INTRODUCTION
Necessity of the topic
COVID-19, caused by the SARS-CoV-2 coronavirus first identified on December 31, 2019, in Wuhan, China, has significantly impacted the world, with over 218 million infections and 4.5 million deaths reported by the end of August 2021 (Worldometers, 2021) The data presented in Figures 1.1 to 1.4 highlights a concerning global increase in new cases and fatalities, particularly in Vietnam, where a dramatic rise began in July 2021 due to the emergence of the Delta variant, leading to an unprecedented wave of infections both locally and worldwide.
Figure 1.1 Total confirmed infected cases in the world
Figure 1.2 Total confirmed deaths in the world
Figure 1.3 Total confirmed cases in Vietnam
Figure 1.4 Total deaths in Vietnam
The global economy and stock markets have faced significant negative impacts due to the emergence of new variants and extended lockdowns caused by the pandemic.
The impact of COVID-19 on the Vietnamese stock market has been significant and varied over time During the pre-lockdown phase of the first wave, there was a notable negative correlation between COVID-19 and stock returns Conversely, the lockdown period that followed led to a substantial positive effect on stock performance (Anh & Gan, 2020) As of September 2021, amidst the fourth wave of the pandemic, which began in late April and involved strict curfews and lockdowns in major cities, the stock markets continued to operate.
The COVID-19 outbreak and subsequent lockdown measures have significantly influenced the Vietnamese stock market over a period of approximately twenty months since the initial global cases were reported This article focuses on the topic "The Impact of COVID-19 on Stock Index Volatility: Empirical Evidence from Vietnam on VNIndex," exploring how the pandemic has affected stock index fluctuations in Vietnam.
The coronavirus epidemic has emerged as a significant threat to the global economy and financial markets, impacting various sectors worldwide On March 12, 2020, the Dow Jones Industrial Average (DJIA) experienced a dramatic decline of 2,353 points, marking the largest single-day drop since the 1987 “Black Monday” crash Within a week, the DJIA plummeted nearly 3,000 points, reflecting widespread market panic Over the course of just one month, major stock indices saw substantial losses: the UK FTSE dropped by 29.72%, Germany's DAX fell by 33.37%, France's CAC decreased by 33.63%, Japan's NIKKEI declined by 26.85%, and India's SENSEX fell by 17.74%.
With the purpose of examining the impact of the pandemic on stock markets, various researches were conducted in many regions and different methods Onali
Research from 2020 indicated that the rising death tolls in Italy and France negatively influenced stock market returns while positively impacting the VIX returns In contrast, Albulescu (2021) emphasized that the ongoing uncertainty stemming from the COVID-19 crisis heightened volatility in U.S financial markets, subsequently affecting the global financial cycle.
In China's stock market, where COVID-19 was first identified, extensive research has been conducted to enhance understanding of the virus Studies indicate a significant negative correlation between stock returns and both the daily increase in confirmed COVID-19 cases and the daily rise in death tolls (Al-Awadhi et al., 2020) Additionally, further analysis by He et al reinforces these findings over an extended timeframe.
(2020) found that the pandemic negatively impacted stock prices on the Shanghai Stock Exchange, whereas it positively impacted the stock prices on the Shenzhen Stock Exchange
While in the emerging economy, particularly India, with negative mean returns, the stock market faces losses during the pandemic However, after the end of March
2020, the prices of two stock indexes took an upward trend gradually (Bora &
Basistha, 2021) Besides, Anh & Gan (2020) stated that the daily increase in the number of confirmed COVID-19 cases negatively impacted stock returns
Nearly two years into the COVID-19 pandemic, ongoing research reveals varying results across different countries In Vietnam, however, there is a scarcity of comprehensive studies addressing the stock market's volatility throughout the entire duration of the epidemic This study aims to provide a thorough analysis of the impact of COVID-19 on the Vietnamese stock market, contributing valuable insights to the existing literature.
19 on stock index in Vietnam through find out empirical evidence on VNIndex.
Research objectives and questions
This research determines the existence of the impact of COVID-19 on the volatility of the stock market If it is confirmed that there is an existence of COVID-
19 impact on volatility, this study is going to find out how VNIndex volatility is affected
This article examines the impact of COVID-19 on the volatility of the Vietnamese stock index, focusing on trends observed before and after the pandemic was confirmed Additionally, it provides a comparative analysis of stock index volatility in Vietnam relative to other economies.
To achieve the research objectives, the following questions must be answered:
Question 1 : Is there the impact of COVID-19 on stock index volatility, particularly VNIndex, when it emerged?
Question 2 : If the existence of COVID-19 does affect volatility, what is the relationship between the information of the pandemic and volatility following the process of COVID-19?
Question 3 : Is the stock index volatility in the Vietnamese market in the period of COVID-19 consistent with empirical findings in other parts of the world?
Research subjects and range
- The impact of COVID-19 on VNIndex volatility
- The difference in volatility of stock market between pre- and post-COVID-
- Return of companies listed on Ho Chi Minh Stock Exhange (HOSE) through VNIndex
This study investigates the effect of COVID-19 on the volatility of the Vietnamese stock index, specifically analyzing the VNIndex closing prices from October 2, 2017, to September 30, 2021 This timeframe encompasses both pre- and post-pandemic periods, although it may lead to challenges in clearly distinguishing between the two sub-periods.
Research methodology
In the financial sector, time-series data exhibit key characteristics such as leptokurtosis, volatility clustering, leverage effects, and co-movement in volatilities These traits indicate that the expected size of errors is not constant and that variance is correlated with its historical values Consequently, it is essential to utilize models from the Generalised Autoregressive Conditional Heteroscedasticity (GARCH) family, which have been employed in previous studies to optimize estimation By providing a comprehensive specification of volatility, these techniques facilitate the examination of the relationship between information and volatility.
Research structure
This paper is structured into distinct sections to address the research questions posed It begins with a literature review that provides an overview of COVID-19's impact on economies and stock markets, along with relevant studies The second section outlines the data and methodology used in the research The third part presents empirical findings on the pandemic's effect on stock index volatility Finally, the conclusion offers insights and recommendations for the stock market.
Chapter 1: Introduction, this section provides the reason why the author selects this topic to take into research and presents the problem statements, objectives of the study, research questions, reseacrh methodology and research contributions of the study
Chapter 2: Theoretical framework and literature review, presents general foundation theory of the issue in this study as well as an overall review about relevant researches regarding to it
Chapter 3: Data and methodology, focuses on the introduction of the dataset and the models that are deployed to investigate the impact of the pandemic on stock index volatility
Chapter 4: Empirical results and discussion, withdraws the emprirical results from the estimations of GARCH family models and make some discussion on the findings
Chapter 5: Conclusion and recommendations, summerizes the whole study and gives some recommendations for individuals and institutions in the stock market.
Research contribution
The full impact of COVID-19 on the Vietnamese stock market remains uncertain; however, this article aims to enhance understanding of this unprecedented event and its financial implications It serves as a valuable resource for researchers, providing essential information for related studies and encouraging further exploration into this critical topic.
The author conducts a study utilizing the latest statistics, providing a valuable reference for investors, researchers, and policymakers This research aims to strengthen the foundation of the stock market, ultimately facilitating informed investments, research initiatives, and policy development over time.
THEORICAL FRAMEWORK AND LITERATURE
Vietnam stock market overview
In just 20 years since its inception, the Vietnam stock market has significantly evolved, establishing itself as a crucial component of the domestic economy The market officially launched with the Ho Chi Minh City Stock Exchange in July 2000, marking a key milestone in Vietnam's economic integration As of September 2021, a total of 1,646 companies are listed across three stock exchanges, with a trading volume of 180 billion shares The market capitalization has surpassed 6.8 million billion VND, representing 133.83 percent of the country's GDP, highlighting the Vietnam stock market's vital role in driving economic growth.
The Vietnam stock market consists of three official exchanges: two listed exchanges, the Ho Chi Minh City Stock Exchange (HOSE) and the Hanoi Stock Exchange (HNX), along with one unlisted exchange, UPCoM Among these, HOSE stands out as the largest exchange, hosting the country's biggest companies.
The HOSE imposes stricter requirements for companies seeking to list, including higher charter capital, longer operational history, and enhanced performance standards Specifically, businesses must have a minimum charter capital of VND 120 billion at the time of listing registration, compared to VND 30 billion on the HNX Additionally, companies are required to have operated as a joint-stock entity for at least two years prior to listing on HOSE, whereas the HNX mandates only one year These rigorous criteria aim to ensure greater transparency and stability in the market.
HOSE mandates that listed companies must demonstrate profitable business operations for the past two years, exceeding HNX's one-year requirement Additionally, HOSE stipulates that companies must maintain at least 300 non-major shareholders who collectively hold a minimum of 20 percent of the voting stock, while HNX requires only 100 shareholders with at least 15 percent ownership Furthermore, companies listed on HOSE are obligated to fully disclose all debts owed to internal parties, major shareholders, and related individuals.
In the first three quarters of 2021, 401 companies were listed on the Ho Chi Minh Stock Exchange (HOSE), with an average trading value exceeding VND 19,500 billion per session and an average trading volume of 682.42 million shares per session HOSE's market capitalization represented nearly 75% of the total market, amounting to 81% of Vietnam's GDP for 2020.
VNIndex is an indice that represents all the stocks listed on HOSE from the first operation day, July 28 th 2000 VNIndex is calculated with fomular:
The VNIndex, which reflects the current price (P1i) and volume (Q1i) compared to the base period price (P0i) and volume (Q0i), is continuously calculated and updated during trading hours This real-time data allows investors and the government to assess the overall state of the Vietnamese economy by comparing it with the previous trading day.
Volatility refers to the fluctuations observed in a phenomenon over time and is defined in modeling and forecasting literature as the conditional variance of an asset's return It serves as a statistical measure of the dispersion around the average prices of securities, including individual stocks and market indices In finance, volatility is commonly associated with standard deviation (σ) or variance (σ²) and is measured by the standard deviation of logarithmic returns Generally, higher volatility indicates a higher level of risk associated with the security.
High volatility stock markets are often perceived as unstable due to their tendency to experience significant fluctuations in performance Conversely, low volatility markets are generally associated with greater stability, making them more predictable for investors.
The Vietnamese stock market has experienced significant fluctuations in return volatility due to various developmental events Long (2007) noted that financial liberalization impacted volatility differently during the first seven years of the VNIndex's operation Initially, the market faced its highest volatility in the first year following its establishment However, as time progressed, volatility decreased, indicating a trend towards greater market stability A subsequent wave of volatility arose when the government implemented a second round of liberalization, coinciding with the IPOs of state-owned enterprises.
The launch of index future trading, particularly represented by the VN30-Index, significantly impacted stock market volatility Research indicates that post-launch, volatility levels increased and became more persistent compared to the period before futures were introduced (Truong et al., 2021).
General impact of COVID-19 on the economy
Since its declaration as a pandemic by the WHO on March 11, 2020, COVID-19 has posed a significant threat to the global economy for nearly two years By January 26, 2021, the number of confirmed COVID-19 cases worldwide exceeded 100 million, highlighting the pandemic's extensive impact.
In 2020, the U.S faced a staggering unemployment rate of 14.7%, the highest since the Great Depression, with 20.5 million individuals unemployed, particularly in the hospitality, leisure, and healthcare sectors A month later, the World Bank warned that COVID-19 would lead to the most severe global recession since World War II.
A March 2020 report revealed that the global economy experienced a staggering loss of at least $2.7 trillion (£2 trillion; €2.5 trillion) in output following the official announcement of COVID-19, which is comparable to the annual gross domestic product of the United Kingdom.
In 2020, the real GDP growth of various countries experienced a significant decline, as illustrated in Figure 2.1, which is predominantly shaded in pink and red This stark representation highlights the severe economic impact of the pandemic's emergence and its ongoing spread.
Figure 2.1 Real GDP growth of countries in 2020
According to Agathe Demarais, Global Forecasting Director of The Economist Intelligence Unit (EIU), the global economy is unlikely to return to pre-COVID levels, with most G7 nations potentially requiring four years to regain economic growth Additionally, developing countries are facing greater challenges in overcoming this crisis due to disparities in vaccine access and equity.
When the Covid-19 pandemic was declared by the WHO, global stock markets plummeted sharply From March 9 to March 16, 2020, the US stock market faced its largest decline since 1987, prompting the activation of the "circuit breaker" mechanism three times Additionally, on March 12, ten countries outside the US implemented trading suspensions, with the Philippines notably halting its stock market on March 17, 2020, in response to the crisis.
In addition, volatility levels of the U.S stock market in the middle of March
2020 was approximate to those last seen in the Great Crash (late 1929), Great Depression (the earrly 1930s), Black Monday (October 1987) and Global finanacial crisis (December 2008) (Baker et al., 2020)
Figure 2.2 illustrates the fluctuations in major global stock indices over the course of a year, highlighting two significant events: the onset of the COVID-19 pandemic and the announcement of the first vaccine Following the pandemic declaration, there was a substantial decline in the indices, with the FTSE and MIB remaining in negative territory until January 2021.
Figure 2.2 The impact of coronavirus on stock markets since the start of the outbreak
COVID-19 made its presence felt in Vietnam even before the WHO declared it a pandemic on January 23, 2020 Since then, the country has experienced four distinct waves of the virus The first wave began on January 23, 2020, and lasted for six months, mirroring the duration of the second wave, which commenced on July 25, 2020 The third wave ran from January 28, 2021, to April 26, 2021, marking significant milestones in Vietnam's ongoing battle against the pandemic.
In 2021, Vietnam faced its most severe and rapid COVID-19 outbreak, marking the third wave of the pandemic that began in late April Each wave was accompanied by varying lengths of lockdowns, with the most recent wave, driven by the Delta variant, resulting in Ho Chi Minh City enduring a four-month lockdown, the longest to date.
The COVID-19 pandemic has significantly affected the domestic economy, as evidenced by the GDP growth of just 3.68 percent in the first quarter of 2020, marking the lowest increase for that quarter from 2011 to 2020.
In 2020, amidst the global economic downturn caused by COVID-19, Vietnam stood out as one of the few countries to achieve positive GDP growth, recording a rate of 2.9 percent, the lowest in a decade Specifically, the first quarter saw a GDP increase of 3.68 percent, while the second quarter reflected the early impacts of the pandemic with a growth rate of just 0.39 percent The economy rebounded in the latter half of the year, with GDP growth rates of 2.69 percent and 4.48 percent in the third and fourth quarters, respectively.
As of December 2020, the Covid-19 pandemic has adversely impacted 32.1 million individuals aged 15 and older, resulting in job losses, reduced working hours, and decreased income Notably, 69.2% of these individuals experienced a decline in income, while 39.9% had to cut back on their working hours Additionally, approximately 14.0% were compelled to take breaks or temporarily halt their work or business operations.
The COVID-19 pandemic significantly impacted all sectors of the economy, particularly the aviation industry, which experienced a drastic decline in passenger traffic, with numbers dropping by 43.4% compared to 2019 Vietnamese airlines faced operational losses amounting to VND 18,000 billion and a staggering reduction in revenue of VND 100,000 billion compared to the previous year.
On the other brighter side, the total import-export value of the whole country in
2020 reached 545.36 billion USD, rising 5.4 percent compared to 2019 In detail, the value of export goods increased 7 percent and reached $282.65 billion while import value was at $262.70 billion, growing 3.7 percent
In 2020, Vietnam successfully managed the pandemic and demonstrated economic growth despite facing numerous challenges However, the emergence of the Delta variant has triggered a fourth wave of COVID-19, leading to forecasts that the country's GDP will only reach between 2% and 2.5%.
2021, about 4 percentage points lower than the world average (Worldometers, 2021)
In the third quarter of 2021, Vietnam's GDP experienced a significant decline of 6.17 percent compared to the same period in the previous year, marking the steepest drop since the country began calculating and reporting quarterly GDP.
As the afore mention, the fourth wave of COVID-19 in Vietnam began on April
Review of previous researches
2.3.1 Researches on the impact of COVID-19 on stock markets volatility
The COVID-19 pandemic has raised significant concerns for the global economy and jeopardized the stability of financial markets Consequently, it is crucial to investigate the impact of the pandemic on stock markets to enhance their resilience, paralleling the urgency of developing effective medical treatments.
He et al (2020) conducted a pioneering study on the market performance and response trends of Chinese industries during the COVID-19 pandemic Utilizing an event study methodology, they analyzed a sample from the Shanghai and Shenzhen A-share market, which included 2,895 listed companies starting from June 3rd.
From 2019 to March 13, 2020, stock prices on the Shanghai Stock Exchange experienced a negative impact due to the epidemic, while the Shenzhen Stock Exchange saw the opposite trend The study highlighted that industries such as transportation, mining, electric and heating, and environmental sectors were significantly affected by the pandemic In contrast, the manufacturing, information technology, education, and health industries, which remained essential during lockdowns, demonstrated a positive response and bolstered confidence in the stock market.
A study by Al-Awadhi et al (2020) examined the impact of COVID-19 on the stock market through panel data analysis, focusing on stock prices, market capitalization, and the market-to-book ratio of companies listed on the Hang Seng Index and the Shanghai Stock Exchange Composite Index during the period starting in January.
From March 10 to March 16, 2020, research indicated a negative correlation between stock market returns and the COVID-19 pandemic, specifically highlighting the daily increases in confirmed cases and deaths.
In the other half of the world, Albulescu (2021) studied the impact of COVID-
The study analyzed the volatility of US financial markets, specifically through the lens of the S&P 500's 3-month realized volatility Utilizing Ordinary Least Squares (OLS) regression, the research highlighted that announcements of new global and US infections significantly amplified market volatility Notably, in contrast to the findings of Al-Awadhi et al (2020), this study revealed that the fatality ratio positively and significantly influenced market behavior.
Research on the pandemic's impact has been conducted across various economies, with Yang & Deng (2021) examining stock market returns and government interventions in twenty OECD countries Utilizing a panel regression model with robust standard errors, their findings indicated that stock returns were negatively impacted by rising COVID-19 case numbers Conversely, government measures such as social isolation, testing, and contact tracing positively influenced stock market returns Interestingly, policies aimed at economic support did not show a statistically significant effect on returns.
A study by Bahrini & Filfilan (2020) examined the stock markets of the GCC countries, including Abu Dhabi, Bahrain, Oman, Saudi Arabia, and Qatar, from April 1 to June 26, 2020 Using panel data regression, the researchers found that, similar to many other economies, these stock markets experienced a negative impact due to COVID-19, particularly correlating with the number of confirmed deaths Notably, the number of reported infections did not seem to influence investor sentiment in these nations.
Regarding Vietnam, Anh & Gan (2020) investigated the influence of COVID-
A study analyzed the impact of COVID-19 lockdowns on the stock market performance of 723 listed firms in Vietnam from January 30 to May 30, 2020, by comparing pre-lockdown and lockdown periods Utilizing panel-data regression models, the research revealed that an increase in daily confirmed COVID-19 cases negatively influenced stock returns Interestingly, while stock performance deteriorated before the official lockdown, it showed a positive trend during the lockdown phase The authors attributed this unexpected outcome to Vietnamese investors' confidence in government strategies to manage the pandemic and the favorable stock prices during that time.
2.3.2 Researches on stock indexes volatility with GARCH family models
Also worth pointing out is studies with the methodology of GARCH, for instance, the paper of Onali (2020) which determined the impact of COVID-19 on
US stock market represented by daily price and volume of VIX, Dow Jones and
S&P500 indices from April 8 th , 2019 to April 9 th , 2020 The results based on GARCH (1,1) indicated that changes in the number of new cases and deaths in the
US and six other countries that had more that 1000 cases of deaths at the end of March
In 2020, the U.S stock market returns remained unaffected, but the analysis revealed significant insights through various models Notably, the VAR models indicated that the rising number of COVID-19 deaths in Italy and France negatively influenced Dow Jones returns while positively impacting the VIX Additionally, Markov-Switching models demonstrated that by the end of February 2020, the negative effect of the VIX on stock market returns had intensified threefold.
Gherghina et al (2021) utilized the GARCH (1,1) model to analyze the Romanian stock market from January 2020 to April 2021, focusing on the Bucharest Exchange Trading (BET) index and COVID-19 case numbers in the USA, Italy, and Romania Their findings revealed that stock market volatility surged in the first quarter of 2020, approaching levels seen during the 2007–2009 global financial crisis, followed by a downward trend in the subsequent two quarters Notably, the study found no causal relationship between COVID-19 variables and the BET index.
Osagie et al (2020) investigated the impact of COVID-19 on the stock exchange performance in Nigeria using GARCH family models Their study analyzed daily all-share price data from March 2, 2015, to April 16, 2020, applying both the exponential GARCH (EGARCH) and quadratic GARCH (QGARCH) models to assess market volatility during the pandemic.
As the result, both of the models suggested a negative impact of the epidemic on the Nigerian stock returns but EGARCH (1,1) with SSTD by incorporating the COVID-
The 19-period model emerged as the most effective among competing models during a significant downturn in India's economy, marked by a steep decline in key stock indices According to Bora & Basistha (2021), the BSE Sensex plummeted by 13.2 percent on March 23, 2020, while the NSE Nifty experienced a nearly 29 percent drop during the same timeframe In response to these challenges, the authors employed the GJR-GARCH methodology in their research to analyze the situation.
BSE Sensex was volatile throughout the pandemic while NSE Nifty has no significant react to COVID-19
In a study of five developed Asian economies—Hong Kong, Japan, Russia, Singapore, and South Korea—only Singapore demonstrated significant market volatility, as analyzed using the GARCH (1,1) model (Sharma, 2020).
Ibrahim et al (2020) utilized the GJR-GARCH model to analyze stock index data from February 15 to May 30, 2020, across Vietnam, other ASEAN nations, and select developed markets Their study explored the relationship between the COVID-19 pandemic, government responses, and stock market volatility The findings indicated that, in most countries, government interventions helped to mitigate market volatility, except in China, where the market remained unaffected, and South Korea, which exhibited a different trend Additionally, a negative correlation was found between COVID-19 confirmed cases and market volatility in China and Thailand, while Japan, Laos, and the Philippines experienced a positive correlation Notably, the study revealed that COVID-19 infection rates did not significantly impact the stock markets of other countries, including Vietnam.
Gaps of previous studies
The Vietnamese stock market, like many around the globe, has faced significant impacts due to COVID-19, highlighting the urgent need for initial research in this area However, the lack of studies specifically addressing this context within the Vietnamese stock market poses challenges for researchers, necessitating reliance on empirical findings from international journals This situation underscores a notable gap in existing literature that the author aims to address.
The Vietnamese economy has experienced significant declines in demand and supply due to social distancing measures, curfews, and extended quarantines As a crucial component of the economy, the stock market has faced challenges stemming from the pandemic and its ongoing developments While stock indices are inherently volatile, the pandemic's impact has intensified these fluctuations.
COVID-19 affects stock index volatility is a concern that needs to be put forward
When utilizing the GARCH model, it is crucial to consider its symmetric assumption, especially given the high volume of trading days and market shocks This model's limitation lies in its presumption of a uniform response to news, which can lead to inaccurate conclusions about the connection between information and volatility when asymmetric responses are present To overcome this challenge, the EGARCH model from the GARCH family is utilized, as it effectively captures the asymmetric response of volatility to news, demonstrating that a price decline induces greater volatility compared to an equivalent price increase.
DATA AND METHODOLOGY
Data
The Ho Chi Minh Stock Exchange, the largest in Vietnam, features the VNIndex, a crucial indicator of the market's overall performance This study aims to analyze the impact of COVID-19 on stock market volatility, specifically using the VNIndex as the focal point The dataset, consisting of 1,000 daily closing prices, was sourced directly from HOSE and meticulously verified against other data sources, covering the period from October 2, 2017, to September 30, 2021 The analysis divides the data into two distinct phases: the pre-COVID-19 period, from October 2, 2017, to January 22, 2020, and the post-COVID-19 period, starting from January 23, 2020, the date of Vietnam's first confirmed case, until September 30, 2021.
Return of VNIndex in period t, 𝑅 𝑡 , will be computed in the logarithmic first difference, which means:
𝑅 𝑡 = 𝑙𝑜𝑔 (𝑉𝑁𝐼𝑛𝑑𝑒𝑥 𝑡 /𝑉𝑁𝐼𝑛𝑑𝑒𝑥 𝑡−1 ) where 𝑉𝑁𝐼𝑛𝑑𝑒𝑥 𝑡 , 𝑉𝑁𝐼𝑛𝑑𝑒𝑥 𝑡−1 is the daily closing price of the VNIndex at 𝑡 and
𝑡 − 1 In this paper, two different methods are going to be employed to capture the volatility of VNIndex.
Methodology
3.2.1 Autoregressive Conditional Heteroskedasticity – ARCH model
In 1982, Engle introduced the Autoregressive Conditional Heteroskedasticity (ARCH) process while studying inflation's means and variances in the UK Prior to ARCH, traditional econometric models only allowed for the analysis of constant one-period forecast variances, neglecting the fact that variance is influenced by recent past data Engle demonstrated that maximum likelihood estimators are more efficient than ordinary least squares (OLS) in the context of the ARCH process, and he utilized the Lagrange multiplier procedure to identify the ARCH effect.
3.2.2 Generalized Autoregressive Conditional Heteroskedasticity – GARCH model
The ARCH model is easy to implement, yet it has notable weaknesses Engle (1982) highlights that a high ARCH order is necessary to accurately capture the dynamic behavior of conditional variance, leading to an increase in parameters that diminishes degrees of freedom Additionally, the excessive number of parameters may result in the violation of the non-negativity constraint for conditional variance.
In 1986, Tim Bollerslev addressed the limitations of the ARCH model by introducing the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model, which allows for longer memory and a more flexible lag structure This innovation has become widely utilized in financial data analysis In this context, let \( \epsilon_t \) represent a real-valued discrete-time stochastic process, and \( \psi_t \) denote the information set (σ-field) encompassing all information available up to time \( t \) The GARCH (p,q) process is defined accordingly.
𝜎 𝑡−1 2 where 𝜎 𝑡 2 is conditional variance; 𝛼 0 , 𝛼 𝑖 and 𝛽 𝑗 are parameters to be estimated For p
The ARCH(q) process simplifies to a scenario where the error term, 𝜖ₜ, represents white noise when both p and q are set to zero In contrast to the ARCH(q) model, which defines the conditional variance solely as a linear function of past sample variances, the GARCH(p,q) process incorporates lagged conditional variances as well Both models maintain a non-negativity constraint, requiring parameters such as 𝛼₀ to be greater than zero and 𝛼ᵢ, 𝛽ⱼ to be non-negative Additionally, they adhere to a stationary constraint, ensuring that the sum of the parameters remains less than one.
GARCH processes stand out from homoskedastic models, which assume constant volatility and are commonly employed in basic ordinary least squares (OLS) analysis While OLS focuses on minimizing the deviations between data points and a regression line to achieve an optimal fit, GARCH models account for changing volatility over time, providing a more accurate representation of financial data.
Asset returns exhibit varying volatility over time, influenced by historical variance, which renders a homoskedastic model suboptimal The GARCH model effectively addresses the limitations of the OLS model, offering improved accuracy without being constrained by the weighted ratio in regression analysis.
The GARCH model outperforms the ARCH model by being more efficient and avoiding overfitting issues It adheres to the non-negative constraint, functioning equivalently to an infinite order ARCH model By reducing the estimated parameters from infinity to just three, the GARCH model effectively allows for conditional variance that relies on both the lag order of the residuals and their previous lags.
The ARCH and GARCH models effectively capture volatility clustering and leptokurtosis; however, their symmetric distribution does not account for asymmetry, leverage effects, and coefficient restrictions To overcome these limitations, various nonlinear extensions of the GARCH model have been introduced, including the Exponential GARCH (EGARCH) model developed by Nelson in 1991.
The EGARCH model, proposed by Nelson in 1991, extends the standard GARCH model by effectively capturing both size and sign effects of shocks This exponential GARCH model is represented by the equation: ln(𝜎 𝑡 2) = 𝛼 0 + 𝛼 1 {|𝜀 𝑡−1|.
If 𝛾 = 0, symmetry i.e no asymmetric volatility
If 𝛾 < 0, negative shocks will increase the volatility more than positive shocks
If 𝛾 > 0, positive shocks increase the volatility more than negative shocks
The use of the log of the variance ensures that the conditional variance remains positive, even with negative parameters The 𝛾 coefficient represents the "leverage effect," highlighting the existence of asymmetries in the data The EGARCH model is particularly effective in capturing these asymmetries, as it demonstrates that an unexpected price drop significantly impacts volatility more than an equivalent price increase Therefore, the value of 𝛾 can be utilized to test for the presence of such asymmetries.
Testing for ARCH effect
Before estimating GARCH family models for data analysis, it is essential to detect the ARCH effect to determine the appropriateness of the ARCH model Generally, two methods are used to test for the ARCH effect.
Method 1: Observation of the ACF of the residual and residual squared determines the effect of the ARCH when the residual has no autocorrelation but the residuals have autocorrelation
Method 2: Perform Lagrange Multiplier LM test
Step 1: Estimate the mean equation (a) by the OLS method:
The equation 𝑌 𝑡 = 𝛽 1 + 𝛽 2 𝑋 𝑡 + 𝑢 𝑡 illustrates the relationship between the dependent variable 𝑌 𝑡 and the explanatory variable 𝑋 𝑡, where 𝛽 1 and 𝛽 2 represent coefficients It is important to note that the explanatory variables can include both the dependent variable itself and other factors influencing 𝑌 𝑡 Additionally, when analyzing sample data, the disturbance term is referred to as the residual 𝑒𝑡.
Step 2: Estimate the following secondary regression equations:
Step 3: Calculate T*R 2 Determine the hypothesis as follows:
H0: 𝛾 1 = 0 and 𝑦 2 = 0 and 𝑦 3 = 0 and … and 𝑦 𝑞 = 0: no ARCH effect
From the regression result, we obtain T*R 2 with Chi distribution 2 (𝑞) where
In statistical analysis, the number of observations is denoted as T, while the number of delays is represented by q, which corresponds to the degrees of freedom derived from the sum of q squared components in the equation When calculating the statistical value χ²(q) as T*R², if this value exceeds the critical χ² value from the table (obtained using the CHIINV(α, d.f.) function in Excel), we reject the null hypothesis H0 Conversely, if the null hypothesis is not rejected, it indicates that the data series in question does not exhibit an ARCH effect.
Research models
With experience from Onali (2020) and Gherghina et al (2021), this study apply GARCH(1,1) and EGARCH(1,1) according to (Osagie et al., 2020)
This article investigates the impact of the COVID-19 pandemic on the volatility of the VNIndex by estimating GARCH and EGARCH models It seeks to determine whether the pandemic affects stock index volatility and, if so, to analyze the relationship between COVID-19 information and VNIndex fluctuations To address this, the author utilizes a conditional variance equation that incorporates a dummy variable, assigning a value of zero for the period before the pandemic and one for the period following its onset.
The analysis reveals that the dummy variable 𝑑𝑣 is statistically significant, indicating that the COVID-19 pandemic has influenced the volatility of the VNIndex Consequently, the study divides the examined timeframe into two distinct sub-periods: one preceding the pandemic and the other following its onset.
In the GARCH model, the term 𝜎 𝑡 2 denotes the conditional variance at time t, while 𝛼 𝑖 signifies the news coefficient and 𝛽 𝑗 indicates the persistence coefficient It is essential to note that both coefficients, 𝛼 𝑖 and 𝛽 𝑗, must be non-negative to satisfy the model's constraints.
This study evaluates the impact of COVID-19 on stock index volatility using GARCH and EGARCH models to identify the most suitable time series models The GARCH (p,q) model, developed by Bollerslev in 1986, effectively captures the volatility clustering observed in financial returns Specifically, the focus is on the GARCH(1,1) model, following the methodologies of Onali (2020) and Gherghina et al (2021).
The symmetric GARCH model may occasionally violate non-negativity constraints during estimation, as noted by Nelson & Cao (2010) While these models effectively capture volatility clustering, they fail to address leverage effects To remedy this, asymmetric models like the Exponential GARCH (EGARCH) should be utilized, as they incorporate asymmetries necessary for accurately capturing leverage effects in financial data.
The investigation into the impact of the COVID-19 pandemic on the VNIndex incorporates a dummy variable in both GARCH and EGARCH models A statistically significant dummy variable indicates that the pandemic has influenced the VNIndex, with values of 0 assigned to the pre-COVID-19 period and 1 to the COVID-19 era The GARCH(1,1) model is selected to analyze stock index volatility, while the EGARCH(1,1) model addresses the limitations of GARCH by effectively capturing the leverage effect present in the dataset.
The GARCH models with a dummy variable is presented below:
The equation √𝜎 𝑡−1 2 + 𝜔𝑑𝑣 + 𝑢 indicates that a statistically significant and negative dummy coefficient suggests a decline in the volatility of the VNIndex due to the pandemic Conversely, a statistically significant and positive dummy coefficient implies that the COVID-19 crisis has led to an increase in the volatility of the VNIndex.
RESULTS AND DISCUSSION
Descriptive statistics
Table 4.1 Descriptive statistics of VNIndex daily closing price from 02/10/2017 to 30/09/2021
Obs Mean Median Max Min Std.Dev
Figure 4.2 VNIndex daily closing price graph
From October 2, 2017, to September 30, 2021, the VNIndex experienced significant growth, increasing by over 750 points over four years The index reached a maximum value of 1420.27 and a minimum of 659.21, indicating substantial fluctuations in stocks listed on the HOSE This volatility is further illustrated in Figure 4.2, with a high standard deviation of 149.9492 among the collected samples.
The VNIndex experienced a consistent upward trend from October 2017 until the end of 2019, prior to the COVID-19 pandemic However, following the initial cases in early 2020, the index sharply declined, reaching its lowest point by the end of March 2020 Since April 2020, the VNIndex has demonstrated a significant recovery and upward movement.
Figure 4.3 VNIndex logarithm return chart
The log return of the VNIndex, as illustrated in Figure 4.3, demonstrates significant volatility throughout the observed period, particularly peaking at the end of January 2020, coinciding with the onset of the pandemic Additionally, the graph indicates that the VNIndex return exhibited increased sensitivity in the post-COVID-19 era.
Table 4.4 Descriptive statistics of VNIndex’s return in two sub-periods
Table 4.4 presents the mean and standard deviation of adjusted returns for the periods before and after the COVID-19 outbreak in Vietnam The pre-COVID standard deviation of returns is 1.0731 percent, while the post-COVID figure rises to 1.4782 percent, indicating increased volatility during the pandemic This suggests that the onset of COVID-19 may have influenced the stock index Additionally, the significant difference in the number of observations between the two sub-periods could lead to misleading comparisons regarding volatility Future analysis utilizing GARCH techniques will provide a more comprehensive understanding of the volatility structure.
Unit root test
Before estimating volatility, it is essential to ensure that the time series data is stationary To determine this, the Augmented Dickey-Fuller (ADF) test is conducted to check for the presence of a unit root, indicating whether the data is stationary or non-stationary.
H0: time series has a unit root
H1: time series has no unit root
The results presented in Table 4.5 indicate that the absolute value of the t-Statistic is significant and exceeds the critical absolute values, leading to the rejection of the null hypothesis (H0) This suggests that the time series is stationary.
Table 4.5 Unit root test – ADF test t-Statistic Prob.*
Augmented Dickey-Fuller test statistic -19.47099 0.0000
Test of ARCH effect on the data set
To ensure the reliability and validity of the data for applying GARCH and EGARCH models, conducting an ARCH effect test is a crucial step This test provides essential insights that support the selection of the GARCH family of models over alternative options.
Table 4.6 Heteroskedasticity Test: ARCH effect at Latency 1
Obs*R-squared 49.05292 Prob Chi-Square(1) 0.0000
Variable Coefficient Std Error t-Statistic Prob
Table 4.7 Heteroskedasticity Test: ARCH effect at Latency 7
Obs*R-squared 136.7201 Prob Chi-Square(7) 0.0000
Variable Coefficient Std Error t-Statistic Prob
The ARCH-Lagrange multiplier (ARCH-LM) test is the formal method used to detect ARCH effects This involves regressing the squared residuals (𝜀 𝑡 2) on a constant and the lagged values of itself The outcomes of this regression analysis are presented in Tables 4.6 and 4.7.
The estimation of the ARMA(1,1) model was initially performed, followed by an ARCH effect test The results indicated that the Prob Chi-Square(1) was statistically significant, with a lag coefficient of 0.2217, significant at the 1 percent level, leading to the rejection of the null hypothesis (H0) of no ARCH effect Upon testing with seven lags, it was suggested that the third-order lag may be optimal, as the regression model's estimation coefficients were significant at the 1 percent level Consequently, the data sample is deemed suitable for the GARCH model due to the presence of an ARCH effect at a 1 percent significance level with one degree of freedom.
Empirical findings and results of estimation of GARCH model
4.4.1 Results on existence of the impact of COVID-19 on VNIndex volatility - GARCH(1,1) with dummy variable
Table 4.8 GARCH(1,1) model with dummy variable
The analysis reveals that the dummy variable is statistically significant at the 1% level, indicating a positive relationship between the emergence of COVID-19 and increased volatility in VNIndex returns Furthermore, the significant parameters of both the ARCH and GARCH terms demonstrate that recent news has notably influenced return volatility The combined coefficients of these parameters reinforce the impact of external factors on market fluctuations.
The sum of 0.11841 and 0.819209 approaches 1, indicating that shocks to the conditional variance are likely to be highly persistent This suggests that significant fluctuations in stock returns are typically followed by further substantial changes, while minor fluctuations are likely to be followed by smaller ones Specifically, the coefficient 𝛼 1 is 0.1187, signifying that 11.87 percent of the volatility in the index's return at time t+1 can be attributed to recent news at time t.
4.4.2 Results of GARCH estimation in two sub-periods
Table 4.9 GARCH(1,1) model in two sub-periods
Variable Coefficient Prob Coefficient Prob
The analysis of P-values indicates that all coefficients are highly significant at the 1 percent level, with positive coefficients confirming that the non-negative constraint is upheld The constant parameters, 𝛼 0, are very low at 0.00000214 and 0.00003 for the pre- and post-COVID periods, respectively The ARCH terms, represented by 𝛼 1, demonstrate a minimal impact of past volatility on current periods, with values of 0.083861 and 0.166494 In contrast, the GARCH terms, indicated by parameter 𝛽 1, are significantly higher at 0.900273 pre-COVID and 0.686393 post-COVID, suggesting that past squared errors are critical in determining present variances The combined sum of 𝛼 1 and 𝛽 1, which approaches 1, signifies that today's shocks will influence variance forecasts for an extended duration.
The EGARCH model is implemented to uncover more practical results compared to the GARCH model, particularly by accounting for the leverage effect, which is a significant advantage of this approach.
Empirical findings and results of estimation of EGARCH model
4.5.1 Results on existence of the impact of futures trading introduction on spot price volatility – EGARCH (1,1) with dummy variable
Table 4.10 EGARCH(1,1) model with dummy variable
From the result of EGARCH(1,1) model, it is evident that the dummy variable is strongly statistically significant at 1% significance level which means the COVID-
The emergence of COVID-19 has significantly influenced VNIndex returns, with a surprising positive coefficient of 0.090372 indicating an increase in stock index volatility This suggests that the pandemic has heightened the fluctuations in the closing prices and returns of the VNIndex, a finding that aligns with the results observed in the GARCH(1,1) model.
The “news” parameter with the coefficient 𝛼 1 (ARCH term) and the
The estimation reveals that the "persistence" parameter with the coefficient 𝛽1 (GARCH term) is significant, indicating its importance in the analysis Additionally, the leverage effect coefficient, 𝛾, is also significant, negative, and different from zero, suggesting the presence of leverage in the sample-period returns and highlighting the asymmetric impact of news Following this, the EGARCH (1,1) model is utilized to further investigate the effects of the pandemic on stock index volatility across two distinct sub-periods.
4.5.2 Results of EGARCH estimation in two sub-period
Table 4.11 EGARCH(1,1) model in two sub-periods
Variable Coefficient Prob Coefficient Prob
All parameters in both periods demonstrate high statistical significance at the 1 percent level, with the ARCH term (𝛼 1) showing a p-value of 0.0007, indicating that the size of the shock significantly affects return volatility Additionally, the leverage effect term (𝛾) reveals an asymmetric influence of news The parameter 𝛽 1 has a p-value of 0.0000, suggesting that past volatility is a valuable predictor of future volatility Similarly, the parameters in the post-COVID period exhibit highly statistically significant coefficients, maintaining consistent signs with those observed in the earlier period.
The ARCH parameters from two distinct periods indicate a positive relationship between past and current variance, suggesting that larger shocks to variance lead to increased volatility Specifically, the coefficient 𝛼 1 is 0.141278 during the pre-COVID period and rises to 0.203924 in the post-COVID period, signifying that new information has influenced closing prices more swiftly after the pandemic's onset The ongoing rise in confirmed COVID-19 cases and deaths, coupled with uncertainty surrounding the virus, has significantly impacted the stock market, particularly the VNIndex.
The GARCH effect shows a decline from 0.949105 to 0.843690 during the COVID-19 pandemic, indicating that past news has a weaker influence on the volatility of the HOSE stock index Negative coefficients suggest that negative news increases volatility more than positive news of the same magnitude, highlighting the leverage effect The pandemic itself, along with reports of new infections and fatalities, has been perceived as "bad news," leading to heightened stock index volatility Such information often triggers uncertainty and panic among investors, contributing to fluctuations in stock prices.
Comparision with previous researches
The COVID-19 outbreak significantly increased the volatility of the VNIndex, as evidenced by this study's findings, which align with Bora's research.
According to Basistha (2021), the BSE Sensex in India experienced increased volatility during the COVID-19 pandemic Similarly, this trend was also observed in the U.S stock market, as indicated by the fluctuations in the S&P 500 index (Albulescu).
2021) and Singapore stock market (Sharma, 2020)
Despite the significant damage inflicted by COVID-19 across various sectors, a GARCH model analysis revealed that the pandemic had no discernible impact on the Dow Jones and VIX indices in the U.S (Onali, 2020; Gherghina et al.).
(2021) was another authors who found that COVID-19 had not affected on stock market and in this case, it was Romanian’s BET index
Different from the implementation of this study, other researches picked the updates of the pandemic as research subjects Among all terms relating to COVID-
The announcement of new COVID-19 cases and confirmed deaths has led to investor panic and negative decisions in stock markets, particularly in twenty OECD countries, China, and Thailand, as highlighted by studies from Yang & Deng (2021), Al-Awadhi et al (2020), and Ibrahim et al (2020) Conversely, the GCC and Vietnam stock markets remained unaffected by such news, while Japan, Laos, and the Philippines experienced positive impacts on their stock indices (Ibrahim et al., 2020) Additionally, research on the stock market's behavior before and after lockdowns in Vietnam revealed a negative reaction prior to lockdowns, contrasting with a positive response during the lockdown period (Anh & Gan, 2020).
CONCLUSION AND RECOMMENDATIONS
Conclusion
The COVID-19 pandemic has impacted every sector worldwide, significantly increasing the volatility of stock indexes, as evidenced by numerous studies.
This study explores the influence of COVID-19 on the volatility of Vietnam's stock index, specifically the VNIndex Utilizing both symmetric and asymmetric GARCH models, the research analyzes data spanning from October 2, 2017, to September 2021.
A study conducted on the VNIndex before and after the first confirmed COVID-19 cases in Vietnam reveals that both GARCH(1,1) and EGARCH(1,1) models, incorporating a dummy variable, indicate that the pandemic has positively influenced the index This influence has resulted in increased volatility in both the index and its returns.
Post-COVID-19 analysis using GARCH(1,1) and EGARCH(1,1) reveals that market information reacts more swiftly than before the pandemic, indicating heightened sensitivity to news—both positive and negative Additionally, EGARCH(1,1) findings highlight an asymmetrical response of volatility to news, demonstrating that abrupt fluctuations in stock indexes influence volatility at time t + 1 by approximately 16.29 percent.
Recommendations
The pandemic has significantly increased volatility in the Vietnam stock market, leading to both higher risks and potential profits for investors Following a sharp decline at the end of the first quarter of 2020, the VNIndex has shown a remarkable upward trend, consistently reaching new value records.
Despite the adverse effects of COVID-19 on the economy, the author believes it has created profitable opportunities for stock investors To develop effective investment strategies and reduce risks, investors should closely monitor news related to the pandemic and the fluctuations of the VNIndex The analysis indicates a leverage effect, with information impacting market prices more swiftly, highlighting the need for investors to remain vigilant about market shocks when making investment decisions.
The remarkable rise of the VNIndex and the overall stock market has drawn a growing number of investors, intensifying competition among securities firms As a result, these companies must enhance their service quality by improving employee professionalism, particularly within their brokerage teams Additionally, it is essential for these firms to continuously upgrade their information technology systems to ensure fast data processing and provide investors with comprehensive functionalities in the increasingly volatile stock market during the COVID-19 era.
The stock market remains vulnerable to the effects of COVID-19, with investors and businesses experiencing significant volatility News announcements, whether positive or negative, continue to cause substantial fluctuations in stock prices.
The pandemic has led to financial and operational challenges for many companies, resulting in a decline in business growth that negatively impacts stock prices and shareholders To ensure inclusive and sustainable growth in both the stock market and the broader economy, it is essential for authorities to implement policies that provide financial assistance to struggling businesses.
The rising stock market and substantial profits may encourage some investors to excessively utilize margin services, increasing the risk of defaults and negatively impacting stock companies during market volatility Therefore, it is essential for policymakers to implement stricter regulations to manage margin activities within the stock industry.
The release of news, particularly negative news, can lead to increased volatility, highlighting the necessity for comprehensive regulations on information disclosure This need extends beyond the financial sector to encompass various aspects of life, ensuring that information is shared responsibly and transparently.
Limitation and further research
The VNIndex serves as a focal point for analyzing the impact of COVID-19 on Vietnam's stock market; however, the author believes that this index alone may not adequately reflect the overall effects on the entire market Therefore, incorporating two additional indices would provide a more comprehensive understanding of the pandemic's influence on Vietnam's stock market.
This study analyzes data from October 2017 to September 2021, but the ongoing Coronavirus pandemic continues to impact the economy and stock market volatility, potentially for years to come Consequently, the author suggests that researchers focus on continuously updated datasets to better understand these effects.
This paper analyzes the stock index fluctuations before and after the emergence of COVID-19, highlighting the need for further research Future studies could focus on the effects of the pandemic on stock index volatility during each wave of COVID-19 in Vietnam, as well as the impact of significant announcements related to the disease on market performance.
This paper utilizes the standard GARCH model to analyze the effects of COVID-19 on market volatility Although various GARCH family models exist to address the limitations of traditional regression or homoscedasticity approaches, time constraints have prevented the author from exploring these alternatives Models such as GJR-GARCH, I-GARCH, and T-GARCH may offer improved estimations, suggesting that relying solely on GARCH and EGARCH might not fully capture the market's volatility.
9 tháng, tổng giá trị giao dịch trên HOSE tăng 290,69% (2021) Truy cập
15/10/2021, từ https://tapchitaichinh.vn/kinh-te-vi-mo/9-thang-tong-gia-tri-giao- dich-tren-hose-tang-29069-339748.html
Trong nửa đầu năm 2020, thị trường chứng khoán thế giới và Việt Nam đã trải qua nhiều biến động đáng kể Bài viết của Dương (2020) phân tích tình hình này và cung cấp cái nhìn tổng quan về sự phát triển của thị trường Đồng thời, nghiên cứu của Nam, Dũng và Loan (2021) cũng chỉ ra tác động của đại dịch COVID-19 đến hoạt động xuất nhập khẩu của Việt Nam, nhấn mạnh những thách thức mà nền kinh tế phải đối mặt trong bối cảnh toàn cầu hóa.
Kinh tế toàn cầu đã chịu tổn thất nghiêm trọng do đại dịch Covid-19, và việc phục hồi sẽ gặp nhiều khó khăn trước năm 2022 Các tác động kinh tế từ đại dịch đã khiến nhiều quốc gia phải đối mặt với thách thức lớn trong việc khôi phục tăng trưởng.
Nguyên, L (2021) 5 vấn đề để hàng không Việt phát triển hình chữ V Truy cập 07/10/2021, từ https://vneconomy.vn/5-van-de-de-hang-khong-viet-phat-trien- hinh-chu-v.htm
Tổng cục thống kê (2020) Báo cáo tình hình kinh tế - xã hội quý IV và năm
2020 Truy cập 10/10/2021, từ https://www.gso.gov.vn/du-lieu-va-so-lieu-thong- ke/2020/12/baocao-tinh-hinh-kinh-te-xa-hoi-quy-iv-va-nam-2020/
Tổng cục Thống kê (2021) đã công bố báo cáo về tác động của dịch COVID-19 đối với tình hình lao động và việc làm trong quý IV và năm 2020 Báo cáo này cung cấp cái nhìn tổng quan về những ảnh hưởng tiêu cực mà đại dịch gây ra cho thị trường lao động, bao gồm sự giảm sút việc làm và thay đổi trong cơ cấu lao động Để tìm hiểu chi tiết, bạn có thể truy cập vào liên kết sau: https://www.gso.gov.vn/du-lieu-va-so-lieu-thong-ke/2021/01/bao-cao-tac-dong-cua-dich-covid-19-den-tinh-hinh-lao-dong-viec-lam-quy-iv-va-nam-2020/.
Tổng cục Thống kê đã công bố báo cáo tình hình kinh tế - xã hội quý III và 9 tháng năm 2021 Báo cáo này cung cấp cái nhìn tổng quan về các chỉ số kinh tế, tình hình sản xuất, tiêu dùng và các yếu tố xã hội trong bối cảnh phục hồi sau đại dịch Để tìm hiểu chi tiết về những diễn biến này, bạn có thể truy cập vào trang web chính thức của Tổng cục Thống kê tại địa chỉ https://www.gso.gov.vn/du-lieu-va-so-lieu-thong-ke/2021/09/bao-cao-tinh-hinh-kinh-te-xa-hoi-quy-iii-va-9-thang-nam-2021/.
The study by Al-Awadhi et al (2020) examines the effects of the COVID-19 pandemic on stock market returns, highlighting the relationship between death rates from contagious diseases and market performance The findings indicate a significant impact of the COVID-19 virus on financial markets, suggesting that investor sentiment and market volatility are closely tied to the pandemic's progression This research contributes to the understanding of behavioral finance by illustrating how health crises can influence economic indicators and investor behavior.
Albulescu, C T (2021) COVID-19 and the United States financial markets’ volatility Finance Research Letters, 38(March 2020), 101699 https://doi.org/10.1016/j.frl.2020.101699
Anh, D L T., & Gan, C (2020) The impact of the COVID-19 lockdown on stock market performance: evidence from Vietnam Journal of Economic Studies, 48(4), 836–851 https://doi.org/10.1108/JES-06-2020-0312
Aslam, F., Mohmand, Y T., Ferreira, P., Memon, B A., Khan, M., & Khan, M
(2020) Network analysis of global stock markets at the beginning of the coronavirus disease (Covid-19) outbreak Borsa Istanbul Review, 20, S49–S61 https://doi.org/10.1016/j.bir.2020.09.003
Bahrini, R., & Filfilan, A (2020) Impact of the novel coronavirus on stock market returns: evidence from GCC countries Quantitative Finance and Economics, 4(4), 640–652 https://doi.org/10.3934/qfe.2020029
Baker, S R., Bloom, N., Davis, S J., Kost, K., Sammon, M., & Viratyosin, T (2020) The unprecedented stock market reaction to COVID-19 Review of Asset Pricing
Studies, 10(4), 742–758 https://doi.org/10.1093/rapstu/raaa008
Bollerslev, T (1986) Generalized autoregressive conditional heteroskedasticity Journal of econometrics, 31(3), 307-327
Bora, D., & Basistha, D (2021) The outbreak of COVID-19 pandemic and its impact on stock market volatility: Evidence from a worst-affected economy Journal of
Public Affairs, August 2020 https://doi.org/10.1002/pa.2623
Engle, R F (1982) Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation Econometrica, 50(4), 987 https://doi.org/10.2307/1912773
Gherghina, Ștefan C., Armeanu, D Ștefan, & Joldeș, C C (2021) COVID-19 Pandemic and Romanian Stock Market Volatility: A GARCH Approach
Journal of Risk and Financial Management, 14(8), 341 https://doi.org/10.3390/jrfm14080341
He, P., Sun, Y., Zhang, Y., & Li, T (2020) COVID–19’s Impact on Stock Prices Across Different Sectors—An Event Study Based on the Chinese Stock Market
Emerging Markets Finance and Trade, 56(10), 2198–2212 https://doi.org/10.1080/1540496X.2020.1785865
Ibrahim, I., Kamaludin, K., & Sundarasen, S (2020) COVID-19, Government Response, and Market Volatility: Evidence from the Asia-Pacific Developed and
Developing Markets Economies, 8(4), 105 https://doi.org/10.3390/economies8040105
Long, V T (2007) Empirical analysis of stock return volatility with regime change: The case of Vietnam stock market Department of Economic development and policies
Nelson, D B., & Cao, C Q (2010) GARCH in Constraints the Univariate Model
Onali, E (2020) COVID-19 and Stock Market Volatility SSRN Electronic Journal,
February, 1–24 https://doi.org/10.2139/ssrn.3571453
Osagie, M., Maijamaa, B., & Owoicholofu, D (2020) On the Effects of COVID-19 outbreak on the Nigerian Stock Exchange performance : Evidence from GARCH Models April, 1–23 https://doi.org/10.20944/preprints202004.0444.v1
Sharma, S S (2020) A Note on the Asian Market Volatility During the COVID-19
Pandemic Asian Economics Letters, 1, 1–6 https://doi.org/10.46557/001c.17661
Truong, L D., Nguyen, A T K., & Vo, D V (2021) Index future trading and spot market volatility in frontier markets: Evidence from Ho Chi Minh Stock Exchange Asia-Pacific Financial Markets, 28(3), 353-366
Yang, H., & Deng, P (2021) The Impact of COVID-19 and Government Intervention on Stock Markets of OECD Countries Asian Economics Letters https://doi.org/10.46557/001c.18646
Worldometers (2021) COVID-19 Coronavirus pandemic Retrieved November 20,
2021, from https://www.worldometers.info/coronavirus/
1 Descriptive statistics of VNIndex daily closing price from 02/10/2017 to 30/09/2021
2 Descriptive statistics of VNIndex’s return in two sub-periods
3 Unit root test – ADF test
4 Heteroskedascity Test: ARCH effect at Lantency 1
5 Heteroskedascity Test: ARCH effect at Lantency 7
6 GARCH(1,1) model with dummy variable
7 GARCH(1,1) model in two sub-periods a) Pre-COVID-19 b) Post-COVID-19
8 EGARCH(1,1) model with dummy variable
9 EGARCH(1,1) model in two sub-periods a) Pre-COVID-19