Testing the evolving efficiency of Arab stock markets

https://doi.org/10.1016/j.irfa.2009.11.004Get rights and content

Abstract

Our goal is to examine whether Arab stock markets are becoming more efficient during the last decade thanks to organizational improvements and agents' learning. To achieve this goal a test of evolving weak-form efficiency using GARCH-M (1,1) approach along with state–space time-varying parameters is implemented for 11 Arab stock markets for periods ending in March 2009, rather than studying their efficiency/inefficiency at a given point of time as commonly done. All markets show high sensitivity to the past shocks and are found to be weak-form inefficient. Moreover, the efficiency does not clearly improve towards the first quarter of 2009 and negatively reacts to contemporaneous crises, except temporary sub-periods of efficiency improvement for the largest markets. This contrasts with mature markets and reveals the ineffectiveness of the reforms so far undertaken and calls to intensify efforts to expand and deepen these markets besides improving their liquidity and transparency and counteracting the shortcomings of the large individual trading by enhancing investment culture and spreading institutional trading.

Introduction

Most of the Arab countries reconsidered the role of stock markets in the early 1990s, by attempting to revitalize dormant existing markets, such as Egyptian, Saudi, Tunisian or Kuwaiti stock markets, or launching new ones, such us Dubai and Abu Dhabi stock markets.1 These actions aimed at developing their financial systems in order to stimulate economic growth and foster international integration. Overall, the pace of changes has been gradual and slow, and capital markets remain dominated by the banking systems. Nonetheless, certain policy measures, including new regulation allowing the entry of foreign investors and wider dissemination of information through the operation of electronic trading systems, resulted in growth in terms of capitalization and caused increase in the number of listed companies.

Moreover, the issue of market efficiency, as introduced by Fama, 1965, Fama, 1970, remains the most important from resource allocation and portfolio investment's point of view. Efficient mature markets are generally found to be weak-form efficient. Hence, conclusions for emerging markets are very mixed and generally support the idea of a departure from weak efficiency (as for Arab stock markets see for example Civelek, 1991, Butler and Malaikah, 1992, Al-Loughani, 1995, El-Erian and Kumar, 1995, Abraham et al., 2002, Onour, 2004, Moustafa, 2004, Smith, 2004, Squalli, 2006, Asiri, 2008, Lagoarde-Segot and Lucey, 2008,2 Marashdeh and Shrestha (2008) and Awad and Daraghma (2009)).

Nonetheless, the limitation to the above literature is twofold. First, it has been argued that conventional efficiency tests are only reliable if the methodology adopted accounts for the institutional features of the emerging markets leading to wrongly accept or reject the efficiency hypothesis (see, for example, Miller et al., 1994 and Antoniou et al., 1997). Thus, the family of GARCH models, developed by Engle (1982) and extended by Bollerslev, 1986, Nelson, 1991, is found to better fit empirical data of stock returns and accommodate for non-linear and infrequent trading caused by thinness, lack of liquidity and regulatory changes and became commonplace in empirical finance. Though, only few have tested GARCH models using daily data from Arab stock markets. Asal, 1998, Mecagni and Sourial, 1999, Tooma, 2003 investigated the Egyptian stock market while Hassan et al. (2003) investigated the Kuwaiti stock market. Both markets exhibit significant departure from the efficient market hypothesis, even though Asal (1998) by investigating efficiency on a yearly basis shows that efficiency improves towards 1997. Likewise, Hassan et al. (2003) show that Kuwaiti market efficiency improves towards the last sub-period of the 1990s. More recently, Alagidede and Panagiotidis (2009) investigated Egyptian, Moroccan and Tunisian stock markets in addition to Kenya, Nigeria, South Africa and Zimbabwe. The random-walk model was rejected in all cases. In addition, they found that the empirical stylized facts of volatility clustering, leptokurtosis and leverage effects are present in all cases, and thus GARCH, GARCH-M and EGARCH-M were fitted to model the conditional variance. Second, it is hardly credible for an infant stock market to be efficient at its early stages of development as it takes time to gradually evolve towards more efficiency as regulatory environment and trading conditions improve and trade participants become more sophisticated. Therefore, the steady-variable approaches testing its efficiency as fixed throughout the entire estimation period could not trace the efficiency evolution and lead to biased results.

To overcome these problems, a new stand of research has developed since Emerson et al., 1997, Zalewska-Mitura and Hall, 1999 examine the evolution of the efficiency of stock exchanges over time rather than assessing it at a given point of time. Thus, Emerson et al., 1997, Zalewska-Mitura and Hall, 1999, Rockinger and Urga, 2000, Rockinger and Urga, 2001, Hall and Urga, 2002, Harrison and Paton, 2004, Posta, 2008 revisited the weak-form efficiency of many European transition stock exchanges by using a GARCH-M (1,1) model of the daily index returns volatility as well as a Kalman filter state–space in estimating the time-varying dependency of the daily returns on their lagged values. This time-varying dependency is expected to become more stable and infinitely small if the market moves towards more efficiency. Changing levels of inefficiency were found in Bulgaria, Hungary, Czech, Poland, Russia and Romania stock markets, some of them showing however a tendency towards becoming more efficient. This methodology allows to explore a continuous and smooth change in the behavior of stock prices and thus captures the evolution of efficiency over time rather than splitting the period into sub-periods on the basis of postulated factors as in Antoniou et al., 1997, Muslumov et al., 2003, Hassan et al., 2003, Abrosimova et al., 2005, Coronel-Brizio et al., 2007, Lim and Brooks, 2009.

This line of inquiry is highly relevant for emerging markets at early stage of development for which previous studies regarding their efficiency should be reconsidered. For instance, Li, 2003, Maghyereh and Omet, 2003, Jeferis and Smith, 2005 implemented the test of evolving efficiency to investigate Chinese, Jordanian and African stock markets respectively. Li (2003) shows that Shanghai and Shenzhen stock markets were inefficient but showed a steady convergence towards efficiency. Maghyereh and Omet (2003) found that Jordanian stock market pricing efficiency is not improving despite the move to electronic trading system. In Jeferis and Smith (2005) Morocco and Egypt stock markets become weak-form efficient towards June 2001, which is not the case for Kenya and Zimbabwe stock markets.

The aim of the present paper is to implement the Evolving Efficiency Test for the most formal Arab stock markets. Therefore, the paper follows the Emerson et al., 1997, Zalewska-Mitura and Hall, 1999 methodology in order to investigate the dynamics of their weak-form efficiency/inefficiency rather than taking a snap shot of it at a given point of time. To this end we use daily data of 11 Arab stock markets that cover a wide time period up to March 2009, allowing us to test for the impact of the many organizational reforms undertaken by the authorities during the last decade. The sample period also allows us to account for the contemporaneous crisis effects on Arab stock market efficiency. The present paper is therefore better able to answer the numerous questions raised by markets' authorities, anxious to know whether the reforms so far undertaken are effective, and investors eager to safely diversify their investments through stock markets. The remainder of the paper is organized as follows. In Section 2, we present the data and discuss the markets' characteristics. Methodology and empirical results are presented in Section 3. Section 4 concludes.

Section snippets

Highlights from data

The data include daily prices of the national indexes of Saudi Arabia, Kuwait, Tunisia, Dubai, Egypt, Qatar, Jordan, Abu Dhabi, Bahrain, Morocco and Oman. In addition, we use data of the AMEX index for comparative purposes.

Table 1 shows a very significant but variable growth from a market to another regarding their size and liquidity. Saudi and Kuwait stock markets are the largest in terms of market capitalization, followed by Egypt and Qatar. Oman and Tunisia markets are far behind. With

GARCH-M (1,1) estimations

First we use GARCH-M (1,1) (Generalized Autoregressive Conditionally Heteroskedastic in Mean) model allowing the variance of the error term to vary over time, in contrast with classical regressions assuming constant variance. Also GARCH-M allows us to test for the presence of risk premium in the stock markets.4 Our GARCH-M (1,1) is stated as follows:rt=β0+β1rt1+δht+etetN(0,ht)ht=α0+α1ht1+α2et12

In addition to the

Conclusion

Arab countries have shown a growing concern in developing their stock markets since the early 1990s, which explains their number and the many reforms undertaken in order to improve their liquidity and efficiency. These markets have shown a fair development regarding their size and liquidity even though the progress in terms of efficiency remains mixed if one refers to the empirical literature. However, conventional efficiency tests often applied to these emerging markets are considered

References (56)

  • A. Zalewska-Mitura et al.

    Examining the first stages of market performance: A test for evolving market efficiency

    Economics Letters

    (1999)
  • A. Abraham et al.

    Testing the random walk behavior and efficiency of the Gulf stock markets

    The Financial Review

    (2002)
  • Abrosimova, N., Dissanaike, G., Linowski, D., 2005, Testing the weak-form efficiency of the Russian Stock Market....
  • N.E. Al-Loughani

    Random walk in thinly traded stock markets: The case of Kuwait

    Arab Journal of Administration Science

    (1995)
  • Anand, A., C. Tanggaard, C., Weaver, D.G., 2006. Paying for Market Quality. Rutgers University. Working...
  • T.G. Andersen et al.

    Modeling and forecasting realized volatility

    Econometrica

    (2003)
  • A. Antoniou et al.

    Market efficiency, thin trading and non-linear behavior: Evidence from an emerging market

    European Financial Management

    (1997)
  • M. Asal

    Are there trends towards efficiency for the Egyptian stock market. School of Economics and Commercial Low

    (1998)
  • B. Asiri

    Testing weak-form efficiency in the Bahrain stock market

    International Journal of Emerging Markets

    (2008)
  • I. Awad et al.

    Testing the weak-form efficiency of the Palestinian securities market

    International Research Journal of Finance and Economics

    (2009)
  • Bessembinder, H., Hao, J., Lemmon, M.L., 2007. Why Designate Market Makers? Affirmative Obligations and Market Quality...
  • M.A. Civelek

    Stock market efficiency revisited: Evidence from the Amman stock exchange

    The Middle East Business and Economic Review

    (1991)
  • M. Demarchi et al.

    Equity trading systems in Europe

    Annales d' Economie et Statistiques

    (2000)
  • M.A. El-Erian et al.

    Emerging equity markets in Middle Eastern countries

    IMF Staff Paper

    (1995)
  • R. Emerson et al.

    Evolving market efficiency with application to some Bulgarian shares

    Economics of Planning

    (1997)
  • R.F. Engle

    Autoregressive conditional heteroscedasticity with estimates of the variance of UK inflation

    Econometrica

    (1982)
  • R. Engle

    GARCH 101: The use of ARCH/GARCH models in applied econometrics

    Journal of Economic Perspectives

    (2001)
  • E. Fama

    The behavior of stock market prices

    Journal of Business

    (1965)
  • Cited by (0)

    View full text