Economic Modelling 29 (2012) 10531063

Contents lists available at SciVerse ScienceDirect

Economic Modelling
journal homepage: www.elsevier.com/locate/ecmod

Evaluating ination targeting as a monetary policy objective for India

Ankita Mishra a, 1, Vinod Mishra b,
a
b

School of Economics, Finance and Marketing, RMIT University, VIC 3000, Australia
Department of Economics, Monash University, VIC 3800, Australia

a r t i c l e

i n f o

Article history:
Accepted 26 February 2012
JEL classication:
E52
E58
E63
Keywords:
India
Ination targeting
CPI
Taylor's rule
Monetary policy

a b s t r a c t
This study formulates a small open economy model for India with exchange rate as a prominent channel of
monetary policy. The model is estimated using the Instrumental Variable-Generalized Methods of Moments
(IV-GMM) estimator and evaluated through simulations. This study compares different cases of domestic and
CPI ination targeting, strict and exible ination targeting, and simple Taylor type rules. The analysis highlights the unsuitability of simple Taylor-type monetary rules in stabilizing the Indian economy and suggests
that discretionary optimization works better in stabilizing this economy. There seems to be a trade-off between output gap stabilization and exchange rate stabilization in exible domestic ination targeting and
CPI ination targeting respectively. However, exible domestic ination targeting seems a better alternative
from an overall macro stabilization perspective in India where nancial markets are still not sufciently
integrated to ensure quick transmission of interest rate impulses and existence of rigidities in the economy.
2012 Elsevier B.V. All rights reserved.

1. Introduction
Ination targeting has emerged as a powerful and effective monetary policy regime since the early 1990s. It has been adopted by a
number of industrial countries starting with New Zealand in 1990,
Canada in February 1991, Israel in December 1991, the United Kingdom in 1992, Sweden and Finland in 1993 and Australia and Spain
in 1994. Many of the empirical studies show that an ination targeting regime has been successful in signicantly reducing ination in
these countries. Bernanke et al. (1999), for example, found that ination remained lower after ination targeting than would have been
the case if forecasted by using Vector Auto-Regressions (VARs) estimated with the data from the period before ination targeting
started. Ination targeting also helped to maintain price stability
once it was achieved. Inspired by the success of ination targeting
in industrialized economies, many Emerging Market Economies
(EMEs) also adopted an ination-targeting approach to monetary
policy, including Chile in 1991, Brazil in 1999, Czech Republic in
1997, Poland in 1998 and Hungary in 2001.
Ination targeting is currently practiced by a group of advanced
economies and several medium to small sized EMEs. The applicability
of this regime to a large, growing, developing economy like India is
still a researchable area. There has been growing interest in analyzing

Corresponding author. Tel.: + 61 3 99047179.

E-mail addresses: ankita.mishra@rmit.edu.au (A. Mishra),
vinod.mishra@monash.edu (V. Mishra).
1
Tel.: + 61 3 99251638.
0264-9993/$ see front matter 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.econmod.2012.02.020

the applicability and suitability of ination targeting as a monetary policy regime for India, primarily because the current multiple indicator
monetary policy approach2 of the Reserve Bank of India (RBI) seems
to have lost its relevance and does not appear to work effectively.3
Many studies in the literature have attempted to analyze India's preparedness for ination targeting. Indeed. several studies have examined
nancial sector reforms as the essential pre-condition for adoption of
ination targeting, and analyzed the preparedness of India for ination
targeting from that perspective (see, Jha, 2008; Kannan, 1999). Following Kannan's (1999) suggestion that implementation of ination targeting in India should wait until nancial sector reforms have been
completed, Singh (2006) argued that the rst phase of nancial sector
reforms is complete and macroeconomic performance in terms of
level of ination and interest rates is satisfactory and stable suggesting
that conditions are favorable in India for the adoption of ination targeting. Singh advocated addressing a few issues, namely, use of both scal
and monetary instruments to control ination, publication of fulledged ination reports, and establishing an ination committee to
bring transparency in its operation before an actual ination targeting
framework could be adopted in India. Khatkhate (2006) asserted that
ination targeting might be a good policy framework for India as the
RBI always has to be on alert to maintain its credibility and authority

2
Post global nancial crisis, nancial stability has become one of the major concerns
of central banks across the world. However, if implemented in a exible manner, ination targeting is perfectly compatible with a nancial stability objective (Walsh, 2009).
3
Refer to D'souza (2003), Shah (2007), Mishra and Mishra (2009), Mishra and Mishra
(2010).

1054

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

in controlling ination, even though the sources of ination in India are

often non-monetary. 4 She suggested that in operational terms India
ought to target headline ination. Mishra and Mishra (2009) analyzed
the preconditions for ination targeting in India, namely, the independence of monetary policy from scal, external, structural and nancial
concerns, and assessing its suitability as a monetary policy framework
for India. They found that the Indian economy satises the preconditions for ination targeting.
Extending the analysis of Mishra and Mishra (2009), this study attempts to answer the question of the probable consequences of shifting to an ination targeting framework of monetary policy, and how
different shocks will affect the economy under this framework by
using a general linear model of the economy with quadratic loss function to be minimized by the central bank for India.
Ination targeting is conducted in conjunction with a Monetary
Policy Rule (MPR). MPR is part of the overall monetary policy of the
central bank or monetary authority, and species how the instrument
of monetary policy is to be changed given the characteristics of the
macro economy and the policy objective of the central bank. This
study compares different cases of strict and exible domestic and
CPI (Consumer Price Index) ination targeting (Optimal MPR) with
simple Taylor type rules (simple MPR) to examine the most suitable 5
ination targeting framework for India. Modeling ination targeting
as the announcement and assignment of a relatively specic loss
function to be minimized by the central bank, this study suggests
that simple Taylor type rules are inadequate in stabilizing the economy, and that optimal rules work better. Further, though there seems
to be a trade-off between output gap stabilization and exchange
rate stabilization in exible domestic ination targeting and CPI ination targeting respectively, exible domestic ination targeting appears to be a better policy option for India from an overall macro
stabilization aspect.
The organization of the rest of this paper is as follows: Section 2
presents a brief review of the models used in literature to examine
the monetary policy rules; Section 3 outlines the structure of the theoretical model and the description of its main equations; Sections 4
and 5 present empirical and simulation results respectively; and the
nal section presents the conclusions and policy implications of this
study.
2. Literature review
There are a wide variety of models developed in the literature to
investigate monetary policy rules. These models differ in size, degree
of openness and degree of forward lookingness assumed. Some of
these models are developed in closed economy settings to examine
the performance of policy rules that are consistent with a monetary
policy regime of ination targeting, for example, Rudebusch and
Svensson (1999), and Clarida et al. (1999) among others. Many of
the studies extended the analysis of monetary policy from closed
economy settings to open economy settings. These open economy
models differ from their closed counterparts as the real exchange
rate affects both aggregate demand and ination. This complicates
monetary management as the impact of exchange rate on real activity
and ination must be accounted for while formulating monetary policy. Some of the notable studies are Ball (2000) 6, Walsh (1999),
Svensson (2000), and Clarida et al. (2001).
The above models study ination-targeting regimes for developed
countries; however, for the purpose of the current study, more relevant are those studies that look at ination targeting as a monetary
policy regime for EMEs. Many of the studies modify the models
4

Such as changes in food output, marketed surplus, lax scal policy etc.
The most suitable implies the framework which is capable of bringing in overall
macro stabilization and not just ination stabilization.
6
For a closed economy version of the model, refer to Ball (1997).
5

developed in the literature to include characteristics mainly present

in EMEs, for example, Moron and Winkelried (2003) modied the
model of Svensson (2000) to incorporate nancial vulnerability characteristics (which are commonly present in EMEs) to examine ination targeting monetary policy rules. Fraga et al. (2003) analyzed
the case for ination targeting in EMEs and asserted that the problems faced by EMEs with respect to their scal, nancial and external
sectors are more acute when compared to developed economies.
They use a small open economy model where imports enter as intermediate goods rather than as consumption goods. This assumption
seems more plausible for the developing economy as capital and intermediate goods form a larger proportion of total imports in developing countries than do consumption goods. Goyal (2008a) adapted
a standard IS-LM-UIP framework to build in the dualistic labor market
and wage price rigidities present in the Indian economy. She found
that in a simple open EME model calibrated to typical institutions
and shocks of a densely populated EME like India, a monetary stimulus preceding a temporary supply shock can help stabilize ination at
minimum output cost, because of the exchange rate appreciation that
accompanies a fall in interest rates and rise in output. In another
study, Goyal (2008b) adapted the dynamic, stochastic, general equilibrium models 7 to analyze optimal monetary policy rules to the
labor market structure of the small open emerging economy of
India. By examining the welfare effects of different types of ination
targeting regimes for India, she concluded that exible CPI ination
targeting without lags 8 works best and even more so when the degree of openness increases. However, due to welfare loss from the
lags of CPI, domestic ination targeting continues to be more robust
and effective.
3. A macro structural model of a small open economy
The formulation of the model of monetary policy used to analyze
different ination targeting monetary policy rules is based on the literature (in particular the models by Batini and Haldane, 1999,
Svensson, 2000 and Goyal, 2008a) and broad conclusions derived by
Mishra and Mishra (2010). Their model suggests that the monetary
policy has real effects. It suggested that a New Keynesian framework
incorporating some form of stickiness in the prices which gives rise to
non-neutral effects of monetary policy is needed to prepare the
framework suitable for the evolution of monetary policy. Second,
since the mid 1990s rate variables have been better at signaling the
stance of monetary policy for India than quantity variables, and this
implies the use of nominal interest rate (rather than money supply)
as an instrument of monetary policy. Third, there is a growing importance of the exchange rate channel in the transmission of monetary
policy in India, with exchange rate shocks playing a central role in
explaining the volatility of ination, interest rate, growth of credit
and money supply in that country. Thus, exchange rate shocks as
well as shocks originating from the rest of the world (transmitted
through exchange rate) are important in conducting monetary policy,
and the model for evaluation of monetary policy should incorporate
this. Against this background we set up a model of monetary policy
7
Notable models of the type include models by Clarida et al. (1999, 2001), Svensson
(2000), Woodford (2003) and Gali and Monacelli (2005).
8
CPI is available only at a monthly frequency in India, with a two month lag. Further,
four distinct CPI measures are compiled for different reference populations. These are
Consumer Price Index for Industrial Workers (CPI-IW), Consumer Price Index for Urban
Non-Manual Employees (CPI-UNME), Consumer Price Index for Agricultural Labourers
(CPI-AL) and Consumer Price Index for Rural Labourers (CPI-RL). These indices differ
from each other in reference population, basket of goods and services and their
weights, geographical areas and base-periods etc. None of these measures presents
all the spectrum of the population of the country. There is a need to develop a harmonized measure of consumer price index which is computed on an all India basis and reects the purchasing power of domestic currency in domestic market and thus
facilitates international comparisons. Therefore, there are both information and price
adjustment lags in the CPI.

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

for India which is then used to evaluate different monetary policy

rules.
All the variables in the model (other than interest rate) are taken
in logs and are expressed as deviations from their respective means.
The aggregate supply 9 in period t is given by the following short
run open economy Phillips equation:
p

t b1 t1 1b1 Et t1 b2 yt b3 qt1 qt2 et :

Here, t denotes domestic or Wholesale Price Index (WPI) ination. We assume that there is some inertia in ination and it is not
completely forward-looking. The degree of forward-lookingness depends on the values of the parameter b1. Thus, ination depends
on its lag, its expected value in period t + 1, output gap (Yt), lag of
(log change in) real exchange rate (qt) (or lag of depreciation rate)
and zero mean i.i.d. ination shock (etp). Ett + 1 as the expected ination rate in period t + 1 is observed from period t. 10 The variable
output gap yt is dened as:
yt

d
p
yt yt :

Here, ytd is (log) aggregate demand and ytp is potential output 11.
The variable qt 1 is the (log) lag real exchange rate. The real exchange rate in the economy is dened by following the Purchasing
Power Parity (PPP) condition:


qt et pt pt :

The variable rt is dened as:

r t it Et t1 :

Here, it is short-run nominal interest rate and is the instrument of

the central bank.
Using Eq. (6) and assuming pt* and it* are normalized to zero, the
real interest rate parity condition is dened as:
e

qt qt1jt t1jt it t :

We dene CPI ination as weighted average of domestic ination

and domestic currency ination of foreign goods, given by the following equation:
c

t 1t t :

Here, is the share of imported goods in the CPI and tf denotes

domestic currency ination of imported foreign goods which in turn
is dened as:
f

t pt pt1
ptf pt et

and

And since, pt* = 0, ptf = et

So, tf = et et 1 and substituting the value of et, from Eq. (3)

Here, pt is the (log) price level of domestic (ally produced) goods,

pt* the (log) foreign price level and et represents (log) nominal exchange rate. The variable et fullls the following interest rate parity
condition:

1055

t qt pt qt1 pt1 t qt qt1

Substituting this value in Eq. (8):
c

et Et et1 it it t

where it and it* are domestic and foreign nominal interest rates respectively, and te is a combined shock in foreign interest rate and
other disturbances in the foreign exchange market, including shocks
to foreign exchange risk premium. 12
The aggregate demand (dened in terms of output-gap) in the
economy in period t is given by the following equation:
y

yt a1 yt1 a2 r t a3 qt1 qt2 t :

t 1t t qt qt1 t qt qt1 :

Thus, the model consists of aggregate supply Eq. (1), aggregate demand Eq. (5), real interest rate parity Eq. (7), and the CPI Eq. (9).
3.1. The Loss function
The optimal monetary policy rules can be derived from central
bank's explicit loss function. We assume the following loss function
of RBI:

5
2

c2

Lt t t y yt i it i it it1 :
The aggregate demand curve is backward looking. It depends on
its own past lag, real interest rate (rt), depreciation rate and demand
shocks. The term ty is a zero mean i.i.d. aggregate demand shock.
9
This Phillips curve is similar to the Phillips curve emerging from a Calvo type staggered price setting framework from the inter-temporal maximization of a representative agent that demands domestic and foreign goods, and also to Fuhrer and Moore
Phillips curve in that ination depends on both lagged ination and future expected ination. Further, the motivation for this empirical version of Phillips curve comes from
the results of Mishra and Mishra (2010). They found that the (change in) exchange
rate affects ination with a lag. External factors like oil prices and foreign interest rates
were also signicant in explaining ination. We assumed that effects of all these external factors would be transmitted to domestic ination via a change in exchange rate.
10
For estimation, the expected ination is taken as a forecast obtained from tting
AR(1) model in WPI ination series.
11
Potential output is obtained with HodrickPrescott lter method. It is thus a longterm trend component in (log) of Index of Industrial Production (IIP) series, which is
the measure of output (aggregate demand) in our model.
12
This assumption is motivated by Batini and Haldane's (1999) model and the response of exchange rate to monetary policy shock as discovered by Mishra and Mishra
(2010) where we see some prolonged appreciation of exchange rate to a positive interest rate shock.

10

Here, all the weights are non-negative; thus, the loss function is
the weighted sum of the respective unconditional variances and dened in the following way:

c
c
ELt  Var t  Var t y Varyt  i Varit 
i Var it it1 :

11

Here, , c , y, i and i are policy parameters that relate to domestic ination, CPI ination, output, interest rate and interest rate
smoothing. The strict domestic ination targeting corresponds to
positive and all other weights are equal to zero. Flexible domestic
ination targeting refers to positive weights to other policy parameters also. CPI ination targeting will have c positive rather than .
Thus, the decision problem of the central bank is to choose the instrument it conditional upon the information available in period t so as
to minimize Eq. (10).

1056

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

The above loss function can be seen as a limit of the inter-temporal

loss function:
min Et

h
i
j
2
c c2
2
2
2
t t y yt i it i it :

12

j0

Here, is the discount factor (fullling 0 b b 1 representing central bank's rate of time preference).
4. Data and empirical estimations
The aggregate demand and aggregate supply equations are estimated with an Instrumental Variable-Generalized Method of Moments (IV-GMM) estimator. 13 We use monthly data for the period
January 1996 to March 2007. The data is sourced mainly from the
IFS (IMF) and the RBI (www.rbi.org.in). All the variables other than
interest rates are transformed as annual changes in log values. Thus,
all the variables denote year-on-year changes in the original series
which take care of seasonality issues in the monthly sample. All the
variables are for the 199394 base period. All the variables entering
into the estimation are stationary. 14 As there are expectation variables in the equations, we estimated the equations by replacing
these expected values with their corresponding realized values and
thereby introducing expectation errors into the equations' composite
disturbances.
4.1. Variable description
Domestic ination and CPI ination are measured as year-to-year
changes in Wholesale Price Index (WPI) and Consumer Price Index
for Industrial Workers (CPI-IW). Output gap is measured as the difference between (log of) Index of Industrial Production (IIP) and
its long term trend, proxied by (log of) HodrickPrescott trend. Real
exchange rate is measured by real effective exchange rate (REER)
trade weighted. The call money rate (CMR) is taken as the short-run
nominal interest rate.
4.2. Estimation of aggregate supply equation
The specication of aggregate supply equation is kept as close as
possible to the theoretical specication as described in Section 3.
The annual WPI ination (WPI) is the dependent variable. The lag
of ination (WPIt1 ) and lag of (log) change in real effective exchange rate (REERt1 ) are the included instruments. The endogenous regressors are output gap ( Y) and expected ination (E
[WPI]). They are instrumented with a number of variables and their
lags (excluded instruments) 15 in the regression.
The estimation results of aggregate supply equation suggest that
both lagged ination rate and expected ination are highly signicant
(at 1% level). The ination in India shows substantial inertia along
with the degree of forward-lookingness. This conrms our assumption in the theoretical section that hybrid Phillips curve could better
explain ination dynamics in India. The lag of (log) change in real
13

The GMM estimator applied here is the two-step efcient generalized method of
moments (GMM) estimator. The efcient GMM estimator minimizes the GMM criterion function J = N * g * W * g, where N is the sample size, g is the orthogonality or moment conditions (specifying that all the exogenous variables or instruments in the
equation are uncorrelated with the error term) and W is the weighting matrix (inverse
of an estimate of the covariance matrix of orthogonality conditions). These moments'
conditions are tested using Hansen J statistic.
14
The unit root tests of stationarity are reported in Appendix Table A.1.
15
List of excluded instruments in aggregate supply equation is given in
Appendix Table A.2. We also estimated alternative specications of IV regressions
using a subset of 12, 8, 6 instruments respectively. The results of these alternative specications were very close (sign, signicance and magnitude) to the results reported in
Table 1, suggesting that the model is robust to alternative specications of choice of
instruments.

effective exchange rate (or lag of depreciation) is also signicant.

The coefcient on output gap is a very small positive number suggesting that the supply curve in India is sufciently elastic and there is an
excess (or unused) capacity in the economy. However, this coefcient
is not signicant which reconrms the ndings of Mishra and Mishra
(2010) that ination in India is not markedly governed by demand
pull factors. The overall goodness of t as measured by R-square is
0.73 (centered), suggesting that the included regressors explain
much of the movements in the ination.
The next step in instrumental variable estimation is to check
the validity of instruments. We performed two tests, that is, the
HansenSargan test of over-identifying restrictions and the LM test
of under-identication (KleibergenPaap rk statistic). The joint null
hypothesis of HansenSargan test is that the instruments are valid instruments, that is, uncorrelated with the error term, and that the excluded instruments are correctly excluded from the estimation
equation. From Table 1 we see that we fail to reject the null of instrument validity. For the under-identication test, null hypothesis is that
the matrix of reduced form coefcients has rank equal to K-1 (such
that K is the number of a regressor). The statistic provides a measure
of instrument relevance and the rejection of null indicates that the
model is identied. We reject null at 1% level (p-value 0.000) of
under-identication, which means that the aggregate supply equation has been properly identied.
The next step was to check whether or not GMM estimator is the
appropriate estimator. In the presence of heteroskedasticity, GMM estimator is more efcient than the simple IV estimator. However, in
the absence of heteroskedasticity, GMM estimator is asymptotically
no worse than the IV estimator. Pagan and Hall (1983) proposed
the test of heteroskedasticity for instrumental variable estimation.
The results of the PaganHall test statistic for our aggregate supply
equation did indicate the presence of heteroskedasticity in aggregate
supply equation and thus justied the use of GMM estimator.
4.3. Estimation of aggregate demand equation
The aggregate demand equation of the model is also estimated
with the IV-GMM estimator. The output gap is the dependent variable in the regression. The lag of the output gap Y t1 ) and lag of
log change in real effective exchange rate (REERt1 ) are included
Table 1
IV-GMM estimates of the aggregate supply equation.
Dependent variable: WPI
Variable

Coefcient

Standard
error

P-value

Y
WPIt1
E(WPI)
REERt1
Constant
Number of observations
Centered R-square
Uncentered R-square
Root MSE
F (4,128)
Underidentication test (KleibergenPaap
rk LM statistic)
Hansen J statistic

0.0109
0.5931
0.4226
0.057
0.0005
133
0.7387
0.7939
.0039
498.80
36.39

0.0078
0.0498
0.0570
0.0072
0.0011

0.168
0.00
0.00
0.019
0.638

Test for the presence of heteroskedasticity

PaganHall test statistic(using levels of IVs)
PaganHall test statistic (using levels
and squares of IVs)
PaganHall test statistic (using level and
cross products of IVs)

0.00
0.00

11.072

0.1976

33.854
76.349

0.0508
0.0013

123.027

0.6550

Note: WPI : WPI ination; Y: Output gap; REER: (change in) real effective exchange
rate; E(WPI) : expected WPI ination. ***, **, * indicate 1%, 5% and 10% signicance
levels respectively.

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

instruments. Variable real interest rate (RIR) is the endogenous regressor, which has been instrumented with a number of exogenous
variables (excluded instruments). 16
The estimation results of the aggregate demand equation suggest
that lag of output gap is signicant at 1% level and the depreciation
rate is signicant at 5% level. The real interest rate coefcient is signicant only at 10% level. The coefcient on real interest rate is a very
small negative number indicating the low interest elasticity of the aggregate demand in the Indian economy. Model selection statistics
such as F-statistic, Root Mean Square Error and R-square are signicant
and support the tight t for the model. The tests of validity of instruments, that is, Hansen J test and KleibergenPaap LM test, suggest that
instruments are valid instruments; thus excluded instruments are correctly excluded from estimation equation and the equation is identied.
The PaganHall test suggests the presence of heteroskedasticity.
5. Simulations
5.1. Results on optimal policies
We simulated the model 17 using the estimates of aggregate demand and supply from Tables 1 and 2 (taking the values of b1 = 0.
5931, b2 = 0. 0109, b3 = 0.057 for aggregate supply and a1 =
0.2437, a2 = 0.0032, a3 = 0.1536 for aggregate demand), an estimate for (), the share of imported goods () in CPI 18 and the variance of various shock parameters in our model, demand shocks
(y2), supply shocks (p2) and exchange rate shocks (q2)). Also we
assumed that the model is subjected to historical shocks, thus the variance of demand shocks (y2) and supply shocks (p2) is assumed to be
1 and the variance of exchange rate shocks (q2) is to be 0.5).

1057

Table 2
IV-GMM estimates of the aggregate demand equation.
Dependent variable: Y
Variable

Coefcient

Standard
error

P-value

Y t1
RIR
REERt1
Constant
Number of observations
Centered R-square
Uncentered R-square
Root MSE
F (3,122)
Underidentication test (KleibergenPaap
rk LM statistic)
Hansen J statistic

0.2437
0.0032
0.1536
0.0446

0.0818
0.0018
0.0653
0.0123

0.00
0.087
0.019
0.00

Test for the presence of heteroskedasticity

PaganHall test statistic (using levels of IVs)
PaganHall test statistic (using levels
and squares of IVs)
PaganHall test statistic (using level and
cross products of IVs)

125
0.4210
0.4210
.03015
12.24
54.363

0.00
0.00

9.433

0.8538

78.945
137.814

0.000
0.000

253.931***

0.000

Note: Y: output gap; REER: (change in) real effective exchange rate; RIR: real interest
rate. ***, **, * indicate 1%, 5% and 10% signicance levels respectively.

ination targeting (domestic or CPI) are dened as weight only

on ination stabilization (domestic or CPI ination depending on
the targeted ination of the central bank), while in the cases of exible ination targeting there is a weight on output stabilization but
with preference (represented by higher weight) toward ination
stabilization.

5.2. Ination targeting cases and Taylor rules

5.3. Unconditional standard deviations
The different cases of ination targeting are identied by the
weights of policy parameters in the loss function of the RBI. We examined four targeting cases depending on whether the instruments
correspond to WPI (domestic ination) or CPI ination along with
two closed economy versions and one open economy 19 version of
Taylor rules. 20 The weights given to the policy parameters under different targeting cases 21 are given in Table 3. The cases of strict
16
The list of excluded instruments in aggregate demand equation is given in
Appendix Table A.3. We also estimated alternative specications of IV regressions
using a subset of 12, 8, 6 instruments respectively. The results of these alternative specications were very close (sign, signicance and magnitude) to the results reported in
Table 2, suggesting that the model is robust to alternative specications of choice of
instruments.
17
We modied a GAUSS code for solving optimal monetary policy under discretion
and simple rules, made available by Paul Soderlind on his website http://home.
tiscalinet.ch/paulsoderlind. The detailed solution of the model is with the authors
and available on request.
18
We took the value of = 0.3. The share of imported goods in CPI is not very high in
India. We also experimented with a number of values for ; these do not affect our results qualitatively.
19
In an open economy version of Taylor rule, exchange rate enters with non-zero coefcient. In particular the coefcient on exchange rate should be less than zero and coefcient on lag of exchange rate should be a small positive number representing partial
adjustment. (for more discussion on this refer to Taylor (2001) and Cavoli and Rajan
(2005)). Using model simulations, recent literature nd the optimal coefcient values
for exchange rate range between 0.45 and 0.25 and lag of exchange rate range between 0.15 and 0.45 (refer to Cavoli and Rajan 2005).
20
We also tested Taylor rules with weight on interest rate smoothing at 0.50, 0.75
and 0.95. We nd that Taylor rules with interest rate smoothing do bring down the volatility in interest rate (the higher the weight on interest rate smoothing, the lesser the
variability in interest rate). However, their performance in terms of overall stabilization is even worse than optimal rules. These results are available from the authors on
request.
21
The case of strict CPI ination targeting does not converge without adding a small
weight on interest rate smoothing. Therefore, for uniformity we added a small weight
of 0.01 to interest rate smoothing parameter under all rules.

The unconditional standard deviation results highlight the unsuitability of simple Taylor type monetary policy rules over optimal
monetary policy rules. Closed economy simple monetary policy
rules with only weight on output and ination result in huge volatility
in exchange rate and interest rate. The performance of simple Taylor
rule CPI ination is worse than the simple Taylor rule domestic ination. The open economy Taylor rule with weight on exchange rate
generates lesser volatility in exchange rate than its closed economy
counterpart. However, the performance of open economy Taylor
rule is worse in stabilizing ination and output in comparison to
closed economy Taylor rules and exchange rate and interest rate in
comparison to optimal monetary policy rules (Table 4).
The results also indicate that there exists a trade-off between ination variability and variability of other macro variables (output, exchange rate and interest rate) in strict ination targeting cases. When
we move from strict (both domestic and CPI) targeting cases to exible targeting cases, the variability of ination rises and the variability
of other macro economic variables reduces. In CPI targeting cases the

Table 3
Targeting cases and Taylor rules.
Optimal MPR
Strict domestic-ination targeting
Flexible domestic ination targeting
Strict CPI targeting
Flexible CPI targeting

= 1
= 1
= 0
= 0

Simple MPR
Taylor rule, domestic ination
Taylor rule, CPI ination
Taylor rule, open economy

it = 1.5t + 0.5yt
it = 1.5tc + 0.5yt
it = 0.5yt + 1.5t 0.4qt + 0.2qt 1

c = 0
c = 0
c = 1
c = 1

y = 0
y = .5
y = 0
y = .5

i = 0
i = 0
i = 0
i = 0

i=.01
i :01
i=.01
i :01

1058

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

Table 4
Unconditional standard deviations.
Targeting cases

yt

tc

qt

it

Strict domestic-ination targeting

Flexible domestic ination
targeting
Strict CPI targeting
Flexible CPI targeting
Taylor rule, domestic ination
Taylor rule, CPI ination
Taylor rule, open economy

2.075
0.946

4.536
5.466

4.487
5.472

0.0039
0.0026

0.0039
0.0026

0.906
0.734
1.308
1.824
3.973

3.972
4.785
2.221
2.949
2.787

3.534
4.908
1.832
1.886
2.779

0.942
0.754
7.818
8.712
4.825

0.905
0.754
4.458
8.848
5.142

volatility of exchange rate and interest rate is higher when compared

to their volatility in domestic ination targeting cases. This suggests
the extensive use of exchange rate in CPI targeting to bring about stabilization. The domestic ination targeting rules in our model are
quite successful in containing variability of exchange rate and interest
rate but at the expense of high volatility in domestic and CPI ination
rates. The volatility of output is also higher in domestic ination targeting rules (both strict and exible) when compared to CPI ination
targeting rules. This creates a trade-off between domestic and CPI

Effect of 1 s.d. dd shock on y

0.5

0
0.1

0.1

1 2 3 4 5 6 7 8 9 10 11 12
Effect of 1 s.d. dd shock on
dom.inflation

0.001

1 2 3 4 5 6 7 8 9 10 11 12
Effect of 1 s.d. dd shock on CPI inflation

1 2 3 4 5 6 7 8 9 10 11 12
Effect of 1 s.d. dd shock on interest
rate

1 2 3 4 5 6 7 8 9 10 11 12
Effect of 1 s.d.
shock on dom. inflation

0
1

1 2 3 4 5 6 7 8 9 10 11 12
Effect of 1 s.d.
shock on CPI inflation

0
0.001

1 2 3 4 5 6 7 8 9 10 11 12
Effect of 1 s.d.
shock on interest rate

0.0005

1 2 3 4 5 6 7 8 9 10 11 12

0.001 Effect of 1 s.d. dd shock on exchange

rate
0.0005
0

0.5

0.0005

Effect of 1 s.d.
shock on y

0.5

0.05
0

0.01

0.005

0.05

ination targeting rules. Domestic ination targeting rules are more

successful in containing the volatility of exchange rate and interest
rate while CPI targeting rules are better at controlling the variability
of output and ination.
The simple Taylor type rules, though somewhat successful in
curbing the volatility of output and domestic and CPI inations, result in huge volatility in exchange rate and interest rate. The volatility of exchange rate and interest rate is higher in simple Taylor type
CPI ination rules than in the simple Taylor type domestic ination
rule.
The above results indicate that a exible form of ination targeting does result in lower variability of major macroeconomic variables
at a small ination volatility cost when compared to their strict targeting counterpart. Therefore some form of exible targeting is preferable for the Indian economy. On the other hand, the simple
rules turned out to be completely unsuccessful in stabilizing the economy in the current volatile open environment. There is a trade-off
between exible domestic ination and CPI ination targeting and
neither of them appears to be very successful from an overall
macro stabilization perspective. However, exible domestic ination
targeting with low to moderate volatility in all the variables seems
to be a better option but at the expense of some higher volatility in ination rates.

0.001

1 2 3 4 5 6 7 8 9 10 11 12
Effect of 1 s.d.
shock on exchange rate

0.0005

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

Fig. 1. Impulse response: domestic ination targeting (cases: strict and exible).

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

Effect of 1 s.d. dd shock on y

0.01

1059

Effect of 1 s.d.
shock on y

0.005
0

0.5

-0.005
-0.01

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

Effect of 1 s.d. dd shock on dom.inflation

0.1

0.05

0.5

Effect of 1 s.d.
shock on dom. inflation

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12
Effect of 1 s.d. dd shock on CPI
inflation

0.1

0.05

Effect of 1 s.d.
shock on CPI inflation

0.5

0
1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

Effect of 1 s.d. dd shock on interest

rate
0.01

0.1

0.005

0.05

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

0.01

Effect of 1 s.d.
shock on interest rate

Effect of 1 s.d. dd shock on exchange

rate

0.1

Effect of 1 s.d.
shock on exchange rate

0.05

0.005

0
1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

Fig. 2. Impulse responses: CPI ination targeting (cases: strict and exible).

5.4. Impulse response functions

Next we discuss the impulse responses for six cases of major macroeconomic variables to one standard deviation shock in the model.
These results help us to analyze how an economy reacts over time
to exogenous impulse shocks under various monetary policy rules,
and thus enables us to further evaluate the performance of different
monetary policy rules in a dynamic context.
In each case the model is subjected to one standard deviation positive demand shock to output gap and one standard deviation ination shock to domestic ination. We analyzed the responses of
ination, output, exchange rate and interest rate to these shocks
across different optimal monetary policy rules and for simple Taylor
rules for 12 time periods in Figs. 1 to 4. The optimal monetary policy
response under different monetary policy rules is examined. The

purpose of this analysis is to identify the monetary policy rule,

which quickly brings the economy back to equilibrium with limited
volatility in the macroeconomic variables in the event of a shock.
Also, these results throw light on the effect of the different shocks
on the economy under alternative ination targeting rules. 22

22
We found that patterns of impulse response functions to demand and ination
shocks are quite alike in strict and exible domestic ination targeting cases and in
strict and exible CPI ination targeting cases. However, these response functions do
differ in magnitude of volatility from strict to exible targeting cases. For example, demand and cost push shocks generate greater volatility in output and exchange rate and
lesser in ination in strict targeting cases when compared to exible ones. Given the
similar pattern of impulse response function in strict and exible cases, while discussing the results, we presented the results of strict domestic ination targeting together
with exible domestic ination targeting and strict CPI ination targeting together
with exible CPI ination targeting.

1060

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

Effect of 1 s.d. shock on y

Effect of 1 s.d. dd shock on y

2
1
0
-1
-2

2
1
0
-1
-2

1 2 3 4 5 6 7 8 9 10 11 12

Effect of 1 s.d. dd shock on dom.

inflation

1 2 3 4 5 6 7 8 9 10 11 12

Effect of 1 s.d. shock on dom.

inflation

-1

-1

-2

-2

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

Effect of 1 s.d. dd shock on CPI

inflation

2
1

1
0

-1

-1

-2

-2
1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

Effect of 1 s.d. shock on CPI

inflation

Effect of 1 s.d. dd shock on interest

rate

2
1
0
-1
-2

1
0
-1

Effect of 1 s.d. shock on interest

rate

1 2 3 4 5 6 7 8 9 10 11 12

-2
1 2 3 4 5 6 7 8 9 10 11 12

Effect of 1 s.d. dd shock on exchange

rate

Effect of 1 s.d. shock on

exchange rate

-2

-1

-4

-2

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12
Fig. 3. Impulse responses: simple Taylor rule domestic ination.

5.4.1. Strict and exible domestic ination targeting

One standard deviation positive aggregate demand shock under
domestic ination targeting monetary policy rule (strict as well as
exible) increases output by 1% above its potential level in period 1.
There is a smaller increase in domestic ination due to these shocks
and it increases by 0.1% in period 1. Monetary policy reacts to this development and there is an increase in nominal interest rate. As a result, real interest rate rises and there is an appreciation of real
exchange rate. Due to this output gap contracts, domestic ination
falls and all the macro economic variables return to their steadystate values in about 56 periods.
For one standard deviation ination shock, domestic (and CPI) inations rise by half a percent in period 1 and there is no response

from output gap in period 1. Reacting to rise in ination, there is a

rise in nominal interest rate and hence real interest rate. But the
increase in real interest rate is very modest when compared to the increase in ination. As a result, real interest rate falls in the subsequent
periods. With this, output gap starts expanding. This expansion of
output gap helps to bring down ination and ination reaches to its
target level in 56 periods.
5.4.2. Strict and exible CPI ination targeting
One standard deviation positive demand shock under CPI ination
targeting cases evokes similar reactions in output gap and ination as
under domestic ination targeting cases. However, it evokes a stronger monetary policy response when compared to domestic ination

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

1061

Effect of 1 s.d. shock on y

Effect of 1 s.d. dd shock on Y

2
1
0
-1
-2

2
1
0
-1
-2

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

2
1
0
-1
-2

Effect of 1 s.d. dd shock on

dom.inflation

2
1
0
-1
-2

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

2
1
0
-1
-2

Effect of 1 s.d. dd shock on CPI

inflation

2
1
0
-1
-2

2
1
0
-1
-2

Effect of 1 s.d. shock on CPI

inflation

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

rate

Effect of 1 s.d. shock on

dom.inflation

2
1
0
-1
-2

Effect of 1 s.d. shock on interest

rate

1 2 3 4 5 6 7 8 9 10 11 12

1 2 3 4 5 6 7 8 9 10 11 12

4
2
0
-2
-4

Effect of 1 s.d. dd shock on

exchange rate

1 2 3 4 5 6 7 8 9 10 11 12

2
1
0
-1
-2

Effect of 1 s.d. shock on

exchange rate

1 2 3 4 5 6 7 8 9 10 11 12

Fig. 4. Impulse responses: simple Taylor rule CPI ination.

targeting monetary policy rule. There is a bigger (as compared to domestic ination targeting rules) rise in nominal interest rate and
hence real interest rate, as well as a larger (compared to domestic ination targeting rules) appreciation in real exchange rate. As a result,
output gap starts contracting and ination starts falling in the subsequent periods. The economy returns to its steady state after about 56
periods.
Similarly one standard deviation ination shock in CPI targeting
cases causes a similar reaction in domestic and CPI ination as
under domestic ination targeting cases, and they increase by half a
percentage point in period 1. Output gap again does not respond to
this shock in period 1. There is a very strong monetary policy reaction
to this and a large increase in nominal interest rate and hence real

interest rate. This causes a large appreciation in real exchange rate.

Following an increase in real interest rate, output gap starts falling
from period 2 onward below its potential level and ination starts decreasing. The economy returns to its steady state equilibrium in
around 78 periods.
5.4.3. Taylor rules
Demand and ination shocks generate much larger variability in
major macroeconomic variables in Simple Taylor Domestic Ination
monetary policy rule when compared to optimal monetary policy
rules. The economy takes more time to recover from these shocks in
this rule. In the event of one standard deviation demand shock in domestic ination Taylor rule, output gap starts rising and ination also

1062

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

increases. Additionally, there is an increase in nominal (hence real interest rate) and an appreciation in real exchange rate. Due to this,
output gap contracts and ination falls. Ination turns negative after
around 78 periods and consequently there is a fall in interest rate
and hence exchange rate depreciates. This leads to the expansion of
output gap which rises for a few periods before returning to its steady
state. Ination also returns to its steady state after around 12 periods.
The supply shocks generate erratic response from all the major variables and the economy takes longer to stabilize.
CPI Taylor type monetary policy rules cause very high variability in
exchange rate. However, they result in very moderate volatility of
output gap and thus are quite successful in curbing the volatility of
the output gap. Under CPI Taylor rules (similar to domestic ination
Taylor rule), the economy takes more time to stabilize after a shock.
In our model framework, CPI Taylor rule results in huge volatility of
real exchange rate in line with Svensson (2000) observation that,
This might illustrate the danger of responding to a current
forward-looking variable, and provides support for the warning in
Woodford 23 (p.23).
6. Conclusions

better alternative from an overall macro stabilization perspective in

the present scenario where nancial markets are still not integrated
enough to ensure quick transmission of interest rate impulses (as
suggested by low sensitivity of demand to interest rate impulses)
and existence of rigidities in the economy (as indicated by at Phillips
curve).
There are two main limitations to the study that provide useful directions and scope for future research. One of the limitations relates
to the formulation of the model, in that the model is a simple three
equation model and foreign variables are not explicitly modeled.
The model is linear with quadratic loss function and there are different sources which can induce non-linearity in the model (like non
negative nominal interest rate or non linear Phillips curve). The
other limitation of the study is the application of general empirical
methodology. We note that the GMM method exploits only part of
the information implied by the model, while the currently more popular likelihood methods (with the Kalman lter) and Bayesian estimation can use all the implications of the DSGE model. It would be
very desirable to test the results of the model using the abovementioned alternative empirical approaches, and thus provides an interesting area for future research.
Appendix

Our small macro model of the Indian economy was built on the
results from the VAR model proposed by Mishra and Mishra
(2010) and based on literature on small open economy models. In
our model the exchange rate has a prominent role to play in both
the determination of the aggregate demand and supply equation of
the economy. Our model specication is different from other models
of open economies in that we assumed both the aggregate supply
and demand equation to be affected by the lag of (change in) exchange rate rather than its contemporaneous value. 24 Estimation results of aggregate supply justify the assumption of hybrid Phillips
curve for India and suggest that backward dynamics (lag of ination) are slightly more important than forward dynamics (expected
ination in the next period) to explain ination outcomes in India.
This further indicates the presence of rigidities 25 in price setting
behavior in the Indian economy resulting in short-run trade-off
between ination and output. We found that the supply curve in
India was atter compared to a mature small open economy and
that there was an excess capacity in the economy. We also found
that the interest rate elasticity of the aggregate demand is low.
Within this framework, the properties of strict vs. exible domestic and CPI ination targeting were examined and compared with the
Taylor rules (domestic and CPI Inations). The analysis highlights the
unsuitability of simple Taylor type monetary rules in stabilizing the
economy and suggests that discretionary optimization works better.
There appears to be a trade-off between output gap stabilization
and real exchange rate stabilization under domestic and CPI ination
targeting respectively. Flexible domestic ination targeting seems a

23
Woodford analyzed the budding issue in monetary policy literature that a desirable
monetary policy would be able to contain ination before much had developed. This
could be conducted by monetary indicators that are supposed to be good predictors
of future ination rather than by concentrating on variables (such as money supply)
that are thought to be probable causes of ination. In his analysis, Woodford suggested
that there were important advantages in nding indicators of the causes of ination
rather than of inationary expectations. For more details, refer to Woodford (1994).
24
The assumption is motivated from the results of Mishra and Mishra (2010) and
further veried by the estimations in this study where lag of exchange rate turned
out to be statistically signicant in both aggregate demand and supply equations.
25
The main sources of real rigidity in the Indian context relate to goods market and
labor market imperfections. Both goods and labor markets are characterized by dualistic consumers and labors respectively, where one type of consumers and labors are at
above subsistence level while others are at subsistence level (for more details refer to
Goyal, 2011).

A.1. Unit root tests

WPI
REER
RIR
CPI
2 CPI
Y
OIL
USIIP
2 USIIP
USCPI
M3
ffrate
f f rate
lrate5
lrate5
lrate10
lrate10
spread1
spread5

ADF test

PhillipsPerron test

3.020 (0.033)
3.99 (0.011)
5.495 (0.00)
1.992 (0.29)
7.513 (0.00)
8.679 (0.00)
3.016 (0.033)
2.160 (0.22)
10.33 (0.00)
2.689 (0.075)
2.622 (0.089)
0.689 (0.84)
5.874 (0.00)
1.815 (0.37)
10.791 (0.00)
1.705 (0.43)
11.009 (0.00)
3.656 (0.0048)
3.670 (0.0046)

2.958
3.614
5.290
2.477
7.304
8.799
2.538
1.749
10.525
2.72
2.616
1.088
5.985
1.836
10.779
1.710
11.009
3.118
3.517

(0.039)
(0.0055)
(0.00)
(0.12)
(0.00)
(0.00)
(0.10)
(0.41)
(0.00)
(0.071)
(0.089)
(0.72)
(0.00)
(0.36)
(0.00)
(0.43)
(0.00)
(0.0252)
(0.0076)

Notes:
1. ***, **,* indicate 1%, 5% and 10% signicance levels respectively.
2. Figures in parenthesis indicate Mackinnon p-value.
3. The number of lagged difference terms included in testing for each series has
been decided on the basis of no autocorrelation in the error terms for the ADF tests.
For PP tests lags have been selected on the basis of NeweyWest criterion.

A.2. List of excluded instruments in aggregate supply equation

Variable

Meaning

Y t1
T t2
RIR
REER
OIL
OILt1
OILt2
OILt4
2 CPI
2 CPI t1
USCPI
USCPIt1
2 USIIP
f f rate
f f ratet1
f f ratet2

Lag one of output gap

Lag two of output gap
Real interest rate
Log change in REER
Oil price ination
Lag one of oil price ination
Lag two of oil price ination
Lag four of oil price ination
Change in CPI ination
Lag one of change in CPI ination
United States (US) CPI ination
Lag one of us CPI ination
Change in Index of Industrial Production (IIP) growth rate in US
Change in federal funds rate (US)
Lag one of change in federal funds rate
Lag two in federal funds rate

A. Mishra, V. Mishra / Economic Modelling 29 (2012) 10531063

A.3. List of excluded instruments in aggregate demand equation

Variable

Meaning

Y t2
Y t4
spread1

Lag two of output gap

Lag four of output gap
Difference between yield on SGL transaction on government
security of 1 year maturity and 10 year maturity.
Difference between yield on SGL transaction on government security
of 5 year maturity and 10 year maturity.
Change in yield on SGL transaction on government security of 5 year
maturity
Lag one of change in yield on SGL transaction on government
security of 5 year maturity
Change in yield on SGL transaction on government security of
10 year maturity
Lag two of change in yield on SGL transaction on government
security of 10 year maturity
Lag one of real interest rate
Lag two of real interest rate
Log change in money supply (M3)
Lag two of log change in M3
Log change in REER
Lag two of log change in REER
WPI ination
Lag one of WPI ination
Lag one of change in federal funds rate
Lag two in federal funds rate

spread5
lrate5
lrate5t1
lrate10
lrate10t2
RIRt 1
RIRt 2
M3
M3t2
REER
REERt2
WPI
WPIt1
f f ratet1
f f ratet2

References
Ball, L., 1997. Efcient rules for monetary policy. NBER Working Paper, w5952.
Ball, L., 2000. Policy rules and external shocks. NBER Working Paper, w7910.
Batini, N., Haldane, A.G., 1999. Forward looking rules for monetary policy. In: Taylor,
John B. (Ed.), Monetary Policy Rules. University of Chicago Press.
Bernanke, B.S., Mishkin, F.S., Posen, A.S., 1999. Ination Targeting: Lessons from International Experience. Princeton University Press.
Cavoli, T., Rajan, Ramkishan S., 2005. Ination Targeting and Monetary Policy Rules for
Small and Open Economies: Simple Analytics with Application to Thailand. School
of Public Policy working papers, National University Singapore. Available at http://
www.spp.nus.edu.sg/docs/wp/wp0505b.pdf.
Clarida, R., Gali, J., Gertler, M., 1999. The science of monetary policy: a New Keynesian
perspective. Journal of Economic Literature 37 (4), 16611707.
Clarida, R., Gali, J., Gertler, M., 2001. Optimal monetary policy in closed versus open economies: an integrated approach. The American Economic Review 91 (2), 248252.

1063

D'Souza, E., 2003. What is monetary policy doing? Economic and Political Weekly 38 (8),
821823.
Fraga, A., Goldfajn, I., Minella, A., 2003. Ination targeting in emerging market economies. NBER Working Paper, w10019.
Gali, J., Monacelli, T., 2005. Monetary policy and exchange rate volatility in a small open
economy. Review of Economic Studies 72 (3), 707734.
Goyal, A., 2008a. Incentives from exchange rate regimes in an institutional context.
Journal of Quantitative Economics 6 (1&2), 101121.
Goyal, A., 2008b. The Structure of Ination, Information and Labour Markets: Implications for Monetary Policy. Available at www.igidr.ac.in/pdf/publication/WP-2008010.pdf.
Goyal, A., 2011. A general equilibrium open economy model for emerging markets: monetary policy with a dualistic labor market. Economic Modelling 28, 13921404.
Jha, R., 2008. Ination targeting in India: issues and prospects. International Review of
Applied Economics 22 (2), 259270.
Kannan, R., 1999. Ination targeting: issues and relevance for India. Economic and
Political Weekly 115122 January 1623.
Khatkhate, D., 2006. Ination targeting: much ado about something. Economic and
Political Weekly 50315033 December 9.
Mishra, A., Mishra, V., 2009. Preconditions for ination targeting in an emerging
economy: the case of India. DRU Working Paper No. DEVDP 0908. Development
Research Unit, Monash University.
Mishra, A., Mishra, V., 2010. A VAR model of monetary policy and hypothetical case of
ination targeting in India. Monash University Discussion Paper. Available at
http://www.buseco.monash.edu.au/eco/research/papers/2010/index.html. 1510.
Moron, E., Winkelried, D., 2003. Monetary policy rules for nancially vulnerable economies. IMF Paper No./03/39. .
Pagan, A.R., Hall, D., 1983. Diagnostic tests as residual analysis. Econometric Reviews 2 (2),
159218.
Rudebusch, G.D., Svensson, L.E.O., 1999. Policy rules for ination targeting. In: Taylor
John, B. (Ed.), Monetary Policy Rules. University of Chicago Press.
Shah, A., 2007. Righting the rupee. Asian Wall Street Journal (18 May), Online publication,
Available at http://www.mayin.org/ajayshah/MEDIA/2007/righting_the_rupee.html.
Singh, K., 2006. Ination targeting: international experience and prospects for India.
Economic and Political Weekly 29582960 July 815.
Svensson, L.E.O., 2000. Open-economy ination targeting. Journal of International
Economics 50 (1), 155183.
Taylor, John B., 2001. The role of the exchange rate in monetary-policy rules. The
American Economic Review 91 (2), 263267.
Walsh, Carl E., 1999. Monetary Policy Trade-offs in the Open Economy. Available at
http://econ.ucsc.edu/~walshc/exchNov99.pdf.
Walsh, Carl E., 2009. Ination targeting: what have we learned? International Finance
12 (2), 195233.
Woodford, M., 1994. Nonstandard indicators for monetary policy: can their usefulness
be judged from forecasting regressions? In: Mankiw Gregory, N. (Ed.), Monetary
Policy. University of Chicago Press, Chicago.
Woodford, M., 2003. Interest and Prices: Foundations of a Theory of Monetary Policy.
Princeton University Press.