Cambridge Journal of Education and Science 1

Transcription

1 Cambridge Journal of Education and Science 1

2 Cambridge Journal of Education and Science No.1. (15), January-June, 2016 VOLUME III Cambridge University Press 2016

3 Cambridge Journal of Education and Science, 1(15), (January - June). Volume III. Cambridge University Press, p. Proceedings of the Journal are located in the Databases Scopus. Source Normalized Impact per Paper (SNIP): SCImago Journal Rank (SJR): Editor-in-Chief: Prof. Ella Brown, D. Hum. Litt. (UK) Executive Editor: Prof. Jesse Hyland, D. Litt. et Phil. (UK) Technical Editors: Julia Mills, Eva Collins (UK) Editors: Prof. Sheila Harbor, Psy. D. (UK) Prof. Amanda Howard, Psy. D. (UK) Prof. Preston Smith, Ed.D. (UK) Prof. David Pearson, D.F. (UK) Prof. Jimmy Lithgow, D. B. A. (UK) Prof. Jeffrey Stevenson, Psy. D. (UK) Prof. James Carpenter, D. Tech. (UK) Prof. Lauren Robinson, D. G. S. (UK) Prof. Adam Heigl, D. I. T. (UK) Prof. Julia Berger, D.Phil. (UK) Prof. Michael Hiller, D. Litt. (USA) Prof. Susan Gould, D. Litt. (UK) Prof. James Hunter, D. S. Sc. (UK) Prof. Sarah Baker, D. A. (UK) Prof. Jessica Walker, D. S. Sc. (UK) Prof. Henry Lennox, D. Hum. Litt. (USA) Prof. Jonathan Lucas, D. Litt. et Phil. (UK) Prof. Adelaide Field, D. S. Sc. (New Zealand) Prof. Christian Mills, D. G. S. (Australia) Prof. Richard Coventry, D.Sc. (Australia) Prof. Eva Galan, D.F.A. (France) Prof. Patrice Lande, D. S. Sc. (France) Prof. Sophie Holden S. J. D. (Canada) Prof. Richard Martin, D. S. Sc. (Canada) Prof. Philip Rice, D. I. T. (Canada) Prof. Denis Cumming, Ed.D. (UK) Prof. Anna Hay, Ed.D. (Canada) Prof. Abigail Lesser, Ed.D. (Canada) Prof. Joshua Savage, D. M. Sc. (UK) Prof. Michel River, Psy. D. (New Zealand) Prof. Emma Allen, Ed.D. (Australia) Prof. David Lim, D. Sc. (Australia) Prof. Lance Hagen, D. M. Sc. (UK) Prof. Charles Winger, D. E. Sc. (UK) Prof. Daniel Varney, D. E. Sc. (UK) Prof. Peter Shield, D. Tech. (USA) Prof. Jonathan Baxley, D. C. S. (UK) Prof. Glenn Richardson, D.Sc. (UK) Prof. Jane Walsh, D. C. S. (UK) Prof. Kevin Rothman, D. Env. (UK) Prof. Vincent Howard, D. I. T. (UK) ISSN: Cambridge University Press, 2016 University of Cambridge, 2016

4 Cambridge Journal of Education and Science 333 Vladimir Tsenkov, South-West University Neofit Rilski, Bulgaria, Assistant Professor Dr., Faculty of Economics, Department of Finance and Accounting, Sonya Georgieva, South-West University Neofit Rilski, Bulgaria, Ph.D. Student, Faculty of Economics, Department of Finance and Accounting Market efficiency in post-crisis period in the case of Central and Eastern Europe Abstract Efficient Market Hypothesis (EMH) defines efficient market as a market in which prices fully reflect all available information, assumption that can be argued especially in the time of crisis. This study aims to investigate the effects of the global financial crisis of 2008 on the efficiency of fourteen stock market indices, nine representing the emerging markets of Central and Eastern Europe (CEE) the Romanian BET, the Hungarian BUX, the Latvian OMXR, the Estonian OMXT, the Lithuanian OMXV, the Czech PX, the Russian RTS, the Bulgarian SOFIX, and the Polish WIG, and the other five representing developed markets - the French CAC 40, the German DAX, the British FTSE 100, the Europe's leading index for the Eurozone EURO STOXX 50, and the U.S. S&P 500. The applied EGARCH models use data for the studied indices from 2004 to 2014 divided in three periods: Pre-crisis, Crisis and Post-crisis. The empirical results reveal that there is no improvement of the market efficiency in the post-crisis period regardless whether in developed or emerging markets. Of all CEE indices only the Polish is present in the group of more efficient markets for all studied periods. The crisis worsens most significantly the market efficiency of the Russian, Latvian, Lithuanian, Estonian and Bulgarian indices. Keywords: Efficient Market Hypothesis, financial crisis, efficiency and information asymmetry, capital markets, EGARCH.

5 334 Cambridge Journal of Education and Science 1. Introduction The theme of informational efficiency of the capital markets obtained stronger relevance during the development of the global financial crisis that started in the USA in The theoretical postulates covering the dynamics of capital markets determine the prediction of future changes of the financial assets prices as useless because they accept that the fundamental changes in the market are a result of new information about them. Such information by definition is unknown, because if it is known, it will already be included in the market dynamics. These assumptions form the information efficiency of a capital market and they are at the base of the Efficient Markets Hypothesis (EMH). Empirical research market data and some practical techniques showed that there could be some violations of EMH, especially if one takes into account their determination as to the level of development of the capital markets. The global financial crisis of 2008 proved its significant impact on all capital markets, in direct contradiction of the theoretical postulates that the correlation between developed and emerging markets by default is very low. This revealed a new opportunity to check the consistency of one of the basic assumptions of capital markets - their information efficiency, reflected by EMH. The main goal of this study is to investigate the effects of the global financial crisis of 2008 on the efficiency of the studied indices. In order to conduct this research fourteen indices are used: five represent the developed markets - the French CAC 40, the German DAX, the British FTSE 100, the Europe's leading index for the Eurozone EURO STOXX 50, and the U.S. S&P 500; the other nine - the emerging markets of Central and Eastern Europe (CEE) the Romanian BET, the Hungarian BUX, the Latvian OMXR, the Estonian OMXT, the Lithuanian OMXV, the Czech PX, the Russian RTS, the Bulgarian SOFIX, and the Polish WIG. Using daily returns from 2004 to 2014, we investigate the level of information efficiency according to EMH, determined by the level of development of the capital market and the impact of the global financial crisis of For the sake of the investigation we have applied EGARCH model to the market data divided in three periods Pre-crisis, Crisis and Post-Crisis. The structure of the research is as follows: Section 2 describes the main theoretical and empirical achievements reflected in the literature on the topic of EMH, its assumptions and violations relating to developed and emerging markets. Section 3 reflects the econometric methodology used in the empirical part. Section 4 provides an

6 Cambridge Journal of Education and Science 335 empirical study and interpretation of the results. Section 5 generalizes the conclusions and findings of the empirical study. 2. Literature Review Random Walk Hypothesis states that prices or returns of the financial assets are independent of each other and change randomly, thus, this cannot be predicted from historical data. Unpredictability of the return changes is based on their rationality. Only rational change will be determined as a result of new market information. Therefore, the random walk will be an expected result for the dynamics of the index which always takes into account all available current information (Efficient Market Hypothesis - EMH). Fama defines efficient market as a market in which prices fully reflect all available information and in such a way there are no possibilities to earn profit based on past information [1]. Empirical studies on the emerging capital markets present evidence of violations of the assumptions associated with EMH. They provide evidence of their information ineffectiveness, questioning the accuracy of the EMH. Significant levels of inefficiency are observed in the capital markets of India, Singapore, Ghana and Mauritania [2,3,4]. Empirical evidence from tests on the hypothesis of a random walk for prices of capital markets in Egypt, Kenya, Morocco and Zimbabwe indicate rejection of the normal distribution assumption of the returns and the presence of positive autocorrelations [5]. Research made by Koutmos about Asian emerging markets showed differences in incorporation of market information, resulting in faster market incorporation of bad news and respectively negative return [6]. The circumstances established by Koutmos are directly related to the presence of statistically significant autocorrelation of volatility and indirect evidence of leverage effect. Using a GARCH, EGARCH and GJR models, Balla and Premarante examine the volatility dynamics of Singapore, Hong Kong, Japan, US and UK stock markets over a 10-year period [7]. Generally, it was found that asymmetry is significant and supported in all five markets. Shocks to the Singapore market tend to linger around for a longer period than it does in other stock markets. This may imply that the Singapore one shows less market efficiency than the other markets as the effects of the shocks take a longer time to dissipate. A study that covered four of the most popular indices of the Egyptian Stock Exchange showed a significant deviation from EMH [8]. Expressions of this inefficiency

7 336 Cambridge Journal of Education and Science are the observed volatility clustering and high kurtosis distribution of returns. Evidence of violation of the EMH can be found at Serbia s capital market. In the study conducted by Miljković and Radović evidence that the Serbian stock market does not show efficiency even in the weak-form of EMH is presented [9]. They found statistically significant levels of autocorrelation in returns with high kurtosis distribution, significantly different from the normal one. Simeonov study dynamics of Bulgarian and Serbian capital markets before and after the 2008 crisis and conclude that the exchange indices reflect long-term expectations, consistent partially with current mistrust of fundamental macroeconomic factors [10]. This type of market beheviour can be determined as an evidence of the presents of inefficiency on the studied markets. The empirically established characteristics of the emerging markets, like type of distribution, statistically significant autocorrelation and the presence of non-linear dependencies are also valid for some developed capital markets - the Swedish index OMX, which includes 30 large companies in Sweden [11]. Maria Borges conducted empirical tests on the weak-form of market efficiency applied to stock market indices of UK, France, Germany, Spain, Greece and Portugal, from 1993 to 2007 [12]. The results show mixed evidence on the EMH. The hypothesis is rejected on daily data for Portugal and Greece, but these two countries have been approaching martingale behavior after France and UK data reject EMH, the tests for Germany and Spain do not allow the rejection of EMH, this last market being the most efficient. A testing for random walks and weak-form market efficiency in European equity markets is done by Worthington and Higgs [13]. Daily returns for sixteen developed markets (Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland and the United Kingdom) and four emerging markets (Czech Republic, Hungary, Poland and Russia) are examined for random walks. The results indicate that only Hungary of the emerging markets is characterized by a random walk and hence is weak-form efficient, while in the developed markets only Germany, Ireland, Portugal, Sweden and the United Kingdom comply with the most stringent random walk criteria. Ali and Afzals empirically revealed, by using EGARCH model, that negative shocks have more pronounced impact on the volatility than positive shocks on Pakistani and Indian stock exchanges [14]. These stock markets also demonstrate the existence of persistent volatility clustering. Sengonul and Degirmen investigated the impact of recent global financial crisis on the weak-form of efficiency of markets of the

8 Cambridge Journal of Education and Science 337 countries from 2004 enlargement of the European Union, Bulgaria and Romania on the one hand, and Turkey on the other hand [15]. The results indicate that Bulgaria, Romania, Estonia, Lithuania, Malta and Slovenia demonstrate a weak-form of market inefficiency during both the pre-crisis and the post-crises periods. The Czech Republic, Cyprus and Latvia clearly departed from weak-form of efficiency after the crisis. Among the studied countries, Hungary, Slovakia and Turkey performed better. Among three of them, Hungary appeared the most efficient while Slovakia and Turkey follow her with slight departing from efficiency. Abdmoulah uses GARCH-M (1,1) approach to test weak-form of efficiency for 11 Arab stock markets for periods ending in March 2009 [16]. All markets show high sensitivity to the past shocks and are found to be weakform inefficient as a negative reaction to contemporaneous crises. Mishra studies the impact of recent global financial crisis on the weak form of EMH in the context of selected emerging and developed capital markets - BOVESPA of Brazil, SENSEX of India, Shanghai Composite Index of China, KOSPI Composite Index of South Korea, RTS of Russia, NASDAQ Composite of US, DAX of Germany and FTSE 100 of UK for the period 2007 to 2010 [17]. The used ADF unit root test and GARCH model estimation provide the evidence that the selected capital markets are not weak-form efficient. However, such informational inefficiency of capital markets often provides the impetus for successful financial innovation by financial firms, thereby, making the market move towards efficiency in the long run. 3. Methodology The study aims at tracing the influence of the informational links between developed and emerging capital markets in the context of the Efficient Market Hypothesis, taking into account the importance of specific factors - the global financial crisis of The data used in this study include daily values of the analyzed indices, for the period (Table 1) and calculated on their return that is based on the formula r t I log t I, where It is the value of an index for the t day. The t 1 period under investigation has been divided into three sub-periods - pre-crisis, crisis and post-crisis. The separation of the sub-periods is made on the basis of achieved the highest and lowest value of the researched indices in the testing period. Econometric modeling is done on the basis of the application of EGARCH (p, q) models to the data of the return of the studied indices for each of the three sub-periods under

9 338 Cambridge Journal of Education and Science consideration. The analyses are based on final selected EGARCH models while testing the following basic combinations and lags - EGARCH (1,1), EGARCH (2.1), EGARCH (1,2), EGARCH (2.2). For the choice of this combination which most successfully models data for the corresponding period and index the tests - Likelihood ratio test (lratiotest) and Akaike's information criterion (AIC) are applied. The used EGARCH (p, q) model represents the conditional volatility of residuals with the respect of leverage effect: P Q Q 2 2 t j t j t j log ht 0 ilog ht i j E j i 1 j 1 ht j h t j j 1 h t j, where: v 1 t j v 2 2 E zt j E h v, with degree of freedom v 2. t j 2 The used model of returns is as follows - YY tt = CC + ii=1 kk ii YY tt ii + εε tt. The autoregression dependencies in the return of the studied indices - YY tt are presented by the coefficient AR (1) - kk ii. The AR(1) measure the impact of the return from the previous day, i.e. measure the impact of the previous day information to the current day return. The information efficiency as an indicator of performance will be determined by the magnitude of the coefficient of persistence, representing the impact of tendencies from prior periods on the volatility in the present period. In the asymmetric EGARCH (p, q) model coefficient of persistence is represented by the coefficient. High values of this coefficient would indicate a low informational efficiency, reflected in slower incorporation of the market information because of a higher influence of the market volatility tendencies, while it is the other way round at low values of the coefficient of persistence. The definition of information efficiency is made according to the efficient markets hypothesis (EMH). The measurement of information asymmetry as an indicator, i.e., measurement of asymmetric adjustment of the EGARCH (p, q) model, is implemented and measured by the coefficient. The analysis of these values can specifically measure the impact on the volatility of different informational impact and make conclusions about asymmetric volatility adjustments to them. pp

10 Cambridge Journal of Education and Science 339 Table 1: Studied indices, periods under consideration and number Index/Period The French CAC 40 The German DAX The British FTSE 100 Euro STOXX 50 The U.S. S&P 500 The Romanian ВЕТ The Hungarian BUX The Latvian OMXR The Estonian OMXT The Lithuanian OMXV The Czesh РХ The Bulgarian SOFIX The Poland WIG The Russian RTS of observations Pre-crisis Crisis Post-crisis Period from/to Observations Period from/to Observations Period from/to Observ ations / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / / Empirical data 4.1. Analysis of autoregression coefficients - AR (1) As Magnus and Abdmoulah point out, if the autoregression coefficient AR(1) in the model of return is statistically significant and different from zero then we reject the weak-form of Efficient Market Hypothesis [18,16]. Table 2: Significant autoregression coefficients (AR(1)) Pre-crisis Crisis Post-crisis Index/Period AR(1) AR(1) AR(1) The French CAC 40-0, ,08539 The U.S. S&P 500-0, ,41003

11 340 Cambridge Journal of Education and Science The Romanian BET 0, , ,05714 The Latvian OMXR -0,15572 The Estonian OMXT 0, ,19555 The Lithuanian OMXV 0,16721 The Bulgarian SOFIX 0, ,05755 The Poland WIG 0, ,06008 The Russian RTS 0, ,06701 Statistically significant auto regression coefficients (AR(1)) are registered as for the indices of developed capital markets - France and the United States, and also for the emerging - Romania, Lithuania, Latvia, Estonia, Bulgaria, Poland and Russia. This statistical significance indicates an inefficiency of the markets that allows us to reject the weak form of the EMH. It is noteworthy that this inefficiency is registered more for the emerging rather than for developed capital markets - the ratio is 9 to 2. The highest absolute value of the AR(1) in the pre-crisis period is registered for the emerging markets - Russia ( ), and in post-crisis period for the developed markets - USA ( ). Out of the three studied periods, it is namely in the crisis period where we observe least statistically significant AR(1) and therefore it can be defined as the most effective in terms of the weak form of the EMH. There is a decrease in the efficiency for France, where the coefficient shows, however slightly, but decreased efficiency (AR(1) from to ), much more significant deterioration is observed in the effectiveness for Romania (AR(1) from to ) and only Estonia improves its efficiency (AR(1) from to ). The least effective is the pre-crisis period, eight of the studied indices show statistically significant AR(1). Values of the coefficients are in the range of (France) to (Russia), as for the both developed markets (France and USA) they are negative and for all the emerging markets positive. This indicates the presence of sustainable market trends for the emerging capital markets for which the positive AR(1) gives a greater weight of the return of the previous period and therefore strengthens the established market trend. Assuming that in this pre-crisis period returns are mostly positive, this leads to formation and acceleration of the positive market trend. This can also be seen as an indication of the presence of irrational following and subsequent acceleration of the positive market trend due to irrational behavior of the investors arising from their assessment of the prospects for the development of these markets. Generally, this behavior could be seen as creating a speculative bubble in these markets. To confirm it, we could point out that AR(1)

12 Cambridge Journal of Education and Science 341 coefficients are in the range of (Poland) to (Russia), three of which are above 0.20 (Bulgaria, Estonia and Russia) and two over 0.10 (Romania and Lithuania). In contrast of that behavior we should consider the negative values of AR(1) for the developed markets as a counter reaction to the established positive market trend from the previous period. Assuming that these markets give a greater weight to the current information influences from the market rather to pursue a long-established trends expressed by the return from the previous periods, we could explain the existence of this negative statistical significant AR(1)s. It must be pointed out that this counter reaction to the positive market trend, being a violation of the weak form of market efficiency according to EMH, is relatively much smaller in size compared with the emerging markets. The values of the AR(1) are in the range of (France) to (USA). In the post-crisis period there are improvements in the market efficiency with statistically significant AR(1) observed only for six indices from which only one represents the developed markets - the US S&P500. For four of them (Romania, Bulgaria, Poland and Russia) a decrease of AR(1) is observed and therefore improvement in their efficiency compared to the pre-crisis period. This applies most strongly for Russia, where we notice a reduction in the coefficient from (before the crisis) to the post-crisis level of The presence of market inefficiency is registered for Latvia, as a counterpoint to the recorded efficiency in the pre-crisis and crisis period. The significant deterioration of efficiency shows the US market as from the pre-crisis level of AR(1) equal in the post-crisis period the value goes to This is a period of pronounced positive returns for the S&P500 in response to the post-crisis recovery of this market. The high negative AR(1) shows an increased tendency to follow positive market impulses at the given moment rather than following a sustainable trend of return from the previous period Data analysis of coefficients of persistence Data analysis of coefficients of persistence of the tested indices is made on the basis of the assumptions of EMH, namely: small values of the coefficients indicate higher degree of informational efficiency as long as they mark slighter presence of clusters of volatility and, hence, the presence of the autocorrelation dependencies for

13 342 Cambridge Journal of Education and Science the return of an index. Thus, lower values of the coefficients indicate to a greater degree the confirmation of the weak form of the EMH. In a direct comparison of empirical data based on the criteria coefficient of persistence we could distinguish the studied indices of two conditional groups: - Indices with relatively high market efficiency (ratio below 1.1); - Indices with relatively low market efficiency (ratio above 1.1) Pre-crisis period The data for the pre-crisis period point out that with a relatively more information efficiency of the developed equity markets are the French, German, British, US indices and index EURO STOXX 50. On the emerging markets of CEE, only two indices show increased information efficiency according to EMH, namely: the Bulgarian and the Polish index. Comparing the values of the coefficient of persistence we can register that the most efficient and with the lowest coefficient value index is the EURO STOXX 50 (1,0052), while at the other extreme is the Bulgarian SOFIX (1,0658). It can be noticed that despite the fact that all indices in this group are relatively information efficient, there is a clear distinction between developed and emerging capital markets. The Bulgarian and the Polish indices have the highest coefficients of persistence and therefore the lowest efficiency in this group. If one adds to this that the leverage factor of the two indices are positive and are relatively very low compared to the developed markets, one could conclude that although they are relatively more efficient than other capital markets in CEE in terms of pre-crisis period. What is more important for the dynamics of the Bulgarian and Polish indices is the short-term market impulses than following sustainable market trends in comparison with the developed capital markets. Low and positive leverage coefficient shows that the predominant positive market returns of the pre-crisis period lead to very low variability correction in the index upwards. In contrast, we could mention the significantly higher values of leverage coefficients in developed markets whose negative sign and weight lead to relatively higher decrease of volatility in terms of positive market returns typical of the pre-crisis period. The group of relatively less efficient markets in terms of the pre-crisis period includes the following indices: Romanian, Hungarian, Latvian, Estonian, Lithuanian, Czech and Russian. Their high coefficients of persistence mark low level of information efficiency, inconsistent with the weak form of the EMH.

14 Cambridge Journal of Education and Science 343 These indices have a tendency to strongly follow short-term market fluctuations to follow long-term trends. This basically applies to the Estonian and lowest for the Romanian index. If we consider the last and relatively very low leverage ratio and its positive sign, we can draw the conclusion that although informational inefficient, i.e., following most closely the immediate market impulses that are positive at this period, they do not cause significant correction in its variability. At the other extreme are the indices with significantly higher (in absolute value) and negative leverage ratio. Combined with high rates of persistence, typical of this group, they point out markets which largely follow cyclical and immediate positive market impulses and contract relatively significant volatility of their returns. This situation could be considered not only as an indication of market inefficiency, according to the EMH, but as an indication of the presence of risk markets, since only the change of the positive returns with negative returns would gain these negative market impulses and significant correction in volatility towards its increase. The most significant one in this regard are the Russian, Latvian and Estonian indices. Table 3: Comparison of empirical data according to the criteria coefficient of persistence in the pre-crisis period Index/Coefficient Coefficient of persistence 1.1 Leverage coefficient The French САС The German DAX The British FTSE Euro STOXX The U. S. S&P The Bulgarian SOFIX The Poland WIG Index/Coefficient Coefficient of persistence >1.1 Leverage coefficient The Romanian ВЕТ The Hungarian BUX The Latvian OMXR The Estonian OMXT The Lithuanian OMXV The Czech РХ The Russian RTS

15 344 Cambridge Journal of Education and Science Crisis period The crisis period is characterized by poor market efficiency expressed in increased coefficients of persistence for all examined indices. In the group of those with relatively weak correction of market efficiency can be pointed out the German, the British index and the EURO STOXX 50 index. What is observed is that the Polish and Romanian indices fall in this group of more efficient indices in the crisis. Compared to pre-crisis period they indicate significant improvement in the market efficiency and the lowest rates of persistence from all investigated indices. What should be added is the fact that these indices demonstrate the most significant correction in their efficiency of the pre-crisis to the crisis period. This increase in information efficiency of these markets could be explained by the presence of enhanced incorporation of negative market impulses typical of the period and the formation of a sustainable and long-term market trend. The cyclical market impulses have less importance for it since their informational influence is already largely included in the long-term market trend. The French and the US indices, representing developed equity markets, are observed in the category of expressed more inefficient market. However, we must do the necessary specification that even with poor efficiency (coefficient of persistence and ), it is a significant degree lower than shown by the emerging CEE indices with coefficients of persistence ranging from to With the poor market efficiency during the crisis we could mention Russian, Latvian, Estonian, Lithuanian and Bulgarian indices. During the crisis a significant increase in the coefficients of persistence and of leverage is observed. This marks the markets that are in the focus of this analysis as highly dependent on immediate and short-term market dynamics, leading to more negative cyclical market impulses accelerating market slowdown and a significant increase in market volatility. Table 4: Comparison of empirical data according to the criteria coefficient of persistence in the crisis period Index/Coefficient Coefficient of persistence 1.1 Growth rate Leverage coefficient The German DAX The British FTSE Euro STOXX The Poland WIG The Romanian ВЕТ

16 Cambridge Journal of Education and Science 345 Index/Coefficient Coefficient of persistence >1.1 Growth rate The French САС The U.S. S&P The Hungarian BUX The Latvian OMXR The Estonian OMXT The Lithuanian OMXV The Czech РХ The Bulgarian SOFIX The Russian RTS Leverage coefficient Post-crisis period The post-crisis recovery of the markets of this survey register an improvement in market efficiency, as the number of relatively efficient markets is greater than those characterized by weaker efficiency. In the latter group we can only indicate the presence of a representative of the developed capital markets - the US and five emerging CEE - Hungarian, Latvian, Czech, Bulgarian and Russian indices. The presence of the US index in the group of markets with weaker efficiency could be analyzed taking into account the high rates of persistence and leverage. In a post-crisis recovery, characterized by positive market impulses, the US market index shows behavior leading to the strengthening of a positive trend, resulting in the most significant limitation of the volatility of all considered indices. However, this could not be said about other less efficient indices. High rates of persistence, combined with relatively low leverage ratios indicate a market heavily dependent on the immediate market situation and not forming long-term market trends with volatility that even in the post-crisis recovery and predominantly positive market impulses does not indicate conditions for its limitation. In a direct comparison between pre- and post-crisis period we can infer that there is no improvement of market efficiency of the indices in question, whether the group of more efficient or those with poor efficiency. If during the pre-crisis period the coefficient of persistence of more efficient indices ranges between to , then during the post-crisis period, it is between to A comparison of the same group of less efficient indices registers an interval of to before and until after the crisis period. On the one hand, it shows that a big number of the markets would have difficulty to meet the requirements of even the weak form of EMH, and on the other hand, the fact that their post-crisis

17 346 Cambridge Journal of Education and Science recovery did not reach the level of market efficiency characteristic of the pre-crisis period. Table 5: Comparison of empirical data according to the criteria Index/Coefficient coefficient of persistence in the post-crisis period Coefficient of persistence 1.1 Growth rate The French САС The German DAX The British FTSE Euro STOXX The Romanian ВЕТ The Estonian OMXT The Lithuanian OMXV The Poland WIG Index/Coefficient Coefficient of persistence >1.1 Growth rate The U.S. S&P The Hungarian BUX The Latvian OMXR The Czech РХI The Bulgarian SOFIX The Russian RTS Leverage coefficient Leverage coefficient 5. Conclusion From the survey the following conclusions can be drawn: Statistically significant AR(1) indicates the presence of inefficiency more for the emerging rather than for developed capital markets - the ratio is 9 to 2. The highest inefficiency in the pre-crisis period is registered for Russia (emerging market), while in the post-crisis for the US (developed market). The least statistically significant AR(1) are observed for the crisis period and therefore it can be defined as the most effective in terms of the weak form of the EMH. The least effective is the pre-crisis period, eight of the studied indices shows statistically significant AR(1) with values in the range of (France) to (Russia), as for the both developed markets (France and USA) they are negative and for all the emerging markets positive. In the post-crisis period there are improvements in the market efficiency with statistically significant AR(1) observed only for six indices from which only one represents the developed markets - the US S&P500. For four of them (Romania,

18 Cambridge Journal of Education and Science 347 Bulgaria, Poland and Russia) there is improvement in the level of their efficiency compared to pre-crisis period. Of all surveyed CEE indices only the Polish index is present for all periods in a group of more efficient capital markets. The Hungarian, Czech, Latvian and Russian indices shows constantly less market efficiency whether for the pre-crisis, crisis or postcrisis period. The crisis worsens most significantly the market efficiency of the Russian, Latvian, Lithuanian, Estonian and Bulgarian indices. In its post-crisis recovery Lithuanian and Estonian indices move to the group of more efficient markets. References: 1. Fama, E. F. (1970). Efficient capital markets: A review of theory and empirical work, The Journal of Finance 25(2), Bekaert, G. and Campbell, H.R. (2002). Research in Emerging Market Finance: Looking to the future. Emerging Markets Review, 3, 4: Bundoo, S.K. (2000). The Mauritius Stock Exchange: An Assessment. Social Sciences & Humanities and Law & Management Research Journal. 3. University of Mauritius. 4. Smith, G. and Jefferis, K. (2002). The Evolving Efficiency of African Stock Markets. Unpublished Research. 5. Mlambo, C., Biekpe, N. and Smit E. vd. M. (2003). Testing the Random Walk Hypothesis on Thinly-Traded Markets. The Case of Four African Stock Markets. The African Journal of Finance. 5. 1: Koutmos, G. and R. Saidi (1995). The leverage effect in individual stocks and the debt to equity ratio, Journal of Business Finance and Accounting, 7, 1995, p Bala, L., Premaratne, G. (2004). Stock Market Volatility: Examining North America, Europe and Asia, Econometric Society 2004 Far Eastern Meetings, Econpapers No Mecagni, M. and Sourial, M.S. (1999). The Egyptian Stock Market: Efficiency Tests and Volatility Effects. International Monetary Fund, Working Paper, No 99/48.

19 348 Cambridge Journal of Education and Science 9. Miljković, V., Radović, O. (2006). Stylized facts of asset returns: case of BELEX, Facta Universitatis, Series: Economics and Organization Vol. 3, No 2, 2006, p Simeonov, S. (2015). Stock Exchange and Economic Activity Indicators Relations and Asymmetry during the Recession in Serbia and Bulgaria, Financial Markets and the Real Economy: Some Reflections on the Recent Financial Crisis, ISBN: , p.: Nässtrom, J. (2003). Volatility Modelling of Asset Prices using GARCH Models, Linkӧping University, Sweden, Reg nr: LiTH-ISY-EX , 13 February Borges, MR. (2010). Efficient Market Hypothesis in European stock markets, The European Journal of Finance, Vol. 16, Issue Worthington, A., Higgs, H. (2004). Random walks and market efficiency in European equity markets, Global Journal of Finance and Economics 1(1): p Ali, R., Afzal, M. (2012). Impact of global financial crisis on stock markets: Evidence from Pakistan and India, Journal of Business Management and Economics. Vol. 3(7), p , June, Sengonul, A., Degirmen, S. (2010). Does the Recent Global Financial Crisis Affect Efficiency of Capital Markets of EU Countries and Turkey? 13th international conference on finance and banking, Ostrava, Czech Republic, October 2011, ISBN: Abdmoulah, W. (2009). Testing the Evolving Efficiency of 11 Arab Stock Markets. Arab Planning Institute Working Paper, no /jodep/ products/delivery/wps0907.pdf. 17. Mishra, P.K. (2011). Weak form market efficiency: evidence from emerging and developed world, The Journal of Commerce, Vol. 3, No. 2 (2011), ISSN: , , pp.: Magnus, F.J. (2008). Capital Market Efficiency: An Analysis of Weak-form Efficiency on the Ghana Stock Exchange, Journal of Money, Investment and Banking, Issue 5, p