TESTING WEAK FORM MARKET EFFICIENCY: EMPIRICAL EVIDENCE FROM SELECTED ASIAN STOCK MARKETS
1Assistant Professor at BRAC Business School, BRAC University, Dhaka, Bangladesh
2Master of Business Administration (MBA) graduate (with dual major in Finance and Marketing) from BRAC Business School, BRAC University, Dhaka, Bangladesh
ABSTRACT
This empirical study examines stock market efficiency of selected fifteen countries from the Asian region using weekly stock returns from the year 2001 to 2017. In order to test the market efficiency, the following statistical methods were conducted on the realized returns: Auto Correlation, Q Statistics, Correlation Matrix, Unit Root Test, and Run Test. It is revealed that the weekly return is not normally distributed as the historical returns from the considering markets are negatively skewed. We came up to a conclusion that weekly returns do not follow the random walk as it rejects the null hypothesis. Therefore, it may be possible for the investors to gain an arbitrage profit by investing in any of the markets in consideration.
Keywords:Weak form efficiency Asian market Random walk Arbitrage profit Unit Root Test Augmented Dickey- Fuller (ADF) Test
ARTICLE HISTORY: Received:4 May 2018. Revised:1 June 2018. Accepted:4 June 2018. Published:7 June 2018.
Contribution/ Originality:This study is one of very few studies which have investigated the weak form market efficiency in selected capital markets from the Asian region. This paper also contributes in the existing literature of finance by testing the existence of random walk theory in the stated context.
1. INTRODUCTION
In the realm of falling interest rate in banking industry investors tend to show their interest in investing in capital market with the hope of gaining higher return. However, the risk is relatively higher in this sector. People are now more concerned of risk minimization and finding appropriate techniques to maximize profit. Investors are testing different markets to know the possibility of earning arbitrage profit using different techniques. In this research, market efficiency were studied and tested from fifteen Asian countries - Bangladesh, India, Pakistan, Vietnam, Thailand, Hong Kong, Singapore, Malaysia, Thailand, Sri Lanka, South Korea, Philippines, Indonesia, China, and Japan.
According to Fama (1970) who provides the theory of efficient market hypothesis (EMH), the efficiency is based on weak form, semi strong form, and strong form. Weak form effi ciency is whether the historical information is reflected in the present market price or not. In Semi strong form efficiency, the past information and publically available information is reflected in the current market price. Therefore individual cannot achieve arbitrage profit. In the strong form of market efficiency, all publically and privately available information is reflected in the market price. Therefore no one can achieve excess profit. When the market is inefficient in random walk, it suggests that we can reject the null hypothesis of random walk. According to the random walk theory we are unable to predict the future stock price by analysing historical information. Abnormal return is generally not possible to achieve on a continuous basis. Therefore technical analysis does not work in that particular scenario. However, excess return can be achieved in some cases by conducting fundamental analysis since they deal with semi strong form of efficiency.
2. LITERATURE REVIEW
Füss (2005) investigated the financial liberation and stock price behaviour in Asian emerging markets by testing random walk hypothesis and market efficiency over seven Asian emerging countries. In that particular study, he used single variance ratio test, multiple variance ratio tests and non-parametric run test for understanding the week form efficiency in those markets. It was revealed that the major Asian emerging markets do not follow the random walk theory and as a result it may be possible to achieve arbitrage profit by conducting proper technical analysis.
Shaik and Maheswaran (2016) examined the random walk hypothesis on ten emerging Asian stock markets using indices from the year 2001 till 2015 through applying six different root tests. The research showed the findings in two ways i) before financial crisis – from the year 2001 to 2007, where a total of ten major stock markets followed the random walk theory according to the unit root test; ii) the post crisis situation where only five out of ten markets followed the random walk theory based on unit root test.
Gündüz and Hatemi (2005) investigated to explore a relationship between stock price and volume figures in some selective stock markets. Granger Causality Test was used to justify the relationship and they came up to a conclusion that there was no causal relationship exists between the stock prices and volume figures in various stock markets of Czech Republic.
Cooray and Wickremasinghe (2005) examined the efficiency in emerging markets. Empirical evidence from South Asian region consisting of monthly returns from 1996 to 2005 were used and applied in the Augmented Dickey and Fuller (1979;1981); Perron (1988) test and Elliot et al. (1996) test. Their objective was to investigate the presence of weak form efficiency in the emerging stock markets. However, their study failed to proof any presence of weak form efficiency in the emerging stock markets.
Chiang and Doong (2001) carried out an empirical study on the stock return and volatility by using historical data from seven Asian Stock Markets based on TAR-GARCH Models. The result suggests that the null hypothesis cannot be rejected for monthly data.
Hamid et al. (2017) tested the weak form market efficiency in fourteen countries from the Asia Pacific region using monthly returns of indices from 2004 to 2009. Autocorrelation, Ljung- Box test, Q statistics test, run test, unit root test, and variance ratio test were applied to test the market efficiency. Their finding indicates that the monthly return do not follow the random walk hypothesis in the stock markets of this particular region. Therefore, the study suggests that investors may earn arbitrage profit from these markets.
Shaik and Maheswaran (2016) examined the market efficiency in five Asian stock markets. Individual variance ratio was applied to test their proposition and they found out that, Cambodia, Lau and Singapore stock markets are weak form efficient where as other markets in the study are not weak form efficient.
Hussain et al. (2015) studied weak form market efficiency and asymmetric relationship among various stock markets from South Asian region by using monthly data during 1998 to 2013. Asymmetric Co-integration and Asymmetric Error Correlation Models were used to understand the relationship. The result showed that Indian stock market tends to influence Pakistan stock market in the long run.
3. DATA AND METHODOLOGY
3.1. Data and Observation
The observation consists of weekly price indices of 15 Asian countries. Data was collected from the period 2009 to 2017 for the considering countries. The market return is calculated as following:
3.2. Descriptive Statistics
The calculated stock return was used in the following descriptive statistics: mean, median, standard deviation, variance, skewness, and range to observe our study.
3.3.Correlation Matrix
The Correlation matrix gives an insight about the relationship of stock markets among different countries. A correlation matrix is developed to observe the relationship among the stock markets of fifteen Asian countries. Direction of relationship (positive or negative) among stock markets of the concerned countries is derived from autocorrelation matrix.
3.4. Auto Correlation and Ljung-Box Statistics
The serial auto correlation is used to justify the relationship between its own values with the series of different legs. If a negative result is drawn from auto correlation, it means the markets follow the null hypothesis, and if it gives a positive result, it emphasizes that the null hypothesis is rejected.
3.5. Run Test
In the observational research method, a “Run” is the series of identical signs where we can see a sequence of observation. The run test evaluates the value of one observation reflected by the value of considering earlier observations. Run test is used to analyse independence of a data series in the return stream. For evaluating a run test in terms of null hypothesis, two approaches can be considered – (i) positive return mean and (ii) negative return mean. Now, if the positive “m” and negative “m” reflecting total positive return (+m) and total negative return (-m) with “m” observations, then the test statistic is normally distributed as following –
3.6. Unit Root Test
Unit root of Augmented Dickey Fuller (ADF) test is used to test the time series of stock price change in indices. The main objective here is to test the stationary of the time series. The equation is following:
4. ANALYSIS AND FINDINGS
In the Table-1 below, descriptive statistics of the fifteen Asian market returns are disclosed. The weekly returns of eleven markets were negatively skewed which indicates that, the empirical returns were more negative than positive in these markets. The rest of the observed returns from the markets of Bangladesh, China, Singapore, and Sri Lanka yielded positive returns. Average weekly return is highest in Pakistan (0.004) with a standard deviation of 3.1%, while China and Japan have the lowest weekly average returns. During the observed period, Bangladesh had a weekly return of 0.1% along with a standard deviation of 2.00%. The Jarque-Bera test rejects the normality on the sample set consisting of weekly returns from the market of the following countries: India, Malaysia, China, Thailand, Singapore, Vietnam, Indonesia, Pakistan, South Korea, Japan, Philippine and Sri Lanka.
According to the findings, it can be observed that, the return series from Bangladesh and Hong Kong during the period from 2001 to 2017 accepts the null hypothesis of normal distribution. For further analysis in justification of randomness, we have studied the autocorrelation and Ljung-Box test. If p value is less than .05 and Q statistics and autocorrelation coefficient both are less than zero, we can reject the null hypothesis at 5% significance level.
Table-1. Descriptive Statistics of fifteen Asian market returns
| Details | IND | BD | MAL | CHI | THA | SIN | VIE | HoK | INDO | PAK | SoK | JAP | PHI | TAI | SRI |
| Mean | 0.002 | 0.001 | 0.001 | 0 | 0.002 | 0.001 | 0.001 | 0.001 | 0.003 | 0.004 | 0.002 | 0 | 0.003 | 0.001 | 0.003 |
| Std. Error | 0.001 | 0 | 0.001 | 0.002 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 |
| Median | 0.003 | 0 | 0.002 | -0 | 0.004 | 0.002 | 0.003 | 0.003 | 0.005 | 0.007 | 0.004 | 0.003 | 0.003 | 0.003 | 0.001 |
| Std. Dev. | 0.023 | 0.02 | 0.013 | 0.034 | 0.024 | 0.019 | 0.029 | 0.026 | 0.031 | 0.031 | 0.031 | 0.031 | 0.023 | 0.022 | 0.026 |
| Sample Var. | 0.001 | 0 | 0 | 0.001 | 0.001 | 0 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0 | 0.001 |
| Kurtosis | 0.828 | 0.35 | 2.281 | 1.834 | 1.135 | 1.494 | 2.01 | 0.855 | 5.893 | 5.716 | 5.885 | 8.307 | 1.464 | 1.515 | 5.902 |
| Skewness | -0.08 | 0.02 | -0.25 | 0.168 | -0.37 | 0.007 | -0.22 | -0.06 | -0.94 | -1.245 | -0.7 | -1.1 | -0.41 | -0.69 | 0.725 |
| Range | 0.187 | 0.12 | 0.105 | 0.271 | 0.155 | 0.134 | 0.233 | 0.201 | 0.349 | 0.31 | 0.4 | 0.393 | 0.173 | 0.164 | 0.293 |
| Jarque- Bera |
42.23 | 6.60 | 51.13 | 18.88 | 42.49 | 10.28 | 11.23 | 4.32 | 80.87 | 199.79 | 91.46 | 65.64 | 37.32 | 9.79 | 92.35 |
| Proba- bility |
0.00 | 0.03 | 0.00 | 0.00 | 0.00 | 0.01 | 0.00 | 0.11 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.01 | 0.00 |
| Minimum | -0.1 | -0.06 | -0.05 | -0.14 | -0.09 | -0.06 | -0.11 | -0.1 | -0.23 | -0.201 | -0.23 | -0.28 | -0.1 | -0.1 | -0.11 |
| Maximum | 0.088 | 0.06 | 0.052 | 0.135 | 0.071 | 0.071 | 0.126 | 0.105 | 0.116 | 0.109 | 0.17 | 0.114 | 0.074 | 0.068 | 0.18 |
| Sum | 0.752 | 0.26 | 0.568 | 0.09 | 0.945 | 0.302 | 0.413 | 0.328 | 2.646 | 3.497 | 1.395 | 0.413 | 1.14 | 0.393 | 2.691 |
| Count | 417 | 221 | 417 | 412 | 417 | 417 | 410 | 417 | 848 | 853 | 855 | 855 | 417 | 411 | 855 |
| Largest(1) | 0.088 | 0.06 | 0.052 | 0.135 | 0.071 | 0.071 | 0.126 | 0.105 | 0.116 | 0.109 | 0.17 | 0.114 | 0.074 | 0.068 | 0.18 |
| Smallest(1) | -0.1 | -0.06 | -0.05 | -0.14 | -0.09 | -0.06 | -0.11 | -0.1 | -0.23 | -0.201 | -0.23 | -0.28 | -0.1 | -0.1 | -0.11 |
| Confidence | 0.002 | 0 | 0.001 | 0.003 | 0.002 | 0.002 | 0.003 | 0.002 | 0.002 | 0.002 | 0.002 | 0.002 | 0.002 | 0.002 | 0.002 |
Source: Processed and developed by authors
Table-2.Auto Correlation and Q statistics
| Details | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | |
| IND | AC | -0.498 | 0.008 | -0.017 | 0.048 | -0.071 | 0.006 | 0.062 | -0.038 | -0.018 | -0.001 |
| Q-Stat | 104.1 | 104.2 | 104.3 | 105.3 | 107.4 | 107.4 | 109.1 | 109.7 | 109.8 | 109.8 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| BD | AC | 0.251 | 0.089 | -0.059 | -0.008 | 0.020 | -0.100 | -0.061 | 0.035 | 0.036 | -0.042 |
| Q-Stat | 14.2 | 16.0 | 16.8 | 16.8 | 16.9 | 19.2 | 20.0 | 20.3 | 20.6 | 21.0 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| MAL | AC | 0.024 | -0.032 | 0.000 | 0.025 | -0.071 | 0.051 | -0.005 | -0.019 | -0.002 | 0.053 |
| Q-Stat | 93.1 | 93.7 | 93.7 | 95.4 | 100.7 | 103.5 | 103.6 | 103.7 | 103.8 | 105.3 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| CHI | AC | -0.540 | 0.084 | -0.072 | 0.069 | -0.053 | 0.054 | -0.054 | 0.029 | -0.023 | 0.057 |
| Q-Stat | 121.2 | 124.1 | 126.2 | 128.2 | 129.4 | 130.6 | 131.9 | 132.2 | 132.5 | 133.8 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| THA | AC | -0.530 | 0.081 | -0.032 | -0.034 | 0.025 | 0.051 | -0.118 | 0.091 | -0.049 | 0.039 |
| Q-Stat | 118.0 | 120.8 | 121.2 | 121.7 | 122.0 | 123.1 | 129.0 | 132.6 | 133.6 | 134.3 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| SIN | AC | -0.508 | 0.116 | -0.157 | 0.057 | -0.034 | 0.055 | -0.058 | 0.069 | -0.058 | -0.006 |
| Q-Stat | 108.5 | 114.2 | 124.6 | 126.0 | 126.5 | 127.8 | 129.2 | 131.3 | 132.7 | 132.7 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| VIE | AC | -0.436 | -0.011 | -0.054 | -0.010 | 0.051 | -0.047 | 0.024 | -0.013 | 0.059 | -0.088 |
| Q-Stat | 78.5 | 78.6 | 79.8 | 79.8 | 80.9 | 81.8 | 82.1 | 82.1 | 83.6 | 86.9 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| HoK | AC | -0.500 | 0.045 | -0.065 | 0.001 | 0.013 | 0.002 | -0.019 | 0.095 | -0.079 | -0.043 |
| Q-Stat | 105.2 | 106.1 | 107.8 | 107.8 | 107.9 | 107.9 | 108.1 | 111.9 | 114.6 | 115.4 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| INDO | AC | -0.555 | 0.065 | 0.001 | -0.055 | 0.102 | -0.051 | -0.029 | 0.069 | -0.111 | 0.116 |
| Q-Stat | 262.0 | 265.5 | 265.5 | 268.1 | 277.0 | 279.2 | 280.0 | 284.0 | 294.5 | 306.1 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| PAK | AC | -0.464 | -0.036 | 0.032 | 0.014 | -0.033 | -0.045 | 0.032 | 0.007 | -0.006 | 0.030 |
| Q-Stat | 183.9 | 185.0 | 185.9 | 186.1 | 187.0 | 188.7 | 189.6 | 189.6 | 189.7 | 190.4 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| SoK | AC | -0.550 | 0.066 | -0.041 | 0.056 | -0.029 | -0.044 | 0.094 | -0.073 | 0.065 | -0.070 |
| Q-Stat | 259.7 | 263.4 | 264.9 | 267.6 | 268.3 | 270.0 | 277.6 | 282.2 | 285.9 | 290.1 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| JAP | AC | -0.533 | 0.092 | -0.095 | 0.043 | -0.019 | 0.020 | -0.010 | 0.009 | -0.005 | -0.001 |
| Q-Stat | 243.8 | 251.1 | 258.9 | 260.5 | 260.8 | 261.1 | 261.2 | 261.3 | 261.3 | 261.3 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| PHI | AC | -0.582 | 0.158 | -0.117 | 0.059 | -0.003 | -0.023 | 0.014 | -0.012 | -0.005 | -0.002 |
| Q-Stat | 142.2 | 152.7 | 158.5 | 160.0 | 160.0 | 160.2 | 160.3 | 160.3 | 160.3 | 160.3 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| TAI | AC | -0.560 | 0.137 | -0.112 | 0.012 | 0.058 | -0.042 | -0.022 | 0.059 | -0.065 | 0.092 |
| Q-Stat | 130.0 | 137.7 | 142.9 | 143.0 | 144.4 | 145.1 | 145.3 | 146.8 | 148.6 | 152.1 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | |
| SRI | AC | 0.272 | 0.114 | 0.068 | 0.046 | 0.058 | 0.013 | -0.033 | -0.023 | 0.066 | -0.047 |
| Q-Stat | 63.6 | 74.8 | 78.7 | 80.6 | 83.5 | 83.6 | 84.6 | 85.1 | 88.9 | 90.8 | |
| Prob | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
Source: Processed and developed by authors
It can be observed from the above table that, our sample data from all the considering markets provided a p value which is almost zero. Therefore, we can conclude that, our findings reject the null hypothesis of random walk. In order to understand the inter-correlation among all the markets, we studied the correlation matrix in Table-3 below where we analysed the auto correlation. The pairwise correlation coefficient in these markets is either positively or negatively related. The positive correlation means the returns from both countries are in same direction and negatively related correlation coefficient reveals that the returns from both countries are inversely related. The observed correlation coefficient range was .71 to -.13. Based on our result, returns from South Korea and Indonesia are highly positively related and those of Indonesia and Hong Kong are highly negatively related.
Table-3. Correlation Matrix
| IND | BD | MAL | CHI | THA | SIN | VIE | HoK | INDO | PAK | SoK | JAP | PHI | TAI | SRI | |
| IND | 1 | ||||||||||||||
| BD | -0 | 1 | |||||||||||||
| MAL | 0.02 | 0.06 | 1 | ||||||||||||
| CHI | 0.03 | -0 | 0.03 | 1 | |||||||||||
| THA | 0.04 | 0.06 | 0.48 | 0.05 | 1 | ||||||||||
| SIN | 0.08 | 0.02 | 0.6 | 0.13 | 0.53 | 1 | |||||||||
| VIE | 0 | 0.06 | 0 | -0.1 | -0 | 0 | 1 | ||||||||
| HoK | 0.01 | 0.02 | 0.59 | 0.19 | 0.56 | 0.7 | -0 | 1 | |||||||
| INDO | 0.14 | -0 | -0.1 | 0.15 | -0.2 | -0.1 | 0.08 | -0.1 | 1 | ||||||
| PAK | 0.14 | -0 | 0.24 | 0.14 | 0.21 | 0.3 | 0.01 | 0.23 | -0.01 | 1 | |||||
| SoK | 0.05 | -0.1 | 0.56 | 0.03 | 0.5 | 0.7 | -0 | 0.71 | -0.06 | 0.1 | 1 | ||||
| JAP | 0.03 | 0.02 | 0.39 | 0.08 | 0.37 | 0.6 | -0 | 0.58 | -0.03 | 0.1 | 0.6 | 1 | |||
| PHI | 0.03 | -0 | 0.52 | 0 | 0.52 | 0.6 | 0 | 0.53 | -0.1 | 0.2 | 0.5 | 0.4 | 1 | ||
| TAI | -0 | 0 | 0.13 | 0.1 | 0.04 | 0.1 | 0.16 | 0.08 | 0.16 | 0.1 | 0 | 0 | 0.1 | 1 | |
| SRI | 0.09 | -0 | 0.14 | 0.03 | 0.19 | 0.1 | 0 | 0.17 | 0.06 | 0 | 0.1 | 0.2 | 0.1 | 0.1 | 1 |
Source: Processed and developed by authors
Table-4. Unit Root Test
| Country | Augmented Dickey- Fuller (ADF) Test At Level |
Augmented Dickey - Fuller (ADF) Test at first Difference |
| India | -0.631 | -20.4 |
| BD | -2.187 | -11.6 |
| Malaysia | 0.545 | -20.1 |
| China | -2.180 | -19.9 |
| Thailand | -0.035 | -21.6 |
| Singapore | -1.683 | -19.1 |
| Vietnam | -2.382 | -19.2 |
| Hong Kong | -2.946 | -20.1 |
| Indonesia | -1.191 | -32.3 |
| Pakistan | -3.593 | -30.5 |
| South Korea | -1.069 | -30.5 |
| Japan | -2.290 | -29.2 |
| Philippines | -0.137 | -21.2 |
| Taiwan | -2.427 | -21.6 |
| Sri Lanka | -0.571 | -22.1 |
Source: Processed and developed by authors
In the above table, we have conducted a Unit Root Test to justify the stationary which is examined to understand the random walk. We can observe the presence of stationary or non-stationary in a considering market by examining Unit root test. Analysing the data from the period of 2001 to 2017 with almost zero probabilities indicates that, it rejects the null hypothesis of random walk. Here, the series of indices are non-stationary at level and stationary at first difference of 5% significance level. If a market is non-stationary, then it is unpredictable or cannot be forecasted. According to the Augmented Dickey- Fuller (ADF) Test at first difference, the time series of indices are stationary. In a stationary situation, a market can be predicated beforehand and there is a possibility of achieving arbitrage profit.
Table-5. Run Test
| Details | IND | BD | MAL | CHI | THA | SIN | VIE | HoK | INDO | PAK | SoK | JAP | PHI | TAI | SRI |
| K=Mean | 0.22 | 0.14 | 0.15 | 0.10 | 0.28 | 0.10 | 0.18 | 0.12 | 0.36 | 0.46 | 0.21 | 0.09 | 0.31 | 0.12 | 0.35 |
| No. of Run | 2.00 | 20.00 | 4.00 | 23.00 | 10.00 | 20.00 | 14.00 | 44.00 | 10.00 | 6.00 | 4.00 | 17.00 | 4.00 | 24.00 | 2.00 |
| Z | -20.37 | -12.36 | -20.16 | -18.18 | -19.56 | -18.61 | -19.01 | -16.25 | -28.53 | -28.84 | -29.07 | -28.17 | -20.17 | -18.05 | -29.21 |
| P-value | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
Source: Processed and developed by authors
In the table above, the p values are almost zero for all the considering countries at 5% significance level which indicates that the actual number of runs are less than the expected number of runs. Therefore it rejects the null hypothesis of random walk.
4.1. Summary of Findings
Table-6. Summary of Findings
| Country | Auto Correlation | Q statist | Unit Root Test | Run Test |
| Indication | If accept Random walk 'YES' if reject Random Walk 'NO' | |||
| India | No | No | No | No |
| BD | No | No | No | No |
| Malaysia | No | No | No | No |
| China | No | No | No | No |
| Thailand | No | No | No | No |
| Singapore | No | No | No | No |
| Vietnam | No | No | No | No |
| Hong Kong | No | No | No | No |
| Indonesia | No | No | No | No |
| Pakistan | No | No | No | No |
| South Korea | No | No | No | No |
| Japan | No | No | No | No |
| Philippines | No | No | No | No |
| Taiwan | No | No | No | No |
| Sri Lanka | No | No | No | No |
Source:Processed and developed by authors
The summary of all findings are compiled and disclosed in the table above. It can be concluded that the considering markets are inefficient according to the results found on the conducted tests. Therefore, excess profit can be earned through analysing.
5. CONCLUSION
The study examines the weak form efficiency in the capital markets from fifteen Asian countries. The considering time period was seventeen years. No one is expected to earn abnormal profit if market follows the random walk. To analyse the normal distribution, we studied the descriptive statistics in Jarque-Bera test and it rejected the null hypothesis in considering markets excluding Bangladesh, Hong Kong, and Thailand. In the Unit Root Test, the markets were found to be stationary which indicates that the considering markets are predictable at first difference. Finally, the findings from Auto-Correlation, Q Statistics and Run Test were unable to show the market efficiency. It indicates that, the considering markets are inefficient and do not follow the random walk. Therefore, technical analysts can achieve arbitrage profit due to market inefficiency. The outcome of this empirical study can be used for conducting further researches, for example, cross market integration, efficiency and cross market integration among various countries and to test the market efficiency among Europe and Asian countries as well.
| Funding: This study received no specific financial support. |
| Competing Interests: The authors declare that they have no competing interest. |
| Contributors/Acknowledgement: Both authors contributed equally to the conception and design of the study. |
REFERENCES
Chiang, T.C. and S.C. Doong, 2001. Empirical analysis of stock returns and volatility: Evidence from seven Asian stock markets based on TAR-GARCH model. Review of Quantitative Finance and Accounting, 17(3): 301-318. View at Google Scholar
Cooray, A. and G. Wickremasinghe, 2005. The efficiency of emerging stock markets: Empirical evidence from the South Asian region. Journal of Developing Areas, 41(1): 171 - 184. View at Google Scholar |
Dickey, D.A. and W.A. Fuller, 1979. Distribution for estimators for autoregressive time series with a unit root. Journal of the American Statistical Society, 74(366a): 427-431. View at Google Scholar | View at Publisher Dickey, D.A. and W.A. Fuller, 1981. Likelihood ratio statistics for autoregressive time series. Econometrica, 49(4): 1057-1072. View at Google Scholar | View at Publisher
Elliot, B.E., T.J. Rothenberg and J.H. Stock, 1996. Efficient tests of the unit root hypothesis. Econometrica, 64(8): 13-36. View at Google Scholar
Fama, E.F., 1970. Efficient capital markets: A review of theory and empirical work. Journal of Finance, 25(2): 383-417. View at Google Scholar | View at Publisher
Fornell, C., S. Mithas, F.V. Morgeson and M.S. Krishnan, 2006. Customer satisfaction and stock prices: High returns, low risk. Journal of Marketing, 70(1): 3-14.View at Google Scholar | View at Publisher
Füss, R., 2005. Financial liberalization and stock price behavior in Asian emerging markets. Economic Change and Restructuring, 38(1): 37-62. View at Google Scholar | View at Publisher
Gündüz, L. and J.A.B.D.U.L.N.A.S.S.E.R. Hatemi, 2005. Stock price and volume relation in emerging markets. Emerging Markets Finance and Trade, 41(1): 29-44. View at Google Scholar
Hamid, K., M.T. Suleman, S.S.Z. Ali, I. Akash and R. Shahid, 2017. Testing the weak form of efficient market hypothesis: Empirical evidence from Asia-Pacific markets.
Hussain, S.S.J., M. Zakaria, S. Ali and N. Raza, 2015. Market efficiency and asymmetric relationship between South Asian stock markets: An empirical analysis. Pakistan Journal of Commerce & Social Sciences, 9(3): 875-889. View at Google Scholar
Islam, M.S., H.A.R. Khan and M.F. Ahmed, 1996. The behavior of stock investment in Bangladesh. Savings and Development, 20(4): 447-460. View at Google Scholar
Kaizoji, T., 2000. Speculative bubbles and crashes in stock markets: An interacting-agent model of speculative activity. Physica A: Statistical Mechanics and its Applications, 287(3-4): 493-506. View at Google Scholar | View at Publisher
Kumar, H. and R. Jawa, 2017. Efficient market hypothesis and calendar effects: Empirical evidences from the Indian stock markets. Business Analyst, 37(2): 145-160. View at Google Scholar
Lim, K.P. and R. Brooks, 2011. The evolution of stock market efficiency over time: A survey of the empirical literature. Journal of Economic Surveys, 25(1): 69-108. View at Google Scholar | View at Publisher
Oprean, C. and C. Tanasescu, 2014. Effects of behavioural finance on emerging capital markets. Procedia Economics and Finance, 15: 1710-1716.< View at Google Scholar | View at Publisher
Perron, P., 1988. Trends and random walks in macroeconomic time series: Further evidence from a new approach. Journal of Economic Dynamics and Control, 12(2-3): 297-332.View at Google Scholar | View at Publisher
Phan, K.C. and J. Zhou, 2014. Market efficiency in emerging stock markets: A case study of the Vietnamese stock market. IOSR Journal of Business and Management, 16(4): 61-73. View at Google Scholar | View at Publisher
Shaik, M. and S. Maheswaran, 2016. Random walk in emerging Asian stock markets. International Journal of Economics and Finance, 9(1): 20-31. View at Google Scholar | View at Publisher
Shiller, R.C., 2000. Irrational exuberance. Philosophy & Public Policy Quarterly, 20(1): 18-23.View at Google Scholar
Tuyon, J. and Z. Ahmad, 2016. Behavioural finance perspectives on Malaysian stock market efficiency. Borsa Istanbul Review, 16(1): 43-61. View at Google Scholar | View at Publisher