The Empirical Study about Introduction of Stock Index Futures on the Volatility of Spot Market

doi:10.4236/ib.2013.53B024

Paper Menu >>

Journal Menu >>

iBusiness, 2013, 5, 113-117

http://dx.doi.org/10.4236/ib.2013.53B024 Published Online September 2013 (http://www.scirp.org/journal/ib) 113

The Empirical Study about Introduction of Stock Index

Futures on the Volatility of Spot Market

Guiliang Tian, Huixiangzi Zheng

School of Business, Hohai University, China.

Email: [email protected]

Received July,2013

ABSTRACT

April 16, 2010, China’ s first four Ind ex Future contr acts have been listed for trading in the stock exch ange. Stock index

futures are the world's fastest growing financial derivative products currently, and the research of them is of signifi-

cance to the development of China's financial market. Therefore, it is particularly important to concern the market op-

eration status of stock index future after its official listed. As we all know that there are linkage effects of the futures

and spot, stock index futures market will have some impact on the spot market. Based on the above point, this study

identified the Sh anghai and Shenzhen 300 stock index futures market as th e research object, focu sing on what volatility

affects the stock index futures have made on the spot market over the past 3 years. Through empirical findings, we can

evaluate the operation conditions of stock index futures market objectively. This paper applies GARCH model mainly,

and introduce dummy variables, collecting daily trading data between 16 April 2007 and 16 April 2013 on the spot

market, and studies the impact of volatility on the spot market in-depth by empirical test. The results showed that over

the past 3 years, introduction of stock index futures has reduced the volatility of the spot market which has brought a

positive impact.

Keywords: Index Future; Spot Market; Volatility; GARCH Model

1. Introduction

Stock index futures, which are based on prior agreement

between buyers and sellers, trade at a particular time in

the future, in accordance with the prior agreement of

both parties to engage in stock index trading shares of a

standardized protocol agreement. April 16, 2010, our

country's first financial derivatives - officially listed in

Shanghai and Shenzhen 300 stock index futures trading.

However, soon after its listing launched a vicious slump

appears unilateral market conditions, driving the stock

market also followed the collapse, and thus some schol-

ars have suggested that the stock index futures market

crash blame. Therefore, it is necessary to explore the

introduction of stock index futures on the volatility of the

spot market produce what kind of impact, in order to

clarify whether the stock index futures is the "culprit."

According to the impaction of stock index futures on

the spot market, scholars hold different views, and each

view has a lot of theoretical research and empirical re-

search support. Chan[1] on 1985-1987 Nikkei Stock Av-

erage daily trading data, with its day yield research ob-

ject, through the establishment of GARCH model, the

results found that the introduction of stock index futures

on the spot market is not significantly affected. Anton-

inus[2] based on FTSE 100 index tested a series of

econometric model, including unit root test, vector error

correction model, the establishment of such GARCH

model and found that the introduction of stock index fu-

tures, increased the volatility of the spo t prices stock, th at

is on the spot markets had a negative impact. Domestic

scholars Tian-cai Xing and Ge Zhang[3] based on FTSE

Xinhua A50 index futures, through the establishment of

GARCH model, analyzes its launch on the CSI 300 stock

index, the results found that the introduction of stock

index futures slightly increases the volatility of the spot

market, but this impact is very small. In 2010, they se-

lected the CSI 300 index futures trading simulation data,

through the establishment of GARCH model to test re-

sidual ARCH effects, etc., confirmed the launch of stock

index futures on the volatility of the spot market is not

greater impact. But because of the data from the simula-

tion trading system, therefore the reliability of the con-

clusions still needs to be fastidious.

From the existing research, the current domestic and

foreign scholars are mainly based on mature markets

abroad to study th e introdu ction of stock index futur es on

the spot market volatility impact on emerging markets

such like our lack of research, especially in China the

The Empirical Study about Introduction of Stock Index Futures on the Volatility of Spot Market

114

CSI 300 index futures launch date just over three years,

the research scholars are mostly simulation-based trans-

action data, and conclusions vary. Therefore, this paper

by choosing real transaction data were established befo re

and after the introduction of index futures GARCH

model, and the introduction of dummy variables, a series

of empirical tests to explore CSI 300 stock index futures

on the impact of volatility in the spot market, with certain

practical significance.

2. Research Methods

2.1. Sample Description

This data is selected April 16, 2007 to April 16, 2013 in

CSI 300 stock index futures contracts and the CSI 300

stock index daily closing price, that is listed on the CSI

300 stock index futures before and after the three years

of the corresponding stock the daily closing price of the

stock index of 1460 data. Among them, the futures data

for the selection of the main contract data, namely the

nearest month futures contracts of the daily closing price

data. In the most recent month before the contract set-

tlement date two or three days, month contract will be

converted into the main contract, and then turned to the

next nearest month futures contract closing data, and so

on, can be a continuous set of stock index futures price

time series. Data mainly comes from China Financial

Futures Exchange, and the News Finance website. In

accordance with the stock index futures launch time the

data is divided into two stages, the first stage before the

introduction of stock index futures, is from April 16,

2007 to April 16, 2010; second stage after the introduc-

tion of stock index futures, from April 16, 2010 to April

16, 2013.

2.2. Data Processing

Firstly, in order to eliminate heteroscedasticity, respec-

tively futures price and spot price series sequence loga-

rithm.

Assuming that the trading day closing price of a fu-

tures contract and the spot index closing price are de-

noted as t and t, then the natural logarithm of fu-

tures contract price series expressed as t

P Qln

P

ttt

, its

logarithmic rate of return is 1



 , the

natural logarithm of stock price series expressed as

, its logarithm yields expressed as

St

ln ln

SQQ



 t

The study object of this paper is the log return of stock

index future s and spot price rate.

2.3. Model Selection

Existence of financial time series data aggregation and

volatility heterosced asticity, OLS regression model is the

assumption that “homoscedasticity”, therefore, in the

study when the price volatility of financial products,

widely used heteroscedasticity model is the most widely

used in the 1982 Engle proposed autoregressive condi-

tional heteroskedasticity model (ARCH model), Boller-

slev deepened the ARCH model, the generalized autore-

gressive conditional heteroskedasticity model (GARCH

model), GARCH model can well reflect the aggregation

of price volatility, in the previous literature, the wide-

spread use of GARCH (1, 1) model reflects the aggrega-

tion of financial products.

This paper uses the GARCH (1, 1) model stock index

futures on the volatility of the spot market, and the in-

troduction of dummy variables in the conditional vari-

ance to reflect the CSI 300 stock index futures on the

impact of the stock market volatility, Before the intro-

duction of stock index futures, D = 0, after the introduc-

tion of stock index futures D = 1. The GARCH (1, 1)

model with the following :

The mean equation: ttt









The conditional variance equation:

11tt



 



 

The GARCH (1, 1) model was respectively established

for the different stages, the model does not add dummy

variables.

3. Empirical Tests

We need to know the basic statistical characteristics of

the return series of CSI 300 index before the empirical

test.

From the Figure 1, we can clearly see that this series

has the cluster effect which most economic sequence

possesses. It means that the current price effected by the

prices of the previous period, which we call the ARCH

effect.

3.1. The Volatility Analysis of the Spot Price

Three Years before the Launch of CSI 300

Index Futures

1) Do the ADF test of return series of the stock index

-.10

-.05

.00

.05

.10

250 500 7501000 1250 150

Figure 1. The return’s volatility of stock index in whole

sample period.

The Empirical Study about Introduction of Stock Index Futures on the Volatility of Spot Market 115

First we do the ADF test of return series of the stock

index, the test result is as follows:

From the Table 1, we can see that the stationary test

results are significant, thus rejecting the null hypothesis

that the return series is stationary. Then we establish

ARMA model.

2) Establish ARMA model

a) Determine the value of p and q

Before the establishment of ARMA model, we should

see autocorrelation and partial autocorrelation chart of

the daily return series of Shanghai and Shenzhen 300

index three years before the introduction of futures. The

figure appears in several places which autocorrelation

and partial autocorrelation functions are significant,

therefore we cannot determine the value of p and q di-

rectly. Then we draw support with the AIC and SC prin-

ciple by imitating ARMA model within a 5-order.

By comparing the value of AIC and SC, we can build

ARMA(2,3) model.

b) Build ARMA(2,3) model

The following table is the output of ARMA (2,3).

As we can see from Table 2, the results of model es-

timation are good. At the 1% confidence level, in addi-

tion to the constant term, the P value of all the coeffi-

cients are less than 1%, passing the test, R ^ 2 = 50.20%.

Value of DW is around 2 which has no autocorrelation.

At this point, we remove the constant term. The expres-

sion of ARMA (2,3) model is as followings:

Table 1. The ADF test results.

t-Statistic Prob.*

ADF test -26.3974 0.0000

1% level -2.5681 0.0000

5% level -1.9412 0.0000

Significance

level

10% level -1.6164 0.0000

Table 2. Output of ARMA (2,3).

Variable Coefficient Std. Error t-Statistic Prob.

C -5.72E-06 1.31E-05 -0.4368 0.6624

AR(1) -1.0280 0.0086 -118.93 0.0000

AR(2) -0.9681 0.0086 -112.03 0.0000

MA(1) 0.0701 0.0074 9.4621 0.0000

MA(2) -0.0499 0.0078 -6.4055 0.0000

MA(3) -0.9817 0.0066 -149.81 0.0000

11112

1.0280 0.96810.0701

0.04990.9817

ttt

RRR







 

 (1)

c) Do the ARCH Tests of ARMA (2,3) model

Before the establishment of GARCH model, we need

to test the ARCH effect of residuals series. We use the

method of ARCH-LM test. The result is as follows:

The result of ARCH-LM test is significant, indicating

that we can establish GARCH (1,1) model.

3) Build GARCH model

a) Build ARMA(2,3)-GARCH(1,1) model

The specific form of ARMA (2,3)-GARCH (1,1)

model is as follows:

Mean equation:

11112

1.0053 0.97450.0342

0.0 1 5 10 .9 8 2 4

ttt







 



(2)

Variance equation:

0.00000538 0.06340.9279











(3)

As we can see from Figure 2: the current yield of the

spot index will be effected by yield of its lagged one pe-

riod at a strength of -1.0053, by yield of its lagged two

period at a strength of -0.97 45. The current vo latility will

be affected by the volatility of previous period at a

strength of 0.9279

b) Do the ARCH-LM test of residuals series.

Table 3. ARCH-LM test of residuals series.

F-statistic 4.7359 Probability 0.0002

Obs*R-squared 23.1165 Probability 0.0003

Table 4. Output of GARCH model.

CoefficientStd. Error z-StatisticProb.

Mean equation

AR(1) -1.0053 0.0060 -167.76 0.0000

AR(2) -0.9745 0.0052 -186.64 0.0000

MA(1) 0.0342 0.0042 8.2244 0.0000

MA(2) -0.0151 0.0061 -3.4693 0.0035

MA(3) -0.9824 0.0033 -295.86 0.0000

Variance equation

C 5.38E-06 2.74E-06 2.4586 0.0302

RESID(-1)^20.0634 0.0126 5.0356 0.0000

GARCH(-1)0.9279 0.0135 68.6766 0.0000

The Empirical Study about Introduction of Stock Index Futures on the Volatility of Spot Market

116

Table 5. Output of ARCH-LM test.

F-statistic 0.3319 Probability 0.8938

Obs*R-squared 1.6697 Probability 0.8927

The value of P is 0.8938, thus the ARCH effect of re-

siduals no longer exist, which indicates that ARMA

(2,3)-GARCH (1,1) model is fully fitted at the return

series of CSI 300 stock index.

3.2. The Volatility Analysis of the Spot Price

Three Years after the Launch of CSI 300

Index Futures

This section of the concrete steps is as above, so we will

not repeat. Establish GARCH model of returns on Spot

Index three years after the introduction of index futures

like follows:

Mean equation:

0.5013





 (4)

Variance equation:

0.000014 0.04770.9036







 





(5)

As we can see: the current yield of the spot index will

be effected by yield of its lagged one period at a strength

of -0.5013. The current volatility will be effected by the

volatility of previous period at a strength of 0.9036.

3.3. Differences Analysis

Firstly, we draw the volatility chart of spot price in the

whole sample period.

From the figure, we can easily see that the volatility of

stock price has weakened three years after the introduc-

tion of stock index futures. So we can preliminary judge

that the introduction of stock index futures reduces the

volatility of the spot market.

Next, we will build a GARCH model based on the

whole sample interval, and introduce dummy variables.

Do positive and negative test of dummy variable coeffi-

cients, and see if it passed the examination to verify

whether the introduction of stock index futures reduces

the volatility of the spot market. Establishing revise

GARCH model likes follows:

Mean equation:

0.5431 0.27870.2774

0.76280.0499

ttt

RRR





 



(6)

Variance equation:

0.00000698 0.0345

0.95810.00000378











 (7)

Figure 2. Volatility of whole sample period.

As we can see: the current yield of the spot index will

be effected by yield of its lagged one period at a strength

of -0.5431, by yield of its lagged two period at a strength

of 0.2787. The current volatility will be affected by the

volatility of previous period at a strength of 0.9581.

The most important, the coefficient of dummy variable

is negative, and passed the test, indicating that the intro-

duction of stock index futures reduces the volatility of

the spot market, which made a positive impact on the

spot market.

4. Conclusions and Recommendations

This article studies the impact of the introduction of

stock index futures on the volatility in the spot market by

establishing GARCH (1,1) model, and the introduction of

dummy variables, and found the stock index futures has a

slight decrease on the volatility of spot market volatility.

In order to promote the development of the stock index

futures market, reduce the negative impact to a minimum

of the future futures on the spot market may, we propose

the following comments and suggestions, mainly dis-

cussed from the two aspects include continue to improve

the spot market and strengthen the risk supervision and

management of the futures market.

4.1. Improve the Spot Market Continually

The development of stock index futures markets takes

the developed spot market as a precondition. Although

China's stock market has been so close to the interna-

tional advanced level in the technical facilities, forming

the aspect of market size, marketing supervision system,

information disclosure is also far behind, therefore, we

should strengthen the standardization of the spot market,

and promote the sharing and coordination management

of the stock market and stock index futures market.

4.2. Strengthen Risk Oversight and Management

of the Futures Market

Volatility is an important indicator to measure market

The Empirical Study about Introduction of Stock Index Futures on the Volatility of Spot Market

117

risk, the size of volatility determines the size of the mar-

ket, although the empirical test the introduction of stock

index futures in our country did not increase the volatility

of the spot market, because of the stock index futures

itself has a larger market risk, in order to continue to

promote the development of the futures market, we still

need to strengthen risk management of the stock index

futures. Combined with the actual situation of China's

market, we should mainly make the following recom-

mendations: A. Perfect market investors, strengthen in-

stitutional investors, leading the reasonable and health

investment philosophy in the market. For now, mainly

take the insurance companies, fund companies, securi-

ties companies as the main body of hedging, and en-

courage the arbitrage behavior of the market, reduce sys-

temic risk along with promote the price discovery func-

tion of futur es markets yield well and effectively. B. Im-

prove the management of various accounts. Since the

futures implement the policy o f margin, it is necessary to

strengthen the supervision of the trading margin and

clearing margin, implement the daily free debt settlement

system. At the same time, we also need to continue to

strengthen the management of the large positions system

and forced open system as well as take effectively care of

risk reserves, etc., and adopt international advanced

real-time monitoring technology to effectively circum-

vent the stock index futures market risks. C. Timely and

targeted introduce the institutional arrangements, im-

prove relevant laws and regulations, make the regulatory

abilities and adhere adapt the speed of development in

order to laid a good foundation for the entire market’s

healthy and stable development.

5. Acknowledgements

This study was funded by the National Natural Science

Foundation of China (Grant No. 41001377, 51279058)

REFERENCES

[1] Chan and Karolyi, “Introduction Volatility in the Stock

Index and Stock Index Futures Review of Financial Stud-

ies,” Vol. 4, No. 3, 1991, pp. 657-684.

[2] Antoninus, “Return and Volatility Dynamics in the

FT-SE 100 Stock Index and Stock Index Futures Mar-

kets,” Journal of Futures Markets, Vol. 15, No. 4, 1995,

pp. 457-488.doi:10.1002/fut.3990150405

[3] Xing Tiancai and Zhang Ge, “The Empirical Study about

Impact of stock index futures on the spot market - Based

on the analysis of the FTSE Xinhua A50,” Research on

Financial and Economic Issues, Vol. 4, No. 7, 2009, pp.

83-88.