No Commonality in Liquidity on Small Emerging Markets? Evidence from the Central and Eastern European Stock Exchanges

Olbrys, Joanna Article No Commonality in Liquidity on Small Emerging Markets? Evidence from the Central and Eastern European Stock Exchanges Comparative Economic Research. Central and Eastern Europe Provided in Cooperation with: Institute of Economics, University of Łódź Suggested Citation: Olbrys, Joanna (2020) : No Commonality in Liquidity on Small Emerging Markets? Evidence from the Central and Eastern European Stock Exchanges, Comparative Economic Research. Central and Eastern Europe, ISSN 2082-6737, Łodz University Press, Łodz, Vol. 23, Iss. 3, pp. 91-109, https://doi.org/10.18778/1508-2008.23.22 This Version is available at: https://hdl.handle.net/10419/259243 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0 Comparative Economic Research. Central and Eastern Europe Volume 23, Number 3, 2020 http://dx.doi.org/10.18778/1508-2008.23.22 Joanna Olbryś No Commonality in Liquidity on Small Emerging Markets? Evidence from the Central and Eastern European Stock Exchanges1 Joanna Olbryś Ph.D., Associate Professor, Bialystok University of Technology Faculty of Computer Science, Department of Theoretical Computer Scnience Bialystok, Poland, e-mail: [email protected] Abstract The goal of this comparative research is to investigate intra-market commonality in liquidity on six small emerging Central and Eastern European (CEE) stock exchanges – in the Czech Republic, Hungary, Slovakia, Lithuania, Estonia, and Latvia. The CEE post-communist countries can be analyzed together as they are geographically close, and the stock markets are relatively similar. Three measures based on daily data are utilized as liquidity/ illiquidity proxies: (1) a modified version of the Amihud (2002) measure, (2) the percentage relative spread, and (3) the Corwin-Schultz (2012) high-low two-day spread estimator. The OLS regression with the HAC covariance matrix estimation and the GARCH-type models are employed to explore the patterns of market-wide commonality in liquidity on the CEE stock exchanges. The main value-added comes from the methodology and the novel empirical findings. To the best of the author’s knowledge, this is the first study that investigates commonality in liquidity in the aforementioned group of countries using three liquidity proxies and the time rolling-window approach to provide robustness tests. The regressions reveal no pronounced evidence of co-movements in liquidity within the CEE markets, taken separately. What is important, the empirical results are homogeneous for all investigated markets. Therefore, no reason has been found to reject the research hypothesis that there is no commonality in liquidity on each individual market. This paper aspires to fill the gap in the knowledge of liquidity patterns on the CEE emerging markets. Keywords: Central and Eastern Europe, commonality in liquidity, GARCH, OLS-HAC, time rolling-window, daily data JEL: C32, C58, G12, G15, O52 1 This research was supported by grant No. 2016/21/B/HS4/02004 from the National Science Centre, Poland. 92 Joanna Olbryś Introduction According tothe literature, the existence ofcommonality inliquidity indicates that individual company liquidity issensitive tochanges inaggregate stock market liquidity. Itis well documented that assessing co-movements inliquidity isimportant for anumber ofreasons. There are some crucial topics that are especially frequently investigated inthis context, including the consideration ofcommonality inliquidity innon-classical asset pricing models since itcould represent asource ofnon-diversifiable risk, the relationship between shareholders structure and individual firm liquidity, the influence ofcommonality inliquidity oninvestment strategies, and the importance ofcommonality inliquidity toregulators and central bankers (Olbryś 2019a, p.252). Narayan etal. (2015) emphasize that empirical evidence ofcommon liquidity movements would assist regulators inimproving stock market design. Bekaert etal. (2007) pointed out that liquidity ismore critical for emerging than developed markets. The six small Central and Eastern European (CEE) stock exchanges are emerging markets, but four ofthem (Slovakia, Lithuania, Estonia, and Latvia) are, infact, frontier markets (Kiviaho etal.2014). Therefore, one might expect them tobe very sensitive tochanges inliquidity. Moreover, Brockman etal. (2009) examined the impact ofdomestic macroeconomic announcements oncommonality inliquidity for individual stock exchanges, and their results revealed that the announcement effects are stronger for emerging markets asagroup than for developed markets. The small post-communist CEE stock exchanges can beanalyzed asagroup ofmarkets because they are geographically close and relatively similar. Taking the above into consideration, one might expect that commonality inliquidity exists onthese markets. Unfortunately, alarge number ofcompanies listed onthe CEE exchanges reveal asubstantial non-trading problem (Olbryś 2018). The non-trading effect means there isalack oftransactions over aparticular period when anexchange isopen for trading, with illiquidity asits consequence. Therefore, inthe current paper, the research hypothesis that there isno commonality inliquidity oneach individual CEE stock market istested. The goal ofthis comparative research isto thoroughly investigate intra-market co-movements inliquidity onsix small CEE stock markets – inthe Czech Republic, Hungary, Slovakia, Lithuania, Estonia, and Latvia. The Polish stock exchange isnot included inthis study because itis large compared tothe other CEE stock exchanges inthe region. However, intra-market commonality inliquidity onthe Warsaw Stock Exchange has recently been quite deeply explored inthe papers (e.g., Olbryś 2019a; Będowska-Sójka 2019). The empirical results revealed rather weak evidence ofco-movements inliquidity onthe WSE, regardless ofthe choice ofliquidity proxy. Itis worth noting that Olbryś (2018) conducted apreliminary study ofcommonality inliquidity onthe small CEE exchanges inthe context ofthe non-trading problem. Amodified version ofthe Amihud (2002) measure was used asadaily liquidity proxy inthe period from January 2, 2012, toDecember 30, 2016. The classical market model 93 No Commonality in Liquidity on Small Emerging Markets… ofliquidity proposed byChordia etal. (2000) was employed. The empirical results revealed noevidence ofcommonality inliquidity onany ofthe investigated CEE stock markets. Toconfirm this phenomenon, inthe current research, the daily percentage relative spread and Corwin and Schultz’s (2012) high-low spread estimator are utilized asadditional liquidity proxies. The common feature ofthe measures used inthe study isthat they are all based ondaily data and are calculated indaily frequency. Moreover, the time rolling-window approach isused totest the stability ofthe empirical findings indifferent sub-periods. Following Olbryś (2018), the classical market model ofliquidity isutilized inthis study. The OLS regression with the HAC covariance matrix estimation (Newey, West 1987) and the GARCH-type models (ifnecessary, asthe OLS-HAC may not fully accommodate the ARCH effect) are employed toinfer the patterns ofcommonality inliquidity. The main contribution ofthis research lies inits thorough assessment ofcommonality inliquidity onsix small CEE stock exchanges. The value-added derives from the methodology and the novel empirical findings. Tothe best ofthe author’s knowledge, this isthe first study that investigates commonality inliquidity inthe aforementioned group ofcountries using three liquidity proxies and the time rolling-window approach toprovide robustness tests. The regressions reveal nopronounced evidence ofcommonality inliquidity onthe CEE stock markets, taken separately. Importantly, the empirical findings are homogeneous for all investigated markets. Itmeans that individual firm liquidity isnot significantly influenced byco-movements inthe liquidity ofall other firms traded onthe same exchange. The findings fill the gap inthe knowledge ofcommonality inliquidity onemerging and frontier stock markets. According tothe literature, the results for the CEE stock exchanges reported inthis paper substantially differ from findings that have been obtained for developed markets around the world. The remainder ofthe study isorganized asfollows. Section 2 provides abrief literature review ofcommonality inliquidity onemerging markets. Section 3 specifies the methodological background concerning the measurement ofcommonality inliquidity. Section 4 describes the data and discusses the empirical results for the six stock exchanges. The paper issummarized inthe presentation ofconclusions, implications for practice, aswell aslimitations and suggestions for further research. Commonality in liquidity on emerging markets The first empirical study ofcommonality inliquidity was conducted byChordia etal. (2000). Beginning with this seminal paper, identifying commonality inliquidity emerged asafast-growing strand ofthe literature onliquidity, especially for the U.S.stock market (e.g., Chordia etal.2000; Kamara etal.2008; Kang, Zhang 2013; Korajczyk, Sadka 2008). Commonality inliquidity has also been explored for other individual emerging and developed equity markets inthe world. Ingeneral, the em- 94 Joanna Olbryś pirical results from various markets are ambiguous. The majority ofresearchers suggest that market structure and trading mechanisms play important roles indifferent effects ofcommonality inliquidity for the observed markets. Moreover, the non-trading problem may substantially affect the findings ofliquidity co-movements onsmall emerging stock exchanges. This isbecause infrequently traded stocks cannot provide reliable information. Kearney (2012) pointed out that although the term “emerging market” isin common usage, there isno agreement oneither the theoretical oroperational definition ofwhat itconstitutes, and the classification ofcountries asemerging markets isconsequently somewhat arbitrary. Asthis research aspires todraw attention toward emerging economies, the analysis ofprevious literature focuses onstudies that relate mostly toemerging stock exchanges. However, the majority ofpapers concern Asian emerging markets, which, ingeneral, are not comparable toEuropean small stock exchanges. For example, Pukthuanthong-Leand Visaltanachoti (2009) assessed the Stock Exchange inThailand, and they confirmed market-wide commonality inliquidity onthis market. Narayan etal. (2015) found strong support for commonality inliquidity onthe Chinese stock exchanges inShanghai and Shenzhen. Syamala etal. (2017) analyzed the Indian stock market, and presented evidence for both supply-side and demand-side factors that contribute toliquidity commonality. Wang (2013) examined the impact ofaset ofcommon factors onliquidity variations oneight emerging equity markets inAsia, namely China, India, Indonesia, Korea, Malaysia, the Philippines, Taiwan, and Thailand. Meanwhile, Sensoy (2016) investigated the influence ofspecific macro-announcements onliquidity commonality inTurkey. Alimited number ofstudies investigate commonality inliquidity for agroup ofequity markets, with Central and Eastern European economies receiving particularly little attention. For example, Brockman etal. (2009) utilized the methodology ofChordia etal. (2000) on47 stock markets from different contingent-based regions, but their database included only two ofthe CEE countries, namely Poland and Hungary. Karolyi etal. (2012) analyzed cross-country commonality inliquidity using daily data for stocks from 40 developed and emerging markets, but their database contained only the Polish stock exchange. Bai and Qin (2015) investigated commonality inliquidity in18 emerging countries, but, aswith Karolyi etal., only Poland was included intheir research. Importantly, the authors pointed out that liquidity co-movements across emerging stock exchanges have apronounced geographic component. This evidence might becrucial inthe case ofCEE countries that are geographically close. Olbryś (2019b) assessed market-wide commonality inliquidity onthe CEE-3 stock exchanges inPoland, the Czech Republic, and Hungary. The empirical findings confirmed weak evidence ofco-movements inliquidity onthe analyzed markets, considered separately. None ofthe aforementioned studies concerns the whole group ofthe CEE countries. Table 1 includes brief information onthe six small stock markets that are investigated inthis research, namely the Czech Republic, Hungary, Slovakia, Lithuania, Estonia, and Latvia. 95 Table 1. The six small Central and Eastern European stock markets highlights Country Stock exchange Index Stock market established Market Cap., EUR billion, Dec 2016 Czech Republic Prague Stock Exchange (PSE) PX 1993 22.19 Hungary Budapest Stock Exchange (BSE) BUX 1991 21.27 Slovakia Bratislava Stock Exchange (BSSE) SAX 1993 5.28 Lithuania NASDAQ Vilnius OMXV 1993 3.50 Estonia NASDAQ Tallinn OMXT 1995 2.29 Latvia NASDAQ Riga OMXR 1995 0.80 Source: National stock exchange websites. Methodology Inthis section, the methodological background concerning the measurement ofcommonality inliquidity ispresented. Selected liquidity/illiquidity proxies derived from daily data and econometric methods applied inthe study are described indetail. Liquidity proxies derived from daily data Aninvestigation ofliquidity iscomplicated byvarious obstacles. Alack ofaccess tointraday data onmost emerging stock markets might beconsidered one such inconvenience, and itis aproblem that iswidely known and amply reported inthe literature (e.g., Bekaert etal.2007; Olbryś 2014). High-frequency data are not freely available for the analyzed CEE stock exchanges. Therefore, inthis study, three liquidity proxies approximated from daily data are utilized tocapture various sources ofmarket liquidity, which is, infact, alatent variable. Table 2 presents the formulas ofthese proxies. Table 2. Definition of daily liquidity/illiquidity proxies utilized in the study Liquidity proxy Definition 1 The modified version of the Amihud (2002) measure t MAmih 1 , 0 0, 0 t t tt t r log whenV MAmih V whenV ìæö ï÷ ïç÷ +¹ ïç÷ ïç÷ ç =íèø ï ï ï= ï î 2 The percentage relative spread %t RS ( ) 200 % HL tt tHL tt PP RS PP ×- =+ No Commonality in Liquidity on Small Emerging Markets… 96 Liquidity proxy Definition 3 The Corwin-Schultz (2012) high-low two-day spread estimator t S ( ) 21 1 t e Se a a - =+ Where: t r isthe simple rate ofreturn ofstock onday t, t V isthe trading volume ofstock onday t, H t P, L t Pare the high and low prices ofstock onday t, respectively, 2 3 22 3 22 Bbg a- =- -- isthe main parameter inthe formula for the t S estimator, and 2 2 1 1 H H t t LL tt P P ln ln PP b+ + éù éù æö æö ÷ ÷ç çêú êú ÷ ÷ =+ ç ç÷ ÷êú ç êú ç÷ ÷ çç èø èø êú êú ëû ëû , ( ) ( ) 2 1 1 max , max , HH tt LL tt PP ln PP g+ + éù æö ÷ ç êú ÷ ç =÷ êú ç÷ ç÷ êú ÷ ç èø ëû . Source: author’s own elaboration based on Karolyi et al. 2012; Olbrys, Mursztyn 2018; Olbryś 2019a; and Corwin, Schultz 2012. Table 2 requires some comments. The value ofthe daily proxy t MAmih isdefined tobe equal tozero when the total daily volume isequal tozero. Inthe literature, the Amihud measure isusually calculated for astock for each month (e.g., Fong etal.2017; Olbryś 2014). However, inthis study, daily time series ofthe modified Amihud proxy are estimated. The percentage relative spread %t RS isameasure ofilliquidity because ahigh value ofthis indicator denotes low liquidity while asmall value ofthe %t RS indicates high liquidity. The t S estimator isquite easy tocompute asit requires only the high and low prices from two consecutive days, t and 1t+ . Itis calculated for astock oneach trading day. However, Corwin and Schultz (2012) emphasize that infrequent trading isacrucial problem ifall trades occur atthe same price, and then HL tt PP= . Infact, the t S measures illiquidity, sousually the higher are the values ofthis indicator, the lower liquidity isobserved onagiven day. Assessing commonality in liquidity Toinvestigate commonality inliquidity, the classical market model ofliquidity proposed byChordia etal. (2000) isthe most frequently employed model inthe literature. Inthis research, amodified version ofthis model, including the Dimson (1979) correction for daily data, isapplied: , ,1,1,0,,1,1, , it i i Mt i Mt i Mt it DL DL DL DLab b b e -- ++ =+ × + × + × + (1) Tabel 2. (continued) Joanna Olbryś 97 No Commonality in Liquidity on Small Emerging Markets… where , it DL for stock i isthe change inliquidity variable L from trading day 1ttot, i.e., 1 1 tt t t LL DL L - - - =. The Dimson correction allows usto mitigate the non-synchronous trading problem. Inthis procedure, the ,1Mt DL - , ,Mt DL , and ,1Mt DL + variables are included inthe model equation. These variables are the lagged, concurrent, and leading changes isacross-sectional average ofthe liquidity variable L , respectively. Itis crucial that incomputing the ‘market’ liquidity proxy M L, stock i isexcluded and the measure M L. isestimated asthe equally-weighted average liquidity for the remaining stocks, for each individual stock market, sothe explanatory variables inthe model (1) are slightly different for each stock regression (Olbryś 2019a, p.262). Positive and statistically significant slope coefficients ,0i b, ,1i b-, and ,1i b+ are especially desired since they indicate commonality inliquidity. Basically, they confirm liquidity co-movements inthe same direction (e.g., Brockman etal.2009; Olbryś 2018; 2019a; 2019b). Model (1) isinitially estimated for each stock bythe OLS regression with the robust HAC estimates (Newey, West 1987), but the OLS-HAC may not fully correct for the influence problems introduced bythe ARCH effect. Insuch cases, estimating model(1) asaGARCH-type model ismore appropriate. Engle’s (1982) test isemployed toinfer the ARCH effect. The GARCH(p, q) model isdefined byEq. (2): ( ) , ,1,1,0,,1,1, , , i,t , 2 i,t ,0 , , , i,t-l 11 , h , ~ 0,1 , h h, it i i Mt i Mt i Mt it it it it qp i ik it k il kl DL DL DL DL z zN aa b ab b b e e e -- ++ - == =+ × + × + × + = =+ + åå (2) where ,0 , , 0, 0, 1, , , 0, 0, , , , 0. i ik il a a k qq b l pp> ³ = ¼ > ³ =¼ ³ The ,it e isthe innovation inalinear regression with ( ) 2 Ves= , while ,it h isthe variance function. The rest ofthe notation isthe same asin Eq. (1) (see for example Olbryś 2018; 2019a; 2019b). Data description and empirical results on the CEE stock exchanges Inthe present study, daily data for stock exchanges from the Czech Republic, Hungary, Slovakia, Lithuania, Estonia, and Latvia, are utilized. Data comes from Bloomberg under alicense agreement between Bloomberg and Bialystok University ofTechnology (grant No.2016/21/B/HS4)2. The database contains the opening, high, low, and 2The database was prepared specifically for the grant and it was purchased from Bloomberg in January 2017. Therefore, to avoid internal inconsistency of the research, all empirical analyses concerning various aspects of liquidity for the six stock markets were conducted for the same period from January 2012 to December 2016 (e.g., Olbryś 2018; 2020). 98 Joanna Olbryś closing prices, aswell asthe volume for each equity over each trading day, from January2, 2012, toDecember 30, 2016. The database holds 1252 (for the PSE), 1240 (for the BSE), 1244 (for the BSSE), 1245 (for the NASDAQ Vilnius), 1251 (for the NASDAQ Tallinn), and 1242 (for the NASDAQ Riga) trading days, respectively. The Warsaw Stock Exchange (WSE) isnot included inthe study because itis large compared tothe other CEE stock markets. For comparison, atthe end of2016, the total number oflisted companies was 881 (WSE), 23 (PSE), 41 (BSE), 71 (BSSE), 34 (NASDAQ Vilnius), 17(NASDAQ Tallinn), and 32 (NASDAQ Riga) (Olbryś 2018, p.72). Itis widely known that alot ofequities listed onemerging stock markets display asubstantial non-trading problem. Toavoid this problem, the companies that exhibited anextraordinarily high number ofnon-traded days within the whole sample period (precisely, above 373 zeros indaily volume, which constituted about 30%ofall trading days), were excluded from the data set. Finally, the database contained 10(Prague), 18(Budapest), 3 (Bratislava), 15 (Vilnius), 12 (Tallinn), and 7 (Riga) companies (65firms intotal) (Olbryś 2018, p.73). Testing for stock exchange-level commonality in liquidity Inthe first step, using the ADF-GLS test (Elliott etal.1996) orADF test (Dickey, Fuller 1981), Itested whether the daily time series are stationary. Itwas proved that the unitroot hypothesis can berejected atthe 5% significance level for all time series utilized inthe study. Inorder toreduce the effects ofpossibly spurious outliers, the data was ‘winsorized’ bythe 1st and 99th percentiles for each time series (e.g., Korajczyk, Sadka 2008; Kamara etal.2008). Inthe second step, the OLS-HAC regression was employed toestimate the parameters ofmodel (1). Intotal, 195 models for the six stock markets and three liquidity proxies (MAmiht, %RSt, and St) were estimated, comprising 30 (Prague), 54(Budapest), 9 (Bratislava), 45 (Vilnius), 36 (Tallinn), and 21 (Riga). For each stock, the daily proportional changes inindividual stock liquidity variables were regressed intime-series onthe changes ofan equally weighted cross-sectional average ofthe liquidity variable for all stocks inthe sample, excluding the dependent variable stock (Olbryś 2019a, p.264). The empirical results showed that the OLS-HAC regressions proved tobe appropriate for 29 models (Prague), 42 models (Budapest), 8 models (Bratislava), 35 models (Vilnius), 30 models (Tallinn), and 20 models (Riga) because the ARCH effect did not appear. Only for 31 models was the ARCH effect inthe residuals detected. Therefore, for those companies, the GARCH(p, q), p, q=1,2 models (2) were estimated. The number oflags p, q, was selected onthe basis ofthe AIC and SC information criteria. The cross-sectional estimation results ofmodels (1) and (2) are presented inTable3. This table contains the number ofpositive significant, positive insignificant, negative significant, and negative insignificant coefficients (atthe 10% significance level), for each stock exchange and each liquidity proxy, separately. 105 No Commonality in Liquidity on Small Emerging Markets… Table 6. The rolling-window findings of testing for stock exchange-level commonality in liquidity on six small CEE stock markets (the St proxy) Coefficient The proportion of positive/negative and statistically significant slope coefficients Window 1 Window 2 Window 3 Prague (10 models) Concurrent β i,0 2/2 1/0 0/0 Lag β i,–1 1/0 1/0 0/0 Lead β i,+1 0/0 0/0 1/2 Budapest (18 models) Concurrent β i,0 3/2 3/2 1/1 Lag β i,–1 1/2 1/1 1/0 Lead β i,+1 0/1 0/2 0/2 Bratislava (3 models) Concurrent β i,0 0/0 0/0 0/2 Lag β i,–1 1/0 0/0 0/0 Lead β i,+1 1/0 2/0 0/1 Vilnius (15 models) Concurrent β i,0 0/0 0/0 0/1 Lag β i,–1 3/0 4/1 3/0 Lead β i,+1 0/0 2/0 3/0 Tallinn (12 models) Concurrent β i,0 0/0 0/1 0/1 Lag β i,–1 0/1 0/0 0/0 Lead β i,+1 1/0 2/0 1/0 Riga (7 models) Concurrent β i,0 1/0 1/0 1/0 Lag β i,–1 0/0 0/0 2/0 Lead β i,+1 0/1 0/1 0/0 Notation as in Table 4. Source: author’s own calculations with the use of STATA 14. Discussion and conclusions The purpose ofthis comparative study was toassess market-wide commonality inliquidity onsix emerging Central and Eastern European stock exchanges, inthe Czech Republic, Hungary, Slovakia, Lithuania, Estonia, and Latvia. The modified version ofthe Amihud proxy, the percentage relative spread bid/ask, and the Corwin-Schultz 106 Joanna Olbryś high-low two-day spread estimator were utilized asdaily liquidity/illiquidity measures for stocks. The OLS regression with the HAC covariance matrix estimation and the GARCH-type models were employed toinfer the patterns ofintra-market commonality inliquidity onthe investigated exchanges. According tothe literature, positive and statistically significant slope coefficients inthe estimated models are especially desired, asthey indicate co-movements inliquidity inthe same direction, and therefore confirm commonality inliquidity. Ingeneral, the estimation results provide noevidence ofco-movements inliquidity onthe CEE stock exchanges because positive and statistically significant coefficients rarely appear, regardless ofthe choice ofthe liquidity estimate. The empirical findings are somewhat homogeneous for all investigated markets. Therefore, noreason has been found toreject the research hypothesis that there isno commonality inliquidity onthe CEE stock markets, taken separately. This isperhaps the most significant finding ofour research. The results are novel and generally consistent with the literature concerning other emerging markets inthe world but are incontrast toprevious studies ofdeveloped markets. The findings fill the gap inthe literature ofcommonality inliquidity onemerging and frontier markets, and therefore, our study contributes tothe body ofknowledge inthat respect. Moreover, this paper proposes attributing the absence ofcommonality inliquidity onthe small CEE stock exchanges mainly tothe non-trading problem. Itis worth noting that commonality inliquidity may depend onthe structure ofthe stock market, and itis less pronounced inorder-driven markets than for dealer orhybrid markets because quote-driven orhybrid systems offer aform ofliquidity supplier. The results ofthis research have important practical implications and may beuseful indecision-making processes. From apractical point ofview, the problem iscrucial because the absence ofcommonality inliquidity influences investment strategies, portfolio management and risk diversification, domestic and international asset pricing, etc. Moreover, empirical findings concerning liquidity co-movements would help regulators and policymakers inimproving stock market design. Undoubtedly, alow level ofcommonality inliquidity has some advantages because itreduces the susceptibility ofacountry’s financial system tothe drying upof liquidity across many securities during periods ofmarket stress and crisis (Karolyi etal.2012). The empirical results presented inthis study certainly cannot provide definitive conclusions asto commonality inliquidity onthe investigated markets. Selected liquidity proxies based ondaily data are utilized. According tothe literature, there are several existing liquidity measures, and different frequencies ofdata are used. Various proxies derived from intraday data are particularly useful and frequently employed inassessing commonality inliquidity (e.g., Pukthuanthong-Le, Visaltanachoti 2009; Narayan etal.2015; Olbryś 2019a). However, high-frequency data are not freely available for the analyzed CEE stock exchanges, and this isthe main limitation ofthe study. 107 No Commonality in Liquidity on Small Emerging Markets… Apossible direction for further investigation could beto identify components ofliquidity onthe CEE stock markets taken separately, applying methods based onprincipal component analysis. Tothe best ofthe author’s knowledge, nosuch research has been undertaken thus far. Another important direction for further research could beacomparative investigation ofcommonality inliquidity onthe same six small CEE stock exchanges before and after the COVID-19 pandemic. Inall likelihood, the less liquid emerging stock markets will beamong the most affected bythe worldwide recession. Many firms will have serious problems surviving the COVID-19 pandemic period, and itis possible that the number ofcompanies listed onsmall stock exchanges will substantially change. However, the non-trading problem will increase. References Amihud, Y.(2002), Illiquidity and Stock Returns: Cross-Section and Time-Series Effects, “Journal ofFinancial Markets”, 5(1), https://doi.org/10.1016/S1386-4181(01)00024-6 Bai,M., Qin, Y.(2015), Commonality inLiquidity inEmerging Markets: Another Supply-Side Explanation, “International Review ofEconomics & Finance”, 39, https:// doi.org/10.1016/j.iref.2015.06.005 Bekaert,G., Harvey, C.R., Lundblad, C.(2007), Liquidity and Expected Returns: Lessons from Emerging Markets, “Review ofFinancial Studies”, 20(6), https://doi.org /10.1093/rfs/hhm030 Będowska-Sójka, B.(2019), Commonality inLiquidity Measures. The Evidence from the Polish Stock Market, “Hradec Economic Days”, 9(1). Brockman,P., Chung, D.Y., Perignon, C.(2009), Commonality inLiquidity: AGlobal Perspective, “Journal ofFinancial and Quantitative Analysis”, 44(4), https://doi.org /10.1017/S0022109009990123 Chordia,T., Roll,R., Subrahmanyam, A.(2000), Commonality inLiquidity, “Journal ofFinancial Economics”, 56(1), https://doi.org/10.1016/S0304-405X(99)00057-4 Corwin, S.A., Schultz, P.(2012), ASimply Way toEstimate Bid-Ask Spreads from Daily High and Low Prices, “Journal ofFinance”, 67(2), https://doi.org/10.1111/j.1540 -6261.2012.01729.x Dickey, D.A., Fuller, W.A.(1981), Likelihood Ratio Statistics for Autoregressive Time Series with aUnit Root, “Econometrica”, 49(4), https://doi.org/10.2307/1912517 Dimson, E.(1979), Risk Measurement when Shares are Subject toInfrequent Trading, “Journal ofFinancial Economics”, 7, https://doi.org/10.1016/0304-405X(79)90013-8 Elliott,G., Rothenberg, T.J., Stock, J.H.(1996), Efficient Tests for anAutoregressive Unit Root, “Econometrica”, 64(4), https://doi.org/10.2307/2171846 Engle, R.F.(1982), Autoregressive Conditional Heteroscedasticity with Estimates ofthe Variance ofUnited Kingdom Inflations, “Econometrica”, 50, https://doi.org/10.23 07/1912773 Fong, K.Y.L., Holden, C.W., Trzcinka, C.(2017), What are the Best Liquidity Proxies for Global Research?, “Review ofFinance”, 21, https://doi.org/10.1093/rof/rfx003 108 Joanna Olbryś Kamara,A., Lou,X., Sadka, R.(2008), The Divergence ofLiquidity Commonality inthe Cross-Section ofStocks, “Journal ofFinancial Economics”, 89(3), https://doi.org/10 .1016/j.jfineco.2007.10.004 Kang,W., Zhang, H.(2013), Limit Order Book and Commonality inLiquidity, “Financial Review”, 48(1), https://doi.org/10.1111/j.1540-6288.2012.00348.x Karolyi, G.A., Lee, K.-H., van Dijk, M.A.(2012), Understanding Commonality inLiquidity Around the World, “Journal ofFinancial Economics”, 105(1), https://doi.org /10.1016/j.jfineco.2011.12.008 Kearney, C.(2012), Emerging Markets Research: Trends, Issues, and Future Directions, “Emerging Markets Review”, 13(2), https://doi.org/10.1016/j.ememar.2012.01.003 Kiviaho,J., Nikkinen,J., Piljak,V., Rothovius, T.(2014), The Comovement Dynamics ofEuropean Frontier Stock Markets, “European Financial Management”, 20(3), https://doi.org/10.1111/j.1468-036X.2012.00646.x Korajczyk,R., Sadka, R.(2008), Pricing the Commonality Across Alternative Measures ofLiquidity, “Journal ofFinancial Economics”, 87(1), https://doi.org/10.1016/j.jfine co.2006.12.003 Narayan, P.K., Zhang,Z., Zheng, X.(2015), Some Hypotheses onCommonality inLiquidity: New Evidence from the Chinese Stock Market, “Emerging Markets Finance & Trade”, 51, https://doi.org/10.1080/1540496X.2015.1061799 Newey, W.K., West, K.D.(1987), ASimple, Positive Semi-Define, Heteroskesticity and Autocorrelation Consistent Covariance Matrix, “Econometrica”, 55(3), https://doi .org/10.2307/1913610 Olbryś, J.(2014), IsIlliquidity Risk Priced? The Case ofthe Polish Medium-Size Emerging Stock Market, “Bank iKredyt”, 45(6). Olbryś, J.(2018), The Non-Trading Problem inAssessing Commonality inLiquidity onEmerging Stock Markets, “Dynamic Econometric Models”, 18, https://doi.org /10.12775/DEM.2018.004 Olbryś, J.(2019a), Intra-Market Commonality inLiquidity. New Evidence from the Polish Stock Exchange, “Equilibrium. Quarterly Journal ofEconomics and Economic Policy”, 14(2), https://doi.org/10.24136/eq.2019.012 Olbryś, J.(2019b), Market-Wide Commonality inLiquidity onthe CEE-3 Emerging Stock Markets, [in:] K.Jajuga, H.Locarek-Junge, L.T.Orlowski, K.Staehr (eds), Contemporary Trends and Challenges inFinance, Springer Nature Switzerland AG, https:// doi.org/10.1007/978-3-030-15581-0_13 Olbryś, J.(2020), Market Tightness onthe CEE Emerging Stock Exchanges inthe Context ofthe Non-Trading Problem, [in:] N. Tsounis, A. Vlachvei (eds), Advances in Cross-Section Data Methods in Applied Economic Research. Springer Nature Switzerland AG, https://doi.org/10.1007/978-3-030-38253-7_36 Olbryś,J., Mursztyn, M.(2018), OnSome Characteristics ofLiquidity Proxy Time Series. Evidence from the Polish Stock Market, [in:] N. Tsounis, A. Vlachvei (eds), Advances inTime Series Data Methods inApplied Economic Research, Springer Nature Switzerland AG, https://doi.org/10.1007/978-3-030-02194-8_13 Pukthuanthong-Le,K., Visaltanachoti, N.(2009), Commonality inLiquidity: Evidence from the Stock Exchange ofThailand, “Pacific-Basin Finance Journal”, 17(1), https:// doi.org/10.1016/j.pacfin.2007.12.004 109 No Commonality in Liquidity on Small Emerging Markets… Sensoy, A.(2016), Commonality inLiquidity: Effects ofMonetary Policy and Macroeconomic Announcements, “Finance Research Letters”, 16, https://doi.org/10.1016 /j.frl.2015.10.021 Syamala,R., Wadhwa,K., Goyal, A.(2017), Determinants ofCommonality inLiquidity: Evidence from anOrder-Driven Emerging Market, “North American Journal ofEconomics and Finance”, 42, https://doi.org/10.1016/j.najef.2017.07.003 Wang, J.(2013), Liquidity Commonality Among Asian Equity Markets, “Pacific-Basin Finance Journal”, 21(1), https://doi.org/10.1016/j.pacfin.2012.05.001 Streszczenie Brak wspólności w płynności na małych rozwijających się rynkach giełdowych? Wyniki dla giełd Europy Środkowo-Wschodniej Celem pracy było badanie komparatywne tzw. wspólności w płynności (commonality in liquidity) na sześciu małych giełdach Europy Środkowo-Wschodniej. Analizowane rynki to: Czechy, Węgry, Słowacja, Litwa, Estonia i Łotwa. Wykorzystano trzy miary płynności/niepłynności aktywów kapitałowych, aproksymowane na podstawie danych dziennych. Próba objęła okres 5 lat, od stycznia 2012 do grudnia 2016. Do oszacowania modeli płynności zastosowano metodę estymatorów odpornych HAC oraz modele typu GARCH (w przypadku wystąpienia efektu ARCH w procesach resztowych). Dodatkowo przeprowadzono analizę stabilności wyników w czasie za pomocą procedury ruchomego okna. Wyniki empiryczne nie ujawniły wyraźnych wzorców w płynności na badanych rynkach oraz okazały się bardzo zbliżone na wszystkich giełdach, analizowanych oddzielnie. Na tej podstawie stwierdzono brak podstaw do odrzucenia hipotezy badawczej o braku wspólności w płynności na każdym z rynków. Badanie wypełnia lukę literaturową dotyczącą płynności na małych giełdach Europy Środkowo-Wschodniej, ponieważ żadne z wcześniejszych opracowań nie analizowało w sposób kompleksowy całej grupy wymienionych rynków. Słowa kluczowe: Europa Środkowo-Wschodnia, wspólność w płynności, GARCH, HAC, ruchome okno czasowe, dane dzienne