INTRODUCTION
The Efficient Market Hypothesis (EMH), introduced by Fama in 1965, has sparked extensive empirical research regarding stock market behavior in both developed and emerging countries Most studies have concentrated on weak form efficiency, the most basic level of EMH, yielding mixed results While some research, such as that by Hoque et al (2007), Abeysekera (2001b), and Lima et al (2004), rejects the notion of weak form efficiency in stock markets, other studies, including those by Chan et al (1997), Lee (1992), and Worthington et al (2004), provide evidence supporting market efficiency in certain countries.
Despite numerous empirical studies on the weak form of the Efficient Market Hypothesis in both developed and emerging stock markets, research focusing on Vietnam's stock market remains limited This study aims to explore the presence of weak form market efficiency in Vietnam's stock returns and identify any existing anomalies Recognizing these anomalous patterns can empower investors to optimize their buying and selling strategies, ultimately enhancing their returns through effective market timing.
Since its inception on July 28, 2000, with the first security trading center in Ho Chi Minh City, known as Hose, and just two listed companies—Refrigeration Electrical Engineering Joint Stock Company (REE) and Saigon Cable and Telecommunication Material Joint Stock Company (SACOM)—the Vietnam stock market has experienced remarkable growth Over the past decade, the number of listed companies has surged to 635, showcasing the market's resilience in overcoming various challenges and difficulties.
As of now, the market capitalization of Vietnam's stock market stands at 650.150 trillion VND, with 523.933 trillion VND on the Hose and 121.217 trillion VND on the HNX The market capitalization to GDP ratio has shown a steady increase, rising from 0.24% in 2000 to 0.37% in 2010 Currently, there are 102 licensed securities companies with a total registered capital of 31.866 trillion VND (approximately 1.528 billion USD) The total number of trading accounts has reached about 1,031,000, including 15,000 accounts held by foreign investors, a significant increase from just 2,908 accounts in 2000 This rapid growth of the Vietnam stock market continues to attract both domestic and foreign investors.
Despite the rapid development and recent liberalization of the Vietnam stock market, it still exhibits several characteristics typical of emerging markets, including significant information asymmetry, low trading volumes, and inadequate institutional infrastructure, all of which contribute to market inefficiencies.
While many emerging markets exhibit inefficiencies, some researchers have identified evidence of weak form efficiency in developing countries For instance, Lima et al (2004) discovered that both Hong Kong and A shares on the Shanghai and Shenzhen stock exchanges demonstrate weak form efficiency.
Dickinson et al (1994) demonstrated that the Nairobi Stock Exchange operates in accordance with market efficiency, while Moustafa (2004) supported the weak form Efficiency Market Hypothesis for the United Arab Emirates stock market.
Given the theoretical and practical importance of the random walk hypothesis, along with its testable implications and conflicting empirical evidence, we are prompted to reevaluate the concept of weak form efficiency specifically within the context of the emerging Vietnam stock market.
This study investigates the weak form market efficiency and anomalies in the Vietnam stock market by analyzing the daily and weekly returns of the Vnindex, as well as real estate and seafood shares We will assess whether successive stock prices or returns are independently and identically distributed, in line with Fama's (1970) assertion that past stock prices do not predict future prices Additionally, we will adjust the data for thin trading, a significant characteristic of the Vietnam stock market, which may bias empirical studies on market efficiency.
The research provides a number of complementary testing procedures for random walk or weak form market efficiency which have been widely used in the literature
We conduct tests to assess market efficiency in its weak form, emphasizing the information derived from historical prices Specifically, we employ the parametric serial correlation test of independence to analyze the relationship between current stock returns and their previous values Additionally, we utilize a nonparametric run test to evaluate the randomness of stock returns.
The variance ratio test, introduced by Lo and Mackinlay (1988), is utilized to determine the presence of uncorrelated increments in the series, assuming both homoscedastic and heteroscedastic random walks Additionally, we employ Ordinary Least Squares (OLS), Autoregressive Conditionally Heteroscedastic (ARCH), and Generalized Autoregressive Conditional Heteroscedasticity (GARCH(1,1)) models, which are extensively used in the literature, to investigate calendar anomalies in the Ho Chi Minh stock market.
Our analysis, utilizing the latest data and multiple robustness checks, aligns with Loc's (2006) findings that the Vietnam stock market exhibits weak form inefficiency with daily data Notably, we observe a reduction in the extent of this inefficiency when employing weekly observations in our study.
Our research investigates the calendar effect, focusing on calendar anomalies within the Vietnamese market The findings indicate that the day-of-the-week effect is absent in the Vietnam stock market during the analyzed period.
Consequentially, this does not support the findings of Loc (2006) that the day of week effect existing in Vietnam stock market as negative Tuesday effect
The first contribution of our research is that this is one of the studies in Vietnam applying new econometrics, new methodology which has been affected the Brooks’
This study utilizes the 2008 methodology and builds upon previously tested models in the literature Additionally, it offers evidence challenging the existence of persistent anomaly patterns in the Vietnam stock market.
Then, this study also enhances the established literature by providing the most recent analysis of our stock market
This study is organized into several sections: Section two discusses the theoretical background and reviews prior empirical evidence on weak form efficiency in both developed and emerging markets Section three outlines the data and methodology used, while Section four presents the findings of the empirical research.
LITERATURE REVIEW
Review of Literature on Weak Form Market Efficiency
The literature on the validity of the random walk hypothesis in stock markets is extensive and expanding, revealing mixed empirical results While early studies generally support the weak form of the Efficient Market Hypothesis in developed capital markets, more recent research indicates that stock market returns may be predictable This section reviews the literature on weak form efficiency across both developed and developing countries.
The weak form efficiency is tested using the random walk model, a method prevalent in existing literature Various statistical techniques, including the runs test, unit root test, serial correlation test, and variance ratio test, are commonly employed for this purpose Notably, the runs test has been referenced in studies by Fama (1965), Sharma and Kennedy (1977), Cooper (1982), Chiat et al (1983), Wong et al (1984), Yalawar (1988), Ko and Lee (1991), Butler and Malaikah (1992), Abraham (2002), Worthington and Higgs (2004), and Squalli (2006).
The serial correlation test of returns has been extensively utilized in financial research, as demonstrated by Kendell (1953), Fama (1965), Fama and French (1988), Worthington et al (2004), and Squalli (2006) Additionally, the unit root test has been applied in various studies, highlighting its significance in analyzing financial data.
In this study, we employ various statistical tests, including the Run test, serial correlation, variance ratio test, regression test, and ARCH, as utilized in previous research by MacKinlay (1988), Worthington et al (2004), and others such as Dockery and Vergari (1997), Grieb and Reyes (1999), Alam et al (1999), Chang et al (2000), Cheung et al (2001), Abraham et al (2002), Seddighi et al (2004), Loc (2006), and Hafiz et al (2007).
GARCH(1,1)) to enhance the findings of this study
Empirical studies in developed markets consistently support weak form efficiency Groenewold (1997) tests the Australian stock market's weak and semi-strong efficiency using aggregate share price indexes, yielding results that align with weak form efficiency Additionally, Hudson et al (1996) discover that while technical trading rules possess predictive power in the United Kingdom market, they are insufficient for generating excess returns.
Lee (1992) utilizes the variance ratio test to analyze the weekly stock returns of the United States and ten industrialized nations—Australia, Belgium, Canada, France, Italy, Japan, Netherlands, Switzerland, United Kingdom, and Germany—over the period from 1967 to 1988 The study concludes that the random walk model remains a suitable representation of the weekly return series for most of these countries.
Ayadi et al (1994) conducted a variance ratio test to assess the efficiency hypothesis of the Korean Stock Exchange from 1984 to 1988 They found that under the assumption of homoscedasticity, the random walk hypothesis was rejected However, when considering heteroscedasticity, they could not reject the random walk for daily data Additionally, the study utilized weekly, monthly, 60-day, and 90-day interval data for a comprehensive analysis.
The findings did not provide sufficient evidence to dismiss the random walk hypothesis.
Chan et al (1997) investigate the weak form and cross-country market efficiency hypothesis across 18 international stock markets, including Australia, Belgium, Canada, Denmark, Finland, France, Germany, India, Italy, Japan, Netherlands, Norway, Pakistan, Spain, Sweden, Switzerland, the United Kingdom, and the United States, covering the period from 1962 to 1992 Their findings indicate that all examined stock markets are individually weak form efficient, while only a limited number demonstrate evidence of co-integration with other markets, as determined by Phillips-Peron (PP) unit root and Johansen’s co-integration tests.
C.Cheung et al.(2001) employ variance ratio tests with both homoscedasticity and heteroscedasticity to examine random walk hypothesis for Hang Seng Index on Hong Kong Stock Exchange for period from 1985 to 1997 They conduct that Hang Seng follows a random walk model and consequently that the index is weak form efficient
Worthington et al (2004) analyze random walk behavior in 16 developed and four emerging stock markets from 1987 to 2003, employing various methods such as serial correlation, runs, unit root tests, and variance ratio tests Their findings reveal that the random walk hypothesis holds in major European markets, particularly in Germany and the Netherlands, which exhibit weak form efficiency In contrast, Ireland, Portugal, and the United Kingdom show efficiency under specific tests, while the remaining markets do not conform to a random walk The ADF and Phillips-Perron unit root tests reject the random walk hypothesis across all 20 markets, whereas the KPSS tests only fail to reject it for the Netherlands, Portugal, and Poland Additionally, the variance ratio test indicates that the null hypotheses of homoscedasticity and heteroskedasticity are not rejected in the United Kingdom, Germany, Ireland, Hungary, Portugal, and Sweden.
The analysis reveals a rejection of the null hypothesis of homoscedasticity, indicating the presence of heteroscedasticity in the data for France, Finland, the Netherlands, Norway, and Spain.
Among the emerging markets, only Hungary satisfies the strictest requirements for a random walk in daily returns
Kima et al (2008) conducted a study on the efficiency of stock prices in various Asian markets, analyzing weekly and daily data from 1990 Their findings, utilizing new multiple variance ratio tests, indicate that the Hong Kong, Japanese, Korean, and Taiwanese markets exhibit weak form efficiency In contrast, the markets of Indonesia, Malaysia, and the Philippines show no signs of market efficiency Notably, the Singapore and Thai markets achieved efficiency following the Asian crisis.
The evidence regarding weak form efficiency in developing markets is mixed, contrasting with findings from developed markets Many developing countries face challenges such as thin trading, which can lead to market manipulation by large traders While it is commonly believed that developing countries are less efficient, empirical studies do not consistently support this notion For instance, research by Lima et al (2004) analyzing daily stock price indexes from the Shanghai, Shenzhen, Hong Kong, and Singapore Stock Exchanges between 1992 and 2000 indicates that both Hong Kong and A shares from the Shanghai and Shenzhen exchanges exhibit weak form efficiency.
Dickinson et al (1994) conducted an analysis of the Nairobi Stock Exchange utilizing autocorrelation and runs tests, focusing on weekly prices of the 30 most actively traded stocks from 1979 to 1989 Their findings reinforce the weak form of the Efficient Market Hypothesis in the Nairobi Stock Exchange.
DATA AND METHODOLOGY
Data Description
This study utilizes daily and weekly time series data from the Vietnam stock market index and stock prices of real estate and seafood processing companies, covering the period from 2007 to 2010 The data, sourced from the electronic database at cophieu68.com, includes 996 daily and 202 weekly observations The Vnindex serves as the representative index for the Vietnam stock market Real estate stocks, known for their sensitivity to economic changes, and stable seafood processing stocks were selected for analysis Notable real estate stocks such as CII, ITA, SJS, and TDH, all listed before 2007, are included, along with established seafood processing stocks like ABT, AGF, TS4, and FMC, also listed prior to 2007 on Hose.
Returns are calculated as R t =ln(P P t / t − 1 ) (3.1) Where R t is return at time t, P t and P t-1 are price at time t and t-1 respectively
This study builds on previous empirical research by utilizing well-established econometric methods to examine the independence of price data We apply both parametric and non-parametric techniques to test the random walk hypothesis, specifically employing a parametric serial correlation test to analyze the relationship between current stock returns and their previous values Additionally, we implement a run test, a nonparametric approach, to assess the randomness of stock returns The variance ratio test proposed by Lo and Mackinlay (1988) will also be conducted to evaluate heteroscedastic random walks Lastly, we incorporate OLS, ARCH, and GARCH(1,1) models to investigate calendar anomalies present in the data.
Ho Chi Minh Stock exchange
Table 3 1 Descriptive statistics of daily returns
Share ABT TS4 FMC AGF TDH SJS ITA CII VNINDEX
Mean -0.0001 0.0005 -0.0014 -0.0013 -0.0009 0.0006 -0.0002 0.0003 -0.0004 Median 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0001 Maximum 0.0489 0.0491 0.0489 0.0488 0.0488 0.0488 0.0488 0.0488 0.0465 Minimum -0.0513 -0.05156 -0.05856 -0.0513 -0.05133 -0.0513 -0.05132 -0.05129 -0.04802 Std Dev 0.0255 0.03434 0.02893 0.02745 0.02866 0.03098 0.02994 0.02764 0.01939 Skewness 0.0018 -0.0113 0.0285 0.0906 0.0801 0.0022 0.0230 0.0192 -0.03244 Kurtosis 2.5786 1.67762 2.08469 2.30382 2.23305 1.90021 2.11803 2.28742 2.92193 Jarque-Bera 7.34030** 72.15447*** 34.72820*** 21.39102*** 25.37381*** 50.04514*** 32.27168*** 21.06988*** 0.42763
Note: ***, ** and * denote a significance level of 1%, 5% and 10% respectively
Table 3 2 Descriptive statistics of weekly returns
Share ABT TS4 FMC AGF TDH SJS ITA CII VNINDEX
Mean -0.0009 0.0017 -0.0078 -0.0067 -0.0042 0.0030 -0.0010 0.0008 -0.0023 Median 0.0000 -0.0087 -0.0073 -0.0091 -0.0095 -0.0063 -0.0016 0.0000 -0.0037 Maximum 0.1455 0.2424 0.2343 0.2113 0.2339 0.2276 0.2372 0.2373 0.1332 Minimum -0.2459 -0.2459 -0.2089 -0.1671 -0.2433 -0.2427 -0.1973 -0.1906 -0.1626 Std Dev 0.0640 0.1058 0.0715 0.0715 0.0765 0.0943 0.0844 0.0742 0.0538 Skewness -0.5419 0.1692 0.0904 0.3352 0.1837 0.1547 0.2754 0.3171 -0.1065 Kurtosis 4.0181 2.6302 3.8685 3.6033 3.8466 2.9262 3.4604 3.6433 3.2305 Jarque-Bera 18.5197*** 2.0939 6.5908** 6.8138** 7.1340** 0.8474 4.3161 6.8338** 0.8291
The significance levels are indicated as follows: ***, **, and * represent significance levels of 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com for the master's thesis.
Table 3.1 summarizes the descriptive statistics of daily returns for Vnindex and eight individual stocks, including sample means, maximums, minimums, standard deviations, skewness, kurtosis, and Jacque-Bera statistics with p-values Notably, all indexes, except for TS4 (0.0005), SJS (0.0006), and CII (0.0003), exhibit negative mean returns FMC records the lowest minimum return at -0.05856, while TS4 achieves the highest maximum return at 0.04905 The standard deviations of returns vary from 0.01939 for Vnindex to 0.03434 for TS4.
The analysis indicates that the returns of the Vnindex and other stocks are not normally distributed, as evidenced by skewness and kurtosis parameters, which represent the standardized third and fourth moments of a distribution These parameters, along with the Jarque-Bera statistic, help determine normality in data sets Skewness reflects the asymmetry of a distribution around its mean, with a normal distribution having a skewness of zero Positive skewness suggests a long right tail, while negative skewness indicates a long left tail According to Table 3.1, all stocks, except for Vnindex and TS4, exhibit positive skewness, albeit not significantly This positive skewness indicates that the return distributions of shares traded on exchanges have a higher likelihood of yielding positive returns due to the presence of large values in the right tail.
Kurtosis quantifies the peakness or flatness of a distribution, with a normal distribution having a kurtosis of three Distributions with kurtosis greater than three are termed leptokurtic, indicating a peaked shape, while those with kurtosis less than three are classified as platykurtic, signifying a flatter distribution Notably, the kurtosis values for all stocks and the Vnindex are below three, suggesting a platykurtic frequency distribution for the returns of these stocks.
The Jarque-Bera statistics and p-values presented in Table 3.1 indicate that the daily distribution of stock market returns does not follow a normal distribution, as all p-values are below the 0.01 significance level, leading to the rejection of the null hypothesis (Chen et al., 2001).
Weekly returns are determined using the closing stock prices from Wednesday If Wednesday's price is unavailable, the Thursday price is used, or the Tuesday price if Thursday is also unavailable If neither Tuesday nor Thursday prices are accessible, the weekly return is marked as missing This methodology is designed to mitigate the impact of weekend trading and reduce the occurrence of holidays Table 3.2 summarizes the descriptive statistics of the weekly returns for Vnindex and eight individual stocks Similar to daily returns, the analysis indicates that weekly returns are not normally distributed.
Methodology
The autocorrelation test is a widely used method for assessing the dependence or independence of random variables It quantifies the correlation coefficient between a random variable's value at time \( t \) and its value in the preceding period, specifically examining the relationship between current stock returns and their previous values This test is frequently utilized in various empirical studies (Mobarek et al., 2000; Abraham, 2002; Dickinson et al., 1994; Groenewold, 1997; Lima et al., 2004; Islam et al., 2005; Loc et al., 2010).
∑ (3.2) tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
The serial correlation coefficient of stock returns at lag \( k \), denoted as \( \rho_k \), is calculated using the number of observations \( N \), the stock return \( r_t \) over period \( t \), the stock return \( r_{t+k} \) over period \( t+k \), the sample mean of stock returns \( r \), and the lag period \( k \).
The test evaluates whether autocorrelation coefficients significantly differ from zero When autocorrelation is zero, the sample of autocorrelations follows an approximate normal distribution with a mean of 0 and a variance of \(1/T\) This sample autocorrelation facilitates significance testing for autocorrelation coefficients within a specified confidence interval, allowing us to determine if they are significantly different from zero Statistically, the hypothesis of weak form efficiency is rejected if stock returns exhibit successive correlations, indicating that \( \rho_k \) is significantly different from zero.
This study employed the Ljung–Box portmanteau statistic (Q) to assess the joint hypothesis that all autocorrelations are simultaneously equal to zero The computation of this statistic was conducted as follows:
The Q-statistic, which assesses the null hypothesis of zero autocorrelation for the first k autocorrelations (\$ρ_1 = ρ_2 = ρ_3 = \ldots = ρ_k = 0\$), follows a chi-squared distribution with degrees of freedom equal to the number of autocorrelations (k) This statistical approach is crucial for analyzing the presence of autocorrelation in time series data.
Another common approach to test the randomness which has been widely used by many authors (Abraham, 2002, Dickinson et al., 1994, Sharma et al., 1977, Squalli,
The run test, as discussed by Oskooe et al (2006) and Loc et al (2010), is a non-parametric method that assesses the independence of successive price changes without requiring normally distributed returns or constant variance This test operates on the principle that in a random data series, the actual number of observed runs should approximate the expected number A "run" refers to a sequence of consecutive price changes that share the same direction, which can be categorized into three types: upward runs (increasing prices), downward runs (decreasing prices), and flat runs (no price change).
The run test can be structured to analyze the direction of change in stock returns, where a positive change indicates a return exceeding the sample mean, a negative change signifies a return below the mean, and a zero change reflects a return equal to the mean Under the null hypothesis of independence in share price changes, the actual number of runs (R) is counted and compared to the expected number of runs (m) based on the assumption of independence.
Where N is the total number of observations (price changes or returns) and n i is the number of price changes (returns) in each category (
For a sample size greater than 30, the sampling distribution of the mean (m) approaches a normal distribution The standard error of the mean (σ m) can be calculated using the appropriate formula.
The standard normal Z-statistics that can be used to test whether the actual number of runs is consistent with the hypothesis of independence is given by:
In the analysis of stock returns, the actual number of runs (R) is compared to the expected number of runs (m), with a continuity adjustment of -0.5 applied when R is greater than or equal to m, and a positive adjustment otherwise The test is two-tailed due to the observed dependence among share returns at extreme values of R A negative Z value suggests positive serial correlation, indicating a positive dependence of stock prices and a violation of random walk theory, while a positive Z value indicates negative serial correlation The Z distribution follows a normal distribution N(0,1), with critical values of ±1.96 at the five percent significance level, as noted by Brooks (2008).
The variance ratio test, developed by Lo et al (1988), is a powerful and reliable method for testing the random walk hypothesis of stock prices It effectively assesses the null hypothesis under both homoscedasticity and heteroscedasticity conditions As a result, this test is widely utilized by academics and practitioners to evaluate market efficiency based on the acceptance of the null hypothesis.
(Ayadi et al., 1994, C.Cheung et al., 2001, Kima et al., 2008, Lima et al., 2004, Loc et al., 2010, Smith et al., 2003, Y.Liu et al., 1991)
The variance ratio test exploits the fact that if the logarithm of price series follows a random walk, then the return variance should be proportional to the return horizon
The test assumes that the variance of increments in a random walk series is linear over the sample interval Specifically, if a return series adheres to a random walk process, the variance of its q differences will equal q times the variance of its first differences (He, June 1991).
Var p −p − =qVar p −p − (3.7) Where q is any positive integer The variance ratio VR(q) is then determined as follows:
The variance of quarterly increments should be three times larger than that of monthly differences If we have (nq+1) observations of (P 0 , P 1 , , P nq ) at equally spaced intervals, where q is any integer greater than one, the formulas for calculating σ²(q) and σ²(1) are provided in the following equations.
And tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
This study also employs the Lo and MacKinlay’s (1988) heteroscedasiticity- robust standard normal test statistics Under the null hypothesis of homoscedastic, the standard normal test-statistics, Z(q) is defined as:
Here nq is the number of observation and∅ ( ) q is the asymptotic variance of the variance ratio under the assumption of homoscedascity
Finance time series frequently exhibit time-varying volatilities and deviations from normality Therefore, in addition to assessing homoscedasticity, this study employs the heteroscedasticity-robust standard normal test statistics proposed by Lo et al (1988) The heteroscedasticity-consistent standard normal test statistic, denoted as Z*(q), is subsequently defined.
∑ (3.16) tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
Where ∅ * ( ) q is the asymptotic variance of the variance ratio under the assumption of heteroscedasticity: And δ ∧ ( ) j is the heteroscedasticity – consistent estimator and computed as follows:
The calendar effect refers to the fluctuations in stock prices that occur due to seasonal trends at specific times of the year These anomalies pose a significant challenge to market efficiency, as they allow investors to potentially outperform the market by recognizing these patterns This phenomenon contradicts the Efficient Market Hypothesis, which asserts that it is impossible to consistently achieve returns exceeding the market average Notably, calendar turning points, such as weekends, have been identified as critical factors contributing to these violations of the random walk hypothesis.
This study investigates market efficiency in the Ho Chi Minh Stock Exchange by analyzing the day-of-the-week effect on the VNIndex and eight selected stocks from the real estate and seafood processing sectors The analysis employs regression models along with ARCH and GARCH (1,1) methodologies, which have been widely used in empirical research.
1990) tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
EMPIRICAL RESULT
Autocorrelation Test
We conducted an autocorrelation test to assess market efficiency, summarizing the results for daily returns of selected stocks with 20 lags in tables 4.1 and 4.2.
The estimation results indicate significant positive autocorrelation coefficients in observed daily market returns at the 1st, 4th, 5th, and 13th lags, suggesting that consecutive daily returns are likely to share the same sign on the 4th, 5th, and 13th days ahead This positive autocorrelation implies a slow market adjustment to new information, where a positive (or negative) return today is typically followed by an increase (or decrease) in returns over the subsequent days Additionally, the Liung-Box Q-statistics confirm that the autocorrelation coefficients across all 20 lags are significant at the 1% level.
Base on the result of the Liung-Box Q-statistics, the null hypothesis of absence of autocorrelation therefore is strongly rejected Hence the randomness for market index is rejected also
Regarding to the result of individual stocks, the results also show the significant autocorrelation coefficients at the first lags for each individual stock returns series
The autocorrelation coefficients for the selected stocks show positive values at lag one for ABT, FMC, and TDH, while AGF has significant coefficients at lags one, two, three, four, five, and eleven SJS exhibits positive autocorrelation at lags one and four, ITA at lags one and sixteen, and CII at lag one However, the Q-statistics do not support the joint null hypothesis that all autocorrelation coefficients from lag one to twenty are significant at the 1% level for the observed return series of these individual stocks.
The findings from daily thin trading adjusted returns reveal that the random walk hypothesis is rejected for the market index and individual stocks such as CII, ITA, and SJS Conversely, the null hypothesis of random walk holds true for stocks ABT, TS4, FMC, AGF, and TDH, as there is no autocorrelation coefficient present from lag 1 to 20.
Table 4.3 presents the results of the autocorrelation test for weekly returns, revealing that the null hypothesis of a random walk is not supported for the market index and six out of eight selected stocks, with the exceptions being ABT and TS4 However, when weekly returns are adjusted for thin trading, the randomness of the market index and individual stocks, except for AGF and CII, is accepted, as the Q-statistics are not significant across all lags.
The autocorrelation test results for daily and corrected daily returns indicate a rejection of the random walk hypothesis for the market index and all selected stocks Additionally, randomness is also rejected in weekly return observations However, when accounting for thin trading in weekly returns, the null hypothesis of random walk is accepted, with the exception of CII and AGF Consequently, the weak form of market efficiency does not hold in the Vietnam stock market, as the presence of stock return autocorrelation—whether positive or negative—plays a crucial role for investors in developing effective trading strategies.
Our autocorrelation test results align with previous studies in developing markets, such as those by Abeysekera (2001a), Awad et al (2009), and Mobarek et al (2000), which identified a significant presence of strong serial correlation in emerging stock market returns This finding suggests the existence of various imperfections in the functioning of these markets.
Table 4 1 Results of autocorrelation coefficients and Ljung-Box Q statistics for daily returns
ABT TS4 FMC AGF TDH SJS ITA CII VNINDEX
Lag AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat
The significance levels in the study are indicated by symbols: ***, **, and * represent significance levels of 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com.
Table 4 2 Results of autocorrelation coefficients and Ljung-Box Q statistics for adjusted thin trading daily returns
ABT TS4 FMC AGF TDH SJS ITA CII VNINDEX
Lag AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat
The significance levels are indicated as follows: ***, **, and * represent significance levels of 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com for the master's thesis.
Table 4 3 Results of autocorrelation coefficients and Ljung-Box Q statistics for weekly returns
ABT TS4 FMC AGF TDH SJS ITA CII VNINDEX
Lag AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat
The significance levels are indicated by ***, **, and * for 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com for the master's thesis.
Table 4 4 Results of autocorrelation coefficients and Ljung-Box Q statistics for thin trading adjusted weekly returns
ABT TS4 FMC AGF TDH SJS ITA CII VNINDEX
Lag AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat AC Q-Stat
The significance levels are indicated by ***, **, and * for 1%, 5%, and 10%, respectively The latest full download of the thesis is available at the provided email address.
Runs test
The run test results presented in Table 4.5 reject the hypothesis of randomness, particularly in panel A for observed daily returns The actual number of runs (R) for all selected individual stocks and the market return is significantly lower than the expected number of runs Additionally, the Z statistics are significant at the 1% level for all selected stocks and the market return, leading to the rejection of the null hypothesis of independence among stock returns for these series.
The run test results based on daily price observations, as reported in panel B, do not support the null hypothesis of a random walk for all selected stocks, with Z statistics significant at the 1% level Additionally, the negative Z values for returns and prices of all selected stocks and the Vnindex indicate positive serial autocorrelation, aligning with findings from autocorrelation tests Furthermore, the run test results based on thin trading adjusted daily returns also fail to support the null hypothesis of a random walk, except for CII, which is significant at the 1% level.
Table 4.6 presents the results of the run test for weekly returns and prices of selected stocks and the Vnindex The observed weekly returns and prices for Vnindex, CII, and TDH are significant at the 1% level, leading to the rejection of the null hypothesis of a random walk for these entities Additionally, the independence hypothesis for weekly stock prices is rejected for SJS at the 5% significance level However, panel C indicates that the thin trading adjusted weekly returns do not reject the null hypothesis of a random walk for both the market index and all selected stocks.
The results of the run test, excluding thin trading adjusted weekly returns, align with findings from various authors studying emerging stock markets Notably, Abeysekera's research indicates that the number of runs for daily returns is significantly lower than expected This observation is further supported by Loc's (2006) study on the Vietnam stock market, reinforcing the consistency of these findings across different analyses.
The run test analysis reveals that the daily share prices and returns of the selected stocks and the VNIndex are not random, as the probabilities associated with the expected number of runs exceed the observed number of runs This finding reinforces the evidence of weak form inefficiency in the Vietnamese stock markets However, when analyzing weekly returns and price observations, the null hypothesis of independence among the stocks is rejected in 33% and 44% of the total observations, respectively.
Table 4 5 Result of run test for daily price & return
Share VNINDEX ABT TS4 FMC AGF TDH SJS ITA CII
Panel C: Observed thin trading adjusted return
The significance levels are indicated by ***, **, and * for 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com for the master's thesis.
Table 4 6 Result of run test for weekly price & return
Share VNINDEX ABT TS4 FMC AGF TDH SJS ITA CII
Panel A: Observed thin trading adjusted return Observations (N) 203 202 201 202 202 202 202 202 202 n(+) 102 105 94 96 97 94 92 98 103 n(0) 0 0 0 0 0 0 0 0 0 n(-1) 101 97 107 106 105 108 110 104 99
The significance levels are indicated by ***, **, and * for 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com for the master's thesis.
Variance ratio test
The variance ratio tests for daily returns and thin adjusted daily returns of all studied stocks and the VNIndex over the full sample period are presented in Tables 4.7 and 4.8 The notation VR(q) indicates the variance ratio of the returns, while Z(q) and Z*(q) denote the variance ratio statistics under the assumptions of homoscedasticity and heteroscedasticity, respectively To test the random walk hypothesis, VR(q), ∅(q), and Z(q) are calculated for intervals (q) of 2, 4, 8, and 16 observations According to the null hypothesis, which posits that stock returns follow a random walk, the variance ratios are anticipated to equal one.
The estimated variance ratios for all selected stocks and the market return significantly exceed the conventional critical value of 2.57 at the one percent level Consequently, the random walk hypothesis is strongly rejected for all selected stocks and the market return at this level of significance.
The rejection of the random walk hypothesis may be attributed to heteroscedasticity To explore this, the heteroscedasticity-consistent variance ratio test Z*(q) was conducted The results, presented in Table 4.7 for lags 2, 4, 8, and 16, provide strong evidence against the null hypothesis of random walk behavior, as the estimated variance ratios for all observations of selected stock and market returns exceed the conventional critical value of 2.57 at the one percent level.
The analysis of daily returns under thin trading conditions reveals that both homoscedasticity and heteroscedasticity assumptions lead to a rejection of the random walk hypothesis for the market index and all selected stocks, as indicated by significant Z(q) and Z*(q) values at the one percent level (Table 4.8).
Table 4.9 and 4.10 present the results of weekly returns and thin trading adjusted weekly returns under both homoscedasticity and heteroscedasticity assumptions for the market index and selected stocks The findings consistently indicate that the null hypothesis of a random walk cannot be accepted for all observed weekly return series This aligns with Loc (2006) regarding the Vietnam stock market and supports previous studies on emerging markets, such as Abrosimova et al (2005), which found non-randomness in the Russian stock market's daily and weekly data through the variance ratio test Additionally, Hoque et al (2007) reported that stock prices in eight Asian countries do not follow a random walk, with exceptions for Taiwan and Korea The non-random walk conclusion, based on the variance ratio test, is further corroborated by findings from serial correlation and run tests conducted over the full sample period.
Table 4 7 Variance ratio test results for daily returns under homoscedasticity and heteroscedasticity
Number q of base observations aggregated to form variance ratio Variance
Number nq of base observation 2 4 8 16
The significance levels are indicated by ***, **, and * for 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com for the master's thesis.
Table 4 8 Variance ratio test results for thin trading adjusted daily returns under homoscedasticity and heteroscedasticity
Number q of base observations aggregated to form variance ratio Variance
Number nq of base observation 2 4 8 16
The significance levels in the study are indicated by asterisks: ***, **, and * represent significance levels of 1%, 5%, and 10%, respectively The latest full download of the thesis is available at the provided email address.
Table 4 9 Variance ratio test results for weekly returns under homoscedasticity and heteroscedasticity
Number q of base observations aggregated to form variance ratio Variance
Number nq of base observation 2 4 8 16
The significance levels are indicated by ***, **, and * for 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com.
Table 4 10 Variance ratio test results for thin trading adjusted weekly returns under homoscedasticity and heteroscedasticity
Number q of base observations aggregated to form variance ratio Variance
Number nq of base observation 2 4 8 16
The significance levels are indicated by ***, **, and * for 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com for the master's thesis.
Day of week effects
The results of the OSL model, presented in Table 4.11, indicate that daily seasonal anomalies occur randomly in the Vietnam stock market Notably, a negative Tuesday effect was identified in ABT stock at a 10% significance level, while FMC exhibited a negative Monday effect at a 5% significance level, along with a positive Wednesday effect at a 10% significance level and a Friday effect at a 1% significance level Importantly, no day-of-the-week effects were observed in other stocks or the market index.
The OSL results overlook time-varying volatility, prompting the use of ARCH effect testing in this study The findings indicate that the ARCH effect, with a lag of 3 for the residuals of standard OSL, is highly significant at 1% for all selected stocks and the market index This strongly suggests the presence of the ARCH effect in the current analysis, indicating that the GARCH model may be more suitable than the OSL for this study.
The GARCH (1,1) model results, presented in Table 4.11, indicate a greater number of stocks exhibiting weak effects compared to the OSL results Notably, a significant negative Tuesday effect is observed in ABT at the 1% level, while negative Monday effects are found in FMC and AGF stocks at the 5% level Additionally, positive Friday effects are identified in FMC, AGF, and CII stocks at the 1%, 5%, and 10% significance levels, respectively A positive Wednesday effect is also noted in FMC at the 5% significance level Importantly, no day-of-the-week effects are detected for Vnindex and the other individual stocks analyzed.
All coefficients in the variance equations are statistically significant across all models, with γ, δ, and ω each significant at the 1% level for the selected stocks This strong significance of all three terms reinforces the validity of GARCH modeling for the data.
In additional, the finding from day of week effect under thin trading adjusted return has been reported in table 4.12 also show no clear evidence of day of week effect
The findings regarding the day of the week effect are inconclusive, lacking clear evidence of the phenomenon traditionally documented in literature Daily effects appear to manifest randomly across different days, with varying results across stocks and models For instance, the negative Tuesday effect was observed only in ABT, while a positive Friday effect was identified for FMC in the OSL model The GARCH (1,1) model revealed day of the week effects in ABT, FMC, AGF, and CII stocks, particularly highlighting positive Friday effects in FMC, AGF, and CII, and a negative Monday effect in FMC and AGF Additionally, ABT exhibited a negative effect on Tuesday, whereas FMC showed a positive effect on Wednesday These findings do not fully align with the results of Loc (2006) and Hau (2010), which reported a negative Tuesday return effect.
Table 4 11 Results from OSL & GARCH (1,1) models for daily returns for the day of week effects
Year ABT TS4 FMC AGF TDH SJS ITA CII VNINDEX
The article presents the p-values associated with each coefficient of the OSL and GARCH models, indicating significance levels of 1%, 5%, and 10% as denoted by ***, **, and * respectively Additionally, it includes references to various resources and contact information for further inquiries.
Table 4 12 Results from OSL & GARCH (1,1) models for thin trading adjusted returns for the day of week effects
Year ABT TS4 FMC AGF TDH SJS ITA CII VNINDEX
The p-values displayed below each coefficient of the OSL and GARCH models indicate their significance levels, with ***, **, and * representing significance at 1%, 5%, and 10%, respectively For further inquiries or to access the latest full thesis, please contact via email at vbhtj mk gmail.com.
CONCLUSION
This study investigates the weak form efficiency of the Vietnam stock market by analyzing daily and weekly returns of the market index, along with eight selected stocks from the real estate and seafood processing sectors, covering the period from 2007 onwards.
In this research, we adopt the model proposed by Miller et al (1994) to address the issue of thin or infrequent trading, which can significantly distort empirical studies on market efficiency Through the application of both parametric and non-parametric tests, our findings indicate that the Vietnam stock market exhibits weak form inefficiency, even when daily data is adjusted for thin trading effects.
The analysis of weekly data indicates that the weak form efficiency hypothesis is upheld in both the run test and autocorrelation test when returns are adjusted for thin trading However, the variance ratio test does not support the null hypothesis of a random walk These results contrast with Loc (2006), who concluded that the Vietnam stock market exhibits weak form inefficiency across both daily and weekly data in all test results.
Our analysis of the day of the week effect reveals that, except for ABT, FMC, AGF, and CII, the other stocks studied do not exhibit this effect Each stock shows varying daily returns, indicating that daily anomalies are not consistent across the selected stocks Additionally, no common daily pattern is observed, suggesting that the identified daily effects are influenced by the data itself Consequently, the findings are inconclusive, providing no clear evidence of a calendar effect in the Vietnam stock markets This contradicts Loc's (2006) study, which indicated a significant negative return on Tuesdays.
Our analysis reveals that the test results from the four sub-periods, encompassing the years 2007 to 2010, align consistently with the findings from the full sample period This consistency reinforces our conclusion of weak form inefficiency in the Vietnam stock market.
This research concludes that the Vietnam stock market exhibits weak form inefficiency, with the calendar effect not being clearly evident Consequently, market prices do not accurately reflect available information, creating opportunities for information intermediaries to profit from the high demand for insights Furthermore, the belief that prices fail to fully incorporate certain information may prompt investors to implement portfolio strategies aimed at capitalizing on these inefficiencies for abnormal returns.
The limitations of our study include the exclusion of all stocks listed on HOSE, suggesting that future research should expand the sample size and explore different data frequencies, such as monthly, to validate the robustness of our findings Additionally, there is a need for a comparative analysis of the models used to determine which one offers the best performance for our market, as each model has its unique strengths and weaknesses.
Abdmoulah, W (2010), Testing the evolving efficiency of Arab stock markets,
International Review of Financial Analysis, Vol 19, No.2010, pp 25-34
Abeysekera, S P (2001), Effcient Markets Hypothesis and the emerging Capital
Market in Srilanka: Evidence from the Colombo Stock Exchange, Journal of Business Finance & Accounting, Vol 28, No.1-2, pp 249-261
Abraham, A (2002), Testing the Random Walk Behavior and Efficiency of the Gulf
Stock Markets, The Financial Review, Vol 37, No.3, pp 469 - 480
Abrosimova, N., Dissanaike, G & Linowski, D (2002), Testing the Weak Form
Efficiency of the Russian Stock Market, Working paper
Antoniou, A., Ergul, N & Holmes, P (1997), Market Efficiency, Thin Trading and
Non-linear Behaviour: Evidence from an Emerging Market, European Financial Management, Vol 3, No.2, pp 175-190
Awad, I & Daraghma, Z (2009), Testing the Weak-Form Efficiency of the
Palestinian Securities Market International Research Journal of Finance and Economics, Vol ISSN 1450-2887, No.32
Ayadi, O F & Pyun, C S (1994), An application of variance ratio test to the
Korean securities market, Journal of Banking & Finance, Vol 18, No.4, pp
Brooks, C (2004), Introductory Econometrics for finance, United State of America,
Cambridge University Press Brooks, C (2008), Introductory Econometrics for Finance, United State of
America, Cambridge University Press tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
Butler, K C & Malaikah, S J (1992), Efficiency and inefficiency in thinly traded stock markets: Kuwait and Saudi Arabia, Journal of Banking and Finance, Vol 16, No.1, pp 197 - 201
C.Cheung, K & Countts, J A (2001), A note on weak form market efficiency in security prices: evidence from the Hong Kong stock exchange, Applied Economics Letters, Vol 8, No.6, pp 407- 410
Campbell, J Y., Lo, A W & MacKinlay, A C (1997), The Econometrics of
Financial Markets Princeton, New Jersey Princeton University Press
Chan, K C., Gup, B E & Pan, M.-S (1997), International Stock Market Efficiency and Integration: A Study of Eighteen Nations, Journal of Business Finance
Chen, G., Kwok, C C Y & Rui, O M (2001), The day of week effect regularity in the stock markets of China, Journal of Multinational Financial Management, Vol 11, No.2, pp 139-163
Cooper, J C B (1982), World Stock Markets: Some Random Walk Tests, Applied
Daraghma, I A Z (2009), Testing the Weak-Form Efficiency of the Palestinian
Securities Market, International Research Journal of Finance and Economics, Vol 1450-2887, No.32, pp 7-16
Dickinson, J P & Muragu, K (1994), Market efficiency in developing countries: a case study of the Nairobi stock exchange, Journal of Business Finance &
Fama, E F (1965), The Behavior of Stock Market Prices, The Journal of Business,
Vol 38, No.1, pp 34-105 tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
Fama, E F (1970), Efficient Capital Markets: A review of theory and Empirical work, Journal of Finance, Vol 25, No.2, pp 383
Gao, L & Kling, G (2005), Calendar Effects in Chinese Stock Market, Annals of
Economics and Finance, Vol 6, No.1, pp 75-88
Groenewold, N (1997), Share market efficiency: tests using daily data for Australia and New Zealand, Applied Financial Economics, Vol 7, No.6, pp 645 -
Hau, L L (2010), Day of Week effects in the Different Stock Markets: New
Evidence on Model - Dependency in Testing Seasonalities in Stock Returns, Working Paper
He, C Y L a J (June 1991), A variance Ratio Test of Random Walks in Foreign
Exchange Rates, The Journal of Finance, Vol XLVI, No.2, pp 774
Hoque, H A A B., H.Kim, J & Pyun, C S (2007), A comparison of variance ratio test of random walk: a case of Asian emerging stock markets, International review of Economics and Finance, Vol.16, No.4, pp 488-502
HS, C & FJ, F (1983), Random walks on the stock exchange of Singapore,
Accounting and Finance, Vol.23, No.2, pp 81-87
Hudson, R., Dempsey, M & Keasey, K (1996), A note on the weak-form efficiency of capital markets: The application of simple technical trading rules to UK Stock prices -1935 to1994, Journal of Banking & Finance, Vol
Islam, A & Khaled, M (2005), Tests of Weak-Form Efficiency of the Dhaka Stock
Exchange, Journal of Business Finance & Accounting, Vol 32, No.7-8, pp
From 1613 to 1634, JL, S & RE, K (1977) conducted a comparative analysis of stock price behavior on the Bombay, London, and New York stock exchanges Their findings were published in the Journal of Financial and Quantitative Analysis, Volume 12, pages 391 to 413.
KA, W & KS, K (1984), The behaviour of Hong Kong stock prices, Applied
KC, B & SJ, M (1992), Efficiency and inefficiency in thinly traded stock markets:
Kuwait and Saudi Arabia, Journal of Banking and Finance, Vol 16, pp 97 -
Kima, J H & Shamsuddinb, A (2008), Are Asian stock markets efficient?
Evidence from new multiple variance ratio tests, Journal of Empirical Finance, Vol 15, No.3, pp 518-532
KS, K & SB, L (1991), A comparative analysis of the daily behavior of stock returns: Japan, the US and the Asian NICs, Journal of Business Finance and Accounting, Vol 18, pp 219 - 234
Lee, U (1992), Do stock Prices follow Random Walk? Some international
Evidence, International review of Economics and Finance Vol 1, No.4, pp
Lima, E J A & Tabak, B M (2004), Tests of the random walk hypothesis for equity markets: evidence from China, Hong Kong and Singapore, Applied Economics Letters, Vol 11 No.4, pp 255 - 258
Lo, A W & Mackinlay, A C (1988), Stock market prices do not follow random walks: Evidence from a simple specttication test, Review of Financial Studies, Vol 1, No.1, pp 41-46
Loc, T D (2006), Equitisation and stock-market development: The Case of
Vietnam Ph.D thesis tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
Loc, T D., Lanjouw, G & Lensink, R (2010), Stock market efficiency in thin- trading markets: the case of the Vietnamese stock market, Applied Economics, Vol 42, pp 3519-3532
Miller, M H., Muthuswamy, J & Whaley, R E (1994), Mean Reversion of
Standard & Poor's 500 Index Basis Changes: Arbitrage-Induced or Statistical Illusion?, Journal of Finance, Vol 49, No.2, pp 479-513
Mobarek, A & Keasey, K (2000), Weak-form market efficiency of an emerging
Market: Evidence from Dhaka Stock Market of Bangladesh., Working Paper
Moustafa, M A (2004), Testing the Weak-Form Efficiency of the United Arab
Emirates Stock Market, International Journal of Business, Vol 9, No.3
Oskooe, S A P., Li, H & Shamsavari, A (2010), The Random Walk Hypothesis in
Emerging Stock Market International Research Journal of Finance and Economics, Vol 50, No.50, pp 51-61
Sharma, J L & Kennedy, R E (1977), A comparative analysis of stock price behaviour on the Bombay, London and New York stock exchanges, Journal of Financial and Quantitative Analysis, Vol 12, No.03, pp 391-413
Smith, G & Ryoo, H.-J (2003), Variance ratio tests of the random walk hypothesis for European emerging stock markets, The European Journal of Finance Vol 9, No.3, pp 290-300
Solnik, B & Bousquet, L (1990), Day-of-the-week effect on the Paris Bourse,
Journal of Banking & Finance, Vol 14, No.2-3, pp 461-468
In his 2006 working paper, J Squalli examines the efficiency of the financial markets in the United Arab Emirates (UAE) The study analyzes various factors influencing market performance and assesses whether the UAE financial markets meet the criteria for efficiency.
Worthington, A C & Higgs, H (2004), Random walks and market efficiency in
European equity markets, Journal of Finance and Economics, Vol 1, No.1, pp 59 - 78
Y.Liu, C & He, J (1991), A variance Ratio Test of Random Walks in Foreign
Exchange Rates, The Journal of Finance, Vol XLVI, No.2, pp 774
Yalawar, Y (1988), Bombay stock exchange: rates of return and efficiency, Indian
The Economic Journal, Volume 35, Issue 4, spans pages 68 to 121 For the latest updates and full downloads of theses, please contact via email at vbhtj mk gmail.com for master's thesis inquiries.
Table A 1 Summary Results of all tests for daily returns in 2007
Stock Autocorrelation Run test Variance ratio test OSL GARCH (1,1)
Positive in Wed and Fri Negative in Mon
AGF *** ** No *** *** Positive in Wed Negative in Mon
ITA No No No *** *** No No
Negative in Mon Positive in Wed and Fri
Note: ***, ** and * denote a significance level of 1%, 5% and 10% respectively
Table A 2 Summary Results of all tests for daily thin adjusted returns in 2007
Stock Autocorrelation Run test Variance ratio test OSL GARCH (1,1)
ABT No *** *** *** Positive in Tue Positive in Tue
Positive in Mon, Wed, Fri
Potive in Wed, Thu, Fri
VNINDEX *** No *** *** No No tot nghiep do wn load thyj uyi pl aluan van full moi nhat z z vbhtj mk gmail.com Luan van retey thac si cdeg jg hg
Table A 3 Summary Results of all tests for daily returns in 2008
Stock Autocorrelation Run test Variance ratio test OSL GARCH (1,1)
TS4 *** *** *** *** *** No Negative in Mon
FMC *** *** *** *** *** No Negative in Mon
AGF *** *** *** *** *** Negative in Mon Negative in Mon
TDH *** *** *** *** *** Negative in Mon Negative in Mon
SJS *** *** *** *** *** No Negative in Mon
VNINDEX *** *** *** *** *** No Negative in Mon
Note: ***, ** and * denote a significance level of 1%, 5% and 10% respectively
Table A 4 Summary Results of all tests for daily thin adjusted returns in 2008
Stock Autocorrelation Run test Variance ratio test OSL GARCH (1,1)
ABT No *** *** *** Negative in Tue Negative in Tue
TS4 * *** *** *** No Negative in Mon
Negative in Mon Positive in Wed,Fri
Negative in Mon Positive in Wed,Thu, Fri
AGF ** *** *** *** Positive in Wed, Fri Positive in Wed, Fri
Negative in Mon Positive in Wed,Fri
The significance levels are indicated by ***, **, and * for 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com for the master's thesis.
Table A 5 Summary Results of all tests for daily returns in 2009
Stock Autocorrelation Run test Variance ratio test OSL GARCH (1,1)
Note: ***, ** and * denote a significance level of 1%, 5% and 10% respectively
Table A 6 Summary Results of all tests for daily thin adjusted returns in 2009
Stock Autocorrelation Run test Variance ratio test OSL GARCH (1,1)
ABT No No *** *** Negative in Tue Negative in Tue
TS4 No *** *** *** Positive in Fri Positive in Fri
Negative in Mon Positive in Wed & Fri
Negative in Mon Positive in Wed &
Thu &Fri AGF ** *** *** *** Positive in Wed & Fri Positive in Wed & Fri
Negative in Mon Positive in Wed & Fri
The significance levels are indicated by ***, **, and * for 1%, 5%, and 10%, respectively For the latest full download of the thesis, please contact via email at vbhtj mk gmail.com.