Academic journal article IUP Journal of Applied Finance

An Analysis of the Predictability of Asset Returns: A Case of Six Emerging Stock Markets of Asia

Academic journal article IUP Journal of Applied Finance

An Analysis of the Predictability of Asset Returns: A Case of Six Emerging Stock Markets of Asia

Article excerpt

(ProQuest: ... denotes formulae omitted.)

Introduction

The efficient market hypothesis is associated with the random walk theory that variation of prices is random in time and the excess returns are unpredictable. The empirical findings of significant autocorrelation in stock returns have led some authors to propose an alternative theory to the random walk hypothesis: the ARIMA model. This is not based on the concept of market efficiency but rather on market microstructure arguments. Ayadi and Pyun (1994) apply the variance ratio test developed by Lo and MacKinlay (1988) to investigate the behavior of prices of stock traded on the Korean stock market between January 1984 and December 1988. The results indicate that the Korean stock market is a random walk market. For emerging markets, Kim (2004) reports the existence of a random walk for Hong Kong, Japan and Korea and rejection of random walk hypothesis for Taiwan and Thailand. The study by Hoque et al. (2007) examines the random walk hypothesis for eight emerging equity markets in Asia: Hong Kong, Indonesia, Korea, Malaysia, Philippines, Singapore, Taiwan and Thailand. They used weekly market prices and covered the period from April 1990 to February 2004. They found that the stock prices of the eight Asian countries do not follow random walk, with the possible exceptions of Taiwan and Korea. The same results were found by Lock (2008) using weekly data covering the period 1990 through 2006 of Taiwan stock market. The study by Charles and Darné (2009) examines the random walk hypothesis for the Shanghai and Shenzhen indexes for Chinese stock exchanges using daily data over the period 1992-2007. They find that the former does not follow the random walk hypothesis, but the first index seems more efficient. According to the variance ratio test, the findings of Segot and Lucey (2008) show the rejection of null hypothesis of random walk for Egypt, Morocco, Lebanon, Jordan and Tunisia.

Efficient Market Hypothesis

One of the most important concepts in modern finance is the efficient market hypothesis associated with the random walk hypothesis. Fama (1970, p. 383) summarizes that "a market in which prices always 'fully reflect' the available information is called 'efficient'". To specify what subset of available information is to be 'fully reflected' in stock prices, Fama (1970) presents the following classification of information sets:

* Weak Form Efficiency: The information set includes only historical prices of returns.

* Semi-Strong Form Efficiency: The information set includes all publicily available information.

* Strong Form Efficiency: The information set includes all privately available information.

In our work, we will be concerned with the weak form efficiency, since our goal is to investigate the predictability of stock returns from the time series of the historical returns.

Random Walk Hypothesis

The random walk model assumes that a security's prices fully reflect the available information and implies independence of successive returns. Formally:

... (1)

where rj,t is the return of jth security at point t. The serial covariance of returns is zero at all leads and lags and the expected returns are equal to the unconditional mean of the distribution f at all times.

Lo and MacKinlay (1988) present a simple specification test aimed at testing the random walk hypothesis outlined above. The random walk for a log stock price pt can be written as:

... (2)

where,

m is the expected one period rate of return on the stock;

et is a sequence of independent and identical residuals.

The principle of the test is that the variance of a qth difference of the process (2) is equal to the sum of the corresponding q first difference variances. To refer to Lo and MacKinlay (1988), the qth difference of (2) can be written as:

It follows that the ratio ... must be equal to one under the null hypothesis of random walk and the test is called the variance ratio test. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.