Finance Dissertation 21071809 UWE Dissertation submitted in partial fulfillment of the Requirement for the MSc in Finance FINANCE DISSERTATION ON FORECASTING THE VALUE AT RISK (VaR) FOR THE STOCK MARK[.]
Introduction
Rationale
The stock market, alongside the banking system, plays a vital role in Vietnam's economy, which is recognized as one of the most dynamic emerging markets globally (World Bank, 2022) The Vietnamese stock exchange (VSE) is relatively young, with the Ho Chi Minh Stock Exchange (HOSE) launching on July 28, 2000, and the Hanoi Stock Exchange (HNX) starting on July 14, 2005 (Truong and Friday, 2021) Currently, the VSE consists of two main exchanges: HOSE, which lists companies with a charter capital exceeding VND 120 billion, and HNX, which lists those with over VND 30 billion (State Securities Commission of Vietnam, 2012) The number of listed companies surged from 5 in 2000 to 750 in 2021 (Kokalari, 2021), marking Vietnam's transition from a frontier to an emerging market Market capitalization nearly tripled from USD 52.43 billion in 2014 to USD 186.01 billion in 2020 (The World Bank, 2022) Stock market volatility, a key concept in finance, serves as an indicator of investment return uncertainty, making the modeling and forecasting of this volatility, particularly through Value at Risk (VaR), essential for managing market risk.
Understanding return volatility is crucial for pricing derivatives and developing effective trading and hedging strategies, as extreme fluctuations can disrupt the financial system and prompt structural or regulatory changes.
Over the past two decades, global stock markets have faced significant fluctuations, increasing investment risks The financial crisis of 2007-2009 led to a dramatic decline in asset prices, with the VN index and HNX index in Vietnam dropping nearly 70% in 2008 Since then, the Vietnamese stock market has remained unstable, further exacerbated by the COVID-19 pandemic declared by the WHO on March 11, 2020, which affected over 222 countries and resulted in nearly 2 million deaths The pandemic has severely impacted stock market performance worldwide, including Vietnam, where a nationwide lockdown from April 1 to April 15, 2020, contributed to a 28% decline in market capitalization, totaling a loss of $37.4 billion Consequently, the economic disruptions caused by COVID-19 have led to delays in investment decisions and reduced business profits, necessitating the forecasting of Value at Risk (VaR) for stock markets.
3 investors participating in the market to help them anticipate the greatest loss and then make investment or hedging decisions
Value at Risk (VaR) is an essential tool for measuring and managing market risk, as mandated by the Basel Capital Accord for banks to determine the minimum capital required for market risk (Hull, 2015; Bessis, 2015) Accurate VaR calculations are crucial for banks to effectively manage risk, particularly in today's volatile financial markets, ensuring operational safety The State Bank of Vietnam (SBV) has initiated stricter risk management regulations in line with Basel II standards, requiring commercial banks to implement the minimum capital adequacy ratio starting January 1, 2020, with market risk capital calculated based on VaR (SBV, 2016) Despite the rapid growth of Vietnam's stock market, its development remains rudimentary Research by Vo and Tran (2021) indicates that factors influencing stock earnings volatility in developed markets do not necessarily apply to Vietnam, while Quach et al (2019) highlight contrasting relationships for these factors in the Vietnamese context Consequently, applying international risk assessment models to the Vietnamese market requires careful consideration and evaluation.
Research Gap
Despite existing research, there is a notable lack of studies focusing specifically on Vietnam, particularly in the context of its emerging stock market The limited literature available highlights the need for more exploration into informed trading within this unique market landscape.
A study conducted in 2010 explored the influence of the Vietnam stock market on and from the stock markets of the United States, Japan, Singapore, and China, revealing significant impacts from Japan and Singapore Vuong (2004) identified herd behavior in the Vietnam stock market using cross-sectional standard deviation methods Additionally, Loc, Lanjouw, and Lensink (2010) assessed the market efficiency of Vietnam's stock market through various tests, concluding it was inefficient in weak-form This highlights a gap in research regarding Value at Risk (VaR) methods specific to Vietnam, prompting the current study to analyze the accuracy of different VaR estimation methods using current market return data This research introduces two novel aspects: utilizing the VN-Index's earnings ratio as a market return input for 99% confidence level VaR estimation, and employing three distinct approaches—Historical Simulation Method (non-parametric) and GARCH Model (parametric)—to determine the most suitable VaR estimation method for Vietnam's stock market.
Objectives
This paper aims to assess three widely used methods for forecasting and backtesting Value at Risk (VaR) in the Vietnamese stock market: the non-parametric Historical Simulation Method and the parametric GARCH Model The goal is to recommend the most effective VaR forecasting techniques for participants in the Vietnamese stock market.
This article systematically explores Value at Risk (VaR) as a risk measure, providing an overview of various models and emphasizing the significance of the VaR backtesting process It also presents an overview of the Vietnamese stock market, followed by an assessment of the accuracy of each model in estimating VaR through the backtesting process.
Scope of the Research
This study evaluates three popular methods for calculating and forecasting Value at Risk (VaR) in the Vietnamese stock market, utilizing data from 2008 to June 2022 The research employs a backtesting process based on a 252-day data series, focusing specifically on the context of Vietnam's financial landscape.
In other words, this study was only conducted in Vietnam
Research Questions
To achieve these objectives, this research then focusing on the research question below
Question 1: Among parametric methods and non-parametric methods, which one is the most reliable in estimating VaR?
Question 2: Which is the best method for practically applying forecasting VaR for the stock market in Vietnam?
Research Structure
This paper will be divided into six chapters
Chapter 1: Introduction This chapter points out rationale, purpose, object, and research question of the research
Chapter 2: Literature Review This chapter provides a general overview of Value at
Risk (VaR) and stock market through previous research The general systematic information about forecasting VaR for the stock market is showed in this chapter as well
Chapter 3: An Overview of the Stock Market in Vietnam This chapter aim to show the general overview of stock market in Vietnam
Chapter 4: Data Collection and Methodology This chapter explains the study methodology, and how the researcher gathers data
Chapter 5: Results The results of forecasting VaR and backtesting among each method are mentioned in this chapter to assess which method match best for Vietnam’ s stock market
Chapter 6: Conclusion Summary of key points will be listed in the conclusion
Literature Review
Definition of Value at Risk (VaR)
Understanding risks allows us to adjust our behavior to mitigate potential losses (Engle, 2004) Value at Risk (VaR) quantifies the potential loss of a risky portfolio over a specified timeframe, providing a probabilistic measure of risk Specifically, it indicates the maximum loss we are X percent confident will not be exceeded within a given period, T, where V represents the VaR amount This metric relies on two key factors: the time horizon and the confidence level, making it essential for effective risk management VaR can be derived from the probability distribution of portfolio returns over time.
Value at Risk (VaR) is a crucial metric used to assess the probability distribution of losses over a specified time period, T When analyzing a five-day timeframe, VaR represents the loss at the 3rd percentile of the gains distribution or, conversely, the loss at the 97th percentile of the losses distribution More specifically, VaR can be calculated as the negative of the gain at the (100 - X)th percentile of the gains distribution, or as the loss at the Xth percentile of the losses distribution Defined as a worst-case scenario on a typical day, VaR has garnered significant attention in financial econometrics due to its importance for financial and insurance institutions For instance, a 95% VaR for a European stock portfolio over one month indicates a maximum loss of €4000, suggesting that there is a 95% confidence level that losses will not exceed this amount, with a 5% probability of incurring greater losses.
The Value at Risk (VaR) for a one-month period at a 95% confidence level is €4,000, indicating potential losses in the worst-case scenarios for the investment portfolio VaR serves as a crucial tool for evaluating investment risk by considering both the portfolio's value and the investor's risk tolerance.
Figure 2.1: Calculation of VaR from the probability distribution of the gain in portfolio value Losses are negative gains; confidence level is X%; VaR level is V
Figure 2.2: Calculation of VaR from the probability distribution of the loss in portfolio value Gains are negative losses; confidence level is X%; VaR level is V
Value at Risk (VaR) consists of five key components and is generally calculated daily throughout the asset's holding period It is commonly assessed with a confidence level of either 95% or 99%.
Value at Risk (VaR) offers several advantages for financial analysis, particularly with liquid portfolios, which have values that fluctuate based on market conditions Unlike illiquid assets such as real estate or artwork, VaR is applicable to any liquid asset, allowing for effective risk assessment using a defined probability distribution Additionally, VaR can be utilized for analyzing risk across various scopes, including individual assets, investment portfolios, or entire businesses It also aids in evaluating the competitive risks posed by other firms in the industry.
Value at Risk (VaR) is determined by analyzing the probability distribution of a portfolio's market value, which typically exhibits a normal distribution characterized by significance level and variance A significant limitation of VaR is its assumption that market conditions remain stable during the defined time period, as highlighted by Hull (2015) and Bessis (2015) This assumption can lead to adverse outcomes when VaR is used to manage trader risks, as evidenced by the failures of several investment banks during the unexpected market volatility of 2007 and 2008 Additionally, the risk profile of portfolios can vary significantly, as illustrated by the comparison between two portfolios where one, depicted in Figure 2.3, carries a higher likelihood of substantial losses compared to another shown in Figure 2.1.
Figure 2.3: Probability distribution for gain in portfolio value during time T
The Development of the VaR
In the late 1980s, the simultaneous expansion of investment portfolios and commercial banks, coupled with high volatility, highlighted the urgent need for effective risk management This led to the development and popularization of Value at Risk (VaR) in the early 1990s by a team of scientists and financial mathematicians at JPMorgan Chase VaR is grounded in established principles of probability theory and statistics, building on previous risk assessment methodologies.
Value at Risk (VaR) is a critical tool used by risk managers to quantify their company's risk positions for the board of directors In the late 1990s, the Securities and Exchange Commission mandated that corporations disclose quantitative market risk in their financial statements, leading to VaR becoming the predominant method for such disclosures Additionally, the Basel Committee on Banking Supervision allowed enterprises and banks to utilize their internal VaR calculations to determine capital requirements, enabling them to maintain minimal reserves for potential risks when their VaR is low In response to the evolving global landscape during the late 1980s and early 1990s, various businesses sought to create more sophisticated risk models Understanding credit risk, which refers to the likelihood of loan default, remains crucial in this context.
The rise of derivatives, mortgages, and credit cards packaged by corporations for investors has heightened concerns on Wall Street regarding the balance of risks in JPMorgan's equities and bond portfolios Key questions arise about whether the risks from these portfolios offset or exacerbate each other, how to compare various risk types, and the impact of volatility, interest rate changes, and currency fluctuations on fixed-income investments These considerations are crucial for assessing Value at Risk (VaR) Traders typically buy and sell shares the following day, recalculating VaRs to evaluate the implications of new trades and the potential risk of loss due to reduced transaction volumes.
Value at Risk (VaR) has gained credibility among top executives as a reliable risk assessment tool, particularly after the 1987 stock market crash and the emergence of the "black swan" event, which initially raised doubts about its effectiveness However, it was ultimately determined that the perceived failure of VaR stemmed from human error rather than flaws in the model itself, highlighting the importance of understanding the risks faced by the entire organization Consequently, VaR has solidified its role as a crucial element in the contemporary financial landscape Regulatory bodies like the Securities and Exchange Commission are increasingly concerned about the risks associated with derivatives, emphasizing the necessity for financial firms to disclose these risks to investors, thereby reinforcing the practical significance of Value at Risk as a measure.
An increase in the number of Value at Risk (VaR) metrics in a firm's annual report suggests heightened risk exposure for the company Regulatory authorities determined that VaR disclosure suffices, avoiding intervention to limit the growth of derivatives markets Since its introduction in 1994, with the Risk Metric application, VaR has gained global acceptance as a standard method for measuring and monitoring financial, particularly market, risk.
Since its inception after splitting from JPMorgan Chase, the concept of "Value at Risk" (VaR) has gained significant traction in economic research, particularly following the 1987 stock market crash VaR has evolved into a crucial subfield, widely adopted for assessing financial risk.
Value at Risk (VaR) plays a crucial role in risk management within the financial system, particularly for predicting potential losses in banks' trading portfolios The Basel II regulatory framework established capital requirements for market, credit, and operational risks, emphasizing the significance of market risk as a primary category Banks utilize the VaR technique to calculate their capital charges, which must cover the market risk associated with their trading activities Additionally, financial institutions are allowed to develop internal VaR models to estimate maximum portfolio losses over a ten-day holding period at a confidence level of approximately 99% These internal models require approval from relevant authorities based on specified backtesting procedures.
The accurate calculation of Value at Risk (VaR) is crucial for banks' risk management, enabling them to predict the maximum potential loss of a portfolio over the next 10 days with 99% confidence This is particularly important given the volatility of financial markets Following the global financial crisis, the Basel Committee on Banking Supervision recognized the pro-cyclicality of capital charges for market risk as a significant flaw in the Basel II framework Furthermore, there has been a fundamental review of trading book regulations to better align market risk capitalization with the lessons learned from past financial challenges.
The financial crisis revealed significant shortcomings in the Basel II regulations, leading to the development of Basel 2.5 as a response This updated framework introduced stricter Value at Risk (VaR) calculations, incorporated expected shortfall metrics, and emphasized more stressed scenarios to provide a conservative estimate of risks associated with trading activities Basel 2.5 aims to enhance the resilience of financial institutions in light of the lessons learned from the crisis.
VaR Forecasting Methods
The rate of return on a financial asset's daily price movement at time t is defined as 𝑟 𝑡 = 𝑙𝑛(𝑃 𝑡 /𝑃 𝑡−1 ), where 𝑃 𝑡 represents the asset's closing value on day t The value at risk (VaR) at time t+1, denoted as 𝑉𝑎𝑅 𝑡+1,𝛼, indicates the maximum potential loss of 𝑟 𝑡+1 at a confidence level of 1 − 𝛼, based on information available at time t, and serves as a prediction for losses on day t + 1 (Hull, 2015) Over the years, both non-parametric methods, like historical simulation, and parametric approaches, such as the GARCH model, have emerged as effective techniques for modeling VaR (Abad et al., 2014) These methodologies significantly impact risk measurement and management, which will be discussed in the following sections.
2.3.1 Non-Parametric Method: Historical Simulation Model
The historical simulation approach is the most widely used method for calculating Value at Risk (VaR) in practice, primarily due to its simplicity as a non-parametric forecasting technique Unlike other methods, it does not require assumptions about key parameters such as the mean or standard deviation To compute the VaR for today, we analyze a time series of market returns, starting from r1 up to rt-1, which represents yesterday's returns.
(1 − 𝛼)-percentile p 1-α of the returns r and calculate the formula below (Khindanova & Rachev,
To calculate a 99% 1-day Value at Risk (VaR), we identify the return at the 1% level For instance, with a historical dataset of 1000 daily returns, the 99% daily VaR corresponds to the 10th largest loss (Dowd, 2013).
Historical simulation offers both advantages and disadvantages in financial analysis Its strengths include ease of computation and straightforward interpretability, along with the ability to capture non-normal distribution characteristics like fat tails and skewness without making parametric assumptions (Ahmadi-Javid, 2012) However, its limitations lie in the reliance on the assumption that historical distributions will persist in the future, which emphasizes past data and raises questions about the appropriate timeframe for historical datasets (Pritsker, 2006).
The parametric normal Value at Risk (VaR) method is based on the premise that asset returns follow a specific distribution This allows for the computation of VaR by estimating two key parameters: the conditional mean (µ) and conditional volatility (σ) In practice, it is often assumed that the conditional mean of financial returns is zero To estimate conditional volatility, the widely utilized GARCH family of models is commonly applied (Javed & Mantalos, 2013; Palm, 1996; Drachal, 2015).
As the degrees of freedom (v) increase, the t-distribution approaches the normal distribution, highlighting the pronounced tails of t-distributions with fewer degrees of freedom, which raises the chances of extreme observations (Ringqvist, 2014) Furthermore, the variation of Value at Risk (VaR) at 95% and 99% levels in relation to degrees of freedom is illustrated, demonstrating the impact of this statistical measure on risk assessment (Ringqvist, 2014).
Figure 2.5: t-VaR at different degrees of freedom
Figure 2.4: t- distributions with different degrees of freedom "df" displaying the convergence to the normal distribution as df to infinity
In 1963, Polish mathematician Benoit Mandelbrot observed that market returns exhibit clustered behavior, where significant changes are often followed by even larger fluctuations, while smaller changes lead to minor adjustments (Huang & Lin, 2004) In light of these findings, Engle (1982) introduced the autoregressive conditional heteroscedasticity (ARCH) model to account for the varying variance in time series data This model was further refined into the ARCH (m) model, emphasizing the long-term significance of variance The conditional variance of return series, denoted as σ²ₜ|ₜ₋₁, reflects the volatility based on previous returns Additionally, the squared return, r²ₜ, serves as an unbiased estimator for σ²ₜ|ₜ₋₁ Consequently, the ARCH model can be represented as a regression model, where conditional volatility acts as the response variable and lagged squared returns serve as covariates, exemplified by the ARCH (1) model.
The parameters α and ω are to be estimated in a model where {εt} represents a sequence of independent random variables with zero mean and unit variance The relationship is defined by 𝜔 = 𝛾 ∗ 𝑉𝐿, with VL indicating the long-run variance rate and γ denoting the assigned weight Engle (1982) introduced the ARCH(q) model, which generalizes this equation by including q lags of the squared returns, where q is referred to as the "ARCH order."
Bollerslev (1986) and Taylor (2008) enhanced the GARCH model by incorporating p lags of the conditional variance to improve future outcome predictions The GARCH (p; q) model is defined by its parameters, where q represents the ARCH order, p signifies the GARCH order, and α, β, and ω are positive parameters that require estimation (Cryer and Chan, 2008).
The GARCH model with the smallest number of parameters is known as the GARCH (1,1) model In this model, we give the long-run average variance rate substantial importance
In the equation 𝜔 = 𝛾 ∗ 𝑉 𝐿, 𝛾 represents the weight assigned to the long-run variance rate, while β1 and α1 denote the weights allocated to volatility and returns, respectively It is essential that the sum of these weights equals one to maintain balance in the model.
The Exponentially Weighted Moving Average (EWMA) model is a specific instance of the GARCH (1,1) model, characterized by parameters where γ = 0, α = 1 - λ, and β1 = λ In the GARCH (1,1) notation, the “(1,1)” signifies that the variance is determined by the latest return rate observation and the most recent variance estimate.
The GARCH (p, q) model is designed to calculate variance using the latest p observations of returns and the most recent q variance estimates Among these models, GARCH (1,1) stands out as the most widely used To ensure stability in a GARCH (1,1) model, it is essential that the sum of α1 and β1 remains less than one.
< 1 (Bali & Theodossiou, 2008) Otherwise, the weight applied to the long-term variance is negative If we set E(r 2 t-1 ) = σ 2 t = σ 2 t-1 = σ in the following equation and solve for σ we get the average, unconditional variance:
The GARCH (1,1) model is easy to implement and often aligns well with data, as noted by Dowd (2013) and Jorion (2009) Its simplicity stems from the minimal number of parameters required for application, making it a favored choice in financial modeling, according to Hansen and Lunde.
In their 2005 analysis of 330 ARCH models, researchers found no evidence that more complex ARCH models outperform the GARCH (1,1) model The GARCH model's nonlinearity necessitates the use of maximum likelihood estimation and numerical optimization for parameter estimation, as noted by Theodossiou (2015) This requirement presents a drawback, highlighting the importance of accurately explaining the distribution of the error term central to the ARCH framework.
The assumption regarding the conditional distribution of financial returns is crucial for estimating the parameters of the GARCH model For instance, if the error term is assumed to follow a normal distribution, it impacts the analysis of returns significantly.
If the time series reaches from t = 1; 2… ;T the joint pdf can be written as
Back-testing VaR
Only when the percentage of data that deviates from a VaR model is equal to or substantially equal to 1 minus the confidence level, or = 1, will the model be considered reliable
To validate the model, backtesting is conducted (Stahl et al., 2006; Taylor, 2008) Assuming our time series consists of daily returns for the upcoming T days, we denote the number of backtesting exceptions, or instances where the VaR threshold is exceeded, as X The failure rate can be calculated as X/T by estimating the 1% of the weakest observations, referred to as the left tail, over the total period of T days.
21 alternatively, the 1-day 99% VaR Both of these measures can be taken into consideration Because
To ensure optimal performance, we aim to maximize the X/T ratio as T continues to rise This approach allows us to evaluate the accuracy of a Value at Risk (VaR) model: a significantly higher observed X/T ratio indicates an inaccurate model, while a notably lower ratio suggests the model is accurate Although X/T is often termed the empirical size, it is more widely recognized as the nominal size (Jorion, 2009).
To evaluate the accuracy of a model's predictions, we can analyze the number of exceptions, X, which adheres to a binomial distribution This distribution arises from interpreting backtesting exceptions as successes or failures across T trials, where the expected value E(x) = ρT and the variance V(x) = ρ(1 − ρ)T By applying the central limit theorem, we can approximate the binomial distribution with a normal distribution as T increases For instance, with a probability ρ of 0.02 and T set to 1000, we anticipate 20 exceptions Under the null hypothesis H0: ρ = 0.02, we would reject H0 if the z-score falls outside the range of -2.33 to 2.33 (1% significance level), indicating that the model may inaccurately estimate risk.
2.4.1 Unconditional Coverage Back - testing : Kupiec Test
Kupiec's (1995) recommended test closely resembles the previously discussed method, with the key distinction being that the cut-off zones are determined using the log-likelihood ratio This ratio is asymptotically distributed as χ²(1) under the null hypothesis that ρ represents the true exception probability.
We reject the null hypothesis when the likelihood ratio test statistic (LRK) is greater than or equal to the inverse of the chi-squared cumulative distribution function (CDF) with one degree of freedom at a specified probability level, Pr, and significance level, αK Table 2.1 presents calculated regions indicating the permissible number of backtesting exceptions, X, that can occur without leading to the rejection of the null hypothesis.
Probability level VaR confidence level T = 252 days T = 510 days T = 1000 days
Table 2.3: Non-rejection regions for number of backtesting exceptions X
The Kupiec test is classified as an unconditional coverage model, as it does not account for temporal changes in data An interesting area of exploration would be to determine whether exceptions in the data are randomly distributed over time or if they tend to cluster, presenting a compelling question for further investigation.
In 23 specific situations, the conditional coverage model should be preferred over the unconditional coverage model, as highlighted by Dowd (2013) This article will further explore the model developed by Christoffersen (1998) in the subsequent section.
2.4.2 Coverage Back - testing : Christoffersen Test
Christoffersen's conditional coverage test, introduced in 1998, builds upon the Kupiec test as its initial evaluation The second evaluation, known as the serial independence test, assesses the independence of backtesting exceptions In this context, Tij is defined as the number of days a state j occurs while in state I from the previous day, with a value of 1 for a backtesting exception and 0 otherwise The likelihood of observing an exception given state I is represented by I The conditional expected Tij-values can be calculated as outlined in Table 2.2, leading to the relevant test statistic, where LRind denotes the log-likelihood ratio for the independence test under the null hypothesis H0: π = π0.
The probability of encountering an exception on any given day is independent of prior days, as represented by the formula π1 = (T01 + T11)/T Christoffersen (1998) demonstrated that this test statistic follows a χ² distribution with 1 degree of freedom in the asymptotic sense It's important to note that the Kupiec test for unconditional coverage also conforms to a χ²(1) distribution, resulting in the joint test being asymptotically χ²(2)-distributed.
We reject the accuracy of the exception probability ρ and the independence of exceptions when the likelihood ratio test statistic (LRK) is greater than or equal to the inverse of the chi-squared cumulative distribution function (CDF) with 2 degrees of freedom at a specified probability level.
Pr, and α CC the significance level of the test
Table 4.2: Exception table with expected number of exceptions
T 01 for example corresponds to the expected number of days where one day had an exception while there was no exception on the previous day
An Overview of the Stock Market in Vietnam
Main Feature of Stock Market in Vietnam
Securities, as defined in Clause 1, Article 4 of Vietnam's Securities Law 2019, encompass a range of assets including stocks, bonds, fund certificates, warrants, covered warrants, share purchase rights, depository certificates, derivative securities, and other government-regulated securities Furthermore, Clause 14 of the same article outlines that the stock market encompasses all activities related to securities, including offering, listing, trading, investing, providing securities services, disclosing information, and other activities specified by the law.
The Vietnamese stock market plays a crucial role in the economy, serving as a vital source of capital and an investment avenue for the public, with fluctuations potentially impacting economic stability (SSC, 2019) It offers significant liquidity, prioritizes capital distribution, and enhances the flow of funds to meet economic development needs Additionally, the stock market enables the government to raise funds without triggering inflation, which is particularly beneficial given the current constraints on state sector investment capital Financial and economic experts note that the stock market accurately reflects economic outlook changes on a semi-annual basis, with rising stock prices indicating positive economic trends.
26 economy is expanding; on the other hand, a fall in stock prices is not a very strong indicator of the prospects for an economy in the near future
3.1.2 Functions of Vietnam's Stock Market and Its Participants
Functions of Vietnam's Stock Market
The Vietnam Stock Exchange plays a crucial role as a financial intermediary by fulfilling five primary functions that significantly contribute to the economy Firstly, it attracts substantial investment capital through the issuance of shares and bonds by listed companies, generating a significant cash flow that enables businesses to grow and expand Secondly, the stock market fosters an environment where the general public can invest their money, allowing individuals of legal age to explore financial investment opportunities.
The stock market plays a crucial role in enhancing liquidity in the securities market, allowing stocks to be easily bought and sold without significantly impacting their prices High liquidity in stocks means they can be traded effortlessly, maintain stable prices, and have a lower risk of price fluctuations, increasing the likelihood of recovering initial investments Additionally, the stock market serves as a barometer for evaluating business performance, with stock indexes like P/B and P/E reflecting the value of companies and the overall economy Furthermore, it creates an environment that aids the government in implementing effective macroeconomic policies.
27 to prevent budget deficits, manage inflation, and direct economic investment toward activities that are most likely to be productive
There are 4 entities participating in the Vietnam stock market, including: issuers, securities investors, organizations trading in the stock market, and organizations related to the stock market (Linh & Ly, 2022)
Issuers, including governments, local governments, and companies, play a crucial role in raising capital through the stock market Governments issue bonds to address budget deficits and fund significant national projects, while local governments utilize local bonds to finance construction and social development initiatives Companies, on the other hand, issue stocks and bonds to secure capital for production expansion and business development.
Stock investors can be categorized into individual and institutional investors Individual investors may be risk-averse, while institutional investors encompass entities such as investment firms, insurance companies, social insurance funds, financial institutions, and commercial banks A shared characteristic among all these investors is their ability to swiftly open online securities accounts for trading on the stock market.
Organizations operating in the stock market consist of three main entities: securities companies, securities investment funds, and financial intermediaries Additionally, various related organizations play a crucial role, including state management agencies, stock exchanges, the Vietnam Association of Enterprises and Securities, the Vietnam Securities Depository Centre, securities computer service companies, securities financing organizations, and credit rating agencies.
3.1.3 Principles of Operation in the Stock Market in Vietnam
The stock exchange in Vietnam operates under three fundamental guidelines: the disclosure principle, the auction principle, and the intermediate principle These principles are essential for regulating market activities and ensuring transparency and fairness in trading.
To maintain fairness in securities transactions and protect investors' interests, all activities on the Vietnamese stock market must be transparent Securities issuers are legally required to fully disclose information regarding financial statements, business results, security quantities, prices, and corporate governance Public information must meet four essential criteria: it must be accurate, timely, easy to understand, and complete.
The stock market exemplifies perfect competition and operates as the freest market, characterized by three auction forms: direct, indirect, and automatic auctions In direct auctions, brokers negotiate prices face-to-face, commonly seen in major exchanges like Tokyo and New York Indirect auctions involve investors negotiating through phone or online platforms, as practiced on the London Stock Exchange Lastly, automatic auctions utilize an internet system where buy and sell orders are matched by the exchange's server, a method prevalent in stock markets such as Thailand and Vietnam.
The principle of intermediaries mandates that all securities transactions must be conducted through a broker, ensuring the market operates regularly, safely, and cost-effectively This is particularly crucial in centralized stock markets, such as the Stock Exchange, where brokers facilitate all buying and selling activities In the primary market, this principle is reflected in the issuance of securities via auction or underwriting processes.
Current Situation Vietnam's Stock Market
In recent years, the Vietnam Stock Exchange has attracted significant attention from international financial markets, with MSCI and FTSE Russell closely monitoring its indices The Vietnam market is currently on FTSE's watch list for a potential reclassification as a second-tier emerging market Since 2005, the total floor areas of two commercial districts have expanded considerably Research findings indicate a steady increase in registered stock codes and shares from 2000 to 2018, with approximately 200 codes and 2.6 billion shares listed in 2006, escalating to over 750 codes and more than 90 billion shares by the end of the observed period.
2018 (Figure 2.6) However, the severe effects of the Covid-19 epidemic have also had a negative
Between 2019 and 2022, the Vietnamese stock market experienced a notable impact, evidenced by a decline of around 40 companies listed on the Hanoi Stock Exchange (HNX), while the Ho Chi Minh Stock Exchange (HOSE) saw only a modest increase in listings This growth on the HOSE is significantly less pronounced compared to the trends observed prior to 2020.
Figure 2.6: The total Vietnamese stock market size from 2000 to 2020
Figure 2.7: Number of enterprises listed for HOSE and HNX from 2015 to 2022
Source: Results obtained from website: investing.com
The market capitalization of the Vietnamese economy has shown significant trends, particularly with HOSE, which soared to 3,000 billion VND in 2018—double its 2015 value—before experiencing a decline due to the Covid-19 pandemic However, the economy has rebounded, with current figures reaching twice the levels seen in 2020 Over the past two years, the Vietnam stock market, especially HOSE, has demonstrated remarkable growth and development.
NUMBER OF ENTERPRISE LISTED, REGISTERED FOR
Covid-19 pandemic, lockdown period in Vietnam
Figure 2.8: Market capitalization in Vietnamese stock market from 2015 to 2022
Source: Results obtained from website: investing.com
The Vietnamese stock market has shown significant growth over its 21-year history, with both market capitalization and primary indexes (VNI, HNXI, and VNIAS) demonstrating a long-term upward trend However, between 2006 and 2022, the market faced three major downturns, notably a decline of around one thousand points following its peak in January 2008, influenced by the global financial crisis and the housing bubble This downturn adversely affected the overall financial system in Vietnam, leading to inefficiencies within the market.
The Covid-19 pandemic, which originated in Wuhan, China, significantly impacted Vietnam, leading to a notable decline in the VN-Index by 3.22 percent on July 24, 2020, due to reports of local transmission after three months of stability The Ho Chi Minh Stock Exchange (HOSE) reflected this downturn, with 368 stocks losing value Additionally, escalating geopolitical tensions between Russia and Ukraine further influenced global stock markets, prompting investors to seek safer investment options such as bonds, gold, and commodities Compounding these concerns, the anticipated interest rate hike by the U.S Federal Reserve is contributing to a cautious market outlook.
Figure 2.9: Indices for stock market in Vietnam
Source: Results obtained from website: investing.com
According to Nhung (2022) and SSI (2022), however, the prospects for the second half of
The outlook for 2022 remains positive due to several supportive macroeconomic factors, such as a stable Consumer Price Index (CPI), rising GDP, and an increase in exports and foreign direct investment (FDI) While there are both optimistic variables and inherent risks, the stock market is expected to present numerous opportunities and promising prospects moving forward.
Credit boom, global financial crisis
Indices in Vietnam stock market
Credit boom, global financial crisis
The FPT trading system significantly enhances market efficiency by effectively addressing order congestion, instilling greater confidence among investors about the market's future This system is crucial for promoting transactions, particularly with the anticipated implementation of odd lots and T0 transactions in early 2022, coinciding with the launch of the new securities trading system from Korea (KRX) This initiative aims to introduce new products while enhancing market liquidity and attractiveness Mr Le Anh Tuan from Dragon Capital supports the market adjustments made in July, noting that despite a 13-14% decline in the VN-Index from its peak, this has led to a maturation of the market He expressed concerns about the current liquidity levels of VND 25,000 to VND 30,000 billion, suggesting that sustaining these levels long-term could be challenging In his assessment, companies on the HOSE with liquidity between VND 15,000 to 17,000 billion represent around 80% of total capital, indicating that the current market consolidation is acceptable.
Data Collection and Methodology
Research Methodology
This research employs qualitative and quantitative methods to analyze the relationship between variables in forecasting Value at Risk (VaR) Utilizing secondary time series data and a rolling window approach with a sample size of 1,000 observations, the initial dataset is used to construct VaR forecasting models These models predict the value of VaR for the following 10 days at a 99% confidence level (α = 0.01), based on the VN-Index movement at the 1,000th observation The sample is then shifted by 22 days for re-estimation, resulting in 257 model estimations and a total of 2,564 VaR predictions The R programming language is employed for backtesting the data through unconditional coverage and independence tests to validate the models and ensure they meet the necessary hypotheses for effective forecasting.
Data Description
This research study analyzes the historical progression of the Vietnam Ho Chi Minh Stock Index (VN-Index) from January 3, 2006, to August 31, 2022 Utilizing a forecasting model that employs backtesting methods, the study relies on secondary data, specifically time series data The data is sourced from reputable platforms, including investing.com, ensuring reliability in the analysis.
Series: VN-Index Sample: 03/01/2006 to 31/8/2022
Table 4.1: Descriptive statistic for VN-Index
Source: Results obtained from R-studio
The earnings ratios of the VN-Index, analyzed from January 3, 2006, to August 31, 2022, reveal significant statistical characteristics, as summarized in Table 4.1 Key metrics such as the mean, median, mode, standard deviation, variance, minimum, and maximum are expressed as percentages The analysis shows a high degree of variance, with a daily standard deviation of 1.44% The maximum volatility recorded is a 4.98% increase, while the minimum is a 6.67% decrease Skewness and kurtosis values suggest the income ratio distribution is left-skewed with a heavy tail, indicating a negative skewness and a deviation from normality The Jarque-Bera test supports this finding, yielding a high test statistic of 282.75 and a p-value of 0.00, confirming that the income ratio distribution significantly deviates from normal distribution.
Results
Movement of Timeseries Dataset VN-Index
Figure 5.1: Movement of VN-Index
Source: Results obtained from R-studio
This study analyzes the volatile nature of the Vietnam stock market by utilizing data from Investing.com, specifically focusing on the VN-Index's closing prices from January 3, 2006, to August 31, 2022 The findings, illustrated in Figure 5.1, highlight the index's volatility and earnings ratio throughout the research period Notably, the chart reveals the impact of the global financial crisis from 2007 to 2009 and the unforeseen effects of the Covid-19 pandemic, both of which led to a significant decline in the VN-Index alongside an increase in variance.
Credit boom, global financial crisis
Backtesting Results
Figure 5.2: 𝑽𝒂𝑹 𝟏% forecast for VN-Index earnings ratio for the models under consideration
Source: Results obtained from R-studio
Method appoarch Violation (%) Kupiec test (p-value)
Note: The table shows the results of backtesting the VaR forecasting for the considered models Source: Results obtained from R-studio
The analysis of the VaR 1% forecast for the VN-Index earnings ratio reveals significant insights into the performance of various models Backtesting results, as shown in Table 5.1, highlight the effectiveness of the historical simulation approach (HistSim1000 and HistSim250) and the GARCH models (GARCH-Norm and GARCH-SGED) The p-values from Kupiec's (1995) unconditional coverage test indicate the accuracy of VaR forecasts in predicting the probability distribution of returns, with Column (2) illustrating forecast violation percentages and Column (3) confirming the independence of these forecasts Notably, the GARCH-Norm model exhibits a high violation rate of 1.724%, leading to its rejection based on the coverage test, which suggests that assuming normal distribution underestimates the VN-Index's risk Conversely, the GARCH-SGED model demonstrates reliable percentile coverage and independent VaR predictions, minimizing the chances of systematic violations.
To accurately assess the risk level in a highly volatile market like Vietnam, it is essential to consider the non-standard distribution of return rates This approach helps prevent erroneous evaluations and ensures a more reliable understanding of market dynamics.
Conclusion
Researching the forecasting methods for Value at Risk (VaR) in the Vietnamese stock market is an emerging topic that has received limited attention This study highlights the significance of VaR in both theoretical and practical contexts, emphasizing its potential to enhance investment decisions for Vietnamese investors By evaluating various forecasting methods, including non-parametric (HistSim) and parametric (GARCH) models, this research aims to predict VaR with a 99% confidence level in real-time The effectiveness of these VaR forecasts is validated through backtesting, ensuring they meet criteria for percentile coverage and independence.
In the first chapter, the researcher outlines the introduction, detailing the rationale, purpose, objectives, and research questions of the study Chapter two provides an overview of Value at Risk (VaR), discussing its development and evaluating various forecasting and backtesting methods The researcher also suggests forecasting methods from previous studies to inform model selection and backtesting Chapter three examines the current state of climate change and economic growth in Vietnam, alongside an overview of the Vietnamese stock market, highlighting its key features and the volatility of market returns as represented by significant indices such as VNI, HNXI, and VNIAS.
Chapter 4 allows the research to focus on analysing and explaining research methodology and the descriptive statistic of time series dataset: VN-Index to run backtesting procedure to see
The study evaluates the effectiveness of various Value at Risk (VaR) forecasting methods, revealing that the GARCH-SGED model, which accommodates non-normal distribution assumptions for VN-Index returns, achieves the highest prediction accuracy In contrast, traditional models that assume a normal distribution tend to underestimate the actual market risk These findings have practical implications for forecasting risk-bearing values and determining minimum capital requirements for market risk at banks, particularly in the context of the Basel II accord, as discussed in Chapter 5.
Abad, P., Benito, S., & López, C (2014) A comprehensive review of Value at Risk methodologies The Spanish Review of Financial Economics, 12(1), 15-32
Ahmadi-Javid, A (2012) Entropic value-at-risk: A new coherent risk measure Journal of
Ahmar, A.S and del Val, E.B (2020) SutteARIMA: Short-term forecasting method, a case: Covid-
The study published in "Science of The Total Environment" highlights the relationship between the stock market in Spain and various environmental factors Meanwhile, Al-Awadhi et al (2020) explore the impact of the COVID-19 pandemic on stock market returns in their article in the "Journal of Behavioral and Experimental Finance," emphasizing how contagious infectious diseases can significantly influence financial markets.
Alfaro, L., Chari, A., Greenland, A.N and Schott, P.K (2020)‘Aggregate and Firm-Level Stock
Returns During Pandemics, in Real Time’Working Paper SeriesWorking Paper [online] National Bureau of Economic Research Available from: https://www.nber.org/papers/w26950 [Accessed 12 September 2022]
Ashraf, B.N (2020) Stock markets’ reaction to COVID-19: Cases or fatalities? Research in
International Business and Finance [online] 54, p.101249
Baker, S.R., Bloom, N., Davis, S.J., Kost, K.J., Sammon, M.C and Viratyosin, T (2020)‘The
Unprecedented Stock Market Impact of COVID-19’Working Paper SeriesWorking Paper [online] National Bureau of Economic Research Available from: https://www.nber.org/papers/w26945 [Accessed 12 September 2022]
Bali, T G., & Theodossiou, P (2008) Risk measurement performance of alternative distribution functions Journal of Risk and Insurance, 75(2), 411-437
BCBS (2006) Basel II: International Convergence of Capital Measurement and Capital
Standards: A Revised Framework - Comprehensive Version Available from: https://www.bis.org/publ/bcbs128.htm [Accessed 12 September 2022]
BCBS (2013) Fundamental review of the trading book Available from: https://www.bis.org/publ/bcbs265.htm [Accessed 12 September 2022]
Berkowitz, J., Christoffersen, P., & Pelletier, D (2011) Evaluating value-at-risk models with desk-level data Management Science, 57(12), 2213-2227
Bessis, J (2015) Risk management in banking 4 th edition John Wiley & Sons
Bollerslev, T (1986) Generalized autoregressive conditional heteroskedasticity Journal of econometrics, 31(3), 307-327
Chang, H L., & Su, C W (2010) The relationship between the Vietnam stock market and its major trading partners–TECM with bivariate asymmetric GARCH model Applied
Christoffersen, P F (1998) Evaluating interval forecasts International economic review, 841-
Coronavirus disease (COVID-19) – World Health Organization (2022) [online] Available from: https://www.who.int/emergencies/diseases/novel-coronavirus-2019 [Accessed 12 September 2022]
Corporate immunity to the COVID-19 pandemic - ScienceDirect (no date) [online] Available from: https://www.sciencedirect.com/science/article/pii/S0304405X21000957 [Accessed
Cryer, J D., & Chan, K S (2008) Time series analysis: with applications in R (Vol 2) New
Ding, W., Levine, R., Lin, C., & Xie, W (2021) Corporate immunity to the COVID-19 pandemic Journal of Financial Economics, 141(2), 802-830
Dong Loc, T., Lanjouw, G., & Lensink, R (2010) Stock-market efficiency in thin-trading markets: the case of the Vietnamese stock market Applied Economics, 42(27), 3519-3532 Dowd, K (2013) Measuring market risk John Wiley & Sons
Duffie, D., & Pan, J (1997) An overview of value at risk Journal of derivatives, 4(3), 7-49
Drachal, K (2015) Review of GARCH model applicability in view of some recent research Journal of Academic Research in Economics (JARE), 7(2), 191-200
Engle, R (2004) Risk and volatility: Econometric models and financial practice American economic review, 94(3), 405-420
Giang, N K., & Yap, L (2020) Vietnam is Asia’s best stock market performer in may
Hansen, P R., & Lunde, A (2005) A forecast comparison of volatility models: does anything beat a GARCH (1, 1)? Journal of applied econometrics, 20(7), 873-889
He, Q., Liu, J., Wang, S., & Yu, J (2020) The impact of COVID-19 on stock markets Economic and Political Studies, 8(3), 275-288
Huang, Y C., & Lin, B J (2004) Value-at-risk analysis for Taiwan stock index futures: fat tails and conditional asymmetries in return innovations Review of Quantitative Finance and
Huang, Y C., & Lin, B J (2004) Value-at-risk analysis for Taiwan stock index futures: fat tails and conditional asymmetries in return innovations Review of Quantitative Finance and
Hull, J (2015) Risk Management and Financial Institutions 4th ed The Willey Finance
Investing (2022) VN Historical Rates (VNI) - Investing.com Available from: https://www.investing.com/indices/vn-historical-data [Accessed 17 September 2022]
Javed, F., & Mantalos, P (2013) GARCH-type models and performance of information criteria Communications in Statistics-Simulation and Computation, 42(8), 1917-1933 Jorion, P (2009) Value at risk [New York]: McGraw-Hill
Khindanova, I N., & Rachev, S T (2019) Value at Risk: recent advances (pp 801-858)
Kokalari, M (2021) What You Need to Know About Vietnam’s Three Stock Exchanges and
Understanding Q1 22 Performance Available from: https://vinacapital.com/wp- content/uploads/2022/04/VinaCapital-Insights-What-You-Need-to-Know-About-
Vietnam%E2%80%99s-Three-Stock-Exchanges.pdf [Accessed 11 September 2022] Kupiec, P H (1995) Techniques for verifying the accuracy of risk measurement models (Vol 95,
No 24) Division of Research and Statistics, Division of Monetary Affairs, Federal Reserve Board
Le, H (2020) Stock market drops over Covid-19 fears Available from: https://e.vnexpress.net/news/business/economy/stock-market-drops-over-covid-19-fears-4135613.html [Accessed 10 September 2022]
Linh, P & Ly, H (2022) Tổng quan thị trường chứng khoán Việt Nam Available from: https://thebank.vn/blog/20261-tong-quan-thi-truong-chung-khoan-viet-nam.html
Lockwood, E (2015) Predicting the unpredictable: Value-at-risk, performativity, and the politics of financial uncertainty Review of international political economy, 22(4), 719-756 Manganelli, S., & Engle, R F (2001) Value at risk models in finance
McAleer, M (2009) The ten commandments for optimizing value‐at‐risk and daily capital charges Journal of economic surveys, 23(5), 831-849
Nguyen, P T L (2020) Essays on the Vietnam Stock Market
Vietnam's stock market is poised for significant upgrades that will serve as a driving force for its future economic growth The enhancements are expected to attract more foreign investments and improve market liquidity As the market evolves, it will foster greater transparency and regulatory compliance, ultimately positioning Vietnam as a competitive player in the global financial landscape These advancements are crucial for sustaining long-term economic development and enhancing investor confidence in the region.
Nieto, M R., & Ruiz, E (2016) Frontiers in VaR forecasting and backtesting International
Nocera, J (2009) Risk mismanagement The New York Times, 2(Jan)
Pritsker, M (2006) The hidden dangers of historical simulation Journal of Banking &
Quach, H., Nguyen, H., & Nguyen, L (2019) How do investors price stocks?—Evidence with real‐time data from Vietnam International Journal of Finance & Economics, 24(2), 828-
Palm, F C (1996) 7 GARCH models of volatility Handbook of statistics, 14, 209-240
Ringqvist, A (2014) Value at Risk on the Swedish stockmarket
The ongoing Russia-Ukraine conflict has significant implications for businesses operating in Vietnam Companies are facing challenges such as supply chain disruptions, increased costs of raw materials, and fluctuating market demands The conflict has also influenced foreign investment strategies, prompting businesses to reassess their operations in the region As Vietnam navigates these challenges, understanding the geopolitical landscape is crucial for maintaining competitiveness and ensuring sustainable growth.
The SBV Circular 41/2016/TT-NHNN outlines the prudential ratios for banks and foreign bank branches operating in Vietnam, ensuring financial stability and regulatory compliance For more information, visit the official document at the provided link.
SSC (2019) Law of security Available from: http://www.ssc.gov.vn/ubck/faces/en/enmenu/enpages_envbpl/securitieslaw;jsessionid=p kY6jpQKv4zyYYcTTdB12pf03dM9NLM2Z2cxRTX4yVr1kVrgyBHT!-
1940394760!744835568?_afrLoop98058689402000&_afrWindowMode=0&_afrWind owId=null#%40%3F_afrWindowId%3Dnull%26_afrLoop%3D398058689402000%26_a frWindowMode%3D0%26_adf.ctrl-state%3Djv3ktvbbp_4 [Accessed 12 September
SSI (2022) Strategy Report Available from: https://www.ssi.com.vn/en/organization- customer/strategy-report [Accessed 17 September 2022]
Stahl, G., Carsten, W., & Zapp, A (2006) Backtesting beyond the trading book Journal of
State Securities Commission of Vietnam 2012 Decree No 58/2012/NÐ-CP Detailing and
Guiding the Implementation of a Number of Articles of the Securities Law and the
Amending Law Available online: https://thuvienphapluat.vn/van-ban/Chung-khoan/ Nghi-dinh-58-2012-ND-CP-huong-dan-Luat-chung-khoan-Luat-chung-khoan-sua-doi- 144157.aspx (accessed on 12 April 2021)
Taylor, S J (2008) Modelling financial time series world scientific
Tett, G (2009) Fool's gold: How the bold dream of a small tribe at JP Morgan was corrupted by
Wall Street greed and unleashed a catastrophe Simon and Schuster
Theodossiou, P (2015) Skewed generalized error distribution of financial assets and option pricing Multinational Finance Journal, 19(4), 223-266
Truong, L.D and Friday, H.S (2021) The Impact of the Introduction of Index Futures on the Daily
Returns Anomaly in the Ho Chi Minh Stock Exchange International Journal of Financial
Studies [online] 9 (3), Multidisciplinary Digital Publishing Institute, p.43
Vo, X V., & Tran, T T A (2021) Higher-order comoments and asset returns: evidence from emerging equity markets Annals of Operations Research, 297(1), 323-340
Vuong, Q H (2004) New empirical results on anomalies and herd behavior: Vietnam stock market 2000-2004 (No zk7ax) Center for Open Science
Viet Nam has transformed from one of the world's poorest nations to a middle-income economy due to economic reforms since 1986, with real GDP per capita increasing over sixfold The country’s economic resilience is evident, with projected growth rates of 6.1% in 2024 and 6.5% in 2025, fueled by global demand and restored consumer confidence Health improvements are notable, as infant mortality rates and life expectancy have significantly enhanced, while universal health coverage now reaches 93% of the population Educational achievements include universal primary education and high secondary enrollment rates, though tertiary education faces challenges in capacity and quality Infrastructure access has dramatically improved, with nearly universal electricity usage and increased clean water access in rural areas Viet Nam aims to become a high-income country by 2045, necessitating sustained economic growth and a commitment to environmental sustainability, including pledges made at COP27 to reduce emissions and achieve net-zero carbon by 2050 The nation must address megatrends like an aging population and climate change while enhancing policy implementation in finance, digital transformation, and low-carbon infrastructure to meet its development goals.
Zhang, D., Hu, M and Ji, Q (2020) Financial markets under the global pandemic of COVID-19