Analysis of NIFTY Futures

Contents`
Introduction...........................................................................................................................................3
Futures Pricing.......................................................................................................................................3
Nifty Futures Explained..........................................................................................................................5
Analysis..................................................................................................................................................6
NIFTY INDEX Futures expiring in September’2010.................................................................................6
t-statistic (p-value).............................................................................................................................6
Adjusted R-square – 99.4%................................................................................................................6
Time Series: Test for Stationary.........................................................................................................6
Comparison of Actual, Theoretical and Regression Model Results..................................................11
NIFTY INDEX Futures expiring on 26th August’2010.............................................................................13
t-statistic (p-value)...........................................................................................................................13
Adjusted R-square – 99.4%..............................................................................................................13
Time Series: Test for stationary.......................................................................................................13
Comparison of Actual, Theoretical and Regression Model Results..................................................17

MDI, Gurgaon Page 1

MDI, Gurgaon Page 2
Introduction
Futures are standardized financial instruments in which two party form a contract to sell and
buy the specified asset in the specified standardized quantity and quality at a future date on a
price that is agreed upon today. This price is called the future price of the asset for the given
period.

These securities are exchange traded and the underlying asset for these securities may vary
from the traditional commodities to financial assets like the currency, securities etc. or it can
be even a reference like equity index or the interest rates. In futures contract both parties are
under the obligation to honor the contract. Since these are the exchange traded securities
these are safe form counter party risk as the exchange act as the counter party for both the
buyer and the seller and at the settlement date it can be settled either through the delivery of
the asset or through the cash settlement. These contracts are marked to margin to remove any
risk for default from the either party.

Futures Pricing
The pricing of futures are determined by the equilibrium between the demand and supply
among the buyers and sellers at the time of trading the contract.

There are different arguments when it comes to the pricing of the futures contract. If the
supply of the underlying asset is in plentiful or can be created freely than the futures are set to
be priced through the arbitrage argument. When the asset are limited then the demand and
supply is the only force that is setting the price of these contracts. So the forward price
represents the expected future value of the underlying discounted at the risk free rate. As any
deviation from the theory will lead to the arbitrage and create the opportunity to create a
riskless profit.

MDI, Gurgaon Page 3

Thus, for a simple, non-dividend paying asset, the value of the future/forward, F(t), will be
found by compounding the present value S(t) at time t to maturity T by the rate of risk-free
return r.

Or, with continuous compounding

When the deliverable commodity is not in plentiful supply (or when it does not yet exist)
rational pricing cannot be applied, as the arbitrage mechanism is not applicable. Here the
price of the futures is determined by today's supply and demand for the underlying asset in
the futures.

In this Study we have tried to use this model to study and see how closely the actual pricing
of a nifty future index follows the Arbitrage argument of rational pricing.

We have taken ln(F(t)) as the dependent variable and the ln S(t) and ln(r(T-t)) as the
independent variable.

Where F (t) = Future price at the time t

S (t) = spot price at time t

T-t = Time to maturity

MDI, Gurgaon Page 4

Nifty Futures Explained

MDI, Gurgaon Page 5

Analysis

NIFTY INDEX Futures expiring in September’2010

As per Regression analysis, it can be seen that value of coefficients of ln_x (which is log of
Spot Price) and RT (product of interest rate and time left to maturity) is 0.999969 and
0.359943.

t-statistic (p-value)
The p-value of the coefficients is less than 0.05 which shows that they are significantly
different from zero.

Adjusted R-square – 99.4%

This basically shows that 99.4% of the variation in the dependent variable (ln_f i.e. the log of
price of futures price of Nifty Index). 99.4% is a very significant value which means that
almost all variables responsible for change in value of future’s price have been considered.

MDI, Gurgaon Page 6

Time Series: Test for Stationary

Above figure shows the ADF test conducted for checking the whether the Time-series data is
necessary that has been used. As we can see the p-value is less than 0.05 which shows that
the hypothesis that Time-series data is stationary has a unit root can be rejected and hence,
the time-series is stationary.

Assumption 1: Mean of error

Mean of residual was a compared to zero and p-value was calculated.

p-value = 0.9985

This shows that Null hypothesis cannot be rejected that mean of error term is zero, which
basically shows that error term is not significantly different from zero. So, this assumption is
satisfied.

MDI, Gurgaon Page 7

Assumption 2: Normality

The p-value was calculated by conducting Histogram-Normality test on residuals, the p-

value was 0.531 which means that the null hypothesis of that error term is normality
distributed cannot be rejected.

MDI, Gurgaon Page 8

Assumption 3 – Hetroscedasticity

Here, p-value > 0.05, so it cannot be said that the regression estimates are Homoscedastic in
nature. i.e. with change in observations, the variations in the error term should not change
and should remain constant and also does not have any tendency of any kind of trend.

Assumption 4: Autocorrelation

Durbin-Watson stat

For 43 observations

d l=1 . 430

d u=1 . 615

In this case, d = 1.95

d u <d < 4−d u

This means that we cannot reject the null hypothesis that there is No positive/negative
autocorrelation.

This shows that autocorrelation does not exist in this case.

MDI, Gurgaon Page 9

Removing Hetroscedasticity

Solving the regression equation using White test so as to remove hetrscedasticity, we get

There is a no change in values of coefficients. Only change that is there is in value of

individual standard errors.

Equation after satisfying all assumptions:

ln ( f )=β 1 ln ( S )+ β2 (r∗t)

ln ( f )=0 . 999969∗ln ( S ) +0 .359943∗β 2 (r∗t )

Which basically shows that the value of future’s is dependent on both the Spot Price as well
as product of interest rate, but in actual scenario the model that is followed by NIFTY index
future depends on various market forces like trends (bullish/bearish) as well as costs like
brokerage cost.

MDI, Gurgaon Page 10

Comparison of Actual, Theoretical and Regression Model Results

5600

5550

5500

5450

5400
Futures Price
Regressed Value
5350
Theoretical Value
5300

5250

5200
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
5 /1 0/1 5/1 8/1 3/1 6/1 1/1 6/1 9/1 3/1 6/1 1/1 6/1 9/1 4/1
2 3 7/ 7/ /1 /1 /2 /2 /2 8/ 8/ /1 /1 /1 /2
6/ 6 / 7 7 7 7 7 8 8 8 8

As we can see the values obtained as per theoretical model are higher when compared to
those with regression model. The reason for this can be explained by saying that theoretical
model doesn’t take care of various market trends and other costs like brokerage costs
involved while entering into the transaction

MDI, Gurgaon Page 11

Analysis

Dependent Variable: LN_F

Method: Least Squares

Date: 08/25/10 Time: 15:30

Sample: 5/28/2010 8/24/2010

Included observations: 63

Variable Coefficient Std. Error t-Statistic Prob.  

LN_S 1.000216 7.10E-05 14081.81 0.0000

RT -0.353349 0.119256 -2.962948 0.0043

R-squared 0.993965     Mean dependent var 8.578898

Adjusted R-squared 0.993867     S.D. dependent var 0.029564

S.E. of regression 0.002315     Akaike info criterion -9.267311

Sum squared resid 0.000327     Schwarz criterion -9.199275

Log likelihood 293.9203     Durbin-Watson stat 0.622251

MDI, Gurgaon Page 12

NIFTY INDEX Futures expiring on 26th August’2010

As per Regression analysis, it can be seen that value of coefficients of ln_x (which is log of
Spot Price) and RT (product of interest rate and time left to maturity) is 01.000216 and
-0.353349.

t-statistic (p-value)
The p-value of the coefficients is less than 0.05 which shows that they are significantly
different from zero.

Adjusted R-square – 99.4%

This basically shows that 99.4% of the variation in the dependent variable (ln_f i.e. the log of
price of futures price of Nifty Index). 99.4% is a very significant value which means that
almost all variables responsible for change in value of future’s price have been considered.

Time Series: Test for stationary

MDI, Gurgaon Page 13

Above figure shows the ADF test conducted for checking the whether the Time-series data is
stationary that has been used. As we can see the p-value is less than 0.05 which shows that
the hypothesis that Time-series data is non-stationary has a unit root can be rejected and
hence, the time-series is stationary.

Assumption 1: Mean of error

Mean of residual was a compared to zero and p-value was calculated.

p-value = 0.9985

This shows that Null hypothesis cannot be rejected that mean of error term is zero, which
basically shows that error term is not significantly different from zero. So, this assumption is
satisfied.

Assumption 2: Normality

MDI, Gurgaon Page 14

The p-value was calculated by conducting Histogram-Normality test on residuals, the p-
value was 0.294 which means that the null hypothesis of that error term is normality
distributed cannot be rejected.

Assumption 3 – Hetroscedasticity

Here, p-value < 0.05, so it can be said that the regression estimates are Homoscedastic in
nature. i.e. with change in observations, the variations in the error term will not change and
remain constant and also does not have any tendency of any kind of trend.

Assumption 4: Autocorrelation

Durbin-Watson stat

For 63 observations

d l=1 .51

d u=1 . 65

In this case, d = 1.507

d <d l

MDI, Gurgaon Page 15

This means that we can reject the null hypothesis that there is No positive/negative
autocorrelation. SO, there is some correlation which has to be removed.

Removing Autocorrelation

Solving the regression equation using Newey West test so as to remove auto-correlation, we
get

Dependent Variable: LN_F

Method: Least Squares

Date: 08/25/10 Time: 15:46

Sample: 5/28/2010 8/24/2010

Included observations: 63

Variable Coefficient Std. Error t-Statistic Prob.  

LN_S 1.000216 7.10E-05 14081.81 0.0000

RT -0.353349 0.119256 -2.962948 0.0043

R-squared 0.993965     Mean dependent var 8.578898

Adjusted R-squared 0.993867     S.D. dependent var 0.029564

S.E. of regression 0.002315     Akaike info criterion -9.267311

Sum squared resid 0.000327     Schwarz criterion -9.199275

Log likelihood 293.9203     Durbin-Watson stat 0.622251

There is a no change in values of coefficients. Only change that is there is in value of

indivisual standard errors.

Equation after satisfying all assumptions:

ln ( f )=β 1 ln ( S )+ β2 (r∗t)

ln ( f )=1.000216∗ln ( S ) +0 . 353349∗β 2 (r∗t)

Which basically shows that the value of future’s is dependent on both the Spot Price as well
as product of interest rate, but in actual scenario the model that is followed by NIFTY index
future depends on various market forces like trends (bullish/bearish) as well as costs like
brokerage cost.

MDI, Gurgaon Page 16

Comparison of Actual, Theoretical and Regression Model Results

5600

5500

5400

5300

Calculated
5200
Theoretical
Actual

5100

5000

4900
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
2 0 1 20 1 20 1 2 01 20 1 20 1 20 1 2 01 2 01 20 1 20 1 2 01 2 01 2 01 20 1 2 01 2 01 2 01 20 1 20 1 2 01
5/ 6/ 6/ 6 / 6/ 6/ 6/ 6 / 7 / 7/ 7/ 7 / 7 / 7 / 7/ 7 / 8 / 8 / 8/ 8/ 8 /
8 /0 2/0 7/0 0/0 5/0 8/0 3/0 8/0 1/0 6/0 9/0 4/0 9/0 2/0 7/0 0/0 4/0 9/0 2/0 7/0 0/0
2 0 0 1 1 1 2 2 0 0 0 1 1 2 2 3 0 0 1 1 2

As we can see the values obtained as per theoretical model are higher when compared to
those with regression model. The reason for this can be explained by saying that theoretical
model doesn’t take care of various market trends and other costs like brokerage costs
involved while entering into the transaction.

MDI, Gurgaon Page 17