رفتار توان هرست و ارزیابی بهره وری سهام بازارهای خاورمیانه و شمال آفریقا
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|4582||2012||18 صفحه PDF||سفارش دهید||7334 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Research in International Business and Finance, Volume 26, Issue 3, August 2012, Pages 353–370
In this paper, we test the evolving efficiency of MENA stock markets. Our empirical approach is founded on the behavior of the Hurst exponent over time. We computed the Hurst exponent using a rolling sample with a time window of 4 years. The empirical investigation has been conducted on the major Middle East and North African stock markets. The sample data covers in daily frequency the period (January 1997 to December 2007). Our empirical results show that all MENA stock returns exhibit long-range memory and certain markets are becoming more efficient. Ranking MENA stock markets by efficiency with our measures of long-range dependence have shown that Israel's, Turkey's and Egypt's markets are the less inefficient markets in this region. Furthermore, we have founded evidence of statistically significant rank correlation between the measure of long-range dependence and average trading costs, market capitalization and anti-self-dealing index, which suggests that these variables play a role in explaining these differences in the stage of inefficiency.
In financial theory, the efficient market hypothesis (EMH) is highly controversial. This hypothesis is based on the idea that prices assets already reflect all known information. In its weak-form, the EMH stipulates that future prices cannot be predicted by analyzing the historical asset price. More explicitly, the WFEMH asserts that all past market prices are fully reflected in securities prices. Then, it is impossible to beat the market by using investment strategies based on historical share prices. More explicitly, the existence of autocorrelation between observations violates the WFEMH. However, if the short autocorrelation is accepted by defenders of the EMH, the long autocorrelation is widely rejected. Thus, there is a large body of the literature focused on the relationship between informational efficiency and long-range dependence. Empirical investigations have been conducted on both developed and emerging stock markets. They provided mixed evidence. Briefly speaking, we can range studies within two empirical approaches. The earlier studies including, Hiemstra and Jones (1997), Liu et al. (1999), Barkoulas et al. (2000), Grau-Garles (2001), and Henry (2002) are founded on a static efficiency for all the sample time series. As a result, a rejection of the WFEMH may reveal sub-periods in which the market is efficient. Then, the second approach is based on the study of changed level market efficiency over time. In fact, there are several factors that may cause several deviations in prices, such as unexpected events, limits to arbitrage, market imperfections and financial reforms. Even more, it is important to estimate an evolution efficiency approval of all changes in rules, regulations and technology markets. For this, many studies assessing stock market efficiency are introduced and important regularities between financial and physical data were down (Stanley et al., 1999 and Plerou et al., 2000). Therefore, concepts and methods of statistical physics are increasingly being applied in economics. The purpose of this paper is, therefore, to contribute to this literature by examining the behavior of efficiency of eleven MENA stock markets. We expect the assertion that the emerging stock markets are becoming less inefficient in these recent years due to the recent financial and institutional development. In this paper, we use the Hurst exponent behavior to study the long-range dependence of volatility's MENA stock markets over time. To our knowledge, this paper represents one of the first studies on Hurst exponent behavior and stock market informational efficiency in the context of MENA countries. All the previous studies were focused on developed and some emerging countries which did not include MENA equity markets. The rest of the paper is structured as follows. In Section 2, we expose a brief review of previous studies. Our attention is focused on the literature that studies long-memory in MENA stock markets. In Section 3, we present the MENA stock markets tendency. Data description and methods are reported in Section 4. Empirical results are displayed in Section 5. The last section gives the summary and provides some concluding remarks.
نتیجه گیری انگلیسی
In this paper, we employed a “rolling sample” approach in order to analyze the behavior of the Hurst exponent over time for emerging markets absolute returns (proxy for volatility). In contrast with some previous empirical studies based on a single static measure of long memory dependence, we employed a rolling sample approach to test whether the MENA stock markets are being more efficient. Our empirical investigation has been conducted on sample covering in daily frequency eleven MENA stock markets and the Hurst's exponent behavior over time is analyzed by a “rolling sample” method with a window of four years observations. The obtained results reveal that the Hurst exponent changes over time and MENA stock market volatilities exhibit a strong evidence of long-range dependence. But, we have shown that some MENA equity markets are being more efficient while some others are being more inefficient at the end of the time sample (1997–2007). In order to check the robustness of these results, we have filtered the stock returns with a GARCH-type process. Our main purpose is to take into account the impact of the short-term memory on the Hurst exponent estimation. The two exponents have shown a similar ranking by median value and so Israel's, Turkey's and Egypt's stock markets are the less inefficient markets in this region while Iran's market is the most inefficient. Moreover, our findings show that these physical concepts are positively correlated with average trading costs and negatively correlated with market capitalization and anti-self-dealing index. Therefore, these variables play an important role in understanding stock market inefficiency.