ABSTRACT:
This study explores the impact of derivatives trading activity on the underlying asset returns in Vietnamese stock market. The study’s ARCH (1) model shows a positive influence of past residuals (random factors) on the volatility of VN30 index returns in both pre- and post-derivatives trading periods. Moreover, the study finds out that the GARCH (1,1) model is optimal in estimating the impact of factors causing the volatility of returns. Furthermore, the OLS model confirms that if the returns of VN30F1M (or VN30F2M) increase and the remaining variables remain unchanged, the average returns of VN30 index will also increase, and vice versa.
Keywords: underlying asset, the Vietnamese stock market, VN30 index.
1. Introduction
The derivatives market plays a highly essential role in global economic integration (Ayşegül ÇİMEN, 2018). Derivatives allow investors to take advantage of investment opportunities in the direction of rising or falling prices of underlying assets by providing an easy-to-use, flexible, and leveraged investment method. The main role of derivatives is to hedge against price risks based on decisions to buy or sell positions through predictions of the increase or decrease of underlying assets, providing opportunities for arbitrage or speculation. Through these features, the derivatives market has attracted a large number of investors to participate and increased liquidity as well as the flow of information between the two markets. Therefore, derivatives trading may cause an increase or decrease in the volatility of the underlying market.
This statement is the foundation for research on the impact of trading in the derivatives market on the volatility of the underlying market in many parts of the world. However, research has mainly focused on countries with developed economies or some large developing economies, and very few studies have been conducted on other developing countries, such as Vietnam. This creates a theoretical gap when the newly introduced derivatives market in these countries in terms of the effects of trading activities may differ from those of countries with well - established and developed derivatives markets (Truong et al, 2022).
In 2017, the derivatives market officially started operating with its first derivative product being the VN30 index futures contract, a market index representing 30 stocks with high market capitalization and liquidity listed on HOSE. The majority of individual investor participation in futures contract trading may have increased the volatility of stock prices in the underlying market because they mostly participate for speculative purposes and arbitrage while having limited knowledge and investment experience (Truong et al, 2022). Additionally, Index futures contract trading in the Vietnamese derivatives market allows using high financial leverage, which may trigger market instability.
So, the research articles "How derivatives trading impacts the returns of the underlying securities in Vietnam?" aims to identify the impact of futures contract trading on the volatility of the underlying market.
2. Research gap
2.1. Research gap
This study uses the ARCH/GARCH model to investigate the impact of the derivatives market on the underlying market in Vietnam.
Moreover, in the process of reference, it is also observed that some previous studies have only collected data after the derivative product (futures contract) was launched, without comparing the volatility of the market before and after the product launch, making it difficult to accurately assess the impact of the derivatives market on the underlying market. This study argues that the data collected in previous studies differ in spatial and temporal factors, which cannot be compared and applied in similar research cases.
2.2. Research hypothesis
In their study, Nguyen Thi Hien et al (2022) identified significant shocks in the past that have a substantial impact on the current returns of the index. According to the research conducted by C. Pilar and S. Rafael (2002), the introduction of the derivative stock market reduces the conditional volatility of the underlying market by using data on the price and volume of the Ibex-35 index. But not all derivatives trades are necessarily to blame for volatility, as demonstrated by the study of A. F. Darrat and S. Rahman (1995). Therefore, the following hypothesis is proposed:
H1: The introduction of futures contract trading increases the volatility of market index returns.
Several studies in Vietnam have shown that past returns have a current impact on returns, such as the studies conducted by Trinh Thi Huyen Trang & Le Thi Thu Thao (2021) and Nguyen Thi Hien et al (2022). Increased investment in VN30F1M stock index futures contract creates opportunities for profit, and increases buying demand in the financial market in the following periods. This can lead to an expectation of positive growth in market index returns. Based on the above analysis, the following hypotheses related to returns variations are proposed:
H2: The returns of the VN30F1M index futures contract increases, causing the VN30 market index returns to increase.
H3: The returns of the VN30F2M index futures contract increases, causing the VN30 market index returns to increase.
3. Methodology
3.1. Model
This paper determined the return of the VN30 Index based on the closing price of this index recorded on the Vietstock.vn website, thereby eliminating the daily price volatility of the index.
3.1.1. ARCH/GARCH Model
The ARCH model, introduced by Robert F. Engle in 1982, is used in economics and finance to represent variations in a time series. This model assumes that the heteroscedasticity of the variance of the time series information can occur, meaning that it can change over time. The ARCH (1) model is estimated as follows:
- Time series dataset from 3/1/2012 to 10/8/2017:
Estimator of mean model: rVN30t = rVN30t-1 + εt
- Time series dataset from 10/8/2017 to 31/12/2022: Applying the same as with the previous dataset, the study comes up with two equations as follows:
Estimator of mean model: ret30t = ret30t-1 + εt
In which:
rVN30t, ret30t: the VN30 index return at trading day t, is estimated by the natural logarithm of the closing price of trading day t (Pt) and t-1 (Pt-1), the specific formula is as follows: rVN30t = ln (Pt/Pt-1) and ret30t = ln (Pt/Pt-1)
rVN30t-1, ret30t-1: the VN30 index return at trading day t-1
εt: residual
Variance model:
In which: α0: constant, α1: regression coefficient, residual square at t - 1.
However, the ARCH model still has many drawbacks such as its inability to model sudden changes in the data series over time and at different levels of volatility. Therefore, the study used the GARCH (1,1) model, is a specific development of the ARCH(p) model, captures the trend of profits better while incorporating variance changes into the model estimation. This model is represented as follows:
- 2012-2017 period:
Average equation: rVN30t = rVN30t-1 + εt
- 2017-2022 period:
Average equation: ret30t = ret30t-1 + εt
Variance equation:
In which: ω > 0, α1 ≥ 0, b1 ≥ 0; rVN30t, ret30t is the VN30 index return at t from 2012 to 2017; rVN30t-1, ret30t-1 is the VN30 index return at t-1 from 2012 to 2017; and εt is residual of return.
As previously mentioned, the coefficient α1 is the ARCH coefficient that measures the degree of impact of the residual or the effect of new market-related information on the underlying stock return. The coefficient b1 represents the impact of the variance of the past return series on the current volatility of the data series.
3.1.2. OLS model
The Ordinary Least Squares (OLS) model is a statistical method used to estimate the relationship between one dependent variable and one or more explanatory variables. The OLS model utilizes statistical methods to determine the reliability of coefficient estimates, and also allows for the testing of hypotheses about the relationship between explanatory and dependent variables.
In this study, the OLS model is employed to estimate the impact of the returns of VN30F1M and VN30F2M index futures contracts on the returns of the VN30 index, based on the hypothesis of a linear relationship between these variables. The OLS equation is then represented as follows: Ret30 = f(RetF1M, RetF2M)
The econometric model for this study as follows:
- Overall regression model: Ret30 = b0 + b1*RetF1M + b2*RetF2M + ui
- The sample regression model:
In which: Dependent variable is Ret30 - The return of the VN30 index; Explanatory variables: RetF1M is the return of VN30F1M futures contract, RetF2M is the return of VN30F2M futures contract. b0 is intercept of the Overall regression model; b1, b2 are regression coefficients with explanatory variables RetF1M, RetF2M; , , , are estimators of b0, b1, b2, ui.
3.2. Data Base
The impact of the derivatives market is tested by using 10-year data surrounding the opening time of the derivatives market, including: closing prices of the VN30 index collected from 01/01/2012 to 31/12/2022 and closing prices of the VN30F1M and VN30F2M index futures contracts from 10/08/2017 to 30/12/2022. Secondary data were collected from two websites, CafeF.vn and Vietstock.vn.
This study applied modern financial models, combining interpretive, synthesizing, and analytical methods with knowledge of econometrics and statistical tools, correlation analysis on Excel and Rstudio software.
3.3. Tests
In the paper, various tests were employed to examine the accuracy and reliability of the obtained results. For the GARCH model, stationary and residual tests were utilized to assess the accuracy of the model. Meanwhile, normality tests, autocorrelation tests, heteroscedasticity tests, and multicollinearity tests were used to examine the reliability of the OLS model.
4. Result
4.1. The estimated results of the GARCH(1,1) model
4.1.1. The result of descriptive statistics
According to Table 1, the average return of the VN30 index decreased from 0.047% to 0.0215% after the introduction of futures trading. At the same time, the standard deviation of the index increased, with values of 1.0418% and 1.3745% before and after the introduction of futures trading, respectively. However, the increase in standard deviation is not enough to conclude that the emergence of futures has increased the level of risk of the VN30 index. This can be explained by the impact of external macroeconomic factors or systemic risk factors.
Table 1. Descriptive statistics of VN30 index results
Source: Estimated results of the model on RStudio software from the study
Both periods have negative skewness, indicating a distribution with a left tail and a peak closer to the right towards the median. This data suggests that the return on the underlying asset tends to decrease and is negative. The period before the introduction of futures had a lower kurtosis value of less than 3, indicating that the stock price was relatively stable. In contrast, the kurtosis value in the period after the introduction of futures was 3.635069, indicating that the distribution of the return series has a heavy tail and does not follow a normal distribution.
The result of the stationarity test: According to the given data series, both rVN30 and Ret30 variables are stationary at the 5% significance level. Therefore, it can be concluded that the variables are suitable for applying the conditional heteroscedasticity model with the general autoregressive conditional heteroskedasticity (GARCH) method.
The result of the ARIMA model: estimation for the two variables corresponding to the respective datasets were used. Specifically, the ARIMA (2,0,3) model was applied to the variable corresponding to the period before futures trading was introduced, and the ARIMA (0,0,0) model was applied to the variable corresponding to the period after futures trading was introduced.
4.1.2. The result of the GARCH (1,1) model
The estimated coefficients of the GARCH (1,1) model are all statistically significant at the 5% level. The coefficients α0, α1 and b1 are all positive, indicating that the model is stable in both cases.
The results showed model standard at a significance level of 5%.
4.2. The result of OLS model
4.2.1. The result of descriptive statistics
From The results of descriptive statistics of variables, it can be seen that the mean values of the three variables are 0.000215, 0.000214, and 0.000207, respectively, with corresponding standard deviations of 0.013745, 0.016313, and 0.015944. With relatively small standard deviations, the volatility of the returns is negligible. Additionally, combined with Correlation matrix of variables, there is almost no evidence of correlation between different explanatory variables. Although RetF1M and RetF2M have a relatively high correlation, this may cause multicollinearity. However, through the multicollinearity test in Multicollinearity test result, the VIF values for both variables are less than 10, leading to the conclusion that there is no multicollinearity.
4.2.2. Result of statistical hypothesis test
The regression model is statistically significant at the 5% significance level with p-value < 2.2e-16, indicating that the used variables can explain the volatility of the return of VN30 index. The value of the multiple r-squared is 0.8313, indicating that the regression model can explain 83.13% of the variation in Ret30 using RetF1M, RetF2M, and other factors. The coefficients of RetF1M and RetF2M are 5.053e-01 and 2.815e-01, respectively, both positive, indicating that the impact of these two variables on Ret30 is in the same direction. This is consistent with the initial hypothesis that the increase in the returns of the VN30F1M and VN30F2M futures contracts increases the returns of the VN30 index. The result of the regression coefficient sign test in Table 2 also confirms this conclusion.
Table 2. Summary of the conclusion of regression coefficient
5. Discussion and recommendation
5.1. Discussion of the estimated result
5.1.1 The result of ARCH - GARCH model
Firstly, the ARCH model results demonstrate that the residuals (random factors) have had a favorable influence on the price movement of the VN30 index in the past, both before and after the derivatives market was introduced. Volatility tends to be in series, which means that when the price increases, it tends to continue to increase for a certain period after that, and vice versa. The ARCH model, on the other hand, does not adequately explain the basis of impacts on the returns volatility of securities. As a result, the study used the GARCH model (1,1) to estimate and determine the factors that affect the volatility of stock returns in the underlying market.
Secondly, coefficients: α1 (ARCH coefficient) and (GARCH coefficient) which respectively represent the influence of new and old information, are statistically significant at the significance level of 5%. Before the introduction of futures contracts, α1 and have values of 0.1405 and 0.7902, which change to 0.1024 and 0.8856 after the introduction of futures trading. It can be seen that the coefficients α1 and both suggest a favorable influence of new and old information on stock returns volatility in the underlying market in both periods.
Thirdly, the coefficient of residual lag (noise) α1 dropped after the futures trading occurred, which shows that the speed of new information reflected in the VN30 index has slowed down compared to the period before the introduction of derivatives market. Meanwhile, the delay coefficient variance increased after futures contracts were introduced, indicating that old information has a higher impact on the return volatility of the VN30 index, and this influence is also maintained for a longer period of time after the appearance of futures trading. In both periods, the α1 coefficient is always lower than , so the return of the VN30 stock group is more affected by old information than by new information.
Fourthly, the sum of the two coefficients α1 and before and after the introduction of the derivative market is 0.9307 and 0.988, respectively. The sum of the two coefficients in both periods is less than 1, meaning that the process of volatility of return around its average value always occurs. Furthermore, an increase in the total of these two coefficients implies that the impact of information on the volatility of the VN30 return is stronger.
5.1.2. The result of OLS model
Firstly, the results obtained from the OLS model and the corresponding tests are consistent with the initial expectations of this study and previous research. From the research results, it can be seen that the return of both VN30F1M and VN30F2M futures contracts have the same direction of impact on the return of the VN30 index, which is consistent with the current situation in the Vietnamese stock market. Due to the expectation of investors about the upward trend of market index stocks, specifically the VN30 index, there is a difference between the futures contract price and the fair value, leading to the stimulation of price arbitrage trades. Therefore, when the return of VN30F1M or VN30F2M futures contracts increases and the other variables remain unchanged, the return of the VN30 index also increases and vice versa.
Secondly, the estimates for all coefficients are positive and statistically significant through tests, confirming that the regression model is consistent with the initial expectations. Although the model has a flaw in the heteroscedasticity, leading to unreliable tests, the consistency of the initial estimates is maintained and not biased through the method of estimating adjusted standard errors, improving the reliability of the tests. Therefore, the initial hypothesis remains unchanged.
Thirdly, futures trading amplifies market volatility as traders can use leverage to control larger positions with smaller amounts of capital. This means that even small changes in the futures contract index can have a significant impact on stock prices, leading to higher stock profits during optimistic market sentiment and lower stock profits during pessimistic market sentiment.
5.2. Recommendation
To ensure that the derivative securities market operates safely, effectively, and dynamically, through the research, some policy implications are proposed as follow:
5.2.1. For Government
To develop the derivatives market, the government needs to implement the following measures:
Firstly, enhance education and awareness of the derivatives market for investors through activities such as seminars, specialized conferences, competitions, and educational campaigns, in order to help investors understand the risks and benefits of derivative products as well as how to use derivative tools effectively.
Secondly, establish a clear and comprehensive legal framework to manage the market, and ensure fair and orderly operation of the derivatives market.
Thirdly, invest strongly in developing the infrastructure of the derivatives market, including trading platforms, clearing house, and payment systems, improving them in a modern, efficient, and able to handle increasing trading volume with strong risk management mechanisms.
Fourthly, encourage innovation and diversity of derivative products to meet the needs of investors in the market.
Fifthly, review and rationalize regulations related to market access for foreign investors participating in the Vietnamese derivatives market.
Sixthly, encourage cooperation among stakeholders in the derivatives market, including regulatory agencies, market participants, and industry associations, through regular meetings, discussions, and consultations to create common solutions to cope with challenges and ensure sustainable development of the market.
5.2.2 For investors
Firstly, investors need to continuously enhance their knowledge and skills in trading derivatives. They should always update themselves on market trends, news, and developments, as well as use fundamental and technical analysis methods to make informed trading decisions and improve their trading performance.
Secondly, effective risk management measures should be implemented, such as establishing suitable stop-loss orders, applying diversification approaches, properly controlling leverage, and so on, need to be implemented for investors to protect capital and contribute to market stability.
Thirdly, by actively participating in the market, investors help increase the market's liquidity, and create favorable conditions for price determination and efficient trading.
Fourthly, investors should stay informed and participate in industry initiatives such as information on market development, changes in regulations, and related initiatives in the derivatives market. Participating in forums in the industry, providing feedback on proposed regulations, and collaborating with other market participants can help shape positive changes in the market.
Finally, investors should use technology effectively to improve trading efficiency, risk management, and market analysis. Using trading platforms, analysis tools, and risk management systems can improve trading performance and contribute to market efficiency.
6. Conclusion
The study found that there is a positive correlation between the VN30 index and the VN30F1M and VN30F2M futures contracts. The ARCH and GARCH models show that there are same-direction effects of past random factors on the volatility of the VN30 index, and that the GARCH model is more effective in estimating the impact of factors on volatility. The study also identified some limitations, including the non-normal distribution of residuals and heteroscedasticity in the OLS model, and the fact that only 83.13% of the influence on the dependent variable was explained by the independent variables used in the model.
Due to limitations in terms of time, space, and resources, the research has some shortcomings. Therefore, future studies should expand the research scope to include bonds and foreign markets. Moreover, the impacts of macroeconomic factors should also be considered when running the model to better explain the influence on the dependent variable.
REFERENCES:
Vietnamese references:
- Phùng Thanh Bình. (2011). Biến động giá tài sản: Các mô hình ARCH và GARCH. Truy cập tại https://vi.vnp.edu.vn/tai-lieu-tham-khao/chuong-15-cac-mo-hinh-gia-tai-san-arch-garch/
- Nguyễn Thị Hiên và các cộng sự, (2022). Ứng dụng mô hình ARCH, GARCH phân tích độ biến động của hợp đồng tương lai VN30F1M trên thị trường chứng khoán phái sinh Việt Nam. Truy cập tại https://sti.vista.gov.vn/tw/Lists/TaiLieuKHCN/Attachments/345307/CTV36S132022029.pdf
- Vương Thị Thùy Linh, Nguyễn Tố Nga, (2019). Vai trò của hợp đồng tương lai chỉ số trên thị trường chứng khoán Việt Nam: Khía cạnh truyền dẫn thông tin và giá. Truy cập tại http://rces.info/files/2020/01/UEB-Working-Paper-Series_Final-cua-Final.pdf#page=68
- Trịnh Thị Huyền Trang, Lê Thị Thu Thảo, (2021). Ứng dụng mô hình ARCH và GARCH dự báo lợi suất cổ phiếu VNM. Tạp chí Khoa học và Công nghệ, 30, 79-85.
English references:
- A. F. Darrat & S. Rahman, (1995). Has future trading activity caused stock price volatility? The Journal of Futures Markets, 15(5), 537-557.
- Ayşegül ÇİMEN. (2018). The impact of derivatives on the volatility of the Turkish stock market. Uluslararası İktisadi ve İdari İncelemeler Dergisi, 17, 857-868.
- Corredor Pilar, Santamaría Rafael, (2002). Does derivatives trading destabilize the underlying assets? Evidence from the Spanish stock market. [Online] Availabile at http://www.unavarra.es/digitalAssets/117/117837_dt26-98.pdf
- Engle, R. F. (1982). Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation. Econometrica: Journal of the econometric society, 987-1007.
- Loc Dong Truong, H. Swint Friday và Anh Thi Kim Nguyen, (2022). The Effects of Index Futures Trading Volume on Spot Market Volatility in a Frontier Market: Evidence from Ho Chi Minh Stock Exchange. Risks 2022, 10(12), 234.
TÁC ĐỘNG CỦA HOẠT ĐỘNG GIAO DỊCH CHỨNG KHOÁN PHÁI SINH ĐẾN TỶ SUẤT SINH LỢI CỦA TÀI SẢN CƠ SỞ TRÊN THỊ TRƯỜNG CHỨNG KHOÁN VIỆT NAM
• TS. LÊ PHƯƠNG LAN1
• TRẦN THỊ THANH HOÀI2
• VŨ NHẬT ANH2
• PHẠM THỊ MINH NGUYỆT2
• VŨ DIỆU LINH2
• LÊ TRẦN PHƯƠNG NHI2
1Giảng viên, Khoa Tài chính Ngân hàng, Trường Đại học Ngoại thương
2Sinh viên, Trường Đại học Ngoại thương
TÓM TẮT:
Nghiên cứu này nhằm đánh giá tác động của hoạt động giao dịch chứng khoán phái sinh đến tỷ suất sinh lợi của tài sản cơ sở trên thị trường chứng khoán Việt Nam. Kết quả nghiên cứu từ mô hình ARCH (1) cho thấy có tác động tích cực của phần dư quá khứ (yếu tố ngẫu nhiên) đến biến động của tỷ suất sinh lợi chỉ số VN30 trong cả giai đoạn trước và sau giao dịch chứng khoán phái sinh. Bên cạnh đó, nghiên cứu xác định được mô hình GARCH (1,1) là tối ưu trong việc ước lượng tác động của các yếu tố gây ra sự biến động của lợi nhuận. Ngoài ra, mô hình OLS khẳng định rằng nếu tỷ suất sinh lợi của VN30F1M (hoặc VN30F2M) tăng lên và các biến còn lại không đổi thì tỷ suất sinh lợi trung bình của chỉ số VN30 cũng sẽ tăng lên và ngược lại.
Từ khóa: tài sản cơ sở, thị trường chứng khoán Việt Nam, chỉ số VN30 Index.
[Tạp chí Công Thương - Các kết quả nghiên cứu khoa học và ứng dụng công nghệ, Số 12 tháng 5 năm 2023]
