Testing the weak form efficiency of viet

Testing the weak form efficiency of viet Testing the weak form efficiency of viet Testing the weak form efficiency of viet Testing the weak form efficiency of viet Testing the weak form efficiency of viet Testing the weak form efficiency of viet Testing the weak form efficiency of viet Testing the weak form efficiency of viet

University of Warsaw Faculty of Economic Sciences Quang Hung Nguyen Testing the weak-form efficiency of Vietnamese stock market, case of Ho Chi Minh Stock Exchange 2006-2010 Quantitative Finance The thesis written under the supervision of Dr Michał Brzozowski Chair of Macroeconomics and International Trade Theory Faculty of Economic Sciences, University of Warsaw Warsaw, 02/2013 Abstract The purpose of the following dissertation is to analyze and determine the viability and ultimately the competitiveness of the emerging Vietnamese Stock Market in comparison to the more wellknown and successful stock markets throughout the world The first step in this analysis and the focus of this essay is to determine whether this Vietnamese Stock Market is weak form efficient at this stage of the game This determination depends on analyzing two different approaches The first the test used is of randomness; second, the test of predictability To fully determine the results of each, the analysis will primarily focus on an examination of the validity and applicability of both implemented approaches There are essentially three tests involved to examine randomness: 1) The portmanteau test of auto-correlations; 2) the unit root tests, and 3) the Lo and MacKinlay's variance ratio test In an attempt to measure random walk hypothesis of this market, these tests were summarily applied They revealed that there are, indeed, substantial deviations from random walk hypothesis of the returns in the stock market in Vietnam Additionally, predictability was the next goal Tests of technical trading rules were used in order to measure predictability of the Vietnamese Market These tests showed that the prices of stock changes in the Vietnamese market can be predictable This predictability means that trading cost nets can be profitably exploited In light of these results, the logical conclusion that can be gleaned is that the Vietnamese stock market is not weak-form efficient Keywords Efficient market Hypothesis, Vietnam stock market, Random walk, emerging stock market, market efficiency TABLE OF CONTENTS CHAPTER I INTRODUCTION 1.1 Introduction 1.2 Background and Objective 1.3 Research Methodology 1.4 Outline CHAPTER II OVERVIEW OF VIETNAM STOCK MARKET 2.1.Overview 2.2 Background of Vietnamese stock market 2.2.1 The shareholding reforms 2.2.2 The Stock Exchange 2.3 The structure and organization 2.3.1 The State securities commission 2.3.2 Securities Trading Center 2.3.3 Securities companies 2.3.4 Requirement for listing 2.3.5 Information disclosure 2.3.6 Trading 2.3.7 VNINDEX CHAPTER III THEORETICAL FRAMEWORK AND EMPIRICAL RESULTS 4 5 6 7 11 11 12 12 13 14 15 15 16 3.1 Efficient Market Hypothesis 3.1.1 The development of Efficient Market Hypothesis 3.1.2 Weak Form 3.1.3 Semi-Strong Form 3.1.4 Strong Form 3.2 Empirical evidence on Efficient Market Hypothesis 3.2.1 Evidence from developed markets 3.2.2 Evidence from emerging stock markets 3.3 Random Walk 3.4 Technical Trading Rule 16 17 17 18 18 19 19 20 23 24 CHAPTER IV DATA AND RESEARCH METHODOLOGY 25 4.1.Data 4.2 Research approach 4.2.1 Deductive and Inductive 4.2.2 Positivism and Interpretivisim 4.2.3 Quantitative and Qualitative research 4.3 Methodology 25 25 25 26 26 27 4.3.1Tests of Randomness Explanation of methods chosen Portmanteau tests BDS Test Unit root test Variance Ratio test 4.3.2 Tests of Technical analysis 27 28 29 30 30 31 32 CHAPTER V EMPIRICAL RESULT AND ANALYSIS 34 5.1 Test of Random walk hypothesis 5.1.1 Autocorrelation test 5.1.2 Unit root test 5.1.3 Variance Ratio test 5.2 Tests of Technical analysis 5.2.1 VMA results 5.2.2 FMA results 5.3 Summary of test‘s results 34 34 35 36 37 37 40 42 CHAPTER VI CONCLUSIONS 43 EXTENSION BIBLOGRAPHY 44 45 Acknowledgment I want to express my gratitude to Dr Michał Brzozowski, my supervisor for his precious guidance during preparation of this thesis I want to thank to my parents and Duong Hoang, my best friend for their love and support Fulfilling this goal would not have been possible without them Abbreviation HOSE HASTC GDP IPO OLS SSC VND HoChiMinh Stock Exchange Hanoi Stock Trading Center Gross Domestic Product Initial Public Offering Ordinary Leased Square State Securities Commission of Vietnam Vietnam Dong (National Currency Unit) CHAPTER I INTRODUCTION 1.1 Introduction Over the past 30 years, the efficient market hypothesis (EMH) has been widely mentioned in the financial literature by many financial economists because it has an important implication in a reality Fama (1970) suggested an ideal about the market in which the price of securities at any time ―fully reflects‖ all the available information which happens at that time If it occurs, we can conclude that is the efficient market According to the Fama‘s work, the efficient market hypothesis has been researched in not only developed markets but also in emerging markets in order to classify them into three levels which are depended on the available of information named weak form, semi-strong form and strong form After doing the initial research about the efficient market hypothesis, I chose the topic for my dissertation is: testing the weak-form efficiency of Vietnamese stock market, case of Ho Chi Minh Stock Exchange 2006-2010 Two periods are pre-crisis and crisis was included in this research 1.2 Background and objective The study into market efficiency has elicited increased interest amongst financial researchers A market is defined as efficient if all the available information is reflected in the prices There are three type of market efficiency: strong market efficiency, semi-strong market efficiency and weak market efficiency Financial analysts seem to have concentrated in the study of weak form market efficiency There are numerous factors that affect market prices, an understanding of which would enable market participants to seize opportunities for investments and arbitrages This is because most markets in the world not exhibit the strong and semi strong forms of efficiency but display different degree of the weak form of market efficiency (Fama 1970) The Vietnamese Stock Market has only been in existence for a little over a decade, and yet it is ranked 24th in the world according to the VN index This is remarkable growth for such an adolescent market During the past decade, this stock market has grown significantly, beginning has only two companies and now with 258 companies listed on the Ho Chi Minh Stock Exchange and 328 companies listed on the Hanoi Stock Exchange (USA Today, 2010) More researchers need to take not of these aforementioned monumental successes in the Vietnam Stock Market Potential investors would more than likely invest in this gold mine if only they had all of the information Because of this tremendous growth, the following is an analysis of this market, specifically, the Ho Chi Minh Stock Exchange This research seeks to see if the evidence point to Vietnam Stock Market is, through 2006-2010 is a week-form efficiency 1.3 Research methodology The problem background of this dissertation is whether the Vietnamese market exhibits weak form of market efficiency This dissertation aims at establishing a better understanding of the Vietnamese stock market Testing random walk hypothesis as a first proposition of weak-form efficiency, secondly the Variable Length Moving Average rules (VMA) and Variable Length Moving Average rules (FMA) will be used to test whether future movement of stock prices can be forecasted and profitability exploited Data obtained from States Securities Committee of Vietnam 1.4 Outline The thesis consists in chapters, Chapter deals with an introduction to the paper which states the objective and methodology of study It also provide brief outline of thesis Chapter is devoted an overview of the Vietnamese stock market This covers the development of the market, the characteristics of the market and the ownership structure of the market Chapter presented the Theoretical Framework and Literature reviews of empirical evidence in EMH field include all three forms, but focus on weak-form efficiency In Chapter provides a data and research methodology An overview of the methodology will be provided as well as the statistical explanation of the data Report and analyze results from empirical tests will be presented in Chapter Conclusion from findings of study in Chapter Chapter II OVERVIEW OF VIETNAM STOCK MARKET 2.1 Overview Vietnam is classified as a developing country with a population of an estimated 86 million people The country‘s economy was been transformed into a capital market in the late eighties from a planned economy The Vietnam stock market was established in 2000 in Ho Chi Minh City with an intention to speed the capitalization of the country The establishment of the market had been a major step made by the government towards the development of a public security market In 1993, the government set a special committee that was tasked with the responsibility of conducting research and the preparation of strategic plan towards the creation of the stock market The State Securities Commission (SSC) is a government body charged with the responsibility of establishing laws, organization and supervising the stock market as at the first day of trading the company had only two listed companies 247 companies being listed in 2010 with an overall capitalization of $ 28.28 billion(VND537.4 trillion) The market limits the foreign ownership of listed companies to 49% This capitalization represents a very small percentage of the total GDP of the country making it one of the smallest economies in southern Asia which has some of its countries recording their average capitalization that equated to up to 130% of their GDP However, the market has recorded a remarkable growth that can also be exhibited by the continued economic growth of the country At first the growth was slow, but had a boom in the market value and the number of companies listed, resulting in Vietnam officially becoming the 150th member of the WTO by 2006 By 2007, the market increased to more that 1000 points Moreover, their listed companies had grown to a whopping 400 Trouble came in 2008, however, as it did with nearly all economies throughout the world Farber et al (2006) claimed that there are a number of factors that directly or indirectly had an impact on the market; namely clusters of limit-hits, the effect of financial policies, and the limit of information transparency This is often referred to as the ―herd effect,‖ and has been a serious challenge to not only the Vietnamese market, but to nearly every market Nonetheless, the Consumer Price Index (CPI) was increasing, and peaked in 2008 at 19.89% (QDND.vn, 2009) The Vietnamese stock market has had great growth and great challenges over ten years However, Yen and Tran (2009) believe that investors, especially foreign investors, lack enough information about the development of the Vietnam Stock Market, as mentioned above There simply have not been sufficient empirical studies on this market 2.2 Background of Vietnamese stock market Not alike with development stock market in other countries where unofficial market and OTC market existed for a certain period before stock exchange officially started, there was no such of informal market like OTC in Vietnam before establishment stock exchange in 2000 The established of stock exchange based on government‘s awareness of necessity for development capital market 2.2.1 The shareholding reforms The development of the Vietnamese stock market was unlike other markets that had the previous unorganized markets or Over the Counter markets It was a government initiative having realized the need for the market for the development of capital markets and the economy long periods of subsidization characterize the war torn country as at 1975 Numerous State Owned Organizations were established during the post war period while the privately owned companies were restricted The launch of the Doi Moi reforms, in 1985, was a government‘s move aimed at reviving the sick economic conditions of the country These reforms were led to encourage private investment and foreign direct investment These would be the heart beat into the revival of the ailing economy The state owned corporations were mostly inefficient in comparison to the privately owned enterprises, hence, could not compete effectively This disparity coupled with the need to have several joint-stock companies led to the development of the stock market New legislations were issued by the government dealing with the equitization of State owned Enterprises into Joint stock companies The equitization legislations, formed in 1998 stipulate the methods for the selection and valuation of SOEs for equitization Vietnam government studied model from neighbour is China in management foreign enterprises For example, investments in banking sector are welcomed but government remain 51% controlling stake Besides, essential areas are electricity production, mineral exploration, telecom and water supply Private companies can be 100% foreign ownership but listed companies on the market, it falls to 49% 2.2.2 The Stock exchange As mentioned previously, this section intent present the development process and market features such as participants, listing and trading profiles, stock ownership structure Ho Chi Minh Securities Trading Centre (HOSTC) was established in July, 2000 Five years later, Hanoi Securities Trading Centre (HASTC) was established on July, 2005, it was projected act as OTC market while HOSTC operated as official stock exchange where listed companies or institution are traded In June, 2004, HoChiMinh index (or VN-INDEX) was 249.7 points, total market 10 returns for each period will be computed separately These mean values will then be examined using the t-statistic t-statistic=( where – /( - represent the mean return - represents the number of buy and sell signals -represents the unconditional mean N- number of observations –variance for the whole sample The t-statistic for the difference for the mean returns is given by – /( ) Chapter V EMPIRICAL RESULT AND ANALYSIS In this section, the empirical results will be separated into period are Pre-Crisis period of 20062008 and Crisis period of 2008-2010 and following by three kinds of test which examine random walk hypothesis: Autocorrelation test, Unit root test, Variance test and Technical analysis According to Malkiel (2011), due to influence by financial crisis, liquidity of market, public information during this period must be different pre-crisis, conduct to EMH validity during precrisis and crisis period is differ 5.1 TEST OF RANDOM WALK HYPOTHESIS 5.1.1 Autocorrelation test The null and alternative hypothesis is formed as: H0: no autocorrelation exist in data H1: autocorrelation exist in data The autocorrelation test performed on the VN-Index and four stock markets has been performed with lags The results are summarized table Table presents the correlation coefficients for the post crisis period It can be observed that the coefficients of the weekly VN-Index returns have a positive sign from the 1st to 5th lags According to the null hypothesis, at a significance level of 37 1%, it is evident that the index returns for all the lags has been rejected For the four stock considered in the analysis, it is noted that that the weekly returns have significant autocorrelation coefficients However, the q statistics fail to show that the market sampled support the null hypothesis for all the lags The random walk has been rejected under the autocorrelation tests for the index and the four stocks 2006-2008 AC Q-Stat 0.328 24.554 0.25 38.905 0.155 44.434 0.206 54.28 0.239 67.54 0.075 68.838 2008-2010 AC Q-Stat 0.089 70.685 -0.013 70.725 0.098 72.993 -0.077 74.391 0.069 75.516 0.031 75.741 Table 3: Autocorrelation test results As the table above presented the Ljung-Box show that four lags are all significant at 5% level Therefore, null hypothesis of no autocorrelation is rejected, conducting that autocorrelations exist in weekly return series According previous section, Ljung-Box is very strong to detect linear dependence of time series data, so with this result, Vietnam stock market is characterized by linear dependence, and we not need to have BDS test of non-linear dependence 5.1.2 Unit root test Both ADF and Phillip-Peron test used to test for unit root in weekly series data Null and alternative hypothesis are: H0: Unit root test exist in data (nonstationary) H1:Unit root test does not exist in data (starionary) 38 p-value 2006-2008 2008-2010 1.201207** 1.436127** 9,0284E-41 3.3751E-30 Table 4: Unit root test results The unit run tests are then conducted and provide the results as given in the in table This test is more powerful when compared to the autocorrelation tests as it is nonparametric This is because the time series observed does not follow a normal distribution The tests also reject the null hypothesis for the VN-Index Alternatively the p-value for the index as shown below also rejects the null hypothesis of a random walk 5.1.3 Variance Ratio test Under LOMAC test which proposed by Lo and MacKinlay (1988), null and alternative hypotheses are: H0: VR (q)=1 Return follow a random walk H1: VR(q)≠1 Return does not follow a random walk VR(q) and z-statistic under homoskedasticity and heteroskedasticity conditions presented below table: Variables obs Nq VNINDEX 224 VR(q) Z(q) Z*(Qq 16 32 0.56 -6.56 -2.92 0.3 -5.59 -2.8 0.2 -4.05 -2.19 0.1 -3.05 -1.71 0.06 -2.21 -1.39 Table 5: Variance ratio for the weekly return on the data 2006 to 2010 The homoscedacity and heteroscedacity have been tested for the null hypothesis in the variance ratio test The test has been applied for q intervals of 2, 4, 8, 16 and 32 39 Variables obs Nq VNINDEX 224 VR(q) Z(q) Z*(Qq 16 32 0.41 -8.82 -3.33 0.22 -6.24 -2.69 0.14 -4.35 -2.08 0.06 -3.18 -1.63 0.04 -2.25 -1.32 Table 5.1: Variance ratio for the weekly return on the data 2008 to 2010 From the result performed in table , null hypothesis of a random walk can be rejected at 5% significance level for studied sample Variance ratio is more than for all cases, Lo and MacKinlay (1988) suggested variance ratio equal to plus first order autocorrelation coefficient weekly returns Therefore, variance ratio are bigger than one, it point out positive autocorrelation for weekly holding period returns Both z(q) and z*(q) statistics increase with time interval that significance of reject become stronger as raw sample variances are compared to weekly variances Combine the results of three tests which performed above, Random walk hypothesis is strongly rejected for Vietnam stock market Further test in predictability of stock price should be conducted to conclude whether rejection of random walk hypothesis is a result of market inefficiency or not 5.2 TEST OF TECHNICAL ANALYSIS In this section, I applied tests which proposed by Brock et al (1992): The VMA and the FMA— the Variable Length Moving Average and the Fixed Length Moving Average The moving averages generated the result from trading strategies, in their research, to explore stochastic properties of stock return, they used bootstrap techniques 5.2.1 VMA Result Table below shows results of VMA trading rules, which are differentiated according to band size and short and long period lengths, for a full sample The difference in the rules is in the duration between the long and short period The rules are stated such as (1, 50, 0) One represents the shortest period, 50, the long period while is the band in percentage form For this dissertation, 10 rules will be used as in the table The results should show results that emit sell or buy decisions The returns of technical trading strategies are similar to unconditional returns of the buy-andhold strategy; since technical analysis cannot apply predict price changes Moreover, it is 40 important to examine the days when returns from rules emitting buy signals are similar to returns from rules sending sell signals Pre-Crisis Test N(Buy) N(Sell) Buy Sell Buy>0 Sell>0 BuySell (1,50,0) 543 561 0.3583 220 122 0.6318 0.3115 (1,150,0) 568 536 0.5317 0.4049 (1,150,0.01) 384 442 0.5469 0.4027 (5,150,0) 572 532 0.5262 0.4098 (5,150,0.01) 380 437 0.5395 0.4233 (1,200,0) 570 534 0.5316 0.4045 (1,200,0.01) 409 424 0.5330 0.3797 (2,200,0) 569 535 0.5272 0.4037 (2,200,0.01) 408 427 -0.001518 (-3.5912)* -0.003878 (-3.6297)* -0.000770 (-2.3547)* -0.000825 (-2.1802)* -0.000479 -1.8891 -0.000358 -1.5117 -0.000594 (-2.0737)* -0.000789 (-2.1041)* -0.000413 -1.7886 -0.000643 (-1.9056)* -0.001027 0.5856 (1,50,0.01) 0.003029 (3.6716)* 0.006446 (5.5201)* 0.002123 (2.2669)* 0.003008 (3.0875)* 0.001832 1.8014 0.002735 (2.7052)* 0.001948 (1.9869)* 0.002490 (2.4442)* 0.001783 1.7181 0.00253 (2.1169)* 0.002765 0.5270 0.3841 0.004546 (6.2897)* 0.010323 (7.6173)* 0.002894 (4.0024)* 0.003832 (4.5754)* 0.002311 (3.1960)* 0.003092 (3.6720)* 0.002543 (3.5165)* 0.003279 (3.9404)* 0.002196 (3.0369)* 0.002896 (3.4846)* 0.003791 Test N(Buy) N(Sell) Buy (1,50,0) 482 509 (1,50,0.01) 190 117 (1,150,0) 478 510 0.001029 -0.000985 (3.2346)* (2.9942)* 0.001123 -0.002769 (5.2740)* (3.7935)* 0.001123 -0.000460 (2.0969)* (2.9634)* Mean Crisis Sell Buy>0 Sell>0 Buy-Sell 0.3656 0.2497 0.003466 (4.4757)* 0.5378 0.2265 0.001323 (6.4753)* 0.3327 0.3950 0.001045 (2.0147)* 41 (1,150,0.01) 340 434 (5,150,0) 504 523 (5,150,0.01) 367 420 (1,200,0) 530 565 (1,200,0.01) 398 423 (2,200,0) 527 554 (2,200,0.01) 396 415 Mean 0.003008 -0.000735 (3.0875)* (2.2602)* 0.001832 -0.000545 1.8014 -1.9521 0.002735 -0.000309 (2.7052)* -1.6465 0.001948 -0.000693 (1.9869)* (2.0267)* 0.002490 -0.000662 (2.4442)* (2.2371)* 0.001783 -0.000372 1.7181 -1.83639 0.00253 -0.000734 (2.1169)* (1.8452)* 0.001957 -0.002047 0.4679 0.4027 0.002832 (3.4724)* 0.3092 0.3584 0.3749 0.3759 0.3095 0.3584 0.001731 (2.97960)* 0.002452 (3.24840)* 0.001483 (3.3467)* 0.3758 0.2759 0.002879 (3.3854)* 0.3945 0.3573 0.3570 0.2573 0.002046 (2.8469)* 0.001945 (3.1476)* 0.002747 Note: * Significant at 5% level Returns are reported in daily N(Buy) and N(sell) are numbers of buy and sell signal respectively during the sample Column and reported mean returns during buy and sell period with tstatistics in parentheses testing equality with unconditional mean And column depicts the differences between the mean buy&sell returns with corresponding t-statistics testing buy-sell difference from zero Table 6: Results of standard test VMA rules Observing the Vietnamese stock market, there are VMA rules test results that indicate its predictability When one holds the index for a sample period, buy returns are positive and have an average daily return of 0.276 percent in pre-crisis period while decrease to 0.195 percent during crisis period Compared to the mean daily return of 0.072 percent, the buy returns are extremely high Conversely, the sell returns have negative value with an average of -0.103 percent for ten tests; this shows a value of -23 percent for the annual rate As Table shows, the differences between all of the returns, including unconditional returns, are very significant Out of twenty tests, fifteen rejected the null hypothesis, which states that buy-and-hold strategy returns equal conditional VMA trading rules returns with a value of percent significance in a two-tailed test On the other hand, other statistics indicate a small margin of significance To view the fraction of buy and sell returns, which are higher than 0, one must carefully inspect the ―Buy>0‖ and ―Sell>0‖ columns The value of the buy fraction for returns is typically more than 50 percent; however, the value of the sell fraction for returns is usually less and between 31 42 and 42 percent These two fractions should be equal because of the null hypothesis which declares that technical trading rules are not useful in the production of signals After conducting a binomial test, the results of the test (shown in the last column) indicated that these different values have great significance in the range of statistics between and 7.6 Therefore, one can easily reject the null hypothesis of equality Because buy-and-hold strategies differ from the returns of VMA rules, they provide varying degrees of prediction Unlike other research findings, which pinpointed significant negative returns are twice as likely to occur as significant positive returns, this research explores the opposite occurrence Most notably, this research finds that the amount of positive returns from buy signals is much higher than the amount of negative returns from sell signals This finding helps to prove that the technical trading rule is both sustainable and profitable and can be applied to the Vietnamese stock market 5.2.2 FMA results Pre-Crisis Test N(Buy) N(Sell) Buy Sell Buy>0 Sell>0 Buy-Sell (1,50,0) 56 55 0.2727 22 13 0.8182 0.1538 (1,150,0) 56 55 0.5714 0.3273 (1,150,0.01) 39 46 0.6667 0.2826 (5,150,0) 56 55 0.5714 0.3273 (5,150,0.01) 38 44 0.6579 0.2955 (1,200,0) 58 53 0.5862 0.3019 (1,200,0.01) 41 43 0.6585 0.3023 (2,200,0) 57 54 0.5789 0.3148 (2,200,0.01) 40 44 -0.013302 (-2.5886)* -0.039873 (-3.3714)* -0.008559 (-1.9808)* -0.008653 -1.8740 -0.008559 (-1.9808)* -0.006826 -1.6279 -0.006074 -1.6418 -0.007910 -1.7421 -0.005903 -1.6304 -0.006785 -1.6231 -0.012844 0.6250 (1,50,0.01) 0.026738 (2.5576)* 0.061009 (4.8992)* 0.022080 (1.9571)* 0.032317 (2.8853)* 0.022080 1.9571 0.031696 (2.7877)* 0.018753 1.5459 0.029606 (2.6254)* 0.019027 1.5726 0.028948 (2.5264)* 0.02755 0.7000 0.3182 0.0400 (4.4567)* 0.1009 (6.0935)* 0.0306 (3.4103)* 0.0410 (3.9771)* 0.0306 (3.4103)* 0.0385 (3.6755)* 0.0248 (2.7606)* 0.0375 (3.6317)* 0.0249 (2.7739)* 0.0357 (3.3251)* 0.0492 Mean Crisis 43 Test N(Buy) N(Sell) Buy Sell Buy>0 Sell>0 Buy-Sell (1,50,0) 46 50 0.1649 17 11 0.6027 0.1305 (1,150,0) 54 51 0.4028 0.3018 (1,150,0.01) 35 41 0.5028 0.2204 (5,150,0) 58 52 0.4958 0.2947 (5,150,0.01) 34 42 0.5038 0.2485 (1,200,0) 52 46 0.4953 0.2648 (1,200,0.01) 37 41 0.5840 0.2850 (2,200,0) 52 47 0.5204 0.2905 (2,200,0.01) 36 41 -0.010174 (-2.7473)* -0.048753 (-3.54723)* -0.009451 (-2.01759)* -0.009365 -1.92759 -0.0095739 (-2.0475)* -0.008462 -1.8426 -0.007372 -1.8265 -0.008830 -1.9642 -0.006829 -1.7271 -0.008265 -1.7478 -0.01593 0.4058 (1,50,0.01) 0.015928 (2.0376)* 0.031009 (3.9385)* 0.010804 (1.2475)* 0.021347 (2.2475)* 0.014850 1.37485 0.011346 (2.3475)* 0.001847 1.4957 0.021475 (2.3583)* 0.013847 1.4759 0.019847 (2.3485)* 0.019474 0.6970 0.2984 0.0309 (4.2045)* 0.0985 (5.7849)* 0.02048 (3.1947)* 0.0349 (3.4857)* 0.0284 (3.1947)* 0.0240 (3.3957)* 0.0210 (2.40583)* 0.0296 (3.4059)* 0.01049 (2.40690)* 0.0204 (3.1094)* 0.0359 Mean Note: * Significant at 5% level Table 7: FMA results According to FMA rules, cross short and long run triggered buy and sell signal , moving average After a 10-day holding period for post signals, one can calculate returns based on a cumulative pattern These results support similar conclusions gathered according to the rules of VMA Following a buy signal, the mean return for a 10-day holding period is equal to 2.152 percent during pre-crisis period while in crisis period mean return is 1.592 percent Next, following a sell signal, the mean return for the same 10-day holding period is equal to -1.048 percent in pre-crisis and -1.304 percent in crisis period Compared to an unconditional cumulative rate of return of 0.405 percent for a 10-day holding period, the returns are quite different At a percent level of confidence, 11 out of 22 tests of significance for buy and sell decisions are significant Once again, the number of significant negative returns of sell signals is less than the number of significant positive returns of buy signals Further, the number of buys greater than is greater than the number of sells greater than in all of the test results Observing the differences between mean buy and sell returns, one can deduce a high level of significance concerning all test statistics higher than 2.6 Given the aforementioned results, it is clear that the null hypothesis of equality should be rejected Below are the summarized testing results for the VMA and FMA trading rules: 44 i For the VMA and FMA rules, the mean returns are quite different than the unconditional returns created according to the buy-and-hold strategy For VMA trading rules, it appears that the mean buy-sell return is 5.2 times greater than the unconditional mean return Yet, for the FMA rules, the mean buy-sell return is 5.7 times greater than the mean return (unconditional) ii According to moving average rules, sell signals are far less significant than buy signals This occurs as the amount of significantly positive buy signals goes beyond the amount of significantly negative sell signals iii For the studied period, these rules, which exclude transaction costs, are useful investment tools that apply to the Vietnamese stork market The predictive ability of the two moving average trading rules is quite evident Therefore, in the Vietnamese market, technical analysis is very helpful in forecasting Applying an efficient market hypothesis in a costly environment, it is debatable whether or not trading rules translate into abnormal returns If one includes transaction costs, it is likely that abnormal returns caused by using trading rules will diminish The combination of transaction cost magnitude and transaction number caused by buy and sells signals affects the level of profits obtained from the aforementioned trading rules A buy, sell, or hold (no signal) happens when a short moving average takes over a long moving average Thus, until the moving averages intersect again, one records a transaction Consequently, when a transaction takes place, transaction costs, which include trading costs and income tax on capital appreciation, incur These two costs are considered small by the Vietnamese stock market Currently, Vietnam does not impose taxation on the trading of securities As a result, the exchange authority strictly regulates brokerage fees, and the Exchange varies brokerage fees to increase stock market participation During the given period of study, a range of 0.25 to 0.4 percent per transaction for the maximum was used to set up brokerage fees Upon observation, the combination of these two factors creates small trading costs per year Moreover, when comparing to the average annual return (buy or sell) between 60.55 and 65.76 percent, both VMA and FMA rules are followed closely With these given factors, a conclusion can be drawn, which declares that the two trading rules have predictive ability Furthermore, on the Vietnamese stock market, these trading rules serve as profitable investment tools for the time period of Jan 2006 to March 2010 5.3 Summary of test’s results Two kinds of tests were selected in order to examine the predictability and random character of data for an investigation of the Vietnamese stock market and its weak-form market efficiency levels Over a time period Jan 2006- March 2010, weekly stock returns were examined by randomness testing for a full sample; over a sub-period , daily data was studied by predictability testing The portmanteau test for autocorrelations, Augmented Dickey-Fuller and Phillips-Perron unit root tests, and the LOMAC variance test were the chosen tests for the measurement of randomness for a weekly series of returns 45 With these three chosen tests, one conclusively can reject the null hypothesis and confirm that the Vietnamese stock market does not exemplify a random pattern In particular, at four lags of 1, 5, 6, and 10, the portmanteau test‘s Ljung-Box statistics show high significance This suggests that weekly returns data on VN-Index is linear dependent Moreover, according to the results of the ADF and Phillips-Perron unit root test, one can easily reject the null hypothesis of the unit root This helps to propose that stationarity could explain the linear dependence of the time series data, as exhibited by the initial test Lastly, the non-random character of the weekly returns series was demonstrated by the LOMAC variance ratio test results Since all of the variance ratios have values greater than 1, it can be ascertained that the returns exhibit a positive serial correlation Given this finding, for the Vietnamese stock market, the random walk hypothesis does not hold, and the first condition for a weak-form market efficiency designation is refuted Due to the moving average empirical results, it is shown that the Vietnamese stock market is quite ineffective Further, VMA and FMA rules pinpoint the predictability of changes stock price in Vietnamese stock market as well as the exploitability of the net transaction costs Accordingly, during the period of study, empirical findings help one to conclude that the Vietnamese stock market is inefficient This finding is similar to previous studies on the efficiency of the market hypothesis in emerging and developing markets Chapter VI CONCLUSIONS This dissertation has detailed description into the Vietnamese market in an aim of determining if the market has exhibited weak forms of efficiency in the periods before and after the crisis According the result from Ljung-Box test , ADF and Phillips Peron, LOMAC, the results conclude that evidence exists against the hypothesis of an efficient market in weak form The rejection of the random walk hypothesis implies that the market in both periods pre-crisis and post-crisis is inefficient in the weak form off market efficiency Nevertheless, the reject of random walk does not necessary imply market inefficiency so the technical analysis was applied to test which used VMA and FMA It presented the techniques used in Vietnam stock market had predictive ability and helped generate significant return net of trading costs Empirical evidence from both of tests are enough to reject hypothesis Vietnam stock market efficiency in weak form Therefore, the hypothesis that the stock prices followed a random walk pattern based on the weak efficient market theory was rejected and that there was a pattern in the price variations The rejection of the null hypothesis called for technical analysis to be conducted on the VNIndex to exhaustively conclude that the market suffers from inefficiency in the weak form The predictability test also implies that the market is inefficient in the weak although the prices of stocks are predictable This has been a conclusion reached by several other researchers who have 46 conducted studies on emerging markets This could be concluded by several reasons such as thinning, low levels of liquidity, ill governance on the system, the irrational behaviour of investors, insider trading and manipulation The Variable Moving Average and the Fixed Moving Average also reveal that the stock prices are predictable and generate substantial returns at the lowest cost The main sources of predictability are investor rationality and numerous market imperfections The application of the two tests rejects the null hypothesis which is understandable as Vietnam is a relatively new market with a life of only 12 years since inception Finally, the stock prices in Vietnam market did not follow a random pattern, and the market was not efficient in weak form for the last period, from Jan 03, 2006 to March 31, 2010 Extension It should be noted that the Vietnam market pays dividends in cash, stocks, or a combination of cash and stocks This study was directed at announcements of the cash dividend payments only It would be of interest to further investigate if announcements of dividend payments would also be influential in the other two methods of payout Further investigations might combine both the Vietnam Stock Exchange along with the Hanoi Trading Center in order to provide larger sample sizes Another investigation could be the examination based on seasons of the year, or possible monthly effects to note if they also indicate statistical connectivity This data might also be analyzed to determine a clearer picture for future price movements in the Vietnam Market 47 BIBLIOGRAPHY Abraham et al.(2002) The Financial Review 37 469–480 Bachelier, L (1900) Theory of Speculation Reprinted in The Random Character of Stock Market Prices Boston: MIT Press , Cootner,1964 Brock, W., et al.(1992) Simple technical trading rules and the stochastic properties of stock returns Bodie, Z et al (2009) Investment (8th ed.) New York: Mc Graw Hill Borges, M.R (2010) Efficient Market Hypothesis in European stock market, The European Journal of Finance, 16/7, 711-726 Burton, G.M (2003) The Efficient Market Hypothesis and Its Critics, Journal of Economic Perspectives, 17/1, 59-82 Bryman, (2001) Social Research Methods Copeland, E., Thomas, Weston J Fred, Financial Theory and Corporate Policy 3rd Edition, 1993, Library of Congress Cataloging-in Publication Data Cooper and Schindler, (2006) Business Research methods Campbell, J Y., Lo, A W., and MacKinlay, A C (1997): The Econometrics of Financial Markets Second Edition Princeton: Princeton University Press Choudhry, T (1994): “Stochastic trends and stock prices: an international inquiry” http://www.tandfonline.com/doi/abs/10.1080/758518670 Coates N., Sloan D (2008) Newcastle Business School Masters Dissertation, Compiled by Coates N and Sloan D McGraw Hill, a Custom Publication., CNX ,2005 http://cnx.org/content/m10676/latest/ Dimson, E and Mussavian, M (2000) ‗Market efficiency‘, The Current State of Business Disciplines, 3, pp 959-970.Economics Vol 10(6) p 579-586 Elder; Alexander "Come into my Trading room" John Wiley & Sons 2002, in Part ("Trading"), see Chapter "Exiting Trades" ISBN 0471225347 Haugen A., Robert, Modern Investment Theory, 5th Edition, 2001, Prentice-Hall, Inc., Upper Saddle River, New Jersey, 07458 48 Fama, E (1965a) The behaviour of stock makket prices, 38,34-105 Fama, E (1965b) Random walk in stock market prices, Finance Analyst Journal, Fama, E (1970) Efficient Capital Market: A review of Theory and Empirical Work, Journal of Finance, 25/2, 383-417 Fama, E (1976) Foundation of Finance: portfolio decisions and securities prices New York: Basic Book Firestone (1987, pp 16) Qualitative Research in the study of Leadership FXStrategy, 2011 http://www.fxstrategy.com/video/forex-trading-indicators/moving-average-ofoscillator-osma-67.html Ljung, G and Box, G (1978) On a Measure of Lack on Fit in Time Series Models Biometrika, Vol 66, p 67-72 Lo, and MacKinlay (1988) Stock market prices not follow random walks: evidence from a simple specification test The Review of Finance studies, 1.41-66 Hajek, J (2002) Weak-form efficiency in the Czech equity market Politicka Ekonomie 50(3), 377-389 Hin Yu Chung (2006) Testing Weak – Form Efficiency of the Chinese Stock Market Thesis (Master) Lappeenranta University of Technology Hassan, K M., Haque, M., and Lawrence, S (2006) “An Empirical Analysis of Emerging Stock Markets of Europe” Quarterly Journal of Business and Economics, 45(1&2), 31-52 Haralambos and Holborn, 3rd ed (1991) 'Sociology: Themes and Perspectives 3rd Edition Hardouvelis et al (1999) EMU and European Stock Market Integration http://ideas.repec.org/p/cpr/ceprdp/2124.html Jose Benzinho et al.(2002) Efficiency tests in the Iberian stock markets http://papers.ssrn.com/sol3/papers.cfm?abstract_id=599926 Labuschange.A (2003) Qualitative Research - Airy Fairy or Fundamental? Lim, K P et al (2003) Weak – form Efficient Market Hyothesis, Behaviour Finance and Episodic Transient Dependencies: The case of the Kuala Lampur Stock Exchange 49 Malkiel G.Burton, A Random Walk Down Wall Street, 6th Edition, 1996, W.W Norton & Company Inc Malkiel G.Burton, (2003).The Efficient Market Hypothesis and Its Critics Journal of Economics Perspectives, Vol.7, No.1, 59.82 Malkiel G.Burton, (2011) The Efficient-Market Hypothesis and the Financial Crisis Maria Rosa Bogers (2008) Efficient Market Hypothesis in European Stock Markets http://ideas.repec.org/p/ise/isegwp/wp202008.html Marcin Kalinowski (2009) Semi-Strong Information Efficiency of the Polish Stock Exchange Market in Various Market Situations http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1342795 Mishra, P.K et al (2009) Empirical Evidence on Indian Stock Market Efficiency in Context of the Global Financial Crisis, Global Journal of Finance and Management, 01/02, 149 – 157 Mobarek, A and Keasey, K (2006) Weak-form Market Efficiency of an Emerging market: Evidence from Dhaka Stock Market of Bangladesh Leeds University Phillips, P and Perron P (1988) Testing for Unit Roots in Time Series Regression Poterba, J.M., and Summers, L.H (1988) Mean Reversion in Stock Prices: Evidence and Implications Journal of Financial Economics, 22, 631-639 Regúlez, M., & Zarraga, A (2002) Common features between stock returns and trading volume Applied Financial Economics 12(12), 885-893 Roberts, H V (1959): “Stock market “patterns” and financial analysis: Methodological suggestions” Journal of Finance, vol 14, 1-10 Robson (2002), Real world research Poshakwele, S (1996), Evidence on Weak Form Efficiency and Day of Week Effect in the Indian Stock Market, Finance India 10(3), 605 -616 Ross, S, et al (2008) Corporate Finance (8th ed.) New York: Mc Graw Hill Truong Dong Loc (2006) Equilisation and Stock Market Development: The Case of Vietnam http://dissertations.ub.rug.nl/FILES/faculties/eco/2006/t.d.loc/00_titlecon.pdf Samuelson Paul A (1965) Proof that Properly Anticipated Prices Fluctuate Randomly QDND.VN (2009), http://www.qdnd.vn/qdndsite/vi-vn/61/67531/print/Default.aspx 50 Shields, Linda and Twycross, Alison (2003) The difference between quantitative and qualitative research Saunders et al (2009) Research Methods for Business Students Saqib Nisar and Muhammad Hanif (2012) Testing Market Efficiency: Empirical Evidence from Developed Markets of Asia Pacific http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1983960 Scott Niblock and Keith Sloan (2007) Are Chinese stock markets weak-form efficient? http://epubs.scu.edu.au/bus_pubs/505/ Graham Smith and Huyn-Jung Ryoo (2003) Variance ratio tests of the random walk hypothesis for European emerging stock markets The European Journal of Finance 9, 290-300 Velimir Sonje et al (2011) Efficient market hypothesis is the Croatian stock market as (in)efficient as the U.S market 51 ... overview of the Vietnamese stock market This covers the development of the market, the characteristics of the market and the ownership structure of the market Chapter presented the Theoretical... profits are also not possible If the market is semi-strong then it implies the weak form efficiency, the strong efficiency, in addition, implies semi-strong and weak efficiency The last form of. .. because most markets in the world not exhibit the strong and semi strong forms of efficiency but display different degree of the weak form of market efficiency (Fama 1970) The Vietnamese Stock Market