2 represent business man,
3 represent salaried persons and 4 represent retired persons investing in mutual funds.
These tables show the occupation level of the mutual fund investors investing in mutual
fund. The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level. The results suggest that almost all occupation people invest in mutual fund.

Table 5.55 Case Processing Summary (Q4)

N Marginal Percentage
1 69 30.0%
2 61 26.5%
Occupation
3 55 23.9%
4 45 19.6%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.56 Model Fitting Information (Q4)

Model Model Fitting

201
 Criteria
Likelihood Ratio Tests
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 16.246
Final 16.246 .000 0 .

Table 5.58 Parameter Estimates (Q4)

Occupation B Std. Error Wald Df Sig.
1 Intercept .427 .192 4.976 1 .026
2 Intercept .304 .197 2.397 1 .122
3 Intercept .201 .201 .997 1 .318
a. The reference category is: 4.

Q5 Marital Status
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here gender is the dependent variable and male and female as
independent or categorical variables. 1 represent married mutual fund investors and 2
represent unmarried mutual fund investors.
These tables show the marital status of the mutual fund investors investing in mutual fund.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level. The results found most of the married mutual fund investors investing in mutual fund.

202
 Table 5.59 Case Processing Summary (Q5)
N Marginal
Percentage
1 190 82.6%
Marital Status
2 40 17.4%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.60 Model Fitting Information (Q5)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Chi-Square Df Sig.
Likelihood
Intercept Only 5.340
Final 5.340 .000 0 .

Table 5.61 Likelihood Ratio Tests(Q5)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Chi-Square Df Sig.
Likelihood of
Reduced Model
Intercept 111.651 106.311 1 .000

Table 5.62 Parameter Estimates (Q5)

Marital Status B Std. Error Wald Df Sig.
1 Intercept 1.558 .174 80.223 1 .000
a. The reference category is: 2

203
Q6 Annual Income
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here annual income is the dependent variable and the different categories
are the independent variables. 1, 2, 3 and 4 represent the annual income levels of the mutual
fund investors.
These tables show the annual income of the mutual fund investors investing in mutual fund.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level.

Table 5.63 Case Processing Summary (Q6)

N Marginal
Percentage
1 45 19.6%
2 70 30.4%
Annual Income
3 38 16.5%
4 77 33.5%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

204
 Table 5.64 Model Fitting Information(Q6)
Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Chi-Square Df Sig.
Likelihood
Intercept Only 16.124
Final 16.124 .000 0 .

Table 5.65 Likelihood Ratio Tests (Q6)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 35.094 18.970 3 .000

Table 5.66 Parameter Estimates(Q6)

Annual Income B Std. Error Wald Df Sig.
1 Intercept -.537 .188 8.195 1 .004
2 Intercept -.095 .165 .333 1 .564
3 Intercept -.706 .198 12.690 1 .000
a. The reference category is: 4.

Q7 Annual Saving
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here annual saving is the dependent variable and the different categories
are the independent variables. 1,2, and 3 represent the annual saving levels of the mutual
fund investors.
These tables show the annual saving of the mutual fund investors investing in mutual fund.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than

205
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level.
Table 5.67 Case Processing Summary (Q7)
N Marginal
Percentage
1 94 40.9%
Annual Saving 2 35 15.2%
3 101 43.9%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.68 Model Fitting Information(Q7)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 10.959
Final 10.959 .000 0 .

206
 Table 5.69 Likelihood Ratio Tests(Q7)
Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 50.072 39.113 2 .000

Table 5.70 Parameter Estimates (Q7)

Annual Saving B Std. Error Wald df Sig.
1 Intercept -.072 .143 .251 1 .616
2 Intercept -1.060 .196 29.193 1 .000
a. The reference category is: 3.

Q8 Location
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here location is the dependent variable and male and female as
independent or categorical variables. 1 represent mutual fund investors in Peshawar, 2
represent investors of Islamabad, 3 represent mutual fund investors of Karachi, 4 represent
investors of Lahore and 5 represents investors belonging to Quetta analyzed for the
research study.
These tables show the location of the mutual fund investors investing in mutual fund. The
relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level.

207
 Table 5.71 Case Processing Summary(Q8)
N Marginal Percentage
1 79 34.3%
2 50 21.7%
3 15 6.5%
Location 4 72 31.3%
14 6.1%
5

Valid 230 100.0%

Missing 0
Total 230
Subpopulation 1

Table 5.72 Model Fitting Information (Q8)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 19.849
Final 19.849 .000 0 .

208
 Table 5.73 Likelihood Ratio Tests (Q8)
Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 111.224 91.375 4 .000

Table 5.74 Parameter Estimates(Q8)

Locationa B Std. Error Wald df Sig.
1 Intercept 1.730 .290 35.609 1 .000
2 Intercept 1.273 .302 17.724 1 .000
3 Intercept .069 .372 .034 1 .853
4 Intercept 1.638 .292 31.433 1 .000

Q9 Do you prefer Investment in Mutual Funds to others Saving Avenue in Future?

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here the preference to mutual fund than other saving avenues is used as
the dependent variable and yes, no and not sure as independent or categorical variables. 1
representing yes, 2 represents no and 3 represent not sure. These tables show the preference
towards mutual fund by investors investing in mutual fund. The relationship has been
shown between non matric dependent variable and matric independent variables. The final
value of chi-square is statistically significant showing that there is relationship between
dependent variable and independent variables. The results suggest that there is no
muilticolinarity issue as the standard error value is less than 2.00.The likelihood ratio test
suggest that there exist a relationship between independent variables and dependent
variable, as the chi-square value is significant at 5% probability level. Most of the investors
investing in mutual fund give preference during the survey results.

209
 Table 5.75 Case Processing Summary (Q9)
N Marginal
Percentage
Prefrence 1 214 93.0%
than 2 10 4.3%
other 6 2.6%
avenues 3
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.76 Model Fitting Information (Q9)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 7.742
Final 7.742 .000 0 .

Table 5.77 Likelihood Ratio Tests(Q9)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 375.778 368.036 2 .000

Table 5.78 Parameter Estimates(Q9)

Preference than other B Std. Error Wald Df Sig.
avenues
1 Intercept 3.574 .414 74.560 1 .000
2 Intercept .511 .516 .979 1 .323
a. The reference category is: 3.

210
Q 10 Generally you prefer
The below tables show the output of multinomial logistic regression representing the relationship of mutual
fund investors investing behavior factors to be considered by the mutual fund investors in non-matric forms
and the independent variables in matric or categorical form. Here the general preference to mutual fund
categories is used as the dependent variable and fund categories as independent or categorical variables. 1
to 8 representing the different categories of mutual fund on mufap. These tables show the preference towards
mutual fund categories by investors investing in mutual fund. The relationship has been shown between non
matric dependent variable and matric independent variables. The final value of chi-square is statistically
significant showing that there is relationship between dependent variable and independent variables. The
results suggest that there is no muilticolinarity issue as the standard error value is less than 2.00.The
likelihood ratio test suggest that there exist a relationship between independent variables and dependent
variable, as the chi-square value is significant at 5% probability level.
Table 5.79 Case processing Summary (Q10)
N Marginal Percentage
1 50 21.7%
2 45 19.6%
3 29 12.6%
Generally you4 20 8.7%
prefer 5 18 7.8%
6 8 3.5%
7 41 17.8%
8 19 8.3%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.80 Model Fitting Information(Q10)

Model Model Fitting Criteria Likelihood Ratio Tests
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 33.200
Final 33.200 .000 0 .
Table 5.81 Likelihood Ratio Tests(Q10)
Effect Model Fitting Criteria Likelihood Ratio Tests
-2 Log Likelihood of Chi-Square df Sig.
Reduced Model
Intercept 90.887 57.687 7 .000

Table 5.82 Parameter Estimates(Q10)

211
Generally you prefer B Std. Error Wald df Sig.
1 Intercept .968 .270 12.890 1 .000
2 Intercept .862 .274 9.932 1 .002
3 Intercept .423 .295 2.053 1 .152
4 Intercept .051 .320 .026 1 .873
5 Intercept -.054 .329 .027 1 .869
6 Intercept -.865 .421 4.212 1 .040
7 Intercept .769 .278 7.680 1 .006
a. The reference category is: 8.

Q11 Which kind of Mutual Fund would you preferred?

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here preference to kinds of funds is the dependent variable and different
funds kind are the independent or categorical variables. 1 representing open ended, 2
represent close ended and 3 represent voluntary pension funds.
These tables show the preference towards kinds of fund being preferred by the mutual fund
investors investing in mutual fund. The relationship has been shown between non matric
dependent variable and matric independent variables. The final value of chi-square is
statistically significant showing that there is relationship between dependent variable and
independent variables. The results suggest that there is no muilticolinarity issue as the
standard error value is less than 2.00.The likelihood ratio test suggest that there exist a
relationship between independent variables and dependent variable, as the chi-square value
is significant at 5% probability level. The results indicate that most of the mutual fund
investors like open ended mutual fund.

212
 Table 5.83 Case processing Summary (Q11)

N Marginal
Percentage
Kinds of1 180 78.3%
fund 2 37 16.1%
preferred 3 13 5.7%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.84 Model Fitting Information(Q11)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 9.624
Final 9.624 .000 0 .

Table 5.85 Likelihood Ratio Tests(Q11)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 216.830 207.206 2 .000

Table 5.86 Parameter Estimates(Q11)

Kinds of fund preferred B Std. Error Wald Df Sig.
1 Intercept 2.628 .287 83.736 1 .000
2 Intercept 1.046 .322 10.525 1 .001
a. The reference category is: 3.

Q12 You preferred Investment in Mutual Fund due to?

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here preference any of five characteristic of the financial market is the

213
dependent variable and different funds kind are the independent or categorical variables.
From 1 to 9 represent the different 9 attributes assigned.
These tables show the preference towards the market characteristics of fund being preferred
by the mutual fund investors investing in mutual fund. The relationship has been shown
between non matric dependent variable and matric independent variables. The final value
of chi-square is statistically significant showing that there is relationship between
dependent variable and independent variables. The results suggest that there is no
muilticolinarity issue as the standard error value is less than 2.00.The likelihood ratio test
suggest that there exist a relationship between independent variables and dependent
variable, as the chi-square value is significant at 5% probability level. The results indicate
the mixed preference level of investors towards the nine characteristics.

Table 5.87 Case Processing Summary (Q12)

N Marginal
Percentage
1 75 32.6%
2 44 19.1%
3 25 10.9%
4 5 2.2%
Preference due
5 7 3.0%
to
6 31 13.5%
7 9 3.9%
8 8 3.5%
9 26 11.3%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.88 Model Fitting Information(Q12)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 35.230
Final 35.230 .000 0 .

214
 Table 5.89 Likelihood Ratio Tests(Q12)
Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 184.499 149.269 8 .000

Table 5.90 Parameter Estimates(Q12)

Preference due to B Std. Error Wald df Sig.
1 Intercept 1.059 .228 21.668 1 .000
2 Intercept .526 .247 4.523 1 .033
3 Intercept -.039 .280 .020 1 .889
4 Intercept -1.649 .488 11.398 1 .001
5 Intercept -1.312 .426 9.496 1 .002
6 Intercept .176 .266 .437 1 .508
7 Intercept -1.061 .387 7.524 1 .006
8 Intercept -1.179 .404 8.499 1 .004
a. The reference category is: 9

Q13 How do you invest in this Option?

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here gender is the dependent variable and male and female as
independent or categorical variables. 1 repressing investment through agents, 2 through
family and 3 investment through own self.
These tables show the investment in mutual fund through different options by mutual fund
investors investing in mutual fund. The relationship has been shown between non matric
dependent variable and matric independent variables. The final value of chi-square is
statistically significant showing that there is relationship between dependent variable and
independent variables. The results suggest that there is no muilticolinarity issue as the
standard error value is less than 2.00.The likelihood ratio test suggest that there exist a

215
relationship between independent variables and dependent variable, as the chi-square value
is significant at 5% probability level.

Table 5.91 Case Processing Summary (Q13)

N Marginal
Percentage
How do1 91 39.6%
you 2 74 32.2%
invest 3 65 28.3%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.92 Model Fitting Information(Q13)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 11.233
Final 11.233 .000 0 .

Table 5.93 Likelihood Ratio Tests (Q13)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 15.726 14.494 2 .006

Table 5.94 Parameter Estimates(Q13)

How do you invest B Std. Error Wald Df Sig.
1 Intercept .336 .162 4.293 1 .038
2 Intercept .130 .170 .582 1 .446
a. The reference category is: 3.

Q14 You Think Fund Performance Record as

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here fund performance record is the dependent variable and highly

216
important, important, somewhat important, not very important and not at all important as
independent or categorical variables. From 1 to 5 represent these categories. These tables
show the performance record of the mutual fund to be liked by investors investing in
mutual fund. The relationship has been shown between non matric dependent variable and
matric independent variables. The final value of chi-square is statistically significant
showing that there is relationship between dependent variable and independent variables.
The results suggest that there is no muilticolinarity issue as the standard error value is less
than 2.00.The likelihood ratio test suggest that there exist a relationship between
independent variables and dependent variable, as the chi-square value is significant at 5%
probability level.

Table 5.95Case Processing Summary(Q14)

N Marginal
Percentage
1 65 28.3%
Fund 2 119 51.7%
performa
3 30 13.0%
nce
record 4 5 2.2%
5 11 4.8%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.96 Model Fitting Information (Q14)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 18.333
Final 18.333 .000 0 .

Table 5.97 Likelihood Ratio Tests(Q14)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 210.179 191.847 4 .000

217
 Table 5.98 Parameter Estimates(Q14)
Fund performance B Std. Error Wald df Sig.
record
1 Intercept 1.776 .326 29.691 1 .000
2 Intercept 2.381 .315 57.095 1 .000
3 Intercept 1.003 .352 8.102 1 .004
4 Intercept -.788 .539 2.137 1 .144
a. The reference category is: 5.

Q15 Fund Reputation Brand Name

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. Here fund reputation brand name is the dependent variable and it has
been judged through highly important, important, somewhat important, not very important
and not at all important as independent or categorical variables. From 1 to 5 representing
all these independent variables categories.
These tables show the fund reputation brand name of mutual fund to be considered by
investors investing in mutual fund. The relationship has been shown between non matric
dependent variable and matric independent variables. The final value of chi-square is
statistically significant showing that there is relationship between dependent variable and
independent variables. The results suggest that there is no muilticolinarity issue as the
standard error value is less than 2.00.The likelihood ratio test suggest that there exist a
relationship between independent variables and dependent variable, as the chi-square value
is significant at 5% probability level. The results suggest that all these investors
investigated for the research purpose give highly importance and importance to fund
reputation brand name. Very few opted for the rest three categories.

218
 Table 5.99 Case Processing Summary(Q15)
N Marginal
Percentage
1 52 22.6%
Fund 2 86 37.4%
reputation
3 45 19.6%
bran
name 4 31 13.5%
5 16 7.0%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.100 Model Fitting Information(Q15)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 20.356
Final 20.356 .000 0 .

Table 5.101 Likelihood Ratio Tests(Q15)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 80.488 60.132 4 .000

Table 5.102 Parameter Estimates (Q15)

Fund reputation brand B Std. Error Wald df Sig.
name
1 Intercept 1.179 .286 16.998 1 .000
2 Intercept 1.682 .272 38.154 1 .000
3 Intercept 1.034 .291 12.621 1 .000
4 Intercept .661 .308 4.616 1 .032
a. The reference category is: 5.

219
Q16 Scheme Portfolio of Investments
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. Here scheme portfolio of investment is the dependent variable and it has
been judged through highly important, important, somewhat important, not very important
and not at all important as independent or categorical variables. From 1 to 5 representing
all these independent variables categories.
These tables show the scheme portfolio of investment of mutual fund to be considered by
investors investing in mutual fund. The relationship has been shown between non matric
dependent variable and matric independent variables. The final value of chi-square is
statistically significant showing that there is relationship between dependent variable and
independent variables. The results suggest that there is no muilticolinarity issue as the
standard error value is less than 2.00.The likelihood ratio test suggest that there exist a
relationship between independent variables and dependent variable, as the chi-square value
is significant at 5% probability level. The results suggest that all these investors
investigated for the research purpose give highly importance and importance to fund reputation
brand name. Very few opted for the rest three categories.
Table 5.103 Case Processing Summary(Q16)
N Marginal
Percentage
1 98 42.6%
Scheme 2 67 29.1%
portfolio
3 30 13.0%
investme
nt 4 18 7.8%
5 17 7.4%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

220
 Table 5.104 Model Fitting Information (Q16)
Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 19.856
Final 19.856 .000 0 .

Table 5.105 Likelihood Ratio Tests(Q16)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 125.218 105.362 4 .000

Table 5.106 Parameter Estimates(Q16)

Scheme portfolio B Std. Error Wald df Sig.
investment
1 Intercept 1.752 .263 44.455 1 .000
2 Intercept 1.371 .272 25.505 1 .000
3 Intercept .568 .304 3.501 1 .061
4 Intercept .057 .338 .029 1 .866
a. The reference category is: 5.

221
Q17 Withdrawal Facilities
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. Here withdrawal facilities of mutual fund are the dependent variable and
it has been judged through highly important, important, somewhat important, not very
important and not at all important as independent or categorical variables. From 1 to 5
representing all these independent variables categories.
These tables show the withdrawal facilities of mutual fund to be considered by investors
investing in mutual fund. The relationship has been shown between non matric dependent
variable and matric independent variables. The final value of chi-square is statistically
significant showing that there is relationship between dependent variable and independent
variables. The results suggest that there is no muilticolinarity issue as the standard error
value is less than 2.00.The likelihood ratio test suggest that there exist a relationship
between independent variables and dependent variable, as the chi-square value is
significant at 5% probability level. The results suggest that all these investors investigated
for the research purpose give highly importance and importance to fund reputation brand
name. A very few opted for the rest three categories.

222
 Table 5.107 Case Processing Summary (Q17)
N Marginal
Percentage
1 71 30.9%
2 91 39.6%
withdrawal facilities 3 30 13.0%
4 25 10.9%
5 13 5.7%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.108 Model Fitting Information(Q17)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 19.900
Final 19.900 .000 0 .

Table 5.109 Likelihood Ratio Tests(Q17)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 116.707 96.806 4 .000

Table 5.110 Parameter Estimates(Q17)

withdrawal facilities B Std. Error Wald df Sig.
1 Intercept 1.698 .302 31.671 1 .000
2 Intercept 1.946 .296 43.072 1 .000
3 Intercept .836 .332 6.343 1 .012
4 Intercept .654 .342 3.657 1 .056
a. The reference category is: 5.

223
Q18 Favorable Rating by Rating Agencies
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. Here the favorable rating of mutual fund being rated by the rating agency
is the dependent variable and it has been judged through highly important, important,
somewhat important, not very important and not at all important as independent or
categorical variables. From 1 to 5 representing all these independent variables categories.
These tables show the favorable rating of mutual fund to be considered by investors
investing in mutual fund. The relationship has been shown between non matric dependent
variable and matric independent variables. The final value of chi-square is statistically
significant showing that there is relationship between dependent variable and independent
variables. The results suggest that there is no muilticolinarity issue as the standard error
value is less than 2.00.The likelihood ratio test suggest that there exist a relationship
between independent variables and dependent variable, as the chi-square value is
significant at 5% probability level. The results suggest that all these investors investigated
for the research purpose give highly importance and importance to fund reputation brand
name. A very few opted for the rest three categories.

224
 Table 5.111 Case Processing Summary (Q18)
N Marginal Percentage
1 24 10.4%
Favorable 2 30 13.0%
rating by3 91 39.6%
rating agency 4 70 30.4%
5 15 6.5%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.112 model fitting tests(Q18)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 19.987
Final 19.987 .000 0 .

Table 5.113 Likelihood Ratio Tests(Q18)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 112.438 92.451 4 .000

Table 5.114 Parameter Estimates(Q18)

Favorable rating by B Std. Error Wald df Sig.
rating agency
1 Intercept .470 .329 2.039 1 .153
2 Intercept .693 .316 4.805 1 .028
3 Intercept 1.803 .279 41.853 1 .000
4 Intercept 1.540 .285 29.313 1 .000
a. The reference category is: 5.

225
Q19 Innovativeness of the Scheme
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. Here the innovativeness of mutual fund being considered by the mutual
fund investors is a dependent variable and it has been judged through highly important,
important, somewhat important, not very important and not at all important as independent
or categorical variables. From 1 to 5 representing all these independent variables
categories.
These tables show the innovativeness of the scheme of mutual fund to be considered by
investors investing in mutual fund. The relationship has been shown between non matric
dependent variable and matric independent variables. The final value of chi-square is
statistically significant showing that there is relationship between dependent variable and
independent variables. The results suggest that there is no muilticolinarity issue as the
standard error value is less than 2.00.The likelihood ratio test suggest that there exist a
relationship between independent variables and dependent variable, as the chi-square value
is significant at 5% probability level. The results suggest that all these investors
investigated for the research purpose give highly importance and importance to fund
reputation brand name. A very few opted for the rest three categories.

226
 Table 5.115 Case Processing Summary(Q19)
N Marginal
Percentage
1 45 19.6%
Innovativ 2 54 23.5%
eness of3 78 33.9%
scheme 4 46 20.0%
5 7 3.0%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.116 Model Fitting Information(Q19)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 19.876
Final 19.876 .000 0 .

Table 5.117 Likelihood Ratio Tests(Q19)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 91.235 71.359 4 .000

Table 5.118 Parameter Estimates(Q19)

Innovativeness of B Std. Error Wald df Sig.
scheme
1 Intercept 1.861 .406 20.974 1 .000
2 Intercept 2.043 .402 25.866 1 .000
3 Intercept 2.411 .395 37.333 1 .000
4 Intercept 1.883 .406 21.536 1 .000
a. The reference category is: 5.

227
Q 20 Products with Tax Benefits
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. Here the fund tax benefits were analyzed as the dependent variable and
it has been judged through highly important, important, somewhat important, not very
important and not at all important as independent or categorical variables. From 1 to 5
representing all these independent variables categories.
These tables show the tax benefit of mutual fund to be considered by investors investing in
mutual fund. The relationship has been shown between non matric dependent variable and
matric independent variables. The final value of chi-square is statistically significant
showing that there is relationship between dependent variable and independent variables.
The results suggest that there is no muilticolinarity issue as the standard error value is less
than 2.00.The likelihood ratio test suggest that there exist a relationship between
independent variables and dependent variable, as the chi-square value is significant at 5%
probability level. The results suggest that all these investors investigated for the research
purpose give highly importance and importance to fund reputation brand name. A very
few opted for the rest three categories.

Table 5.119 Case Processing Summary(Q20)

N Marginal
Percentage
1 73 31.7%
Product 2 31 13.5%
with tax3 53 23.0%
benifits 4 37 16.1%
5 36 15.7%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

228
 Table 5.120 Model Fitting Information(Q20)
Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 20.822
Final 20.822 .000 0 .

Table 5.121 Likelihood Ratio Tests(Q20)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 45.033 24.211 4 .000

Table 5.122 Parameter Estimates(Q20)

Product with tax B Std. Error Wald df Sig.
benefits
1 Intercept .707 .204 12.049 1 .001
2 Intercept -.150 .245 .372 1 .542
3 Intercept .387 .216 3.207 1 .073
4 Intercept .027 .234 .014 1 .907
a. The reference category is: 5.

Q21 Entry & Exit Load

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. In these tables the fund entry and exit load has been analyzed considering
the exit and entry load as the dependent variable and it has been judged through highly
important, important, somewhat important, not very important and not at all important as
independent or categorical variables. From 1 to 5 representing all these independent

229
variables categories. These tables viewing the results regarding the entry and exit load of
mutual fund to be considered by investors investing in mutual fund. The relationship has
been shown between non matric dependent variable and matric independent variables. The
final value of chi-square is statistically significant showing that there is relationship
between dependent variable and independent variables. The results suggest that there is no
muilticolinarity issue as the standard error value is less than 2.00.The likelihood ratio test
suggest that there exist a relationship between independent variables and dependent
variable, as the chi-square value is significant at 5% probability level. The results suggest
that all these investors investigated for the research purpose give highly importance and
importance to entry and exit load. A very few opted for the rest three categories.

Table 5.123 Case Processing Summary(Q21)

N Marginal
Percentage
1 105 45.7%
2 31 13.5%
Entry and exitload 3 45 19.6%
4 37 16.1%
5 12 5.2%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.124 Model Fitting Information(Q21)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 19.932
Final 19.932 .000 0 .

230
 Table 5.125Likelihood Ratio Tests(Q21)
Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 118.441 98.509 4 .000

Table 5.126 Parameter

estimate(Q21)
Entryandexitload B Std. Error Wald df Sig.
1 Intercept 2.169 .305 50.667 1 .000
2 Intercept .949 .340 7.793 1 .005
3 Intercept 1.322 .325 16.551 1 .000
4 Intercept 1.126 .332 11.489 1 .001
a. The reference category is: 5

Q22 Minimum Initial Investment

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. In these tables minimum initial investment of fund has been analyzed
considering it as the dependent variable and it has been judged through scaling highly
important, important, somewhat important, not very important and not at all important as
independent or categorical variables. From 1 to 5 representing all these independent
variables categories.
These tables viewing the results regarding the minimum initial investment characteristic of
mutual fund to be considered by investors investing in mutual fund. The relationship has
been shown between non matric dependent variable and matric independent variables. The
final value of chi-square is statistically significant showing that there is relationship
between dependent variable and independent variables. The results suggest that there is no
muilticolinarity issue as the standard error value is less than 2.00.The likelihood ratio test
suggest that there exist a relationship between independent variables and dependent
variable, as the chi-square value is significant at 5% probability level. The results suggest
that all these investors investigated for the research purpose give highly importance and

231
importance to minimum initial investment. A very few mutual fund investors opted for
the rest three categories i.e somewhat important, not very important and not at all
important.

Table 5.127 Case Processing Summary(Q22)

N Marginal
Percentage
1 78 33.9%
Minimum 2 53 23.0%
initial
3 62 27.0%
investme
nt 4 19 8.3%
5 18 7.8%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.128 Model Fitting Information(Q22)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 20.228
Final 20.228 .000 0 .

Table 5.129 Likelihood Ratio Tests(Q22)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 87.257 67.029 4 .000

Table 5.130 Parameter Estimates(Q22)

Minimum initial B Std. Error Wald df Sig.
investment
1 Intercept 1.466 .261 31.446 1 .000
2 Intercept 1.080 .273 15.670 1 .000
3 Intercept 1.237 .268 21.338 1 .000
4 Intercept .054 .329 .027 1 .869
a. The reference category is: 5.

232
Q23 Disclosure of NAV on every day
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. In these tables the disclosure of fund has been analyzed considering it
as the dependent variable and it has been judged through scaling highly important,
important, somewhat important, not very important and not at all important as independent
or categorical variables. From 1 to 5 representing all these independent variables
categories.
These tables viewing the results regarding the disclosure of net asset value of mutual fund.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level. The results suggest that all these investors investigated for the research purpose gave
highly importance and importance to disclosure of NAV on every day. A very few mutual
fund investors opted for the rest three categories i.e somewhat important, not very
important and not at all important.
Table 5.131 Case Processing Summary(Q23)
N Marginal
Percentage
1 82 35.7%
2 72 31.3%
Disclosur
3 39 17.0%
e of NAV
4 19 8.3%
5 18 7.8%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

233
 Table 5.132 Model Fitting Information(Q23)

Model Model Fitting Likelihood Ratio Tests

Criteria
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 20.121
Final 20.121 .000 0 .

Table 5.133 Likelihood Ratio Tests(Q23)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 99.188 79.067 4 .000
.

Table 5.134 Parameter Estimates(Q23)

Disclosure of NAV B Std. Error Wald df Sig.
1 Intercept 1.516 .260 33.938 1 .000
2 Intercept 1.386 .264 27.674 1 .000
3 Intercept .773 .285 7.363 1 .007
4 Intercept .054 .329 .027 1 .869
a. The reference category is: 5.

Q24 Disclosure of Derivation from the Original Pattern:

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. In these tables the disclosure of derivation of fund from original pattern
has been analyzed considering it as the dependent variable and it has been judged through
scaling i.e. highly important, important, somewhat important, not very important and not
at all important as independent or categorical variables. From 1 to 5 representing all these
independent variables categories.
These tables viewing the results regarding the disclosure of net asset value of mutual fund.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that

234
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level.
Table 5.135 Case Processing Summary(Q24)
N Marginal
Percentage
1 92 40.0%
Disclosur 2 35 15.2%
e 3 61 26.5%
derivation4 7 3.0%
5 35 15.2%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.136 Model Fitting Information(Q24)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 19.640
Final 19.640 .000 0 .

Table 5.137 Likelihood Ratio Tests(Q24)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 116.992 97.352 4 .000

Table 5.138 Parameter Estimates(Q24)

Disclosure derivation B Std. Error Wald df Sig.
1 Intercept .966 .199 23.681 1 .000
2 Intercept .000 .239 .000 1 1.000
3 Intercept .556 .212 6.863 1 .009
4 Intercept -1.609 .414 15.110 1 .000
a. The reference category is: 5.

235
Q25 How did you come to know about Mutual Fund investment Scheme?
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors, in non-matric forms and the independent variables in matric or
categorical form. In these tables the source of investment in mutual fund scheme has been
analyzed considering it as the dependent variable and it has been judged through Mail,
Radio, TV, Broker and financial magazine as independent or categorical variables. From 1
to 5 representing all these independent variables categories.
These tables viewing the results regarding the source through which investors have
invested in mutual fund. The relationship has been shown between non matric dependent
variable and matric independent variables. The final value of chi-square is statistically
significant showing that there is relationship between dependent variable and independent
variables. The results suggest that there is no muilticolinarity issue as the standard error
value is less than 2.00.The likelihood ratio test suggest that there exist a relationship
between independent variables and dependent variable, as the chi-square value is
significant at 5% probability level.
Table 5.139Case processing summary(Q25)

N Marginal
Percentage
Informati 1 12 5.2%
on about2 7 3.0%
the 3 8 3.5%
investme 4 135 58.7%
nt 5 68 29.6%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

236
 Table 5.140 Model Fitting Information(Q25)
Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 17.610
Final 17.610 .000 0 .

Table 5.141 Likelihood Ratio Tests (Q25)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 274.864 257.254 4 .000

Table 5.142 Parameter estimates(Q25)

Information about B Std. Error Wald df Sig.
investment
1 Intercept -1.735 .313 30.690 1 .000
2 Intercept -2.274 .397 32.807 1 .000
3 Intercept -2.140 .374 32.782 1 .000
4 Intercept .686 .149 21.267 1 .000
a. The reference category is: 5.

Q26 Do you think Mutual Fund investing is a best alternative to easily investing?
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here the investment in mutual fund has been shown as best alternative to
easily investment as dependent variable. In these tables the mutual fund investment as
alternative to easily investing has been analyzed considering it as the dependent variable
and it has been judged through Yes, No and Do not know.. From 1 to 3 representing all
these independent variables categories.
These tables viewing the results regarding the easy alternative to investment of mutual
fund. The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that

237
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level. Most of the mutual fund investors considered mutual fund investment as a source of
easily investing.
Table 5.143 Case Processing Summary(Q26)
N Marginal
Percentage
Alternativ1 187 81.3%
e to easily2 35 15.2%
investme 8 3.5%
nt 3
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.144Model Fitting Information(Q26)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 9.129
Final 9.129 .000 0 .

Table 5.145 Likelihood Ratio Tests(Q26)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 251.555 242.425 2 .000
Table 5.146 Parameter Estimates(Q26)
Alternative to easily B Std. Error Wald df Sig.
investment
1 Intercept 3.152 .361 76.204 1 .000
2 Intercept 1.476 .392 14.184 1 .000
a. The reference category is: 3.

238
Q27 What kind of communication you like to know about your Fund performance?

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here communication with investors by mutual fund is the dependent
variable and automated response, personal visit, Telephone and no preferences. as
independent or categorical variables. 1 to 4 represents all these independent variables.
. The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level.
Table 5.147 Case Processing Summary(Q27)
N Marginal
Percentage
1 60 26.1%
Communicatio 2 20 8.7%
n preference 3 40 17.4%
4 110 47.8%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

239
Table 5.148 Model Fitting Information(Q27)
Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 15.571
Final 15.571 .000 0 .

Table 5.149 Likelihood Ratio Tests(Q27)

Effect Model Fitting Criteria Likelihood Ratio Tests
-2 Log Likelihood of Chi-Square df Sig.
Reduced Model
Intercept 92.117 76.546 3 .000

Table 5.150 Parameter Estimates(Q27)

Communication preference B Std. Error Wald df Sig.
-.606 .160 14.264 1 .000
1 Intercept
-1.705 .243 49.181 1 .000
2 Intercept
-1.012 .185 30.018 1 .000
3 Intercept

a. The reference category is: 4.

Q 28 Benefits Realization of Stock

The below tables show the output of multinomial logistic regression representing the
Benefits realized to be considered by the mutual fund investors, in non-matric forms and
the independent variables in matric or categorical form.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level.

240
 Table 5.151 Case Processing Summary (Q28)
N Marginal
Percentage
Benefits 1 182 79.1%
realizatio 2 31 13.5%
n of stock 3 17 7.4%
Valid 230 100.0%
Missing 0
Total 230
Subpopulation 1

Table 5.152 Model Fitting Information(Q28)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 9.724
Final 9.724 .000 0 .

Table 5.153 Likelihood Ratio Tests(Q28)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 217.064 207.340 2 .000

Table 5.154 Parameter Estimates(Q28)

Benefits B Std. Error Wald Df Sig.
realization of
stock
1Intercept 2.371 .254 87.389 1 .000
2Intercept .601 .302 3.963 1 .047
a. The reference category is: 3.

Q29 Mutual Fund investing gives a definite positive Return

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or

241
categorical form. In these tables the definite positive return of mutual fund investment is
the dependent variable and Yes, No and Do not know are the independent or categorical
variables. 1 to 3 represents all these independent variables. The relationship has been
shown between non matric dependent variable and matric independent variables. The final
value of chi-square is statistically significant showing that there is relationship between
dependent variable and independent variables. The results suggest that there is no
muilticolinarity issue as the standard error value is less than 2.00.The likelihood ratio test
suggest that there exist a relationship between independent variables and dependent
variable, as the chi-square value is significant at 5% probability level. Most of the investors
believed that mutual fund investment give definite positive return.
Table 5.155 Case Processing Summary(Q29)

N Marginal Percentage
1 180 78.6%
Definite
2 38 16.6%
positive return
3 11 4.8%
Valid 229 100.0%
Missing 1
Total 230
Subpopulation 1

Table 5.156 Model Fitting Information(Q29)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 9.490
Final 9.490 .000 0 .

Table 5.157 Likelihood Ratio Tests(Q29)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model

242
 Intercept 222.685 213.194 2 .000

Table 5.158 Parameter Estimates(Q29)

Definite positive return B Std. Error Wald Df Sig.
1 Intercept 2.795 .311 80.987 1 .000
2 Intercept 1.240 .342 13.110 1 .000
a. The reference category is: 3.

Q30 Mutual Fund Returns and Principals are fully protected and granted by MUFAP
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. In these tables the return and principals of funds being protected by
MUFAP is the dependent variable and Yes, No and Do not know are the independent or
categorical variables. 1 to 3 represents all these independent variables.
. The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level. Most of the investors believed that mutual fund investment is protected by MUFAP.

243
 Table 5.159 Case Processing Summary(Q30)
N Marginal
Percentage
Protectio 1 117 51.1%
n by2 102 44.5%
MUFAP 10 4.4%
to Fund3
return
Valid 229 100.0%
Missing 1
Total 230
Subpopulation 1
Table 5.160 Model Fitting Information(Q30)
Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 9.951
Final 9.951 .000 0 .

Table 5.161 Likelihood Ratio Tests(Q30)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 128.365 118.415 2 .000

Table 5.162 Parameter Estimates(Q30)

Protection by MUFAP B Std. Error Wald df Sig.
to Fund Return
1 Intercept 2.460 .329 55.732 1 .000
2 Intercept 2.322 .331 49.119 1 .000
a. The reference category is: 3.

Q31 Bank sponsored Mutual Fund give more Return than Bank Fixed Deposit
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. In these tables the mutual fund return to bank fixed deposit is the

244
dependent variable and Yes, No and Do not know are the independent or categorical
variables. 1 to 3 represents all these independent variables.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant hence there
is relationship between dependent variable and independent variables. The results suggest
that there is no muilticolinarity issue as the standard error value is less than 2.00.The
likelihood ratio test suggest that there exist a relationship between independent variables
and dependent variable, as the chi-square value is significant at 5% probability level. Most
of the investors believed that it gives more return than bank fixed deposit.

Table 5.163 Case Processing Summary(Q31)

N Marginal Percentage

135 59.0%
1
Bank sponsored
Mutual fund to fixed 65 28.4%
deposit 2

3 29 12.7%
229 100.0%
Valid
1
Missing
230
Total
1
Subpopulation

Table 5.164 Model Fitting Information(Q31)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 10.698
Final 10.698 .000 0 .

Table 5.165 Likelihood Ratio Tests(Q31)

245
 Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 87.615 76.917 2 .000

Table 5.166 Parameter Estimates(Q31)

Bank sponsored MF to B Std. Error Wald df Sig.
bank fixed deposit
1 Intercept 1.538 .205 56.466 1 .000
2 Intercept .807 .223 13.063 1 .000
a. The reference category is: 3.

Q32 Entry and Exit out of Mutual Fund is easy

The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. In these tables the entry and exit load is the dependent variable and Yes,
No and Do not know are the independent or categorical variables. 1 to 3 represents all these
independent variables.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level.. In the mutual fund investors survey most investors believed that entry and exit load
is easy.

246
 Table 5.167 Case Processing Summary(Q32)
N Marginal
Percentage
1 161 70.3%
Entry and
2 15 6.6%
exit load
3 53 23.1%
Valid 229 100.0%
Missing 1
Total 230
Subpopulation 1

Table 5.168 Model Fitting Information(Q32)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
Intercept Only 10.016
Final 10.016 .000 0 .

Table 5.169 Likelihood Ratio Tests(Q32)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 162.841 152.824 2 .000

Table 5.170 Parameter Estimates(Q32)

Entry and exit load B Std. Error Wald df Sig.
1 Intercept 1.111 .158 49.227 1 .000
2 Intercept -1.262 .292 18.627 1 .000
a. The reference category is: 3.

247
Q33 Due to Professional investment, a better Return is expected from Mutual Fund

The below tables show the output of multinomial logistic regression representing the relationship of mutual
fund investors investing behavior factors to be considered by the mutual fund investors in non-matric forms
and the independent variables in matric or categorical form. In these better return due to professional
investment is the dependent variable and Yes, No and Do not know are the independent or categorical
variables. 1 to 3 represents all these independent variables.
The relationship has been shown between non matric dependent variable and matric independent variables.
The final value of chi-square is statistically significant showing that there is relationship between dependent
variable and independent variables. The results suggest that there is no muilticolinarity issue as the standard
error value is less than 2.00.The likelihood ratio test suggest that there exist a relationship between
independent variables and dependent variable, as the chi-square value is significant at 5% probability level.
Most of the investor said That due to professional investment fund gives better return. The
results show that most of the investors

248
 Table 5.171 Case Processing Summary(Q33)
N Marginal Percentage

1 187 81.7%
Better return due to
professional 2 9 3.9%
investment
3 33 14.4%
Valid 229 100.0%

1
Missing

Total 230

1
Subpopulation

Table 5.172Model Fitting Information(Q33)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 9.191
Final 9.191 .000 0 .

Table 5.173 Likelihood Ratio Tests(Q33)

Effect Model Fitting Criteria Likelihood Ratio Tests
-2 Log Likelihood of Chi-Square df Sig.
Reduced Model
Intercept 250.465 241.274 2 .000

Table 5.174 Parameter Estimates(Q33)

Better returns due to B Std. Error Wald Df Sig.

professional
investment
1 Intercept 1.735 .189 84.398 1 .000
2 Intercept -1.299 .376 11.938 1 .001
a. The reference category is: 3.

249
Q34 Ups and down neither of Stock Market will nor affect the Return from Mutual
Fund.
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. In these tables the ups and down effect of stock market on mutual fund
return is the dependent variable and Yes, No and Do not know are the independent or
categorical variables. 1 to 3 represents all these independent variables.
. The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level.. In the mutual fund investors survey the investors were found indifferent.

Table 5.175 Case Processing Summary(Q34)

N Marginal
Percentage
Ups and 1 87 38.0%
down of 2 98 42.8%
stock 44 19.2%
market
and
mutual 3
fund
return
Valid 229 100.0%
Missing 1
Total 230
Subpopulation 1

Table 5.176 Model Fitting Information(Q34)

250
 Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 11.084
Final 11.084 .000 0 .

Table 5.177 Likelihood Ratio Tests(Q34)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
of Reduced Model
Intercept 34.334 23.250 2 .000

Table 5.178 Parameter Estimates(Q34)

Ups and down of stock B Std. Error Wald Df Sig.
market and mutual fund
return
1 Intercept .682 .185 13.580 1 .000
2 Intercept .801 .181 19.472 1 .000
a. The reference category is: 3.

Q35 MUFAP helps in protecting the Mutual Fund Industry and the unit holders.
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. In these tables the MUFAP protection to Mutual Fund Industry and unit
holders is the dependent variable and Yes, No and Do not know are the independent or
categorical variables. 1 to 3 represents all these independent variables.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than 2.00.The
likelihood ratio test suggest that there exist a relationship between independent variables
and dependent variable, as the chi-square value is significant at 5% probability level.. In the
mutual fund investors survey the investors were found indifferent.

251
 Table 5.179 Case Processing Summary(Q35)
N Marginal
Percentage
1 180 78.6%
MUFAP
2 10 4.4%
protection
3 39 17.0%
Valid 229 100.0%
Missing 1
Total 230
Subpopulation 1

Table 5.180 Model Fitting Information(Q35)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 9.422
Final 9.422 .000 0 .

Table 5.182 Parameter Estimates(Q35)

MUFAP protection B Std. Error Wald Df Sig.
1 Intercept 1.529 .177 74.978 1 .000
2 Intercept -1.361 .354 14.742 1 .000
a. The reference category is: 3.

Table 5.181Likelihood Ratio Tests(Q35)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 225.216 215.794 2 .000

Q36 How often you switch the Schemes of Mutual Fund in a year.
The below tables show the output of multinomial logistic regression representing the
relationship of mutual fund investors investing behavior factors to be considered by the
mutual fund investors in non-matric forms and the independent variables in matric or
categorical form. Here switching scheme has been taken the dependent variable and

252
different occurrence per year as independent or categorical variables. 1 to 3 represent the
different categories assigned to switching.
These tables show the switching of the mutual fund by investors investing in mutual fund.
The relationship has been shown between non matric dependent variable and matric
independent variables. The final value of chi-square is statistically significant showing that
there is relationship between dependent variable and independent variables. The results
suggest that there is no muilticolinarity issue as the standard error value is less than
2.00.The likelihood ratio test suggest that there exist a relationship between independent
variables and dependent variable, as the chi-square value is significant at 5% probability
level. It is found that most of the investors do not switch in a year.

Table 5.183 Case Processing Summary(Q36)

N Marginal
Percentage
1 135 59.0%
Switching
2 35 15.3%
in a year
3 59 25.8%
Valid 229 100.0%
Missing 1
Total 230
Subpopulation 1

Table 5.184 Model Fitting Information(Q36)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 10.788
Final 10.788 .000 0 .

Table 5.185 Likelihood Ratio Tests(Q36)

Effect Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 79.756 68.968 2 .000

253
 Table 5.186 Parameter Estimates(Q36)
Switching in a year B Std. Error Wald df Sig.

.828 .156 28.130 1 .000

1 Intercept
-.522 .213 5.990 1 .014
2 Intercept

a. The reference category is: 3.

Q37 After investment, how frequently you monitor the Performance of Mutual Fund.
The below tables show the output of multinomial logistic regression representing the relationship of mutual
fund investors investing behavior factors to be considered by the mutual fund investors in non-matric forms
and the independent variables in matric or categorical form. Here monitoring mutual fund performance is the
dependent variable and weekly, monthly, once and rarely as as independent or categorical variables. 1
representing weekly, monthly, once and rarely. These tables show the performance monitoring of
mutual fund by investors investing in mutual fund. The relationship has been shown
between non matric dependent variable and matric independent variables. The final value
of chi-square is statistically significant showing that there is relationship between
dependent variable and independent variables. The results suggest that there is no
muilticolinarity issue as the standard error value is less than 2.00.The likelihood ratio test
suggest that there exist a relationship between independent variables and dependent
variable, as the chi-square value is significant at 5% probability level.

254
 Table 5.187 Case processing Summary (Q37)
N Marginal
Percentage
1 50 21.8%
Monitoring the2 71 31.0%
mutual fund
performance 3 95 41.5%
4 13 5.7%
Valid 229 100.0%
Missing 1
Total 230
Subpopulation 1

Table 5.188Model Fitting Information(Q37)

Model Model Fitting Likelihood Ratio Tests
Criteria
-2 Log Likelihood Chi-Square Df Sig.
Intercept Only 15.393
Final 15.393 .000 0 .

Table 5.189 Likelihood Ratio Tests(Q37)

Effect Model Fitting Likelihood Ratio Tests

Criteria
-2 Log Likelihood Chi-Square df Sig.
of Reduced Model
Intercept 90.099 74.706 3 .000

Table 5.190Parameter Estimates(Q37)

Monitoring the B Std. Error Wald Df Sig.
mutual fund
performance
1 Intercept 1.347 .311 18.722 1 .000
2 Intercept 1.698 .302 31.671 1 .000
3 Intercept 1.989 .296 45.236 1 .000
a. The reference category is: 4.

5.3 Reliability Analysis

255
Table 5.191 Reliability Statistics
Cronbach's Alpha N of Items

.720 12

The above results of the reliability show that alpha value of the variable job Performance
is 0.720 which is within an acceptable range Total sevens items analyzed here for the
reliability purposes. As many researchers believe that the components or items of a variable
should have combined alpha value of 0.60 the current scale value of the variable suggest
that the scale is reliable and can be used for this research and its analysis will be reliable
being based on the reliable value.

5.4 Factor Analysis and Validity

Table 1.92 KMO and Bartlett's Fund characteristics

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. .801

Approx. Chi-Square 80.665

Bartlett's Test of Sphericity
Df 10
Sig. .000

Table 5.193 component matrix Fund characteristics

Component
1
Q 13 .587

256
 Q 14 .679

Q 14 .789

Q 15 .656

Q 16 .587

Q 17 .567

Q 18 .577

Q 19 .587

Q 20 .523

Q 21 .543

Q 22 .574

Q 23 .601

Q 24 .557

The above table shows the facto analysis of the fund characteristics perceived by the
Mutual Fund investors. As the KMO value is above .8, so the results indicate that the
sample is good enough for the analysis of this study. The recurring value is .801 suggesting
the adequacy of the sample of this research. All the factors of this variable loaded at a well
desired value i.e .5 or above. Which means there is enough interrelation among all
questions?

257
5.5 Research Findings
The Research findings are grouped under two headings i.e. Performance analysis and
investor investing behavior in mutual funds as follow.
5.5.1 Performance Analysis
1. The results found that all three models are capable to predict and explain the mutual fund
performance in Pakistan. In order to identify the best model which predicts the mutual
funds return, better amongst the three, GRS test was applied.. The GRS test showed the
following mean absolute alphas.
a. CAPM reported 0.05 mean absolute alpha
b. Fama French 3-Factor reported 0.088 mean absolute alpha
c. Carhart 4-Factor showed 0.086.

On the bases of mean absolute alphas, the CAPM ranks above than the other, generated
through GRS. The mean absolute alpha of CAPM is closer to zero, thereby documenting
that CAPM is the preferred model for the performance analysis and evaluation of mutual
funds in Pakistan.
2. The results found that Fund expense ratio has positive impact on the Fund adjusted
return. It documents that as fund expense ratio increases for a particular mutual fund it
corresponds increase in funds return. The penal regression table 5.7 illustrates this positive
impact of expense ratio on fund’s performance.
3. The results also indicated that Fund cash flow has positively significant effect on the
Fund’s adjusted return. It documents that as fund show greater amount of cash flows over
the period of time, it generate maximum return.
4. The research analysis revealed that both Fund age and Fund size have positive significant
impact on Fund’s adjusted Return. Table 5.7 illustrated the positive results; document a
positive impact of fund size and fund age on fund performance. The findings suggest that
the fund managers need to focus on the same old funds rather than launching new funds
again and again, as with passing age funds’ performance with time and size. The fund size

258
has been found positively effecting, the fund returns, documenting that as the mangers
succeed to encourage more investment in its fund, it will help in maximizing the funds
return.
5. The results found that both Fund’s Family and liquidity have negative significant impact
on the adjusted Return of mutual Funds. This negative effect of both factors is described in
table 5.7 chapter 5. Since funds family showed negative impact on funds adjusted return,
therefore the assets and fund’s managers must work out to ensure to keep the funds family
limited because of its negative impact on fund’s adjusted return. Liquidity has also shown
negative impact on fund’s adjusted returns, meaning that as funds hold more cash for the
redemption of units or other strategic reasons, it negatively affects the fund’s performance.
6. The analysis reveals persistence in mutual Fund performance of different portfolios
sorted upon their performance. The significance of lagged alpha indicated that past
performers also do well at present.

5.5.2 Investors investing behavior in Mutual fund

1. The investor’s responses indicates following demographics:
a. Most of the investors are male.
b. Most of the investors are above 40 years of age.
c. Most of the investors are educated having graduation and above graduation.
d. Investors belong to all occupations.
e. Investors range from all income groups.
f. Most of the investors are married.
2. Most of the Mutual Fund investor’s preferred, investing in Mutual Funds different
Categories equally as Choice and most of the investors liked investment in open
ended Mutual Fund.
3. Most of the mutual fund Investors liked the safety and professional management of
mutual Fund and most of them invest through agents, family member and own self.

259
4. Most of the investors thought fund record, reputation, scheme, portfolio of
investment, reputation of the manager, withdraw facility, favorable rating,
innovativeness of the scheme, entry and exit load and minimum initial investment
as highly important and important.
5. Disclosure of net asset value was liked by most of the investors during the
investor’s survey, investing in Mutual Fund.
6. Most of the investors surveyed told, they are known about mutual fund through
brokers and financial magazines mainly.
7. Most of the mutual fund investors thought investing in mutual Fund as source of an
easy investment and the investors in mutual funds liked different means of
communication.
8. Most of the investors agreed with the fact that investment in mutual fund realizes
the benefits of stock market.
9. Most of the investors agreed with the fact that mutual fund give a definite positive
return and Funds return are protected by MUFAP.
10. Most of the investors in Mutual Funds liked bank sponsored Mutual Funds, liked
investing in mutual fund as entry and exit out is easy and argued that due to
professional investment a better returns are expected.
11. Different responses were obtained about the ups and down in stock market effects
on Mutual Fund Return. They agreed that stock market variation do affect mutual
funds return.
12. Most of the investors in mutual funds were found, do not want switching mutual
Funds most often and investors were found very indifferent regarding monitoring
the performance of mutual fund after different intervals.

260
 CHAPTER 6

CONCLUSION, RECOMMENDATIONS AND SUGGESSIONS

FOR FUTURE RESEARCH

6.1 Conclusion
Mutual Fund is an investment choice for small investors across the world. Mutual Fund
provides an opportunity to the small investors who have not enough skills, information and
knowledge of investing in capital market. This study was aimed at knowing about the Pakistan
Mutual Fund industry and the investors investing behavior in mutual fund. Mutual Fund
performance of Pakistan industry was first ascertained with asset pricing models. As the single
and multi-factor model show different results across the world. Some researchers believe that
single factor beta is the most favorable and viable risk Factor, which determine the Returns.
However many researchers believe that multi factor asset pricing models better explain the
risk adjusted Return. This research study was focused on CAPM, Fama French 3-factor model
and Carhart 4-Factor model. These models were applied to Mutual fund data in Pakistan from
2009 to 2014. The open ended mutual Fund data of 153 were collected on monthly basis and
created sorted portfolios and ranked all portfolios on the bases of their returns. The results
revealed that CAPM better perform as most of its portfolios co-efficient were found significant
as compared to Fama-French 3- factor and Carhart 4-factor model. The GRS test was applied
to find out which model among the three competing models better explain and predict mutual
fund performance. The GRS test results documented that CAPM better explain the mutual
fund performance in Pakistan as compare to Fama French-3 Factor and Carhart 4-Factor
Model. After analyzing the Mutual Fund performance through the three assets pricing models,
annual data of Fund Characteristics i-e Expense ratio, Fund age, Fund Family, Cash flow, Fund
Size, and liquidity was analyzed for the period 2010 to 2014 through regression on funds
adjusted return. The results found that expense ratio has positive impact on Fund return as
having positive co-efficient meaning that as the Fund increases its expenses, it definitely

261
 enhances the Fund Return. The research study found Fund age having positive significant
impact on Fund adjusted Return, documenting that as fund age increases, then Fund show
better performance. The fund family has shown negative co-efficient, means that as the Fund
Family increases in Pakistan Mutual fund industry, it affect negatively the Fund’s
performance. The results revealed that Fund Cash flow has positive significant impact on
Fund’s return. The study also found that Fund Size has positive significant impact on Fund
performance. Liquidity was found having negative significant impact on Fund Return,
showing that as the fund increases its liquidity and holding cash in Pakistan Mutual Fund
industry, it affects their return negatively. The analysis revealed that there is persistence in
mutual fund performance, as the lagged CAPM alpha is significant showing that past
performer better perform in present as well. The overall conclusion of the performance
analysis shows that CAPM do well in explaining the mutual fund performance in Pakistan
Mutual Fund industry as compared to the other two assets pricing model i.e. Fama French 3-
factor and Carhart 4-factor model. The investor’s behavior in mutual fund was analyzed
through descriptive and multinomial logistic regression. The responses of the mutual fund
investors were shown in the form of table, bars, pies, and curves. Total 37 questions asked
from the mutual fund investors and analyzed through descriptive and multinomial logistic
Regression. The investor’s behavior of mutual fund results showed that most of the mutual
fund investors still like to stay with mutual fund investment and like the prevailing
characteristics of funds due to significant value predicted by multinomial regression.

6.2 Recommendations
Research study makes following recommendations.
1. The Asset Pricing Models applied on the data of Pakistani open ended mutual fund,
declared that most of the portfolios of high return do not have significant association with
the market variations, therefore it is suggested that the fund’s managers should properly
analyze the market variations, so that to ensure the outperforming practices by the mutual
funds over the market return. They can do so by closely and continuously monitoring the

262
 market variations. Better understanding of the market variations by the managers will
improve funds’ performance.
2. The results indicate a negative impact of funds family and liquidity on funds adjusted
return, therefore it is suggested that the fund’s manager should avoid increasing the number
of funds and should mature the existing fund’s portfolio. . As the fund family tends to
increase in size than funds managers diversify their concentration and yield poor results in
term of performance. The managers should stay with the same old funds, rather than
launching new funds. The liquidity shows a negative impact on funds’ performance,
keeping the negative impact, it is suggested that these open ended mutual funds should
minimize its cash holding for the redemption of its units of the investors or other strategic
reasons. As excessive cash holding impact the funds return negatively. The managers
should invest the excessive cash in different portfolios to generate return.
3. The results indicate a positive impact of funds age and size on funds adjusted return.
Keeping in view the positive impact, it is suggested that the managers of these funds should
try not to liquidate or stop investing in any of its fund and should continuously try to
motivate and encourage investment in funds with a view increase its size in terms of assets.
4. The results also revealed a positive impact of fund expense ratio on fund adjusted return.
It is therefore suggested that, if the fund’s managers increase their expenses, they should
ensure increase in the funds return. It is suggested for the mutual funds investors that they
should not hesitate from those funds charging high expenses for their professional
management.
5. The results indicate a positive impact of cash flow on fund performance. It is suggested
that fund managers should encourage positive cash flow to motivate investors and enhance
funds return.

6.3 Suggestions for Future Research

263
This study cover the open ended mutual fund data, so both open and close ended mutual fund,
if tested will give more clear and improved results of Pakistan mutual fund industry.Besides
the tested characteristics of mutual funds in this research some other characteristics like
management fee, front load, back load and service fee can also be tested instead of the
composite expense ratio, knowing its impact on funds’ performance.

264
REFERENCE
Ahmed, P., & Nanda, S. (2005). Performance of enhanced index and quantitative equity
funds. Financial Review, 40(4), 459-479.
Afza, T., &Rauf, A. (2009).Performance evaluation of Pakistani mutual funds. Pakistan
economic and social review, 199-214.
Agudo, L. F., &Magallón, M. V. (2005). Empirical evidence of performance persistence in
a relatively unexplored market: the case of Spanish investment funds. Applied
Financial Economics Letters, 1(2), 85-88.
Ahire, S. L., Golhar, D. Y., & Waller, M. A. (1996). Development and validation of TQM
implementation constructs. Decision sciences, 27(1), 23-56.
Alexander, G. J., Jones, J. D., &Nigro, P. J. (1998). Mutual fund shareholders:
Characteristics, investor knowledge, and sources of information. Financial
Services Review, 7(4), 301-316.
Alexakis, P., &Tsolas, I. E. (2011).Appraisal of mutual equity fund performance using data
envelopment analysis. Multinational Finance Journal, 15(3/4), 273-296.
Ansari, M. L., & Moid,(2013). Factors affecting investment behavior among young
professionals.
Anderson, S. C. (2006). Investment Costs and the Mismanagement Issue. Investment
Management and Mismanagement: History, Findings, and Analysis, 85-97.
Allen, D. E., & Tan, M. L. (1999). A Test of the Persistence in the Performance of UK
Managed Funds. Journal of Business Finance & Accounting, 26(6), 559-593.
Andersen, L. C., & Carlson, K. M. (1970). A monetarist model for economic stabilization.
Federal Reserve Bank of St. Louis.
Avramov, D., & Wermers, R. (2006).Investing in mutual funds when returns are
predictable. Journal of Financial Economics, 81(2), 339-377.
Barber, B. M., Odean, T., & Zheng, L. (2003).Out of sight, out of mind: The effects of
expenses on mutual fund flows. Out of Mind: The Effects of Expenses on Mutual

265
 Fund Flows.
Bauer, R., Koedijk, K., & Otten, R. (2005).International evidence on ethical mutual fund
performance and investment style. Journal of Banking & Finance, 29(7), 1751-
1767.
Barua, A., Kriebel, C. H., & Mukhopadhyay, T. (1991).An economic analysis of strategic
information technology investments. MIS quarterly, 313-331.
Bekaert, G., Harvey, C. R., & Lumsdaine, R. L. (2002). Dating the integration of world
equity markets. Journal of Financial Economics, 65(2), 203-247.
Benson, K. L., & Faff, R. W. (2006). Conditional performance evaluation and the relevance
of money flows for Australian international equity funds. Pacific-Basin Finance
Journal, 14(3), 231-249.
Bollen, N. P. (2007). Mutual fund attributes and investor behavior. Journal of Financial
and Quantitative Analysis, 42(03), 683-708.
Brown, K. C., & Harlow, W. V. (2002).Staying the course: The impact of investment style
consistency on mutual fund performance.
Brown, S. J., & Goetzmann, W. N. (1997).Mutual fund styles. Journal of financial
Economics, 43(3), 373-399.
Barberis, N., & Shleifer, A. (2003).Style investing. Journal of financial Economics, 68(2),
161-199.
Bergstresser, D., Chalmers, J. M., & Tufano, P. (2009).Assessing the costs and benefits of
brokers in the mutual fund industry. Review of Financial Studies.
Berk, J. B., & Green, R. C. (2002). Mutual fund flows and performance in rational markets.
National Bureau of Economic Research.
Berk, J., & Xu, J. (2004).Persistence and fund flows of the worst performing mutual
funds. Unpublished Working Paper, University of California, Berkeley.
Bernhardt, D., & Davies, R. J. (2005).Painting the tape: Aggregate evidence. Economics
Letters, 89(3), 306-311.
Barber, B. M., Odean, T., & Zheng, L. (2000). The behavior of mutual fund investors.

266
 Unpublished working paper.
Bhuvaneswari, C. (2014). A Study on Investor’s Perception towards Equity/Tax Saving
Mutual Funds. Journal of Applied Research,(ISSN 2321-4090). Peer Reviewed
International Journal Vol. No. II Issue, (12).
Blake, D., & Timmermann, A. (1998). Mutual fund performance: evidence from the
UK. European Finance Review, 2(1), 57-77.
Bliss, R. T., & Potter, M. E. (2002). Mutual fund managers: does gender matter?. The
Journal of Business and Economic Studies, 8(1), 1.
Bilson, C., Frino, A., & Heaney, R. (2005). Australian retail fund performance
persistence. Accounting & Finance, 45(1), 25-42.
Bogle, J. C. (1992).Selecting equity mutual funds. The Journal of Portfolio
Management, 18(2), 94-100.
Blake, D., Lehmann, B. N., &Timmermann, A. (1999).Asset al.location dynamics and
pension fund performance. The Journal of Business, 72(4), 429-461.
Bollen, N. P. (2007). Mutual fund attributes and investor behavior. Journal of Financial
and Quantitative Analysis, 42(03), 683-708.
Bollen, N. P., & Busse, J. A. (2001).On the timing ability of mutual fund managers. The
Journal of Finance, 56(3), 1075-1094.
Bollen, N. P., &Busse, J. A. (2005).Short-term persistence in mutual fund
performance. Review of Financial Studies, 18(2), 569-597.
Bordini, R. H., Hübner, J. F., & Wooldridge, M. (2007). Programming multi-agent systems
in Agent Speak using Jason. John Wiley & Sons, (Vol. 8).
Bossaerts, P., Ghirardato, P., Guarnaschelli, S., & Zame, W. (2010). Prices and allocations
in asset markets with heterogeneous attitudes towards ambiguity. Review of
Financial Studies, 23(4), 1325-1359.
Borensztein, E., & Gelos, G. (2001). A panic-prone pack? The behavior of emerging market
mutual funds.
Carlson, R. S. (1970). Aggregate performance of mutual funds, 1948–1967.Journal of

267
 Financial and Quantitative Analysis, 5(01), 1-32.
Cashman, G. D., Deli, D. N., Nardari, F., & Villupuram, S. V. (2006, September).On
monthly mutual fund flows. In FMA Annual Meeting, Orlando, FL. Retrieved
from (Vol. 69, No. 2.130).
Claessens, S., Dasgupta, S., & Glen, J. (1995). Return behavior in emerging stock
markets. The World Bank Economic Review, 9(1), 131-151.
Capon, N., Fitzsimons, G. J., & Prince, R. A. (1996).An individual level analysis of the
mutual fund investment decision. Journal of financial services research, 10(1),
59-82.
Carhart, M. M. (1997). On persistence in mutual fund performance. The Journal of
finance, 52(1), 57-82.
Chordia, T. (1996).The structure of mutual fund charges. Journal of financial
Economics, 41(1), 3-39.
Chang, E. C., & Lewellen, W. G. (1984). Market timing and mutual fund investment
performance. Journal of Business, 57-72.
Cheong Tang, T. (2007). Money demand function for Southeast Asian countries: an
empirical view from expenditure components. Journal of Economic
Studies, 34(6), 476-496.
Chalmers, J., Edelen, R. M., & Kadlec, G. B. (1999).An analysis of mutual fund trading
costs. Available at SSRN 195849.
Chance, D. M., & Ferris, S. P. (1991). Mutual fund distribution fees: An empirical analysis
of the impact of deregulation. Journal of Financial Services Research, 5(1), 25-
42.
Chevalier, J., & Ellison, G. (1998). Career concerns of mutual fund managers. National
Bureau of Economic Research.
Cheema, M., & Shah, S. (2006). Corporate Governance in Developing Economies: The
Role of Mutual Funds in Corporate Governance in Pakistan.Hong Kong LJ, 36,
341.

268
Cook, W. D., & Hebner, K. J. (1992). A multicriteria approach to mutual fund
selection. Financial Services Review, 2(1), 1-20.
Cuthbertson, K., Nitzsche, D., & O'Sullivan, N. (2008). UK mutual fund performance:
Skill or luck?. Journal of Empirical Finance, 15(4), 613-634.
Cutler, D. M., & Gruber, J. (1996).The effect of Medicaid expansions on public insurance,
private insurance, and redistribution. The American Economic Review, 86(2),
378-383..
Chordia, T. (1996).The structure of mutual fund charges. Journal of financial
Economics, 41(1), 3-39.
Daniel, K., Grinblatt, M., Titman, S., & Wermers, R. (1997).Measuring mutual fund
performance with characteristic based benchmarks. The Journal of
finance, 52(3), 1035-1058.
Dimson, E., & Minio-Paluello, C. (2002). The closed-end fund discount.
Duca, J. V. (2005). Why have US households increasingly relied on mutual funds to own
equity?. Review of income and wealth, 51(3), 375-396.
Dasgupta, S., Laplante, B., & Mamingi, N. (2001).Pollution and capital markets in
developing countries. Journal of Environmental Economics and
management, 42(3), 310-335.
DiBartolomeo, D., & Witkowski, E. (1997). Mutual fund misclassification: Evidence based
on style analysis. Financial Analysts Journal, 53(5), 32-43.
Del Guercio, D., & Tkac, P. A. (2002). The determinants of the flow of funds of managed
portfolios: Mutual funds vs. pension funds. Journal of Financial and Quantitative
Analysis, 37(04), 523-557.
Droms, W. G., & Walker, D. A. (1994). Investment performance of international mutual
funds. Journal of Financial Research, 17(1), 1-14.
Dahlquist, M., Engström, S., & Söderlind, P. (2000). Performance and characteristics of
Swedish mutual funds. Journal of Financial and quantitative Analysis, 35(03),
409-423.

269
Droms, W. G., & Walker, D. A. (1996). Mutual fund investment performance. The
Quarterly Review of Economics and Finance, 36(3), 347-363.
Dowen, R. J., & Mann, T. (2007). Do Investors Benefit from 12b-1 Fees? American
Journal of Business, 22(1), 21-30.
Detzler, M. L. (1999). The performance of global bond mutual funds. Journal of Banking
& Finance, 23(8), 1195-1217.
Drew, M. E., & Veeraraghavan, M. (2002). A closer look at the size and value premium in
emerging markets: Evidence from the Kuala Lumpur Stock Exchange. Asian
Economic Journal, 16(4), 337-351.
Edwards, F. R., & Zhang, X. (1998). Mutual funds and stock and bond market stability (pp.
75-100). Springer US.
Elton, E. J., Gruber, M. J., & Blake, C. R. (2003).Incentive fees and mutual funds. The
Journal of Finance, 58(2), 779-804.
Edelen, R. M. (1999). Investor flows and the assessed performance of open-end mutual
funds. Journal of Financial Economics, 53(3), 439-466.
Elton, E. J., Gruber, M. J., & Blake, C. R. (1996).Survivor bias and mutual fund
performance. Review of Financial Studies, 9(4), 1097-1120.
Fama, E. F., & French, K. R. (1992). The cross section of expected stock returns. The
Journal of Finance, 47(2), 427-465.
Fama, E. F., & French, K. R. (1998). Taxes, financing decisions, and firm value. The
Journal of Finance, 53(3), 819-843.
Ferson, W. E., & Schadt, R. W. (1996). Measuring fund strategy and performance in
changing economic conditions. The Journal of Finance,51(2), 425-461.
Fama, E. F. (1998). Market efficiency, long-term returns, and behavioral finance. Journal
of financial economics, 49(3), 283-306.
Ferson, W. E., & Harvey, C. R. (1994).Sources of risk and expected returns in global equity
markets. Journal of Banking & Finance, 18(4), 775-803.
Ferson, W. E., &Schadt, R. W. (1996).Measuring fund strategy and performance in

270
 changing economic conditions. The Journal of Finance,51(2), 425-461.
Ferreira, M. A., Keswani, A., Miguel, A. F., & Ramos, S. B. (2012). The determinants of
mutual fund performance: A cross-country study. Review of Finance, rfs013.
Ferson, W. E., &Warther, V. A. (1996).Evaluating fund performance in a dynamic
market. Financial Analysts Journal, 52(6), 20-28.
Froot, K. A., &Ramadorai, T. (2008). Institutional portfolio flows and international
investments. Review of Financial Studies, 21(2), 937-971.
Fletcher, J., & Forbes, D. (2002).An exploration of the persistence of UK unit trust
performance. Journal of Empirical Finance, 9(5), 475-493.
Gaspar, J. M., Massa, M., & Matos, P. (2006).Favoritism in mutual fund families? Evidence
on strategic cross‐fund subsidization. The Journal of Finance, 61(1), 73-104.
Gilbazo,Javier, & ruizverdú,. pablo (2009). The relation between price and performance in
the mutual fund industry. The Journal of Finance, 64(5), 2153-2183.
Green, R. K., &Wachter, S. M. (2007).The housing finance revolution. University of Penn,
Inst for Law & Econ Research Paper, (09-37).
Golec, J. H. (1996). The effects of mutual fund managers' characteristics on their portfolio
performance, risk and fees. Financial Services Review, 5(2), 133-147.
Grinblatt, M., & Titman, S. (1989). Mutual fund performance: An analysis of quarterly
portfolio holdings. Journal of business, 393-416.
Gharghori, P., Mudumba, S., & Veeraraghavan, M. (2007). How smart is money? An
investigation into investor behaviour in the Australian managed fund
industry. Pacific-Basin Finance Journal, 15(5), 494-513.
Gaspar, J. M., Massa, M., & Matos, P. (2006). Favoritism in mutual fund families?
Evidence on strategic cross fund subsidization. The Journal of Finance, 61(1),
73-104.
Grinblatt, M., & Titman, S. (1994). A study of monthly mutual fund returns and
performance evaluation techniques. Journal of Financial and Quantitative
Analysis, 29(03), 419-444.

271
Goetzmann, W. N., & Ibbotson, R. G. (1994). Do winners repeat?. The Journal of Portfolio
Management, 20(2), 9-18.
Grinblatt, M., & Titman, S. (1992).The persistence of mutual fund performance. The
Journal of Finance, 47(5), 1977-1984.
Gruber, M. J. (1996). Another puzzle: The growth in actively managed mutual funds. The
Journal of Finance, 51(3), 783-810.
Grinblatt, M., & Titman, S. (1993). Performance measurement without benchmarks: An
examination of mutual fund returns. Journal of Business, 47-68.
Goriaev, A. P., Nijman, T. E., &Werker, B. J. (2002). The dynamics of the impact of past
performance on mutual fund flows. Tilburg University.
Grinblatt, M., & Titman, S. (1989). Mutual fund performance: An analysis of quarterly
portfolio holdings. Journal of Business, 393-416.
Guedj, I., &Papastaikoudi, J. (2003, October). Can mutual fund families affect the
performance of their funds?. In EFMA 2004 Basel Meetings Paper.
Harless, D. W., & Peterson, S. P. (1998).Investor behavior and the persistence of poorly-
performing mutual funds. Journal of Economic Behavior & Organization, 37(3),
257-276.
Hamilton, S., Jo, H., & Statman, M. (1993). Doing well while doing good? The investment
performance of socially responsible mutual funds. Financial Analysts
Journal, 49(6), 62-66.
Heffernan, S. (2001). All UK investment trusts are not the same.
Hendricks, D., Patel, J., &Zeckhauser, R. (1993). Hot hands in mutual funds: Short run
persistence of relative performance, 1974–1988. The Journal of Finance, 48(1),
93-130.
Howard, J. A., &Sheth, J. N. (1969). The theory of buyer behavior New York: Wiley. (Vol.
14).
Hwang, B. H., & Kim, S. (2009). It pays to have friends. Journal of Financial
Economics, 93(1), 138-158.

272
Henriksson, R. D. (1984). Market timing and mutual fund performance: An empirical
investigation. Journal of Business, 73-96.
Huij, J., & Verbeek, M. (2007).Cross-sectional learning and short-run persistence in mutual
fund performance. Journal of Banking & Finance,31(3), 973-997.
Ippolito, R. A. (1989). Efficiency with costly information: A study of mutual fund
performance, 1965-1984. The Quarterly Journal of Economics, 1-23.
Jain, P. C., & Wu, J. S. (2000). Truth in mutual fund advertising: Evidence on future
performance and fund flows. The Journal of Finance, 55(2), 937-958.
Jambodekar, M. V. (1996). Marketing Strategies of Mutual Funds–Current Practices and
Future Directions. UTI–IIMB Centre for Capital Markets Education and
Research, Banglore.
Jegadeesh, N., & Titman, S. (1993). Returns to buying winners and selling losers:
Implications for stock market efficiency. The Journal of Finance,48(1), 65-91.
Jayadev, M. (1996). Mutual fund performance: an analysis of monthly returns. Finance
India, 10(1), 73-84.
Jensen, M. C. (1968). The performance of mutual funds in the period 1945–1964. The
Journal of finance, 23(2), 389-416.
Jones, M. A., Lesseig, V. P., & Smythe, T. I. (2005). Financial advisors and mutual fund
selection. Journal of Financial Planning, 18(3), 64-70.
Kahn, R. N., & Rudd, A. (1995). Does historical performance predict future
performance?. Financial Analysts Journal, 51(6), 43-52.
Kon, S. J. (1983). The market-timing performance of mutual fund managers. Journal of
Business, 323-347.
Kon, S. J., & Jen, F. C. (1979). The investment performance of mutual funds: An empirical
investigation of timing, selectivity, and market efficiency. Journal of Business,
263-289.
Khorana, A., & Servaes, H. (1999).The determinants of mutual fund starts. Review of
Financial Studies, 12(5), 1043-1074.

273
Khorana, A., & Servaes, H. (1999). The determinants of mutual fund starts. Review of
Financial Studies, 12(5), 1043-1074.
Kumar, S. P., & Rajkumar, S. Awareness and Knowledge of Mutual Fund among the
Investors with Special Reference to Chennai–A Critical Study. Education, 31(30),
28.
Lakonishok, J., Shleifer, A., &Vishny, R. W. (1992).The impact of institutional trading on
stock prices. Journal of Financial Economics, 32(1), 23-43.
Lee, C. F., & Rahman, S. (1990). Market timing, selectivity, and mutual fund performance:
An empirical investigation. Journal of Business, 261-278.
Lehmann, B. N., & Modest, D. M. (1987). Mutual fund performance evaluation: A
comparison of benchmarks and benchmark comparisons. The journal of
Finance, 42(2), 233-265.
Livingston, M., & O'Neal, E. S. (1998).The cost of mutual fund distribution fees. Journal
of Financial Research, 21(2), 205-218.
Li, C. W., Tiwari, A., & Tong, L. (2015). Investment Decisions under Ambiguity: Evidence
from Mutual Fund Investor Behavior. In AFA 2014 Philadelphia Meetings.
Ippolito, R. A. (1989). Efficiency with costly information: A study of mutual fund
performance, 1965-1984. The Quarterly Journal of Economics, 1-23.
Matallín‐Sáez, J. C. (2006). Seasonality, market timing and performance amongst
benchmarks and mutual fund evaluation. Journal of Business Finance &
Accounting, 33(9/10), 1484-1507.
Malhotra, D. K., & McLeod, R. W. (1997). An empirical analysis of mutual fund
expenses. Journal of Financial Research, 20(2), 175-190.
Malkiel, B. G. (1995). Returns from investing in equity mutual funds 1971 to 1991. The
Journal of Finance, 50(2), 549-572.
Malkiel, B. G., & Fama, E. F. (1970). Efficient capital markets: A review of theory and
empirical work. The Journal of Finance, 25(2), 383-417.
Markowitz, H. M. (1989). Mean—Variance Analysis (pp. 194-198). Palgrave Macmillan

274
 UK.
Massa, M. (2003). How do family strategies affect fund performance? When performance-
maximization is not the only game in town. Journal of Financial
Economics, 67(2), 249-304.
Morey, M. R. (2003). Kiss of Death: A 5-Star Morningstar Mutual Fund Rating?. Journal
of Financial Economics, 67(2), 123-244.
Miller, T. W., &Gressis, N. (1980). Nonstationarity and evaluation of mutual fund
performance. Journal of Financial and Quantitative Analysis, 15(03), 639-654.
Mahmud, M., &Mirza, N. (2011). An Evaluation of Mutual Fund Performance in an
emerging economy: The case of Pakistan. The Lahore Journal of Economics, 16,
301.
Murthi, B. P. S., Choi, Y. K., & Desai, P. (1997). Efficiency of mutual funds and portfolio
performance measurement: A non-parametric approach.European Journal of
Operational Research, 98(2), 408-418.
Nazir, M. S., & Nawaz, M. M. (2010).The determinants of mutual fund growth in
Pakistan. International Research Journal of Finance and Economics, 54(2010),
75-84.
Nanda, V., Wang, Z. J., & Zheng, L. (2004). Family values and the star phenomenon:
Strategies of mutual fund families. Review of Financial Studies.

Nafees, B., Ahmad, S., & Khan, Z. A.(2012. Risk Analysis of Open And Closed End
Mutual Funds.
Odean, T. (1998). Do investors trade too much?. Available at SSRN 94143.Odean, T.
(1998). Do investors trade too much?. Available at SSRN 94143.
Otten, R., &Bams, D. (2004). How to measure mutual fund performance: economic versus
statistical relevance. Accounting & finance, 44(2), 203-222.
Otten, R., &Bams, D. (2002).European mutual fund performance. European Financial
Management, 8(1), 75-101.

275
Osman, R. (2014). Mutual Funds: Determinants of Selection and Performance
Measurement: The Case of Egyptian Mutual Fund (Doctoral dissertation, Arab
Academy for Science and Technology).
Pástor, Ľ., &Stambaugh, R. F. (2002).Mutual fund performance and seemingly unrelated
assets. Journal of Financial Economics, 63(3), 315-349.
Plantinga, A., & de Groot, S. (2001). Risk-adjusted performance measures and implied
risk-attitudes.University of Groningen.
Pushner, G., Rainish, R., & Coogan, D. (2001). Performance of European focused mutual
funds. American Business Review, 19(1), 39.
Panda, T. K., & Tripathy, N. P. (2002). An application of multidimensional scaling model
towards brand positioning of mutual funds: A case study of tax saving scheme.
Finance India, 16(3), 991.
Quigley, G., &Sinquefield, R. A. (2000).Performance of UK equity unit trusts. Journal of
Asset Management, 1(1), 72-92.
Rajeswari, T. R., &Ramamoorthy, V. E. (2001). An Empirical Study on Factors Influencing
the Mutual Fund/Scheme Selection by S mall Investors.
Rouwenhorst, K. G. (1999). Local return factors and turnover in emerging stock
markets. The Journal of Finance, 54(4), 1439-1464.
Saraoglu, H., &Detzler, M. L. (2002).A sensible mutual fund selection model. Financial
Analysts Journal, 58(3), 60-72.
Santini, D. L., & Aber, J. W. (1998). Determinants of net new money flows to the equity
mutual fund industry. Journal of Economics and Business, 50(5), 419-429.
Sawicki, J., & Ong, F. (2000). Evaluating managed fund performance using conditional
measures: Australian evidence. Pacific-Basin Finance Journal,8(3), 505-528.
Schadt, R. W. (1996). Measuring fund strategy and performance in changing economic
conditions. The Journal of Finance,51(2), 425-461.
Schiffman, L. G., & Kanuk, L. L. (2000).Consumer behavior, 7th.
Shah, S. A., Hijazi, S. T., & Hamdani, N. H. (2005).Performance Evaluation of Mutual

276
 Funds in Pakistan [with Comments]. The Pakistan Development Review, 863-
876.
Shankar, V. (1996).Retailing Mutual Funds: A consumer product model. The Hindu, 24.
Shu, P. G., Yeh, Y. H., & Yamada, T. (2002). The behavior of Taiwan mutual fund investors’
performance and fund flows. Pacific-basin finance journal,10(5), 583-600.
Sharpe, W. F. (1966). Mutual fund performance. The Journal of business,39(1), 119-138.
Sharma, P. (2011). Factors That Influence Investment Decision In Indian Mutual Fund
Industry. Editorial Board Chief Bebe Factor.
Sipra, N. (2006). Mutual fund performance in Pakistan, 1995-2004. Centre for
Management and Economic Research (CMER), 1-14.
Siggelkow, N. (2003). Why Focus? A Study Of Intra‐Industry Focus Effects.The Journal
of Industrial Economics, 51(2), 121-150.
Sing, T. C. (2007). Effects of expenditures and size on mutual fund performance. Singapore
Management Review, 29(1), 31.
Sood, D., & Kaur, N. (2014). An analytical study on investors perception regarding
different investment schemes with reference to Punjab. ZENITH International
Journal of Business Economics & Management Research, 4(2), 145-157.
Sipra, N. (2006). Mutual fund performance in Pakistan, 1995-2004. Centre for
Management and Economic Research (CMER), 1-14.
Singh, J., &Chander, S. (2004). An empirical analysis of perceptions of investors towards
mutual funds. Finance India, 18(4), 1673.
Sirri, E. R., &Tufano, P. (1998). Costly search and mutual fund flows. The journal of
finance, 53(5), 1589-1622.
Spitz, A. E. (1970).Mutual fund performance and cash inflows. Applied Economics, 2(2),
141-145.
Suppa-Aim, T. (2010). Mutual fund performance in emerging markets: the case of
Thailand (Doctoral dissertation, The University of Birmingham.
Stiglitz, J. E. (1989). Discussion: mutual funds, capital structure, and economic

277
 efficiency. Theory of Valuation: Frontiers of Modern Financial Theory, 1, 342-
356.
SyamaSundar, P. V. (1998). Growth prospects of mutual funds and investor perception with
special reference to Kothari Pioneer mutual fund. Project Report, Sri
SrinivasVidyaParishad, Andhra University, Visakhapatnam.
Sapp, T., & Tiwari, A. (2004). Does stock return momentum explain the “smart money”
effect?. The Journal of Finance, 59(6), 2605-2622.
Ter Horst, J., &Verbeek, M. (2000).Estimating short-run persistence in mutual fund
performance. Review of Economics and Statistics, 82(4), 646-655.
Tonks, I. (2005). Performance Persistence of Pension‐Fund Managers*. The Journal of
Business, 78(5), 1917-1942.
Teo, M. (2009).The geography of hedge funds. Review of Financial Studies, hhp007.
Treynor, J., & Mazuy, K. (1966). Can mutual funds outguess the market.Harvard business
review, 44(4), 131-136.
Treynor, J. L. (1965). How to rate management of investment funds. Harvard business
review, 43(1), 63-75.
Vidya Shankar, S. (1990). Mutual funds: Emerging trends in India. Chartered
Secretary, 20(8), 639-640.
Warther, V. A. (1995). Aggregate mutual fund flows and security returns.Journal of
financial economics, 39(2), 209-235.

Wermers, R. (1997). Momentum investment strategies of mutual funds, performance

persistence, and survivorship bias. Unpublished Working Paper, University of
Colorado,[downloaded from http://bus. colorado. edu/faculty/wermers/.].
Wermers, R. (2000). Mutual fund performance: An empirical decomposition into stock‐
picking talent, style, transactions costs, and expenses. The Journal of
Finance, 55(4), 1655-1703.
Woerheide, W. (1982).Investor response to suggested criteria for the selection of mutual

278
 funds. Journal of Financial and Quantitative Analysis,17(01), 129-137.
Wolfers, J., &Zitzewitz, E. (2006). Prediction markets in theory and practice (No. w12083).
National bureau of economic research.
Wilcox, R. T. (2003). Bargain Hunting or Star Gazing? Investors’ Preferences for Stock
Mutual Funds. The Journal of Business, 76(4), 645-663.
Yadav, V. N., & Massa, M. (2012). Do Mutual Funds Play Sentiment?. SSRN Working
Paper Series.
Zhao, Z. (2004). Review on the Progress for Mutual Fund Performance Dynamic
Evaluation Theory. Journal of Systems Science & Information,2(4)
Zhao, J. (2006). Quant jocks and tire kickers: Does the stock selection process
matter. Tucson, AZ: University of Arizona working paper. Available at
http://finance.eller. arizona. edu/documents/doctoral/JZhao. JobMarketPaper.pdf.
Zheng, L. (1999). Is money smart? A study of mutual fund investors' fund selection
ability. The Journal of Finance, 54(3), 901-933.

279
Appendixes
Appendix A Questionnaire
I am a Ph.D. scholar at Islamia College Peshawar and doing my research on the topic,
Mutual fund performance analysis and investors investing behavior in mutual fund. Please
fill up the questions below. The information provided will be kept highly confidential and
will be used only for academic purposes.
Please tick the following which suite you.
1. Gender
Male 
Female 
2. Age
Less than 30 
Between 30 to 40 
Above 40 

3. Education
Post Graduate 
Graduate 
Under Graduate 
Diploma 
High school 
Less than high school 

4. Occupation
Professional 
Business 
Salaried 
Retired 

5. Marital status
Married 
Un married 

6. Annual Income
Less than 100,000 
100,000 to 500,000 
500,000 to 10,00,000 
Above 10,00,000 

280
7. Location
Peshawar 
Islamabad 
Karachi 
Lahore 
Quetta 

8. Do you prefer investment in mutual funds to other saving avenue in future?

Yes  No  Not sure 

9. Generally you prefer (1 to 8)

Equity fund Fund of Fund 
Capital protected  Income fund 
Money market  Balanced fund 
Asset al.location  Islamic fund 
fund

10. Which kind of Mutual fund would you prefer?

Open ended scheme 
Close ended scheme 
Voluntary pension fund 

11. You prefer investment in mutual funds due to (rank from 1 to 9 down)
Safety Liquidity 
Tax benefits Flexibility 
Cap appreciation Return 
Prof Management Benefit 
Diversification 

12. How do you invest in this option?

Agent 
Family and friends 
Self 

Fund Characteristics

13. You think fund performance record as:

281
  Highly important  Not very
 Important important
 Somewhat  Not at all
important important

14. You think Fund reputation brand name as:

 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

15. You think Schemes portfolio of investment as:

 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

16. You think Reputation of the fund manager as:

 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

17. You think Withdrawal facilities as:

 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

18. You think Favorable rating of mutual fund by rating agencies as:
 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

19. You think the Innovativeness of the mutual fund scheme as:
 Highly important  Somewhat  Not very
 Important important important

282
  Not at all
important

20. You think the funds as Products with tax benefit:

 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

21. You consider the funds Entry & exit load as:
 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

22. You consider the Minimum initial investment as:

 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

23. You consider the disclosure of NAV on every day as:

 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

24. You consider the disclosure of derivation from the original pattern as:
 Highly important  Not very
 Important important
 Somewhat  Not at all
important important

25. How did you come to know about mutual fund investment scheme?
 Mail
 Radio  Brockert/Agent  Financial
 TV magazine

283
26. Do you think mutual fund investing is a best alternative to easily investing?
 Yes  No  Don’t know

27. What kind of communication you like to know about your fund performance?
 Automated response  Telephone
 Personal visit  No preferences

28. Investment in mutual fund helps you realize the benefits of stock market investing?
 Yes  No  Don’t know

29. Mutual fund investing gives a definite positive return?

 Yes  No  Don’t know

30. Mutual Fund returns and principals are fully protected and granted by MUFAP?
 Yes  No  Don’t know

31.Bank sponsored MF give more return than bank fixed deposit?

 Yes  No  Don’t know

32. Entry and exit out of mutual fund is easy?

 Yes  No  Don’t know

33. Due to professional investment, a better return is expected from Mutual Fund?
 Yes  No  Don’t know

34. Ups and down neither of stock market will nor affect the return from Mutual Fund?
 Yes  No  Don’t know

35. MUFAP helps in protecting the Mutual Fund industry and the unit holders.
 Yes  No  Don’t know

36.How often you switch the schemes of Mutual Fund in a year?

 Never  three or four times
 one or two times

284
37. After investment, how frequently you monitor the performance of Mutual fund?

 Weekly
 Once a month
 Once
 Rarely

Appendix -B Regression of Fama French 3 Factor model adjusted return

(1) (2)
CAPM_lagged
VARIABLES CAPM alphas

lff_alpha 0.601***
(0.072)
exp_ratio -0.197 0.032
(0.596) (0.551)
Lage 0.838*** 0.716**
(0.314) (0.345)
Lfam 0.112 -0.182
(0.165) (0.166)
Cflow 0.465** 0.247
(0.202) (0.201)
Ltna 0.113 0.017
(0.075) (0.072)
-
Turnover 0.181*** -0.065
(0.054) (0.052)
year2

year3

year4

285
Constant -0.814 -0.239
(1.813) (1.728)

R-squared 0.161 0.497

Standard errors in
parentheses
*** p<0.01, ** p<0.05, * p<0.1

286
Appendix- C Carhart Alpha Regression
(1) (2)
CAPM_lagged
VARIABLES CAPM alphas

lcarhat_alpha 0.431***
(0.080)
exp_ratio 0.124 0.326
(0.637) (0.663)
Lage 0.747** 0.945**
(0.335) (0.409)
Lfam 0.058 -0.220
(0.176) (0.199)
Cflow 0.507** 0.374
(0.216) (0.242)
Ltna 0.081 0.019
(0.080) (0.086)
Turnover -0.192*** -0.130**
(0.058) (0.062)
year2

year3

year4

Constant 0.343 0.464

(1.937) (2.069)

R-squared 0.140 0.336

Standard errors in
parentheses
*** p<0.01, ** p<0.05, * p<0.1

287
288