Autoras: Andreio Dionísio (Departamento de Gestão/ECS, Universidade de Évora), Wahbeeah Mothi e Isabel Vieira
Market efficiency is a major area of interest in finance. Investors are more interested in seeing trends, while researchers study developed markets more than small dimension. Emerging markets are assumed to be less efficient than developed markets, and frontier markets seem less likely to be efficient then emerging ones due to the lack of freely available financial information. In this study, our focus is on frontier markets, characterized by smaller markets with fewer traders and lower trading volumes. The criterion for these markets is quite challenging. Two of the renowned world indexes, MSCI and Standard & Poor, define frontier markets as follows. According to S & P “a market’s turnover, number of listing and whether it has attracted a minimum amount of foreign investors and also the market development prospects”. In addition to market size and liquidity, MSCI also takes political stability and economic conditions into account.
Therefore, frontier markets are poor according to GDP per capita, and are characterized by political mismanagement and economic instability. These criteria prevent such countries from acquiring the status of emerging countries. In fact, this is not true in every case and using these criteria to differentiate among frontier markets is ambiguous and misleading. According to the IMF 2015 ranking, Argentina and Nigeria, although included in frontier markets, are among the richest countries in the world with a GDP of $585.623 billion and $490.207 billion USD per capita. Bangladesh and Pakistan are rapidly growing economies. Moreover, frontier markets have done much to improve their political stability and corruption levels in the last few years in countries such as Estonia, Lithuania, Slovenia, Jordan, Mauritius, Bahrain and Croatia, and other countries are improving over time. In addition, many markets are considered frontier markets because they are small and have limited foreign ownership in their exchange listed companies.
This study investigates the weak form efficiency in frontier markets. We analyze stock market indexes of 23 countries and use mutual information and detrended fluctuation analysis (DFA) to examine serial dependence and non linear dependence in the series. The results from the former approach indicate the existence of non-linearity in the return series.
This research contributes to the literature in different ways. Frontier economies have been growing in the past few years and researchers suggest including frontier markets in the global portfolio, e.g. see Jayasuriya and Shambora (2009). For portfolio diversification, examining market efficiency is an essential prerequisite, and therefore, this study investigates confirmation of any weak form efficiency in frontier markets. On the basis of data availability, twenty-three indices from different regions are included. This study also contributes in terms of methodology by applying Detrended Fluctuation Analysis (DFA) proposed by Peng et al. (1994) to investigate the multi-scale autocorrelation The main advantage of this method is avoiding the spurious detection of long-range dependence in non-stationary data. Furthermore, this study also applied Mutual Information, which goes back to Shanon (1948), originating in the theory of communication and measuring the common information between two or more random variables. In the context of time series, mutual information investigates dependence or correlation over time. Several authors such as Darbellay and Wuertz (2000), Menezes et al. (2012) and Ferreira and Dionísio (2014) used Mutual Information in their studies and concluded that information theory measures are very effective and global.
First, we investigated the stationarity of the data by applying unit root with break points, because we cannot get robust results if structural breaks exist in the time series. The results of mutual information and global correlation indicate that non-linearity exists in the return series. After applying filtration to separate non-linear dependence, most markets’ results show that time dependence still exists in the returns.
DFA results reveal that frontier markets have long-term dependence. Sri Lanka, Bulgaria, Serbia, Estonia, Mauritius and Kazakhstan have more pronounced long-term dependence with persistent behavior. Some frontier markets including Argentina, Lebanon, Bahrain, Tunisia, Bangladesh and Vietnam show long-term dependence with anti-persistent behavior. Several frontier stock markets are less liquid and economically fragile, which might explain these results.
Our findings support some previous studies on developed, emerging and frontier markets carried out on different regions (Worthington and Higgs, 2004, Mobarek, et al., 2008, Nisar and Hanif, 2012), but contradict others (Worthington and Higgs, 2005).
We studied only frontier stock markets because most of them have been growing in the past few years. Although some are economically weak and with little potential, but still these markets need to be studied. Despite this limitation, this study is helpful and important for market participants in different ways. First, these markets have dependence in return series that might be because of a lack of information disclosure requirements and private information in the markets. Inefficient regulatory bodies and poor law enforcement standards are one of the major causes of distortion in less developed stock markets. However, it is necessary to concentrate on the legal aspect of frontier markets to protect the international as well as local investor’s interest and to prevent markets from irregularities. Secondly, there is dire need of proper and improved regulatory disclosure rules for agents who participate directly in stock markets, such as financial analysts and investors, so that they can make accurate predictions about future returns. The most important thing is that long-term dependence (seen in our results) may reduce investors’ confidence and decreases the market’s depth and breadth. This study lays the emphasis on improving stock markets’ conditions, because all frontier markets may fail to deliver the desired results without proper structural and regulatory changes. These countries need better trading mechanisms along with a central depository system and banking industry to improve stock market performance. Moreover, there is a great need of effective regulation to restrain market manipulation and give full disclosure to investors. The empirical results of this study are also helpful for policy-makers in these countries, who should draw up policies that help to improve market efficiency and attract investors. The findings will also be helpful for funds managers to give them a better understanding of how these markets function in their region. As these markets are now an attractive location for funds managers and they can get benefit from these results to draw and implement portfolio diversification strategies.
Key-words: Mutual Information, Detrended Fluctuation Analysis, Non-linear dependence, weak form efficient, frontier markets
