Sie sind auf Seite 1von 39

Regression

Analysis
ANURAG MISHRA
Regression

Dependent variable

Independent variable (x)

Regression is the attempt to explain the variation in a dependent variable using


the variation in independent variables.
Regression is thus an explanation of causation.
If the independent variable(s) sufficiently explain the variation in the dependent
variable, the model can be used for prediction.
Simple Linear Regression

Dependent variable (y)


y’ = b0 + b1X ± є
є

B1 = slope
= ∆y/ ∆x
b0 (y intercept)

Independent variable (x)

The output of a regression is a function that predicts the dependent variable


based upon values of the independent variables.

Simple regression fits a straight line to the data.


Simple Linear Regression

Observation: y

Dependent variable Prediction: y^

Zero
Independent variable (x)

The function will make a prediction for each observed data point.
The observation is denoted by y and the prediction is denoted by y. ^
Simple Linear Regression

Prediction error: ε

Observation: y
Prediction: y^

Zero

For each observation, the variation can be described as:

y=y+ε ^

Actual = Explained + Error


Regression

Dependent variable

Independent variable (x)


A least squares regression selects the line with the lowest total sum of squared
prediction errors.
This value is called the Sum of Squares of Error, or SSE.
Calculating SSR

Dependent variable Population mean: y

Independent variable (x)

The Sum of Squares Regression (SSR) is the sum of the squared differences
between the prediction for each observation and the population mean.
Regression Formulas

The Total Sum of Squares (SST) is equal to SSR + SSE.

Mathematically,

SSR = ∑ ( y – y ) ^(measure
2
of explained variation)

SSE = ∑ ( y – y ) ^
(measure of unexplained variation)

2
SST = SSR + SSE = ∑ ( y – y ) (measure of total variation in y)
The Coefficient of Determination

The proportion of total variation (SST) that is explained by the regression (SSR) is
known as the Coefficient of Determination, and is often referred to as R .
2

R = =
2 SSR SSR SST
SSR + SSE

The value of R can range between 0 and 1, and the higher its value the more
2
accurate the regression model is. It is often referred to as a percentage.
Standard Error of Regression

The Standard Error of a regression is a measure of its variability. It can be used in


a similar manner to standard deviation, allowing for prediction intervals.

y ± 2 standard errors will provide approximately 95% accuracy, and 3 standard


errors will provide a 99% confidence interval.

Standard Error is calculated by taking the square root of the average prediction
error.

SSE
Standard Error =
√ n-k

Where n is the number of observations in the sample and


k is the total number of variables in the model
The output of a simple regression is the coefficient β and the constant A. The
equation is then:

y=A+β*x+ε

where ε is the residual error.

β is the per unit change in the dependent variable for each unit change in the
independent variable. Mathematically:

∆y
β=
∆x
Multiple Linear Regression

More than one independent variable can be used to explain variance in the
dependent variable, as long as they are not linearly related.

A multiple regression takes the form:

y = A + β X + β1 X1 + … + β k Xk + ε
2 2

where k is the number of variables, or parameters.


Multicollinearity

Multicollinearity is a condition in which at least 2 independent variables are


highly linearly correlated. It will often crash computers.

Example table of
Correlations
Y X1 X2
Y 1.000
X1 0.802 1.000
X2 0.848 0.578 1.000

A correlations table can suggest which independent variables may be


significant. Generally, an ind. variable that has more than a .3 correlation with
the dependent variable and less than .7 with any other ind. variable can be
included as a possible predictor.
Nonlinear Regression

Nonlinear functions can also be fit as regressions. Common choices include


Power, Logarithmic, Exponential, and Logistic, but any continuous function
can be used.
Regression Output in Excel

SUMMARY OUTPUT

Regression Statistics
Multiple R 0.982655 Y = B0 + B1 X1 + B2X2 + B3X3 - - - +/- Error
R Square 0.96561 Total = Estimated/Predicted +/- Error
Adjusted R Square 0.959879
Standard Error 26.01378
Observations 15

ANOVA
df SS MS F Significance F
Regression 2 228014.6 114007.3 168.4712 1.65E-09
Residual 12 8120.603 676.7169
Total 14 236135.2

Coefficients
Standard Error t Stat P-value Lower 95%Upper 95%
Intercept 562.151 21.0931 26.65094 4.78E-12 516.1931 608.1089
Temperature -5.436581 0.336216 -16.1699 1.64E-09 -6.169133 -4.704029
Insulation -20.01232 2.342505 -8.543127 1.91E-06 -25.1162 -14.90844

Estimated Heating Oil = 562.15 - 5.436 (Temperature) - 20.012 (Insulation)


Risks Faced by a Bond Investor
Default risk
Interest rate risk (price risk)
Reinvestment risk
Call risk
Inflation risk
Foreign exchange risk
Liquidity risk

16
Rating
Category Moody’s S&P
------------------------------------------
High Grade Aaa AAA
Aa AA
-------------------------------------------
Investment A A
Grade Baa BBB
-------------------------------------------
Speculative Ba BB
B B
-------------------------------------------
Default Caa CCC
Ca CC
C C
D

17
Interest Rate Risk
Example: Two bond issues of ABC Co.
N1=1 yr N2= 10 yrs r = 5%

Bond Value
Market Rate of First Issue: Second Issue:
Interest N = 1 yr N = 10 yrs
5% 100.00 100.00
6% 99.06 92.64
7% 98.13 85.95
8% 97.22 79.87

As term to maturity increases, value of the


bond becomes more sensitive to movements
in market interest rate.

18
Bond Value and Coupon Rates
Example:Two issues of ABC Co. n=20 yrs, r1=10%, r2=6%

Market Bond 1 Percent Bond 2 Percent


Interest Rate R=10% change R=6% change
8% 119.64 80.36
9% 109.13 -8.78% 72.61 -9.64%
10% 100.00 -8.36% 65.95 -9.17%
11% 92.04 -7.96% 60.18 -8.75%
12% 85.06 -7.58% 55.18 -8.31%

Low coupon bonds are more sensitive to changes in market interest


rates

19
Value of a Bond in Time
Example: Market rate stays at 10%, values of
two bonds with coupon rates of 8% and 12%
as the term to maturity approaches:
Maturity Bond 1 Bond 2
R=8% R=12%
5 92.42 107.58
4 93.66 106.34
3 95.03 104.97
2 96.53 103.47
1 98.18 101.82
0 100.00 100.00

Assuming that interest rates remain the


same,
bond value approaches to par over time as
term to maturity shortens.

20
Term Structure of Interest Rates
Relationship between yield and time to
maturity.

Example: n=1 i=6%


n=5 i=8%
n=20 i=9%

Yield Curve

Maturity

21
Possible Explanations of the Term Structure
1. Expectations Hypothesis

1 + in =[(1+ i1)(1+ 1i2)…….(1+n-1 in)]1/n

Example: i2=8% i1=6% 1i2=?

1 + 0.08 = [(1+ 0.06)(1+ 1i2)]1/2


1i2 = 0.1004 or 10%

2. Liquidity Preference Hypothesis

Slope of the yield curve is higher than


specified in expectations hypothesis

3. Segmented Markets Hypothesis

22
Duration
Volatility in bond price is directly proportional
to term to maturity but inversely proportional
to coupon payments. Duration of a bond is a
measure that incorporates both factors that
affect volatility.

n
(t )Ct
D V
t 1 (1  i )
t

23
Duration Example
n=5 yrs, r=8%, i=10%
(1) (2) (3) (4) (5) (6)
Year PMT PVIF (2)x(3) (4)/V (1)x(5)
1 8 0.9091 7.27 0.0787 0.0787

2 8 0.8264 6.61 0.0715 0.1430

3 8 0.7513 6.01 0.0650 0.1950

4 8 0.6830 5.46 0.0591 0.2364

5 108 0.6209 67.06 72.57 3.6284

Total 92.41 4.28

Bond Value = $92.41


Macaulay Duration = 4.28 years

24
Hedging Interest Rate Risk
$12 $12 $12 $12 $12 $12 $112
|___|____|____|____|____|...…..|___ |
0 1 2 3 4 5 9 10

V0=$84.94 when i=15%

After i declines to 12%, V = $100


V when term to maturity is 4 years:
V6 = $100

Future value of the first 6 coupon payments


reinvested at 12%:
12 x PVIFA 12%,6 = $97.38
Total savings = $100 + $97.38 = $197.38

$84.94 in 6 years grows to $197.38


Annual growth of 15%.
25
Immunization Example
$1,000 $2,000 $2,500 $2,000 $1650
|_____|______|______|______|______|
0 1 2 3 4 5

Total Premiums = Assets = $6,830.82


Market rate = 10% Flat yield curve

Strategy 1: Invest in 1-yr bills with 10% interest

6830.82 -> 7513.90


(1000.00)
6513.90 --> 7165.29
(2000.00)
5165.29 --> 5681.82
(2500.00)
3181.82 ->3500
(2000)
1500 ->1650
(1650)
26
Immunization Example (Cont’d)
However, if interest rates fall, assets will
be short of liabilities

Strategy 2: Invest in 3-yr zero coupon bonds


yielding 10%

Duration of Liabilities:

1 1000 909.09 0.133 0.133


2 2000 1652.89 0.242 0.484
3 2500 1878.29 0.275 0.825
4 2000 1366.03 0.200 0.800
5 1650 1024.52 0.150 0.750
2.990

Duration = 2.99 years


27
Immunization Example (Cont’d )
Market rate 10%, V = $6,830.82
M = $9,091.82 Duration = 3 years

If interest rates fall from 10% to 8%,


V= $9,091.82 x PVIF 8%,3 = $7,217.38

7217.38 ->7794.77
(1000.00)
6794.77 ->7338.35
(2000.00)
5338.35->5765.42
(2500.00)
3265.42->3526.66
(2000.00)
1526.66->1650
(1650)

28
Modified Duration
D
MD = -----------
(1 + i)

In the example above, MD = 4.28/1.10 = 3.89

Approximate Change in V = -MD x Change in


yield

Example:
If the yield decreases from 10% to 8%

% Change in V= -4.28 x (-2) = 8.56%

In fact when i=10% V = $92.41


i=8% V = $100 increase 8.21%

29
Convexity
Price-Yield Relationship

Yield

The shape of the curve depends on


the coupon rate and term to maturity

High coupon + Short term -----> Linear


Low coupon + Long term ------> Convex

30
Convexity (Cont’d )
Higher convexity means that when interest
rates go up, bond value declines slowly; but
when rates decline, increase in bond price is
large

Therefore high convexity is a desirable


feature.

Factors that increase convexity:

* Low coupon
* Long term to maturity
* Low yield

31
Convexity (Cont’d )
d 2V
Convexity  2 V
di
d 2V 1 n
Ct
2 
 (t 2
 t)
di 2
(1  i) t 1 (1  i) t

(1) (2) (3) (4) (5)


Year Ct PVIF(8%,n) (1) x (2) t2 + t (3) x (4)
1 8 0.9091 7.27 2 14.55
2 8 0.8264 6.61 6 39.67
3 8 0.7513 6.01 12 72.13
4 8 0.6830 5.46 20 109.28
5 108 0.6209 67.06 30 2011.79
92.42 2247.41

Convexity = [1/(1.10)2][2247.41][1/92.42]
= 20.10
Appox. Change in V = -MD x i + K x (i)2

32
Alternative Measures of
Yield
Current Yield = rM / V
Yield-to-maturity
◦ Bond is held until maturity
◦ All coupon and principal repayments are made on time
◦ Bond is not called before maturity
◦ Coupon payments are reinvested at yield-to-maturity

Yield-to-call
Holding period yield

Vt+1 - Vt + rM
HPY = --------------------
Vt

33
Approximate yield-to-maturity
M V
rM 
i n
V M
2
Example V= $877.11 n=3 yrs r=8% M=$1000

1000  877.11
80 
i 10  0.0983
877.11  1000
2

34
Bond Investment Strategies
I. Passive Strategies

Investing $100 in 1925


T-bill
Deposits
Stock Market
AAA Corporate Bonds
Gold
Inflation

Passive Strategies are better when:


Interest rate risk is low, and
Inflation is low and stable

35
II. Active Strategies

Strategies based on maturity structure


◦ Maturity matching - duration
◦ Spreading the maturity
◦ Investing only in short term bills and long term bonds

Strategies based on forecasting interest


rate movements
◦ Interest rate fluctuations
◦ Buy when rates are high, sell when low
◦ Increase duration if higher rates are forecast, reduce duration otherwise

36
- Riding the yield curve
Investing in bonds assuming that the
yield curve will not shift

i
A
B

Maturity

Eg. 1 year bill i=6% V1 = $943.40 B


2 year zero coupon i=8% V2 = $857.34 A

Buy the 2-year bond at $857.34, sell it next year


at $943.40

HPY = (943.40 - 857.34) / 857.34 = 10.04%


37
Strategies based on lack of
market efficiency
Junk bonds
Bond swaps
◦ Yield swap : same coupon, rating, maturity and industry, different yield
◦ Exchange swap: same rating, maturity, industry, yield, different coupon.
Exchange current yield for capital gains
◦ Tax swap: Selling a bond to realize a loss, and replacing it with a similar bond
◦ Swapping bonds with different tax status: eg. AAA corporate bond vs.
municipal bond

38
Strategies based on lack of market efficiency (cont’d)
Possible shortcomings of bond swaps:
◦ time to execute the swap
◦ taxes
◦ transaction costs
◦ risk level of bonds

Portfolio rebalancing: adjusting the bond portfolio for


the changes in market conditions

39

Das könnte Ihnen auch gefallen