TESTING NON-STATIONARY BEHAVIOR OF STOCK PRICES: AN EMPIRICAL EVIDENCE FROM MAJOR ASIAN STOCK MARKETS

Transcription

1 TESTING NON-STATIONARY BEHAVIOR OF STOCK PRICES: AN EMPIRICAL EVIDENCE FROM MAJOR ASIAN STOCK MARKETS Sumit Kumar 1, Dr. Mohammad Anees 2 1 Research Scholar (SRF), 2 Assistant Professor, Department of Business Administration, University of Lucknow, Lucknow(India) ABSTRACT Purpose of the study - The present study examines the Weak form of efficiency of three major stock exchanges of Asia BSE-30 India, SSEC China, N-225 Japan, through the study of Non stationary (Existence of Unit root) behavior of the stock prices. Data and Methodology - The stock prices of BSE 30, N225, SSEC, on the daily bases over the period of 2009 to 2013 has been taken from the website of yahoo finance which provides all the data related to the major stock market in all region of the world. The non stationary behavior series are tested with the help of the unit root testing using Augmented Dickey Fuller (ADF-1979, 1981), the Phillips-Perron (PP-1988), on the market price index series at level and first differences, the existence of unit root at level and absents in first order differences confirm that series are non stationary. Findings- The study clearly reveal that all the three stock market of Asia showing the Non-stationary behavior (existence of unit root) i.e. all these markets are showing weak form of market efficiency. Limitation- The study is conducted only in three market namely BSE, N-225 and SSEC, and the data are of five year only so the result may be vary if other market may be included and the more data from past year may be taken for study. The study is limited to price behavior of stock market, the return of stock is not considered in present study. Contribution to the extant Literature For investor- Present paper give clear cut indication to the investor that on the basis of past prices and information and with the help of any kind of analysis they cannot able to predict the future stock prices of the market and no one have any opportunity to earn extra return by just properly managing historical market information and trading strategy. For Scholar/ academician- In future study on the Market efficiency will be checked for the further higher levels of the semi strong and strong form of efficiency i.e the Informational efficiency of the Markets. Keywords: Market Efficiency, Asian Stock Market, Random Walk, Non Stationary, Unit Root. 184 P a g e

2 I. INTRODUCTION In recent decades the role of stock markets are realizing important in the development of the economy of the developing economies the actual level of economic growth and development of a country is fully reflected in the stock prices traded in these markets. Market efficiency has keen interest among the investor as their profit in investment depends on this and they have always doubt that whether the stock market where he is going to invest is an efficient market or not. The term market efficiency in stock market is used to measure the extent of logical, ethical and desirable trading operation on the basis of available relevant information. The concept of Efficient Market Hypothesis (EMH) is based on the arguments of Samuelson (1965) that predictable price of an asset swings in random manner. In finance, the efficient-market hypothesis (EMH) assures that financial markets are "informational efficient. That is, one cannot consistently earn returns in excess of average market returns on a risk-adjusted basis, on the base of the information publicly available at the time the investment is made. With regards to EMH theory in corporate finance Fama (1970) has put a landmark through his work by extending EMH in three major versions of the hypothesis: weak, semi-strong, and strong and proposed three models for testing market efficiency, The Fair Game model, the Sub martingale model, and the Random Walk model. Testing the weak form of the market efficiency has crucial importance as it is the beginning of the hypothesis if the market fails to support the weak form it is not necessary to test the hypothesis in other levels of the markets (Wong and Kwong,1984). In weak form of efficient market, current price reflects all the information contained in past price i.e. there are no linear as well as non-linear dependences with the lagged values and price process has no memory on the other hand in weak form inefficient markets information of the past are not completely absorbed in the current prices and an investor has the opportunity to make an estimate of current prices on the basis of past information. II. TESTING OF NON STATIONARY (UNIT ROOT) The theory behind ARMA estimation is based on stationary time series. A series is said to be (Weakly or covariance) stationary if the mean and auto covariance s of the series do not depend on time. Any series that is not stationary is said to be non stationary. A common example of a non stationary series is the random walk A test of stationarity (or nonstationarity) that has become widely popular over the past several years is the unit root test (stochastic) process that is a time series Yt = ρyt 1 + ut 1 ρ (21.4.1) where ut is a white noise error term. We know that if ρ = 1, that is, in the case of the unit root, (21.4.1) becomes a random walk model without drift, which we know is a non stationary stochastic process. Therefore, why not simply regress Yt on its (one period) lagged value Yt 1 and find out if the estimated ρ is statistically equal to 1? If it is, then Yt is non stationary. This 185 P a g e

3 is the general idea behind the unit root test of stationarity. For theoretical reasons, we manipulate (21.4.1) as follows: Subtract Yt 1 from both sides of (21.4.1) to obtain: Yt Yt 1 = ρyt 1 Yt 1 + ut (21.9.1) = (ρ 1)Yt 1 + ut which can be alternatively written as: Yt = δyt 1 + ut (21.9.2) where δ = (ρ 1) and _, as usual, is the first-difference operator. In practice, therefore, instead of estimating (21.4.1), we estimate (21.9.2) and test the (null) hypothesis that δ = 0. If δ = 0, then ρ = 1, that is we have a unit root, meaning the time series under consideration is non stationary. Before we proceed to estimate (21.9.2), it may be noted that if δ = 0, (21.9.2) will become Yt = (Yt Yt 1) = ut (21.9.3) Since ut is a white noise error term, it is stationary, which means that the first differences of a random walk time series are stationary. (Damodar N. Gujrati) III. LITERATURE REVIEW With the globalization and integration of stock markets, emerging Asian stock markets of India, china, Japan, has achieving reforming regulation and laws and offering ample opportunity to investors to invest and gains with this characters these major Asian markets have capture the attention of academics, practitioners, and policy makers. There is large number of academic literature focused on measuring the benefits and efficiency of these markets exists. Jae H. Kim (2004) has tested the martingale (or random walk) hypothesis in the stock prices of a group of Asian countries. The selected countries represent well-developed markets (Hong Kong and Japan) as well as emerging markets (Korea, Taiwan and Thailand) with the help of joint variance ratio test based on the wild bootstrap method and concluded that the stock prices of Japan, Korea, and Hong Kong are support the martingale hypothesis, indicating that their stock markets have been efficient where as the stock prices of Taiwan and Thailand do not support the hypothesis as they are weak form inefficient. A. C. Worthington and H. Higgs (2006) inspect the weak-form of market efficiency of ten emerging Asian market of China, India, Indonesia, Korea, Malaysia, Pakistan, the Philippines, Sri Lanka, Taiwan and Thailand and five developed markets Australia, Hong Kong, Japan, New Zealand and Singapore. The result reveals that none of the emerging Asian markets are fallowing the random walks model and hence are not weak-form efficient, where as except the Australian market only the developed markets in Hong Kong, New Zealand and Japan are strongly support the random walk hypothesis.nikunj R. Patel, NiteshRadadia and J(2012) have studies the weak form of market efficiency of four selected Asian stock markets for the periods of1st January 2000 to 31st March concluded that BSE Sensex and NIKKEI are in the weak form inefficient form on the other hand HANSENG and SSE Composite fallows weak form of efficiency. Saqib Nisar and Muhammad Hanif (2012) examined the weak form of efficient market hypothesis on the four major stock exchanges of South Asia including, India, Pakistan, Bangladesh and Sri Lanka. Historical index values on a monthly, weekly and daily basis for a period of 14 Years ( ) for this purpose they applied four statistical tests including runs test, serial correlation, unit root and variance ratio test and concluded that 186 P a g e

4 none of the four major stock markets of south-asia follows Random-walk and hence all these markets are not the weak form of efficient market. Jae H. Kim(2004) tests for the martingale (or random walk) hypothesis in the stock prices of a group of Asian countries. The selected countries represent well-developed markets Hong Kong and Japan as well as emerging markets Korea, Taiwan and Thailand. They found that the stock prices of Japan, Korea, and Hong Kong are found to follow the martingale, indicating that their stock markets have been efficient. The results of these three researchers are similar. The contradictory results in reference of emerging economics were founds by the Arusha V. Cooray, G. Wickramasighe(2007) as they use different econometric tests of Augmented Dickey Fuller (ADF-1979, 1981), the Phillips-Perron (PP-1988), the Dicky-Fuller Generalized Least Square (DF-GLS-1996) Elliot-Rothenberg- Stock (ERS 1996) and found that the stock markets of India, Sri Lanka, Pakistan and Bangladesh. During the period of 1996 to2005 are in Weak form efficient. Hin Yu Chung(2006)has examines the random walk hypothesis for two major stock markets in China. Using Daily returns from February 21, 1992 to December 30, 2006 for the Shanghai A-share, Shanghai B-share, Shanghai Composite and Shenzhen Composite and from October 5, 1992 to December 30, 2005 for Shenzhen A-share and Shenzhen B-share, he concluded that both the Chinese stock markets are weak-form inefficient, a contradictory results for this study is done by thekian-ping et.al.theystudy the random walk behavior of Shanghaiand Shenzhen Stock Exchangesand founds that both of themarkets are consistent with random walk for long periods of time, after the 1997 there exists a serial dependency between both the market which reveals that the both the market reacts in similar fashion to the information related topolitics, economic, social andinstitutional changes, continuation with this Faiq Mahmood et.al.(2010) has studies the behavior of chiness stock market before and after recession by using ADF, DFGLS,PP and KPSS tests on for both the market Shenzhen and Shanghai stock exchanges separately.they concluded that Chinese stock market is consistently weak form efficient and the globalfinancial crisis has no significant impact on the efficiency of Chinese stock market. From the above discussion In respect of the chiness stock market it is founds that most of the researchers has founds the similar results which shows that the market is weak form efficient, there is no opportunity to the investors to beat the market return just with exploiting the privet information or using other trading practices. Ravi kumar Gupta (2014) has tested the efficiency of Indian stock market. BSE index, SENSEX was used in the study to represent the Indian stock market. The daily closing points were taken for the sample period of ten years from January 2003 to December Different statistical tools like Unit Root test, Runs test and Kolmogorov Smirnov test (K S test) and found that Indian stock market are weak form inefficient which means information of the past are not completely absorbed in the current prices. Haritika Arora (2013) attempts to verify weak form of efficient market hypothesis and random walk hypothesis using daily data for the index of Indian Stock Market specifically S&P CNX Nifty (Index of National Stock Exchange) for the period of 1 January 2000 to 31 Dec Statistical analysis is done with help of Augmented Dickey and Fuller (ADF) test, Auto-correlation test (Breusch-Godfrey Serial Correlation LM Test), Ljung-Box Q test, Auto-regression, ARIMA model, portmanteau BDS test and GARCH(1,1) model. Results exhibited that returns series are characterized by linear as well as nonlinear dependences and a high persistence of volatility 187 P a g e

5 clusters over the sample period. The hypothesis of random walk for the series has become redundant. Hence, it can be concluded that Indian Stock market do not show evidence of weak from of market efficiency. R. Vaidyanathkant, kumargali(1994).have tests the weak for efficiency of the Indian capital market for the period 1989 to1994.on the bases of the ten actively traded share prices on Bombay stock exchange,by using Run test and serial correlation they argues that market fallows the weak form of Efficient Market Hypothesis.Similar results are founds by the Rakesh Gupta andjunhao Yang (2011)for two major Indian equity markets BSE and NSE for the sample period 2007 to 2011, but for period 1997 to 2007 it did not support the weak form efficiency. Sunil Poshakwale(1996). Argues that over period of BSE exhibitsthe evidence of day of the week effect and weak form inefficient, the market conditions provide the opportunity to the investorby using different buy and hold strategy issues they get extra returns., similar results were found by the Alan Harper ZhenhuJin for July 1997 to 2011,the Indian stock market are not weak form efficient which provide the opportunity to the investors to earn extra return from exploiting the trading strategy. byaugment using Dickey-Fuller (1979) test and the Phillips-Perron (198l), P. Srinivasan (Nov 2010) concluded that the daily stock market returns of two major indices, S&P CNX NIFTY and the SENSEX does not fallow the random walkbehavior,divyang J Joshi, 2012using Runtest he concluded that 6 major Indian Stock marketindices, BSE 30, BSE 100,200,500, BSE SMALL CAP and BSE MIDCAP for the period from 1st January 2001 to 31st December 2010, do not fallows the random walk model and the markets are in weak form inefficient. In relation with the Indian stock market most of the researches supported that these markets are in weak form inefficient that means the returns in these market are predictable up to some extent, and investor have an opportunity to get extra return where as in respect of the Japanese the short literature and its markets are compared with the other stock markets there are also similar conclusions arrived by the researches as Chinese markets Market efficiency is the topic for the discussion among the researchers, several researches have examined market efficiency over the decades, and found the conflicting results some of the argues inefficient and others are efficient but there are few of the cases proven the efficient form of the market around the Asia, my study is also an attempt to investigate the empirical results and try to test weak form of market efficiency during the current period just after the great global recession. To our knowledge this is the contemporary study which covers the emerging Asian markets for the period after the global financial crises of , hence as insignificant evidence on market efficiency of this kind is being contributed in the literature. Such a study may be useful for present and potential local and global capital market investors IV. DATA ANALYSIS AND RESEARCH METHODOLOGY The stock prices of BSE 30, N225, SSEC, on the daily bases over the period of 2009 to 2013 has been collected from the website of yahoo finance which provides all the data related to the major stock market in all region of the world. The non stationary behavior series are tested with the help of the unit root testing with the help of two major tests applied on the logarithm of the market price index at level and first differences, the existence of unit root at level and absents in first order differences confirm that series are non stationary. 188 P a g e

6 V. THE AUGMENTED DICKEY-FULLER (ADF) TEST The standard DF test is carried out by estimating the equation ΔYt =(ρ 1)Yt 1 + ut, ΔYt = δyt 1 + ut where δ=( -1) The null and alternative hypotheses may be written as, H0: δ=0, H1:δ<0 Dickey and Fuller (1979) show that under the null hypothesis of a unit root, this statistic does not follow the conventional Student s t-distribution, and they derive asymptotic results and simulate critical values for various test and sample sizes. More recently, MacKinnon (1991, 1996) implements a much larger set of simulations than those tabulated by Dickey and Fuller. In addition, MacKinnon estimates response surfaces for the simulation results, permitting the calculation of Dickey-Fuller critical values and p-values for arbitrary sample sizes. VI. THE PHILLIPS-PERRON (PP) TEST Phillips and Perron (1988) propose an alternative (nonparametric) method of controlling for serial correlation when testing for a unit root. The PP method estimates the non-augmented DF test equation, and modifies the α- ratio of the coefficient so that serial correlation does not affect the asymptotic distribution of the test statistic. (E -views user guide ii) 6.1 Empirical Results and Analysis Table-1 Augmented Dickey-Fuller: For level Null Hypothesis: BSE30, N225, SSEC has a unit root Exogenous: Constant Lag Length: 0 (Automatic - based on SIC, maxlag=22) Observations:1228 A DF test statistic t-statistic Prob.* BSE N SSEC Test critical value At 1% % % *MacKinnon (1996) one-sided p-values Table-2 Augmented Dickey-Fuller: For 1 st Difference Null Hypothesis: BSE30,N225,SSEC has a unit root Exogenous: Constant, Lag Length: 0 (Automatic - based on SIC, maxlag=22) 189 P a g e

7 Observation:1228 Augmented Dickey-Fuller test statistic t-statistic Prob.* BSE N SSEC Test critical value At 1% 5% 10% *MacKinnon (1996) one-sided p-values Table no 1and 2 represents the output of ADF test From the table no 1 the value of t-statistics lies within the acceptance regions that mean the p value for all the three markets are higher than the 0.05 (5 % level of significance), so the hypothesis of unit root cannot be rejected at level in the next step table no 2 t statistics for all the series at first difference are falls in rejection region which clearly indicates that our null hypothesis of existence of unit root is rejected, i.e. all the market indices are stationary at first difference which shows that and these markets are represents non stationary behavior. Table-3. Phillips-Perron test: for Level Null Hypothesis: BSE30,N225, SSEC has a unit root Exogenous: Constant Bandwidth: 11,7,5 (Newey-West automatic) using Bartlett kernel Observations:1228 Phillips-Perron test statistic t-statistic Prob.* BSE N SSEC Test critical value At 1% 5% 10% Table -4. Phillips-Perron test: For 1 st Difference Null Hypothesis: BSE30,N225,SSEC has a unit root Exogenous: Constant, Bandwidth: 11,7,6 (Newey-West automatic) using Bartlett kernel Observation:1228 Phillips-Perron test statistic t-statistic Prob.* BSE N SSEC P a g e

8 Test critical value At 1% % % MacKinnon (1996) one-sided p-values Table no 3 represents the output of Phillip-perron test with constant using (Newey-West automatic) using Bartlett kernel, at the bandwidth of 11 for BSE 7 for N225,5 for SSEC, the value of t-statistics falls with the acceptance region at the 5% level of significance, and the MacKinnon (1996) one-sided p-values is greater than the 0.05 i.e our null hypothesis of existence of unit root is accepted at level, table no 4 shows, the output of Phillip-perron test for the first differences of the three series. The null hypothesis of unit root is rejected for all, on the bases of these two tables the market price series of three markets fallowing the non stationary characteristics. VII. CONCLUSION All the financial needs of corporate for short terms as well as long terms are fulfilled by the Capital market of the country from old period. The performance and growth of the industries and corporate are accurately reflected in the price of their stocks traded in the listed stock market. For healthy economic environments it is essential that assets are accurately priced over the capital market and markets operate in an efficient and transparent way. Market efficiency has its role for the investor for making investment in the market. The results of the study reveals that the all three stock indices exhibits the Non stationary characteristics ( existence of unit root),which proves the series under investigation are random walk series i.e. three major Asian stock markets are weak form efficient, it support the work done by Rakesh Gupta and Junhao Yang in earlier time (20110 on the other hand the results contradict with the study of Nikunj R. Patel, Nitesh Radadia(2011) for BSE and SSEC. it is clear cut indication to the investor that on the bases of past prices and information and with the help of any kind of analysis they cannot able to predict the future stock prices of the market and no one have any opportunity to earn extra return by just properly managing historical market information and trading strategy. The Market efficiency will be checked for the further higher levels of the semi strong and strong form of efficiency i.e the Informational efficiency of the Markets. REFERENCES [1]. Arora H. (June 2013) Testing Weak Form of Efficiency of Indian Stock Market, Pacific Business Review International Volume 5 Issue 12 [2]. CoorayA.V.andWickramasighe G. (2007). The Efficiency of Emerging Stock Markets: Empirical Evidence From The South Asian Region, Journal Of Developing Areas, 41 (1): [3]. Chung H.Y. (2006). Testing Weak-Form Efficiency of the Chinese Stock Market, Master s Thesis. Lappeenranta University of Technology, pp [4]. E-views7 user guide II chp-30 univariate Time series Analysis,p P a g e

9 [5]. Gupta K. R.(2014) Weak form efficiency of India stock market with reference to BSE, International Journal of Research in Business Management (IMPACT: IJRBM) (P): Vol. 2, Issue 9 [6]. Gupta R. and Yang J. (2011). Testing Weak form Efficiency in the Indian Capital Market, International Research Journal of Finance and Economics,(75): [7]. Gujrati. N Damodar Basic econometrics, The Mc Graw-Hill companies 2004 edt 4 th chp 21 the unit root test p814. [8]. Harper A. and Jin Z. (.) Examining market efficiency in India: an empirical analysis of the Random Walk Hypothesis, Journal of Finance and Accountancy, 1-6. [9]. Joshi J. Divyang (2012) Testing Market Efficiency of Indian Stock Market, Journal of Scientific and Research Publications, Volume 2, Issue 6, pp:1-4 [10]. Kian.P(2010). http/jpkc.jnu.edu.cn/2010/cbygl/ doc [11]. Kim J.H.(2004).Testing for the martingale hypothesis in Asian stock prices: evidence from a new joint variance ratio test, Department of Econometrics and Business Statistics Monash University, Caulfield East, VIC 3145, Australia. pp [12]. Mahmood F., Xinping X., ShahidH.and Usman M. (2010). Global Financial Crisis: Chinese Stock Market Efficiency, Asian Journal of Management Research, ISSN , [13]. Nisar S. and Hanif M. (2012). Testing Weak Form of Efficient Market Hypothesis: Empirical Evidence from South Asia, World Applied Sciences Journal 17(4): [14]. Poshakwale S. (1996). Evidence on Weak Form Efficiency and Day of the Week Effect in the Indian Stock Market, Finance India, X (3): [15]. Patel N.R., Radadia. N. and Dhawan J. (2012). An Empirical Study on Weak-Form of Market Efficiency of Selected Asian Stock Markets, Journal of Applied Finance & Banking, 2(2): [16]. R.vadiyanath kant, kumar Gali(1994). /accessed on 23/11/2014/ [17]. Srinivasan P. (2010). Testing Weak-Form Efficiency of Indian Stock Markets, Sri Krishna International Research & Educational Consortium, 1 (2): 1-7. [18]. Worthington, AC and Higgs, H (2006) Weak-form market efficiency in Asian emerging and developed equity markets: Comparative tests of random walk behavior, Accounting Research Journal, 19(1), pp Available at: [19]. accessed on 23/11/2014. [20]. [21]. [22] P a g e