Sự xuất hiện của hợp đồng tương lai chỉ số và sự biến động của thị trường: Nghiên cứu thực nghiệm tại Việt Nam

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

INDEX FUTURES INTRODUCTION AND STOCK MARKET

VOLATILITY: EMPIRICAL STUDY IN VIETNAM

SỰ XUẤT HIỆN CỦA HỢP ĐỒNG TƯƠNG LAI CHỈ SỐ

VÀ SỰ BIẾN ĐỘNG CỦA THỊ TRƯỜNG: NGHIÊN CỨU

THỰC NGHIỆM TẠI VIỆT NAM

MA, Nguyen Ngoc Tram

National Economics University

tramnn@neu.edu.vn

Abstract:

This paper aims at answering the question whether the VN30 index futures introduction

has an impact on stock market volatility in Vietnam. Apply GARCH model of volatility with ad-

ditive dummy variable from 28/7/2000 to 10/9/2020, the result shows that when the first listed

index futures contract appears, it makes the volatility of VNIndex increases. The result is still ro-

bust after excluding the turmoil period of Vietnam stock market. This paper implies that policy

maker should be more careful in promoting derivatives market in Vietnam.

Keyword: GARCH, VN30 index futures, Vietnam

Tóm tắt

Bài báo trả lời câu hỏi liệu rằng sự xuất hiện của hợp đồng tương lai chỉ số VN30 có tác

động đến biến động trên thị trường cơ sở tại Việt Nam hay không. Áp dụng mô hình GARCH với

biến giả trong giai đoạn từ 28/7/2000 đến 10/9/2020, kết quả nghiên cứu cho thấy khi hợp đồng

tương lai chỉ số niêm yết đầu tiên xuất hiện, nó khiến cho biến động của chỉ số VNIndex gia tăng.

Kết quả nghiên cứu vẫn vững chắc khi loại giai đoạn biến động của thị trường ra khỏi mẫu

nghiên cứu. Bài báo hàm ý rằng những nhà hoạch định chính sách cần cận trọng khi thúc đẩy

sự phát triển của thị trường phái sinh tại Việt Nam.

Từ khóa: GARCH, Hợp đồng tương lai chỉ số VN30, Việt Nam.

1. Introduction

From the first trading in 2000, Vietnam stock market has witnessed spectacular growth. In

2017, with the first index futures contract, Vietnam has become the fifth country in ASEAN (after

Singapore, Indonesia, Malaysia and Thailand) which has derivatives market. Since then, deriva-

tives market has drawn enormous attention from investors. For 3 years since the introduction of

VN30 index futures, by the end of July 2020, 67.9 million futures contracts were traded. Vietnam

derivatives market is particularly active when the underlying market is strongly volatile. The liq-

uidity in derivatives market continuously surpassed the previous levels. In 2019, average trading

volume on the derivatives market reached 88,740 contracts per session, increase by 12.6% com-

pared to previous year.

VN30 index futures contract, though having attracted many investors, raises the concern

1129

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

of increasing volatility in the spot market. Wang, Lin, Lin & Lai (2020) has pointed out that index

futures become one of the mose popular speculative instruments nowadays. Bologna and Cavallo

(2002) state that futures market promotes speculation which leads to increasing volatility in the

underlying spot market. However, another group of literature argues that the futures market con-

tributes to price discovery process, hence it has positive effect on underlying market. As the index

futures become more widespread coupled with upcoming derivatives instrument in Vietnam mar-

ket, the investigation of futures trading on stock market volatility is essential.

This paper investigates the impact of VN30 index futures introduction on Vietnam stock

market volatility (representing by VNIndex volatility). Specifically, this paper first examines

whether there is any difference in volatility before and after the trading of VN30 index futures.

Then this paper tests whether the introduction of futures contract has an negative or positive in-

fluence on the stock market volatility. In this empirical analysis, this paper applies Generalized

Autoregressive Conditional Heteroskedasticity (GARCH) to model volatility. Most of previous

literature have been conducted in developed market (USA, UK, etc.) while a small number of

studies relates to other countries. This is the first paper to examine this matter in Vietnam.

The paper structure is as follows: section 2 presents brief review of previous literature, sec-

tion 3 describes data collection and methodology, section 4 presents results, section 5 provides

discussion and conclusion on this study.

2. Literature review

There are two main strands in the theoretical researches debating about the impact of futures

on the spot market.

One strand of literature argues that index futures introduction has negative influence on

the spot market. Stein (1987) states that futures market has high degree of leverage, therefore, it

attracts many uninformed traders. Such traders create noises in the market (Black, 2001), which

make information level of futures traders lower compared to cash market traders, then increases

the market volatility. Cox (1976), Finglewski (1981), Stein (1987), Cagan (1981) and Harris

(1989) agree with this argument.

Another strand in the literature supports the argurments of futures trading benefits including

price discovery (Schwart and Laatsch, 1991), increasing market depth and informativeness (Pow-

ers, 1976), hence promoting market efficiency (Stoll and Whaley, 1988). Danthine (1978) implies

that futures trading increases market depth, therefore, stabilizes the market. Bray (1981) and Kyle

(1985) also support the opinion that futures trading lowers spot market volatility and enhances

market effficiency.

From the previous literature, both arguments are supported and the empirical question is

getting more difficult to answer. Depending on each country condition, futures market introduc-

tion can have negative or positive influence on the spot market volatility. Figlewski (1981) when

investigating the GNMA futures maket find that the market becomes more voltile after the futures

introduction while Froewiss (1978) has come to an opposite conclusion. Following those re-

searchers, many studies have been conducted on financial futures and its impact on spot market.

Most of the studies in this topic apply ARCH/GARCH family model and add dummy vari-

1130

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

able for index futures introduction, such as Darrat and Rahman (1995), Pericli and Koutmos

(1997), Antonioua and Holmes (1995), Illueca and Lafuente (2003), Panayiotis (2011), and so

on. Whereas many researches have been conducted in a specific country, Gulen and Mayhew

(2000) use country level data. They apply serveral GARCH models and add both additive and

muplicative dummies. They find that in USA and Japan, index futures trading has a positive im-

pact on market volatility while they can not find any significant results in other countries. In the

contrary, Bologna and Cavallo (2002) investigate this issue in Italian market and show that futures

trading stabilize the market.

While many researches have been done on developed as well as developing markets, the

question about the impact of futures trading on spot market volatility is left unanswered in frontier

markets. Vietnamese researchers have been familiar with ARCH/GARCH model and apply this

type of method in forecasting the stock market (Vương Quân Hoàng, 2004; Hồ Thủy Tiên, Hồ

Thu Hoài & Ngô Văn Toàn, 2017; Phạm Chí Khoa, 2017; and Lê Văn Tuấn & Phùng Duy Quang,

2020). However, there is no empirical question employing this method to investigate the futures

market and its impact on stock market. This paper contributes to the existing literature in the way

of considering Vietnam market, a frontier market which is at early stage of derivatives market

development.

3. Data and methodology

This paper uses VNIndex as proxy for stock market in Vietnam. Daily closing prices of

VNIndex from 28/7/2000 (the establishment of Vietnam stock market) till 9/10/2020 are collected.

The final sample after data cleaning is consisted of 4580 observations of daily returns. All data

are retrieved from the website http://www.hsx.vn. Data is processed using R software. For

GARCH modelling, this paper uses R code provided by Perlin, Mastella, Vancin and Ramos

(2020).

This paper uses continuous compounded rate of return as dependent variable in the mean

model. Specifically, the rate of return is calculated as the difference of natural logarithm of two

consecutive spot index prices.

To examine whether volatility in the underlying market has changed after the futures in-

troduction, this paper applies ARCH family of models. According to Engle (1982), the ARCH

process gives the explanation of difference between conditional and unconditional variance.

ARCH model allows the conditional variance to be time-varying while unconditional variance

remains constant. ARCH (q) model is as follows:

1131

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

Where , Xt is a vector including the information set , is error term, is the conditional

volatility. Bollerslev (1986), student of Engle, suggests a generalized ARCH, which is called

GARCH. In the GARCH (p,q) model, the conditional variance is specified as:

GARCH model restricted that , and . provide information about the extent to which past

returns can be used to explained current volatility. The sum (a+c) captures the volatility persist-

ence and is restricted to be smaller than 1. If (a+c) > 1, the volatility is explosive. It means that

a shock in volatility leads to even larger shock in the next period. Hồ Thủy Tiên et al (2017),

Phạm Chí Khoa (2017), and Lê Văn Tuấn & Phùng Duy Quang (2020) find evidence that GARCH

(1,1) is suitable to explain stock market volatility in Vietnamese market. In this paper, I will use

the same approach in Vietnam context.

To explore the structural break in volatilty in Vietnam stock market when VN30 futures is

introduced, dummy variable is added to the GARCH (1,1) model:

DUM is represented for dummy variable which takes value of 0 in the period of no index

futures trading and value of 1 in the period of index futures trading. If the coefficient of DUM is

significantly different from 0, the VN30 index futures introduction has influenced the stock mar-

ket. Moreover, the sign of DUM’s coefficient let us know whether this impact is positive or neg-

ative. The flow of this empirical analysis is summarized as follows:

1132

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

4. Result

(Source: author’s calculation)

Figure 1: VNIndex close prices from 28/7/200 to 10/9/2020

(Source: author’s calculation)

Figure 2: VNIndex continuous compounded rate of return from 28/7/200 to 10/9/2020

Vietnam stock market has officially established in 1998 with the appearance of HOST and

HASTC. The first trading session occurred in 28/7/2000 with the trading of two stocks (REE and

SAM) in Ho Chi Minh City Securities Trading Center. For 20 years, Vietnam stock market has

1133

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

gone through many ups and downs: the explosive growth in 2007, negative impact from the

worldwide financial crisis in 2008, an epic year 2017 with the doubled market capitalization stock

market (Vietnam stock market capitalization is US$155 billion or VND 3.52 quadrillion, which

equals to 72 percent of total GDP) and then the fear of a new crisis ahead. Through many good

and bad times, it is undeniable that the development of Vietnam stock market has made great

contribution to Vietnam economic growth. In 10/8/2017, the first index futures with VN30 as the

underlying asset has been introduced in Vietnam stock market, providing market participants an-

other instrument to hedge as well as to speculate.

Figure 1 demonstrates the development of VNIndex from the establishment of Vietnam

stock market till 2020 with two peaks of the index in 2007 and 2017. Figure 2 illustrates VNIndex

returns through 20 years. It can be seen from Figure 2 that a large change in returns is followed

by a larger change. It is the sign of volatility clustering in VNIndex returns.

Table 2: Summary statistics

Total sammple

Pre-futures period Post-futures period

(2/7/2000 -10/9/2020

(28/7/2000 to

10/8/2017)

(10/8/2017 - 10/9/2020

Obs

mean

std

4850

4079

770

0.000450

0.014945

-0.076557

-0.005775

0.000451

0.007474

0.077414

0.000502

0.015417

-0.076557

-0.006086

0.000315

0.007751

0.077414

0.000181

0.012163

-0.064820

-0.004295

0.001194

0.00620

min

25%

50%

75%

max

770.000000

(Source: author’s calculation)

Table 2 presents summary statistics for VNIndex for the whole period (from 28/7/2000 to

10/0/2020) as well as two sub periods. The investiaged sample is divided into two sub-samples:

pre-futures period and post-futures period. The volatility of the index measured by standard de-

viation of index returns has changed after the introduction of VN30 index futures. Specifically,

the standard deviation of VNIndex returns has decreased from 0.0151 to 0.0121. However, more

evidence is needed to conclude that the index futures introduction has stabilized the stock market.

Table 3. Lagrange Multiplier test for ARCH effect in VNIndex returns

Lag

1

2

3

4

Chi-squared

1422.7

p-values

0.0000

1622.5

1709.2

1778.7

1848.9

5

(Source: author’s calculation)

1134

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

This study tests the ARCH effect in VNIndex returns by performing Lagrange Multiplier

test (LM test). The result is presented in Table 3. Null hypothesis of LM test is that there is no

ARCH effect in the investigated sample. It can be seen that the p-values of all five lags are sig-

nificant, the null hypothesis is rejected. It can be confirmed that there is ARCH effect in VNIndex

returns.

By using two goodness-of-fit which are AIC and BIC, ARMA (2,2) ~ GARCH (1,1) are

selected because this model yields the smallest AIC and BIC. However, when fitting the GARCH

(1,1), coefficient of AR(2) is insignificant. ARMA (1,2) has been applied and showed a better re-

sults. Dummy variable is then added to GARCH (1,1) model to test the impact of futures intro-

duction to stock market. The results presented in Table 5 show that coefficient of DUM is positive

and statistically significant. This is the evidence that the VN30 index futures introduction has an

impact on the stock market volatility. However, this paper result shows that when VN30 index

futures appears, the stock market volatility increases. Instead of stabilizing the market, VN30

index futures trading seems to make the stock market more volatile than before.

Table 4: Selecting ARMA and GARCH model

lag_ar lag_ma lag_arch lag_garch

AIC

BIC

model_name

0

1

2

3

0

1

2

3

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

1

2

3

4

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

1

2

3

4

-6.08344 -6.07809

-6.08600 -6.07797

-6.08662 -6.07592

-6.08691 -6.07354

-6.08716 -6.07111

-6.12150 -6.11348

-6.12179 -6.11109

-6.12269 -6.10932

-6.12271 -6.10666

-6.12194 -6.10321

-6.12986 -6.11916

-6.13021 -6.11683

-6.13101 -6.11496

-6.13124 -6.11251

-6.13080 -6.10940

-6.12917 -6.11580

-6.12953 -6.11348

-6.13034 -6.11162

-6.13047 -6.10908

ARMA(0,0)+GARCH(1,1)

ARMA(0,0)+GARCH(2,2)

ARMA(0,0)+GARCH(3,3)

ARMA(0,0)+GARCH(4,4)

ARMA(0,0)+GARCH(5,5)

ARMA(1,1)+GARCH(1,1)

ARMA(1,1)+GARCH(2,2)

ARMA(1,1)+GARCH(3,3)

ARMA(1,1)+GARCH(4,4)

ARMA(1,1)+GARCH(5,5)

ARMA(2,2)+GARCH(1,1)

ARMA(2,2)+GARCH(2,2)

ARMA(2,2)+GARCH(3,3)

ARMA(2,2)+GARCH(4,4)

ARMA(2,2)+GARCH(5,5)

ARMA(3,3)+GARCH(1,1)

ARMA(3,3)+GARCH(2,2)

ARMA(3,3)+GARCH(3,3)

ARMA(3,3)+GARCH(4,4)

1135

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

3

4

5

3

4

5

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

-6.12963 -6.10555

-6.13101 -6.11496

-6.13068 -6.11196

-6.13142 -6.11002

-6.13164 -6.10756

-6.13059 -6.10385

-6.12966 -6.11093

-6.13005 -6.10865

-6.13085 -6.10677

-6.13237 -6.10562

-6.13083 -6.10140

ARMA(3,3)+GARCH(5,5)

ARMA(4,4)+GARCH(1,1)

ARMA(4,4)+GARCH(2,2)

ARMA(4,4)+GARCH(3,3)

ARMA(4,4)+GARCH(4,4)

ARMA(4,4)+GARCH(5,5)

ARMA(5,5)+GARCH(1,1)

ARMA(5,5)+GARCH(2,2)

ARMA(5,5)+GARCH(3,3)

ARMA(5,5)+GARCH(4,4)

ARMA(5,5)+GARCH(5,5)

(Source: author’s calculation)

Table 5. GARCH (1,1) with dummy variable from 28/7/2000 to 10/9/2020

Coefficients

Estimates

0.000003***

(0.000004)

0.000002***

(0.00000)

DUM

0.197114***

(0.002344)

0.801668***

(0.023029)

Assuming normal distribution. The standard errors are reported in parentheses. *** denotes

significance at 1%.

(Source: author’s calculation)

In order to check the robustness of this result, the empirical analysis is performed again

using the sample from 30/7/2007 (the day on which HOSE applied continuous order matching)

till 10/9/2020. This sample choice is based on Lê Văn Tuấn & Phùng Duy Quang (2020)’s ap-

proach in the context of Vietnam.

According to Table 6, the coefficient of DUM is still positively and statistically significant.

It confirms our previous result of positive impact of VN30 index futures introduction on market

volatility in Vietnam.

1136

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

Table 6. GARCH (1,1) with dummy variable from 30/7/2007 to 10/9/2020

Coefficients

Estimates

0.000004***

(0.000002)

0.000000***

(0.00000)

DUM

0.138258***

(0.020793)

0.843789***

(0.051417)

Assuming normal distribution. The standard errors are reported in parentheses. *** denotes

significance at 1%.

(Source: author’s calculation)

5. Discussion

This paper provides evidence about the influence of index futures introduction to stock

market volatility. Follow previous literature of modelling volatility in Vietnam context, this paper

employs GARCH (1,1) model with dummy variable to test for the structural break in volatility.

While standard deviation in the pre-futures period is larger than post-futures period, GARCH

model results show that VN30 index futures introduction has a significant positive impact on the

stock market. The result is robust after turmoil period is excluded from the investigated period.

It means that the index futures trading makes the stock market more volatile. Though futures

market attracts more investors, policy makers should be careful and take into account its impact

on spot market when promoting derivatives market development.

REFERENCES

Antoniou, A., & Holmes, P. (1995), ‘Futures trading, information and spot price volatility:

evidence for the FTSE-100 stock index futures contract using GARCH’, Journal of Banking &

Finance, 19(1), 117-129.

Black, Fischer (2001), ‘Noise’, The Journal of Finance. Vol.41(3)

Bollerslev, T. (1986), ‘Generalized autoregressive conditional heteroscedasticity’, Journal

of Econometrics, 31, 307-27.

Bologna, Pierluigia & Cavallo. L., (2002), ‘Does the introduction of stock index futures

effectively reduce stock market volatility? Is the `futures effect’ immediate?’ Evidence from the

Italian stock exchange using GARCH. Applied Financial Economics. Vol. 12: 183-192

Bray, M. (1981), ‘Futures trading, rational expectations and the efficient market hypothe-

sis’, Econometrica. Vol. 49:96-575

1137

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

Cagan, P. (1981), ‘Financial futures markets: is more regulation needed?’, Journal of Fu-

tures Markets.Vol.1: 90-169

Cox, C. C. (1976), ‘Futures trading and market information’, Journal of Political Economy.

Vol. 84:37-1215.

Danthine, J. (1978), ‘Information, futures prices and stabilizing speculation’, Journal of

Economic Theory. Vol. 17: 79-98.

Darrat, Ali and Shafiqur Rahman (1995), ‘Has futures trading activity caused stock price

volatility?’, Journal of Futures Markets. Vol. 15 (5): 537-557

Engle, R. F. (1982), ‘Autoregressive conditional heteroscedasticity with estimates of the

variance of United Kingdom inflation’, Econometrica, 50, 987- 1007

Figlewski, Stephen. (1981), ‘Futures Trading and Volatility in the GNMA Market’, The

Journal of Finance. Vol. 36 (2): 445-456

Froewiss, K. (1978) GNMA futures: stabilizing or destabilizing? Federal Reserve Bank of

San Francisco Economic Review, 20-9.

Gulen, H., & Mayhew, S. (2000), ‘Stock index futures trading and volatility in international

equity markets. Journal of Futures Markets: Futures, Options, and Other Derivative Products’,

20(7), 661-685.

Harris, L. (1989), ‘S&P 500 cash stock price volatilities’, Journal of Finance. Vol.44: 155-

175.

Hồ Thủy Tiên, Hồ Thu Hoài & Ngô Văn Toàn (2017), ‘Mô hình hóa biến động thị trường

chứng khoán: Thực nghiệm từ Việt Nam’, Tạp chí Khoa học ĐHQGHN: Kinh tế và Kinh doanh,

Tập

33,

Số

3

(2017)

1-11

Lê Văn Tuấn, Phùng Duy Quang (2020), ‘Áp dụng mô hình GARCH dự báo ảnh hưởng của đại

dịch Covid-19 đến thị trường chứng khoán Việt Nam’.

Illueca, M., & Lafuente, J. (2003), ‘The effect of spot and futures trading on stock index

market volatility: A nonparametric approach’, Journal of Futures Markets: Futures, Options, and

Other Derivative Products, 23(9), 841-858.

Kyle, A. S. (1985),’Continuous auctions and insider trading’, Econometrica: Journal of the

Econometric Society, 1315-1335.

Pericli, A., & Koutmos, G. (1997),’Index futures and options and stock market volatility’,

The Journal of Futures Markets (1986-1998), 17(18), 957.

Perlin, M. S., Mastella, M., Vancin, D. F., & Ramos, H. P. (2020),’A GARCH Tutorial with

R’, Revista de Administração Contemporânea, 25(1), e200088-e200088.

Powers, M. J. (1976), ‘Does futures trading reduce price fluctuations in the cash markets?’,

In The Economics of Futures Trading (pp. 217-224). Palgrave Macmillan, London.

Phạm Chí Khoa (2017) ‘Dự báo biến động giá chứng khoán qua mô hình Arch – Garch’,

Tạp chí Tài chính, Kỳ 2, 2017, số 6, tr38-39.

1138

INTERNATIONAL CONFERENCE FOR YOUNG RESEARCHERS IN ECONOMICS & BUSINESS 2020

ICYREB 2020

Schwarz, T. V., & Laatsch, F. E. (1991).’Dynamic efficiency and price leadership in stock

index cash and futures markets’, The Journal of Futures Markets (1986-1998), 11(6), 669.

Stein, J. C. (1987), ’Informational externalities and welfare-reducing speculation’, Journal

of political economy, 95(6), 1123-1145.

Stoll, H. R., & Whaley, R. E. (1988). ‘Volatility and futures: Message versus messenger’,

Journal of Portfolio Management, 14(2), 20.

Vương Quân Hoàng (2004), ‘Hiệu ứng GARCH trên dãy lợi suất thị trường chứng khoán

Việt Nam 2000-2003’, Tạp chí Ứng dụng toán học tập II, số 1, 2004

Wang, C. H., Lin, C. C., Lin, S. H., & Lai, H. Y. (2020),‘A new dynamic hedging model

with futures: The kalman filter error-correction model’, Journal of Operational Risk, 22(4).

1139