© 2012 Nicolaus Copernicus University Press. All rights reserved. http://www.dem.umk.pl/dem DYNAMIC ECONOMETRIC MODELS Vol. 12 (2012) 19−32 Submitted January 25, 2012 ISSN Accepted October 8, 2012 1234-3862 Dorota Witkowska, Krzysztof Kompa, Aleksandra Matuszewska-Janica* Analysis of Linkages between Central and Eastern European Capital Markets† A b s t r a c t. The aim of the research is analysis of short- and long-term international relations between stock exchanges in Central and Eastern Europe. The analysis is provided in 3 stages. In the first step the order of the variables integration is examined. In the second stage short-run relationships for pairs of indexes are analyzed using Granger causality test. In the last step long- run relationships for pairs of indexes are examined applying Johansen cointegration method. K e y w o r d s: Emerging Markets, Equity CEE Markets, cointegration, Granger Causality, long- run relationships, short-run relationships. J E L Classification: G15. Introduction The analysis of common stock market movements is important for effective portfolio diversification and a possible starting point to examine the functioning of the global financial system. Therefore international market linkages has at- tracted investors and policy-makers’ attention. Consequently, international equi- ty market integration is a topic often discussed in literature, especially many researchers have investigated the short-term and long-term interrelationships among worldwide financial markets. The theory review, evidence and implica- tions of international equity market integration are presented in (Kearney, Lucey, 2004; Bailey, Choi, 2005) among others. Various aspects of equity mar- kets relationships have been considered, including: * Correspondence to: Dorota Witkowska, Department of Econometrics and Statistics, Warsaw University of Life Sciences, ul. Nowoursynowska 166, 02-787 Warszawa, Poland, e-mail: doro- ta_witkowska@sggw.pl † Scientific research with the financial support of the Polish Ministry of Science and Higher Education No N N111 43 1837. Dorota Witkowska, Krzysztof Kompa, Aleksandra Matuszewska-Janica DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–32 20 − volatility spillovers across markets (e.g. Engle, Susmel, 1993; Kearney, 2000; Koutmos, Booth, 1995; Ng, 2000); − market correlation structures (e.g. Koedijk et al., 2002; Longin, Solnik, 1995) and − financial crises contagion (e.g. Claessens, Forbes, 2001; Rigobon, 1999). Empirical investigations discussed in literature can be classified into 3 major classes due to following criteria: − regions and periods of provided analysis, − length of the return intervals, − methods of analysis. Empirical analysis considering relations among mature markets has been provided since the end of the 20-th century (Eun, Shim 1989; Hamao et al., 1990; Kasa 1992; Engle, Susmel 1993; Lin et al., 1994; Longin, Solnik 1995, 2001; Koutmos, Booth, 1995; Kim, Rogers, 1995; Karolyi, Stulz, 1996; Choudhry, 1996; Koutmos, 1996; Serletis, Booth et al., 1997; King, 1997; Rigobon, 1999; Witkowska, 1999; Kearney, 2000; Ng, 2000; Claessens, Forbes, 2001; MacDonald, 2001; Shachmurove, Witkowska, 2001; Forbes, Rigobon, 2002; Koedijk et al., 2002; Serwa, Bohl, 2005; Sharkasi et al., 2004; Kearney, Lucey, 2004; Baur, 2004; Phylaktis, Ravazzolo, 2005). While investigation of mutual market linkages for emerging markets has shorter history, especially consideration for post-communist countries. Syriopoulos (2007) notices that despite the growing importance of the emerging Central and Eastern European stock markets (see Fig. 1), the relevant body of research remains surprisingly limited. Furthermore, the empirical findings on this topic appear rather ambigu- ous and contradictory. For emerging markets we should mention research pro- vided for: − ASEAN (Janakiraman, Lamba, 1998; Gosh et al., 1999; Masih, Masih, 2001; Siklos, Ng, 2001), − Middle and South Americas (Phylaktis, Ravazollo, 2005; Diamandis, 2009) and − Central and Eastern Europe (Voronkova, 2004; Gilmore et al., 2008; Syl- lignakis, Kouretas, 2010). Taking into account length of the investigated samples we notice that 10- year or longer periods are very often considered, for instance (Caporale, Spagnolo, 2010; Gilmore et al., 2008; Sharkasi et al., 2004). However shorter periods are also used in comparable analysis as Dubinskas, Stunguriene (2010) who consider 2-years period or Gilmore et al. (2008) who use rolling windows approach. The length of the returns interval is also crucial and influences the results of investigation. In fact different intervals are used, for instance daily and weekly returns are discussed by Caporale, Spagnolo (2010), monthly – in Baur (2004), and 5-minutes intraday data – in Hanousek et al. (2008) and Hanousek, Kočen- da (2009). Analysis of Linkages between Central and Eastern European Capital Markets DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–33 21 Figure 1. CEE Equity Markets Capitalization as a percentage of FESE Equity Markets Capitalization Investigations has been provided applying different methods of analysis, the most popular are: − correlation measures (as in: Panton et al., 1976; Watson, 1980; Meric, Meric 1989; Bailey, Stulz, 1989; Fisher, Palasvirta, 1990; Longin, Solnik, 1995), − causality analysis (for instance Kwan et al., 1995; Roca, 1999; Huang et al., 2000; Narayan et al., 2004; Matuszewska-Janica, 2010), − VAR models and cointegration analysis (Eun, Shim, 1989; Kasa, 1992; Richards, 1995; Hassan, Naka, 1996; Choundry, 1997, Gosh et al., 1999; Witkowska, 1999; Shachmurove, Witkowska, 2001; Masih, Masih, 2001; Siklos, Ng, 2001; Chen et al., 2002; Pascual, 2003; Yang et al., 2004; Gil- more et al., 2008; Kuçukcolak, 2008; Matuszewska-Janica, 2011), − GARCH models (Baele, Vennet, 2001; Voronkova, 2004; Li, Majerowska, 2008), − taxonomic methods as Kompa (2010). The aim of the research is identification of short- and long-term internation- al relations between stock exchanges in Central and Eastern Europe. The analy- sis is provided in 3 stages in which: 1. the order of the variables integration, 2. short-run relationships for pairs of indexes, using Granger causality test and 3. long-run relationships for pairs of indexes, applying Johansen cointegration method are investigated. 1. Data Description The research is provided for quotations of 14 indexes from the capital mar- kets in Central and Eastern Europe (CEE) – Table 1, from the period: January 2000 – November 2010. In our research we consider daily, weekly and monthly (for the last quotation in the week and month respectively) data. The observa- tions are transformed into natural logarithms and logarithmic rates of return. Dorota Witkowska, Krzysztof Kompa, Aleksandra Matuszewska-Janica DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–32 22 Table 1. Analyzed indexes LP Index Type of index Stock Exchange (SE) 1 ATX price, blue-chip index Vienna SE 2 PX price, blue-chip index Prague SE 3 PXGLOB price, broad index Prague SE 4 BUX performance, blue-chip index Budapest SE 5 SBI20 price, broad index Ljubljana SE 6 SAX total return, blue chip index Bratislava SE 7 BET price, blue-chip index Bucharest SE 8 SOFIX total return, broad index Bulgarian SE 9 OMXBB performance, 32 companies from Baltic market – benchmark OMX Group, Baltic countries 10 OMXT total return, all share index Tallin SE, OMX Group 11 OMXR total return, all share index Riga SE, OMX Group 12 OMXV total return, all share index Vilnius SE, OMX Group 13 WIG performance, all share index Warsaw SE (WSE) 14 WIG20 price, blue-chip index Warsaw SE (WSE) It is worth mentioning that for 2 indexes: SOFIX and SBI20 the data are available only from January 2001 till October 2010, therefore analysis is pro- vided for 2 samples as it is shown in Table 2 where time ranges, symbols of samples and numbers of observations are presented. Missing data are completed by repeating the last observation (i.e. foregoing the lacking one). Table 2. The considered periods, number of observations and notation of samples Indexes date of first observation date of last observation Frequency of data daily weekly monthly A B A B A B OMXBB, OMXT, OMXR, OMXV, ATX, SAX, BUX, PX, PXGLOB, BET, WIG, WIG20 2000-01-03 2010-11-05 PD1 2790 556 PW1 130 PM1 SOFIX, SBI20* 2000-12-29 2010-10-14 PD2 2515 501 PW2 118 PM2 Note: A – symbol of samples, B – number of observations, * – quotation of SBI20 was stopped in October 2010. 2. Results In the first step the order of the variables integration is identified applying augmented Dickey-Fuller (ADF) test1. The results indicate, that all examined time series of indexes are nonstationary while all returns are stationary so in- dexes are I(1). 1 For technical details see e.g. Maddala, Kim (1998), Elder, Kennedy (2001). Analysis of Linkages between Central and Eastern European Capital Markets DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–33 23 In the second step, short-run relationships between all indexes are examined employing Granger causality test2. For two indexes (X and Y) we denote the direction of Granger causality by the arrow (i.e. X→Y means that X causes changes in Y, and Y→X the opposite). Causality analysis is provided for 182 mutual relations (for 14 indexes), considering from 1 to 10 lags for each inves- tigated relation. Hypotheses are verified at the significance level 0.05. Table 3 contains the results presented as percentage of cases when the null hypothesis is rejected. More detailed results are presented in Tables A1–A3 in the Appendix. For daily data the greatest percentage of rejections is obtained for following relations: WIG20→Y (93%), BUX→Y (92%), WIG→Y (87%), X→SOFIX (93%) and X→SBI20 (82%). On another words, daily changes of WIG20, WIG and BUX are the most often causes of changes in other investigated indexes, while SOFIX and SBI20 are the most sensitive indexes. The smallest number of H0 rejections is observed for following relations: SAX→Y (22%), OMXR→Y (23%), X→WIG20 (18%) and X→WIG (35%). Thus the changes of SAX and OMXR influence other indexes very rarely while the less sensitive to changes of other indexes are WIG and WIG20. It can be explained by the fact that War- saw Stock Exchange is the biggest market in CEE region and it reacts due to the world biggest markets changes. Table 3. Results of the Granger causality test – percentage of rejection H0 Relation % rejections of the H0 Relation % rejections of the H0 PD1/PD2 PW1/PW2 PM1/PM2 PD1/PD2 PW1/PW2 PM1/PM2 ATX→Y 78% 55% 59% X→ATX 66% 45% 2% BET→Y 48% 71% 23% X→BET 72% 55% 4% BUX→Y 92% 44% 33% X→BUX 57% 59% 9% OMXBB→Y 75% 36% 17% X→OMXBB 75% 74% 48% OMXR→Y 23% 41% 2% X→OMXR 72% 42% 42% OMXT→Y 71% 25% 15% X→OMXT 74% 85% 64% OMXV→Y 64% 43% 20% X→OMXV 65% 45% 37% PX→Y 86% 82% 43% X→PX 65% 25% 6% PXGLOB→Y 85% 83% 45% X→PXGLOB 66% 27% 6% SAX→Y 22% 12% 0% X→SAX 55% 56% 5% WIG→Y 87% 64% 25% X→WIG 35% 53% 15% WIG20→Y 93% 53% 15% X→WIG20 18% 54% 22% SOFIX→Y 31% 76% 15% X→SOFIX 93% 60% 38% SBI20→Y 42% 42% 7% X→SBI20 82% 48% 23% Taking into account number of lags (Table A1) we notice that the biggest number of cases, when H0 is rejected for all 10 lags, is obtained for relations: WIG20→Y (11 times for 13 considered cases), WIG→Y (11) ATX→Y (10), X→SOFIX (12) and X→SBI20 (10). The biggest number of cases, when H0 is 2 See Charemza, Deadman (1997), Osińska (2008). Dorota Witkowska, Krzysztof Kompa, Aleksandra Matuszewska-Janica DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–32 24 not rejected for any considered lag, is observed for relations: SOFIX→Y (6), OMXR→Y (6), SAX→Y (5), X→WIG20 (8) and X→WIG (7). For weekly data the greatest percentage of rejections is obtained for follow- ing relations: PXGLOB→Y (83%), PX→Y (82%), X→OMXT (85%), X→OMXBB (74%). Such results denotes that weekly changes of PXGLOB and PX cause weekly changes of other analysed indexes most often. OMXT and OMXBB are the most sensitive to changes of other indexes. The smallest num- ber of H0 rejection is observed for following relations: SAX→Y (12%), OMXT→Y (25%), X→PX (25%) and X→PXGLOB (27%). The weekly changes of index SAX cause (in Granger sense) the changes of other indexes most rarely. The same result we obtain for daily data but, in contradistinction to daily changes, PX and PXGLOB are the less sensitive to weekly changes of other indexes appear (Table A2). With change from daily to weekly data, number of relations when H0 is rejected for all lags (from 1 to 10) is decreasing, and number of cases when H0 is not rejected for any lag is increasing. It can be interpreted that by broadening of the time interval for returns calculations, the number of causal relations (in Granger sense) is reduced. For monthly data the greatest percentage of rejections is obtained for the following relations: ATX→Y (59%), PXGLOB→Y (45%), PX→Y (43%), X→OMXT (64%), X→OMXBB (48%). While the smallest number of H0 rejec- tions is observed for: SAX→Y (0%), OMXR→Y (5%), SBI20→Y (7%), X→ATX (2%) and X→BET (4%) – Table A3. In comparison to results, ob- tained for weekly data, number of causal (in Granger sense) relations is decreas- ing. As it is visible in Tables A1 – A3, we obtain similar results for WIG and WIG20 since both indexes cause changes of other investigated indexes while they do no influence WIG20 and WIG for daily data. However there are two exceptions for: − weekly data since WIG20→SAX but ¬WIG→SAX, and BUX→WIG20 but ¬BUX→WIG, − monthly data since WIG→SBI20 but ¬WIG20→SBI20, and OMXR→WIG20 but ¬OMXR→WIG. Therefore it does not matter if Warsaw Stock Exchange is represented by WIG (performance, all share index) or WIG20 (price, blue-chip index). One could also notice that (Granger) causal short-run relation between WIG and WIG20 for weekly and monthly returns is bilateral (WIG↔WIG20) while for daily observation only changes of WIG20 cause changes of WIG (WIG20→WIG). The last statement could be explained by high capitalization of companies represented by WIG203. 3 Capitalization of WIG20 is 69.5% of whole market capitalization represented by WIG (www.gpw.pl, September 23, 2011). Analysis of Linkages between Central and Eastern European Capital Markets DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–33 25 The next step of investigation is cointegration analysis provided by Johan- sen method4. As it was mentioned, cointegration analysis is applied in order to check if effective international portfolio (risk) diversification between two capi- tal markets from CEE region is possible. Number of cointegrating vectors are presented in Tables 4–6. Table 4. Number of cointegrating vectors for pairs of indexes – daily data BET 1* BUX 0 1 OMXBB 0 1 0 OMXR 1 1* 0 0 OMXT 0 1 0 1* 0 OMXV 0 1 0 1 0 1 PX 0 0 0 0 1* 0 0 PXGLOB 0 0 0 0 1* 0 0 0 SAX 0 0 0 0 0 0 0 0 0 WIG 0 0 1* 0 0 0 0 0 0 0 WIG20 0 1 0 0 0 0 0 0 0 0 2 SOFIX 0 1 0 0 0 0 0 0 0 0 0 0 SBI20 0 0 0 0 0 0 0 0 0 0 1 1 0 X \ Y ATX BET BUX OMXBB OMXR OMXT OMXV PX PXGLOB SAX WIG WIG20 SOFIX Note: Hypotheses are verified at the significance level α=0.05; * represents statistical significance at the 0.1. For daily data all indexes but one, i.e. SAX, are cointegrated with other indexes from CEE capital markets. The greatest number of long-run relations is observed for BET (8 cases from 13 analyzed). Thus we can conclude that Bul- garian Stock Exchange could be the most sensitive market for international shocks (in the region) so it creates the less number of diversified portfolios. Two cointegrating vectors are observed in relations WIG – WIG20, therefore we suppose that indexes from WSE have stronger relationship among them- selves then with other indexes. For weekly data we observe smaller number of long-run relations than for daily data. Indexes OMXR and BET build the biggest number of cointegrating relations, 4 and 3, respectively. Indexes BUX, OMXBB, SOFIX and SBI20 are not cointegrated with other analyzed indexes. We observe that WIG and WIG20 are cointegrated only between themselves with 2 cointegrating vectors only in shorter period. It seems to be two reasons of this phenomenon. Firstly, Johansen tests results are sensitive on investigation period (see Gilmore et al., 2008 and Pascual, 2003 among others). Secondly, WSE could be not influenced by changes that appear on other CEE capital markets. 4 Usually Johansen tests statistics λtrace and λmax yield the same results but in some cases they are different. In such situation it is accepted λmax test indication (the λmax test is considered superi- or to the λtrace test, see Kennedy, 2003, p. 355). Dorota Witkowska, Krzysztof Kompa, Aleksandra Matuszewska-Janica DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–32 26 Table 5. Number of cointegrating vectors for pairs of indexes – weekly data BET 0 BUX 0 0 OMXBB 0 0 0 OMXR 1 0 0 0 OMXT 0 1 0 0 0 OMXV 0 1 0 0 0 0 PX 0 0 0 0 1* 0 0 PXGLOB 0 0 0 0 1* 0 0 0 SAX 0 1* 0 0 1* 0 0 0 0 WIG 0 0 0 0 0 0 0 0 0 0 WIG20 0 0 0 0 0 0 0 0 0 0 2 a SOFIX 0 0 0 0 0 0 0 0 0 0 0 0 SBI20 0 0 0 0 0 0 0 0 0 0 0 0 0 X \ Y ATX BET BUX OMXBB OMXR OMXT OMXV PX PXGLOB SAX WIG WIG20 SOFIX Note: Hypotheses are verified at the significance level α=0.05; * represents statistical significance at the 0.1. a – 2 cointegrating vectors are obtained for sample PW2, for sample PW1 H0 that cointegration does not exist is not rejected. Table 6. Number of cointegrating vectors for pairs of indexes – monthly data BET 0 BUX 0 0 OMXBB 0 0 0 OMXR 0 0 1* 1 OMXT 0 1* 0 0 0 OMXV 0 0 0 0 0 0 PX 0 1* 0 0 1* 1 0 PXGLOB 0 1* 0 0 1* 1 0 0 SAX 0 0 0 0 0 0 0 0 0 WIG 0 0 0 0 0 0 0 0 0 0 WIG20 1 0 0 0 0 1* 0 0 0 0 2 SOFIX 0 0 0 0 0 0 0 0 0 0 0 0 SBI20 0 0 0 0 0 0 0 0 0 0 0 0 0 X \ Y ATX BET BUX OMXBB OMXR OMXT OMXV PX PXGLOB SAX WIG WIG20 SOFIX Note: Hypotheses are verified at the significance level α=0.05; * represents statistical significance at the 0.1. For monthly data we observe smaller number of long-run relations than for daily data however number of linkages is bigger than for weekly data. Indexes OMXR, OMXT and BET build the biggest number of cointegrating relations – 3 each. Indexes OMXV, SAX, SOFIX and SBI20 are not cointegrated with other investigated indexes. WIG and WIG20 are cointegrated with 2 vectors. In opposite to WIG, only WIG20 is cointegrated with other foreign indexes (i.e. ATX and OMXT). Analysis of Linkages between Central and Eastern European Capital Markets DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–33 27 In literature it is remarked that Johansen tests power does not increase when the higher frequency data are used but with the time span of the data (see Hakkio, Rush, 1991; Diamandis 2009). Hence we can suppose that indication of the Johansen test is more trustworthy for weekly or monthly data than for daily ones. So taking into consideration two capital markets from CEE region, inter- national portfolio risk diversification can be achieved. But we also have to take into account sensitiveness on the changes that appear at the world biggest mar- kets since these shocks are quickly transmitted into global market. In the last step of the analysis, VECM models for relations WIG or WIG20 with other foreign index are estimated. The obtained results are presented in Table 7. Table 7. Selected results of the estimation of the VECM models (for WIG and WIG20 indexes) Pair of indexes Variable order in the model Equation ECM parameter WIG i BUX daily WIG, BUX First for WIG Second for BUX –0.0036 * 0.0031 WIG i SBI20 daily SBI20, WIG First for SBI20 Second for WIG –0.0021 *** 0.0045 WIG20 i BET daily BET, WIG20 First for BET Second for WIG20 –0.0010 *** 0.0003 WIG20 i SBI20 daily SBI20, WIG20 First for SBI20 Second for WIG20 –0.0014 *** 0.0010 WIG i SBI20 weekly SBI20, WIG First for SBI20 Second for WIG –0.0106 *** 0.0021 WIG20 i SBI20 weekly SBI20, WIG20 First for SBI20 Second for WIG20 –0.0070 *** 0.0052 WIG20 i ATX monthly WIG20, ATX First for WIG20 Second for ATX – 0.0070** 0.0253 WIG20 i OMXT monthly WIG20, OMXT First for WIG20 Second for OMXT –0.1301*** –0.0542 Note: ECM parameters significant *** – at the level 0.01, ** – at the level 0.05, * – at the level 0.1. Error correction mechanism is significant for all presented cases. For daily observations, index WIG in relation to BUX has the highest speed of adjustment (circa 0.3% of the discrepancy in these two indexes from the previous day is eliminated in present day). While for weekly data the highest speed of adjust- ment has index SBI20 with relation to WIG, and for monthly data the highest speed of adjustment has index OMXT with relation to WIG20. We can observe that restoring the equilibrium is quicker (from period to period) for monthly data. Conclusions Integration of financial markets has important implications since highly integrated markets are not isolated from international shocks. It could be also Dorota Witkowska, Krzysztof Kompa, Aleksandra Matuszewska-Janica DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–32 28 the reason that the effective portfolio risk diversification between integrated markets cannot be achieved5. For investigated time series many short-run Granger causal relationships are found out. It is also noticed that with broadening of intervals (for which returns are computed) many relations disappear, thus the data frequency does matter6. It is also worth mentioning that for the 10 years period of observation only a few long-run relationships are diagnosed, similarly to the results obtained by other researchers (e.g. Gilmore et al, 2008; Pascual, 2003). The results show that changes at the Warsaw Stock Exchange cause (in Granger sense) changes at other capital markets from CEE region. While WSE indexes are less sensitive to the changes that appear at other investigated mar- kets. Thus it could be considered as a premise that WSE is the most developed capital market in CEE region7 since situation at WSE influences other markets in the region while it is not sensitive on shocks that appear in other CEE mar- kets. References Bailey, W., Choi, J.J. (2005), International Market Linkages, Journal of Economics and Business, 55, 399–404. Bailey, W., Stulz, R. (1989), The Pricing of Stock Index in a General Equilibrium Model, Journal of Financial and Quantitative Analysis, 24, 1–12. Baur, D.G. (2004), What is Co-movement? EUR Working Paper No. 20759 http://ssrn.com/ abstract=570585 (17.10.2011). Booth, G. G.; Martikainen, T. Tse Y. (1997), Price and Volatility Spillovers in Scandinavian Stock Markets, Journal of Banking and Finance, 21, 811–823. Caporale, G.M., Spagnolo, N. (2010), Stock Market Integration between Three CEECs, Econom- ics and Finance Working Paper Series, Working Paper No. 10–09, http://bura.brunel.ac.uk/bitstream/2438/5056/1/1009%5B1%5D.pdf (17.10.2011). Charemza, W., Deadman, D.F. (1997), New Directions in Econometric Practice, PWE, Warsaw [Polish edition]. Chen, G., Firth, M., Rui, O. (2002), Stock Market Linkages: Evidence from Latin America, Jour- nal of Banking and Finance, 26(6), 1113–1141. Choudhry, T. (1996), Interdependence of Stock Markets: Evidence from Europe During the 1920s and the 1930s, Applied Financial Economics, 6, 243–249. Choudhry, T., (1997), Stochastic Trends in Stock Prices: Evidence from Latin American Markets, Journal of Macroeconomics, 19, 285–304. Claessens, S., Forbes, K.J. (ed.) (2001), International Financial Contagion: How it Spreads and how it Can be Stopped, Kluwer Academic Publishing, Boston. 5 See e.g. Hassan, Naka (1996), Phylaktis, Ravazzolo (2005), Syriopoulos (2007). Numerous studies concerning long-run relationships among stock exchanges are quoted in the last mentioned paper. 6 For daily data H0 is rejected for every lag (from 1 to 10) in 74 cases in investigated 182 rela- tions, for weekly data we obtained only 39 such cases while taking into consideration monthly data only 15 such cases are observed. 7 Another explanation of mentioned observation is clamming that matured and great capital markets transmit changes to emerging markets. Analysis of Linkages between Central and Eastern European Capital Markets DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–33 29 Diamandis, P.F. (2009), International Stock Market Linkages: Evidence from Latin America, Global Finance Journal, 20, 13–30. Dubinskas, P., Stunguriene, S. (2010), Alterations in the Financial Markets of the Baltic Countries and Russia in the Period of Economic Downturn, Technological and Economic Develop- ment of Economy, 16(3), 502–515. Elder, J., Kennedy, P.E. (2001), Testing for Unit Roots: What Should Students Be Taught?, Jour- nal of Economic Education, 31(2), 137–146. Engle, R.F. Sumsel R. (1993), Common Volatility in International Equity Markets, Journal of Business and Economic Statistics, 11, 167–176. Eun, C., Shim, S. (1989), International Transmission of Stock Market Movements, Journal Fi- nancial and Quantitative Analysis, 24, 241–256. Fischer, K.P., Palasvirta, A.P. (1990), High Road to a Global Marketplace: The International Transmission of Stock Market Fluctuations, The Financial Review, Eastern Finance Asso- ciation, 25(3), 371–94. Forbes, K.J., Rigobon, R. (2002), No Contagion, Only Interdependence: Measuring Stock Market Comovements, Journal of Finance, American Finance Association, 57(5), 2223–2261. Ghosh, A., Saidi, R., Johnson, K.H. (1999), Who Moves the Asia-Pacific Stock Markets: U.S. or Japan? Empirical Evidence Based on the Theory of Cointegration, Financial Review, 34, 159–170. Gilmore C.G., Lucey B.M., McManus G.M. (2008), The Dynamics of Central European Equity Market Comovements, The Quarterly Review of Economics and Finance, 48, 605–622. Hakkio, G.S., Rush, M. (1991), Cointegration: How Short is the Long Run?, Journal of Interna- tional Money and Finance, 10, 571–581. Hamao, Y., Masulis, R. W., Ng, V. (1990), Correlations in Price Changes and Volatility Across International Stock Markets, Review of Financial Studies, 3, 281–307. Hanousek J., Kočenda E. (2009), Intraday Price Discovery in Emerging European Stock Markets, GERGE-EI Working Paper Series 382, http://www.cerge-ei.cz/pdf/wp/Wp382.pdf (17.10.2011). Hanousek J., Kočenda E., Kutan A.M. (2008), The Reaction of Asset Prices to Macroeconomic Announcements in New EU Markets: Evidence from Intraday Data, GERGE-EI Working Paper Series 349, http://www.cerge-ei.cz/pdf/wp/Wp349.pdf (17.10.2011). Hassan, M.K., Naka A. (1996) Short-Run And Long-Run Dynamic Linkages Among Internation- al Stock Markets, International Review of Economics and Finance, 5(4), 387–405. Huang, B.N., Yang C.W., Hu J.W.S. (2000), Causality and Cointegration of Stock Markets among the United States, Japan and the South China Growth Triangle, International Re- view of Financial Analysis, 9(3), 281–297. Janakiraman, S., Lamba, A.S. (1998), An Empirical Examination of Linkages between Pacific- Basin Stock Markets, Journal of International Financial, Markets, Institutions and Money, 8, 155–173. Karolyi, G.A., Stulz, R.M, (1996), Why Do Markets Move Together? An Investigation of U.S.- Japan Stock Return Comovements, Journal of Finance, American Finance Association, 51(3), 951–86. Kasa K. (1992), Common Stochastic Trends in International Stock markets, Journal of Monetary Economics, 29, 95–124. Kearney C., Lucey B.M. (2004), International Equity Market Integration: Theory, Evidence and Implications, International Review of Financial Analysis, 13, 571–583. Kearney, C. (2000), The Determination and International Transmission of Stock Market Volatili- ty, Global Finance Journal, 11, 1–22. Kennedy, P. (2003), A Guide to Econometrics, Blackwell Publishing. Kim, S.W. Rogers, J.H. (1995), International Stock Price Spillovers and Market Liberalization: Evidence from Korea, Japan, and the United States, Journal of Empirical Finance, 2(2), 117–133. Dorota Witkowska, Krzysztof Kompa, Aleksandra Matuszewska-Janica DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–32 30 Koedijk, K.C.G., Campbell, R.A.J., Kofman, P. (2002), Increased Correlation in Bear Markets. Financial Analysts Journal, 58(1), 87–94. Kompa, K. (2010), Capital Markets in Transitional Countries. Comparison of Stock Exchanges in Europe, in Witkowska, D., Nermend, K. (ed.), Regional Analysis: Globalization, Integra- tion, Transformation, Wydawnictwo Uniwersytetu Szczecińskiego, Szczecin, 125–152. Koutmos, G. (1996), Modelling The Dynamic Interdependence of Major European Stock Mar- kets, Journal of Business Finance and Accounting, 23, 975–988. Koutmos, G., Booth, G.G. (1995), Asymmetric Volatility Transmission in International Stock Markets, Journal of International Money and Finance, 14, 747–762. Kuçukcolak, N. (2008), Co-Integration of The Turkish Equity Market With Greek and other European Union Equity Markets, International Research Journal of Finance and Econom- ics, 13, 58–73. Kwan, A.C.C., Sim, A.H.B., Cotsomitis, J.A. (1995), The Causal Relationships between Equity Indices on World Exchanges, Applied Economics, 27, 33–37. Li, H., Majerowska, E. (2008), Testing Stock Market Linkages for Poland and Hungary: A Multi- variate GARCH Approach, Research in International Business and Finance, 22, 247–266. Lin, W.-L., Engle, R. F., Ito, T. (1994), Do Bulls and Bears Move Across Borders? International Transmission of Stock Returns and Volatility, Review of Financial Studies, 7, 507–538. Longin, F. M., Solnik, B. (1995), Is the Correlation in International Equity Returns Constant: 1960–1990, Journal of International Money and finance, 14, 3–26. Longin, F., Solnik, B., (2001), Extreme Correlation of International Equity Markets, Journal of Finance, 56(2), 649–676. MacDonald, R. (2001), Transformation of External Shocks and Capital Market Integration, in Schroder, M. (ed.), The New Capital Markets in Central and Eastern Europe, Springer Verlag, 210–245. Maddala, G.S., Kim, I.-M. (1998), Unit Roots, Cointegration, and Structural Change, Cambridge University Press, Cambridge. Masih, R., Masih, A.M.M. (2001), Long and Short Term Dynamic Causal Transmission Amongst International Stock Markets, Journal of International Money and Finance, 20, 563–587. Matuszewska-Janica, A. (2010), Short-Run Relationship between Selected Central European Stock Exchanges: Causality and Exogeneity Analysis, Uniwersytet Szczeciński, Zeszyty Naukowe nr 612(28), tom Rynek kapitałowy, 519–532. Matuszewska-Janica, A. (2011), Long-run Relationships between Selected Central European Indexes, International Advanced in Economic Research, 17, 157–168. Meric, I., Meric, G. (1989), Potential Gains from International Portfolio Diversification and Inter- temporal Stability and Seasonality in International Stock Market Relationships, Journal of Banking and Finance, 13(4/5), 627–640. Narayan, P., Smyth, R., Nandha M. (2004), Interdependence and Dynamic Linkages between the Emerging Stock Markets of South Asia, Accounting and Finance, 44(3), 419–439. Ng, A. (2000), Volatility Spillover Effects from Japan and the US to the Pacific-Basin, Journal of International Money and Finance, 19, 207–233. Osińska, M. (2008), Econometric analysis of causal relationships, Wydawnictwo Naukowe Uniwersytetu Mikołaja Kopernika, Toruń. Panton, D.B., Lessing, P.V. Joy, M.O. (1976), Co-movement of International Equity Markets: a Taxonomic Approach, Journal of Financial and Quantitative Analysis, 11, 415–32. Pascual, A.G. (2003), Assessing European Stock Market Integration, Economic Letters, 78, 197– 203. Phylaktis K., Ravazzolo F. (2005), Stock Market Linkages in Emerging Markets: Implications for International Portfolio Diversification, International Financial Markets, Institutions and Money, 15, 91–106. Richards, A.J. (1995), Comovements in National Stock Market Returns: Evidence of Predictabil- ity, but not Cointegration, Journal of Monetary Economics, 36(3), 631–654. Analysis of Linkages between Central and Eastern European Capital Markets DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–33 31 Rigobon, R., (1999), On the Measurement of the International Propagation of Shocks, NBER Working Papers 7354, http://www.nber.org/papers/w7354.pdf (17.10.2011). Roca, E.D. (1999), Short-Term and Long-Term Price Linkages between the Equity Markets of Australia and its Major Trading Partners, Applied Financial Economics, 9, 501 –511. Serletis, A., King, M. (1997), Common Stochastic Trends and Convergence of European Union Stock Markets, The Manchester School, 65(1), 44–57. Serwa, D., Bohl, M.,T. (2005), Financial Contagion Vulnerability and Resistance: A Comparison of European Stock Markets, Economic Systems, 29(3), 344–362. Shachmurove, Y., Witkowska, D. (2001), Dynamic Interrelation Among Major World Stock Market: A Neural Network Analysis, International Journal of Business, 6(1), 1–22. Sharkasi, A., Ruskin, H., Crane, M. (2004), Interdependence between Emerging and Major Mar- kets", in Antoch, J. (ed.), COMPSTAT 2004 – Proceedings in Computational Statistics 16th Symposium Prague, Czech Republic, Physica-Verlag, Springer, 1783–1790. Siklos, P.L., Ng, P. (2001), Integration among Asia-Pacific and International Stock Markets: Common Stochastic Trends and Regime Shifts, Pacific Economic Review, 6, 89–110. Syllignakis, M.N., Kouretas G.P. (2010), Dynamic Correlation Analysis of Financial Contagion: Evidence from the Central and Eastern European Markets, International Review of Eco- nomic and Finance, 20, 717–732. Syriopoulos T. (2007), Dynamic Linkages between Emerging European and Developed Stock Markets: Has the EMU any Impact?, International Review of Financial Analysis, 16, 41– 60. Voronkova S. (2004), Equity Market Integration in Central European Emerging Markets: A Coin- tegration Analysis with Shifting Regimes, International Review of Financial Analysis, 13, 633–647. Watson, J. (1980), The Stationarity of Inter-Country Correlation Coefficients: A Note, Journal of Business Finance and Accounting, 7, 297–299. Witkowska, D. (1999), Neural Networks Application to Analysis of Daily Stock Returns at the Largest Stock Markets, in Szczepaniak (ed.), Computational Intelligence and Applications, Physica-Verlag Heidelberg - New York, 351–364. Yang, J., Kolari, J.W., Sutanto, P.W. (2004), On the Stability of Long-Run Relationships between Emerging and US Stock Markets, Journal of Multinational Financial Management, 14(3), 233–248. Analiza relacji pomiędzy rynkami kapitałowymi Europy Środkowej i Wschodniej Z a r y s t r e ś c i. Celem analizy jest ocena związków krótkookresowych (w zakresie przyczynowości) i długookresowych (kointegracja) pomiędzy rynkami kapitałowymi Europy Środkowej i Wschodniej, a w szczególności pomiędzy giełdą w Warszawie i pozostałymi rynka- mi. Analizie poddano dzienne, tygodniowe i miesięczne stopy zwrotu indeksów notowanych na tych rynkach. Badania obejmują okres od stycznia 2000 do listopada 2010. S ł o w a k l u c z o w e: giełdy Europy Środkowej i Wschodniej, rynki rozwijające się, kointegracja, przyczynowość w sensie Grangera, krótkookresowe i długookresowe relacje pomiędzy rynkami. Dorota Witkowska, Krzysztof Kompa, Aleksandra Matuszewska-Janica DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–32 32 APPENDIX Table A1. Lags for H0 rejection in Granger causality test: daily data (PD1, PD2) X \ Y ATX BET BUX OMXBB OMXR OMXT OMXV PX PXGLOB SAX WIG WIG20 SOFIX SBI20 ATX X 1-10 1-10 1-10 1-10 1-10 1-10 1-10 1-10 7 1-10 1-10 BET 1-10 X 1-7 1-3 2 1-10 1-10 4;5;7 1-10 1-9 BUX 1-10 1-10 X 1-10 1;2; 4-10 1-10 1-10 1-10 1-10 1;4-10 4-10 5-10 1-10 1-10 OMXBB 8-10 3-10 4-10 X 1-10 1;2;4-10 1-10 4-10 4-10 2-10 1;4-10 1 1-10 1-10 OMXR 8-10 5-10 7-10 X 8-10 8-10 1-10 6 OMXT 4-10 5-10 4-10 1-10 1-10 X 1-10 4-10 4-10 2-10 1-10 1-10 OMXV 2;7-10 3-10 2-10 1;6-10 1-10 1-10 X 7-10 7-10 2-6;8-10 1-10 1-10 PX 1-10 2-10 1-7 1-10 1-10 1-10 1-10 X 1-3 4-10 2-10 2;5-10 1-10 1-10 PXGLOB 1-10 2-10 1-7 1-10 1-10 1-10 1-10 1 X 4-10 2-10 2;3; 5-10 1-10 1-10 SAX 3 3-10 3-7 5-10 1;3 X 1-4 1 3 WIG 1-10 1-10 1-10 1-10 1-10 1-10 1-10 1-10 1-10 4;9;10 X 1-10 1-10 WIG20 1-10 1-10 1-10 1-10 1-3 1-10 1-10 1-10 1-10 3-10 1-10 X 1-10 1-10 SOFIX 1-10 6-10 6-10 6-10 6-10 1-10 X 1-10 SBI20 3-10 3-10 7 9 4-10 2;5;6 10 4-10 4-10 1-10 X Note: Bolded are cases when the H0 is rejected for all considered lags, shaded - when no H0 is rejected. Table A2. Lags for H0 rejection in Granger causality test: weekly data (PW1, PW2) X \ Y ATX BET BUX OMXBB OMXR OMXT OMXV PX PXGLOB SAX WIG WIG20 SOFIX SBI20 ATX X 1-5; 8-10 2;3;8;9 1-10 1-10 9 8 8;9 2-7;10 3-10 3-10 1-10 7-10 BET 2-10 X 2;3;6-8 2-10 6;7 2-10 4-9 2-10 2-10 3-10 2-4;6-8 2-10 1-10 8;9 BUX 1-8; 10 X 1-10 4-7 1-10 5 6 2-10 1 4-10 5;7-10 OMXBB 3 1 2-8 X 1-10 1-4; 8-10 1 1-4;10 2-5;8 2-7 1;3 1-3 OMXR 1-3; 9;10 2-10 1-9 X 4-10 1 1-5;7; 9;10 1-5;7; 9;10 7;8 1;2; 6;8 OMXT 2-4 8-10 1;3-10 X 1;2;8;10 2-5; 7-10 2-7;10 OMXV 3;6;7; 9;10 9;10 2-10 2-7 1;2;4; 5;7 1-10 X 1-3 2-4; 8;9 2;3;8;9 1-4 1-3 PX 1;4-10 1-10 1-10 1-10 4-8 1-10 4-10 X 1-10 1-10 1-10 1-10 3-10 PXGLOB 1;4-10 1-10 1-10 1-10 4-8 1-10 2;4-10 X 1-10 1-10 1-10 1-10 3-10 SAX 1-10 X 3-5; 7;9;10 WIG 1;4-6 1-8 1-10 1-10 4-7 1-10 1-9 X 2-10 1-10 1;2; 4-10 WIG20 4-8 1-10 1 1-10 4-7 1-10 1-5;6;8 2-4 2-10 X 1;2; 4-10 1;2;4 SOFIX 1-5; 7-9 1-10 1-10 2-10 1-10 1-10 1-4 1-5;7 1-10 1-8 1-7 X 1-7 SBI20 5-10 6-10 3-10 4-10 1-10 1-10 3-4 3-6; 8;10 X Note: Bolded are cases when the H0 is rejected for all considered lags, shaded - when no H0 is rejected. Analysis of Linkages between Central and Eastern European Capital Markets DYNAMIC ECONOMETRIC MODELS 12 (2012) 19–33 33 Table A3. Lags for H0 rejection in Granger causality test: monthly data (PM1, PM2) X \ Y ATX BET BUX OMXBB OMXR OMXT OMXV PX PXGLOB SAX WIG WIG20 SOFIX SBI20 ATX X 1-10 1-10 1-10 1-10 1;4-10 5-10 2;5-10 1-10 1-4;9;10 BET X 1-5;8 2-5 1-6 2;4;8-10 1-3 1-6 BUX X 1-10 1-6;8 1-10 1-5 3 3;4 1-4;6 1;3;4 OMXBB 3 X 1-4;7-9 1-10 1 6-8 OMXR 1 X 3 1 OMXT 3;4;6 3;4 1-10 3;7-10 X OMXV 1;2 2 1-4;9 X 2 1;3;5 1-7 1;6 1;3;4; 6;7 PX 1-10 1-4 1-10 1-10 X 3;6-9 3-9 1-6 2-5 PXGLOB 1-10 1-5 1-10 1-10 X 3;6-9 3-9 1-7 2-5 SAX X WIG 1-3;7 1;3 1-10 1-4;7-9 X 1-4 1-3;4 1;2 WIG20 1;2 1 1-4; 7;9;10 1-3 1-4;10 X 1;2 SOFIX 1-8 1;2;5-7 1;8;9 X 1;9;10 SBI20 2 2 2 1-6 X Note: Bolded are cases when the H0 is rejected for all considered lags, shaded - when no H0 is rejected. Introduction 1. Data Description 2. Results Conclusions References APPENDIX