EXCHANGE-TRADED FUNDS INVESTING IN THE EUROPEAN EMERGING MARKETS

We examine ETFs investing in the equity of emerging European countries. Our sample contains 364 ETFs in developed Europe and 11 emerging European ETFs from 2005 to 2019. Compared to developed Europe’s ETFs, the emerging European equity ETFs are significantly smaller and younger with significantly higher fees. The low correlation of their returns with developed countries and lack of flow sensitivity to the US market volatility suggests that they may be underutilized means of international diversification by investors from developed countries.


INTRODUCTION
Exchange-traded funds (ETFs) registered tremendous growth in the last decade. The key drivers of their success are cheap diversification, transparency, intra-day liquidity, and easy access to international markets. In 2020, the ETFs assets under management worldwide totaled approximately 7.74 trillion U.S. dollars (Worldwide ETF assets). Many ETFs concentrate on the equity of emerging economies, making these markets accessible to investors from developed countries. According to the U.S. Portfolio Holdings of Foreign Securities report by the Department of the Treasury, Federal Reserve Bank of New York, U.S. investors held 922 billion U.S. dollars of common stock in emerging markets in 2019. Although these holdings are increasing every year, they still represent only 4 percent of emerging market equity.
Many arguments have been made for the benefits of international portfolio diversification (Grubel, 1968, Bailey and Stulz, 1990, Divecha et al., 1992, DeFusco et al., 1996, Gilmore and McManus, 2002. Investors in developed countries may reduce their portfolio risk, especially by diversification into emerging markets because of their moderate correlations and absence of longterm co-movements with equity returns in developed markets. At the same time, investors may be reluctant to invest in these economies because of the risks associated with the instability of their markets. Many emerging economies, especially those in Latin America and Asia, tend to experience political instability, currency risks, and other macroeconomic issues. In contrast, other emerging economies such as those in Central and Eastern European tend to be more politically and economically stable. These post-communist countries have undergone extensive liberalization after the fall of the Soviet Union and have created functioning public equity markets.
In this paper, we examine ETFs investing in emerging European equity. Our objective is to provide a description of these funds and investigate what factors affect flows into them. We find that ETFs investing in emerging European equity are significantly smaller and younger than funds investing in developed Europe. They tend to have a significantly smaller number of holdings and significantly higher fees. In contrast to developed Europe, ETFs investing in emerging European equity predominantly use swap-based synthetic replication techniques. In fact, more than 60 percent of the Emerging European ETFs use synthetic replication compared to only 15 percent of funds investing in developed European equity.
Our paper is related to several streams of literature. First, we contribute to the extensive literature on international diversification. While developed markets are becoming more cointegrated, cointegration between developed and emerging markets is generally lower, providing an opportunity for larger risk-reduction through diversification (Campbell and Hamao, 1992, Arshanapalli and Doukas, 1993, Tsetsekos, 1997, Kanas, 1998, and Bracker and Koch, 1999. During the 1990s, the economies of Central and Eastern European countries underwent economic liberalization characterized by marketization, privatization, and integration into the global markets (Gal and Schmidt, 2017). In 2004 and 2007, several Central and Eastern European nations -the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, Slovenia, Bulgaria, and Romania -became members of the E.U. Participation in the E.U. helps increase these regions' attraction for foreign investment and also equity market participation. Kalotay (2010) documented essential changes to the Central and Eastern European economies thanks to foreign investments after becoming E.U. member states. In addition, the emerging European stock markets also benefit from the increasing demand from E.U. and U.S. investors who seek international investment opportunities. Prats and Sandoval (2020) documented that stock market development after E.U. accession contributes to the economic growth in Central and Eastern European nations.
However, becoming E.U. member states and integrating into the international markets can make the emerging European markets more vulnerable to developed market conditions, harming the benefits from diversification. Using daily stock market data, Syriopoulos (2006), Aslanidis, and Savva (2011) documented increasing stock market integration between Central and Eastern Europe and the developed euro area. Syllignakis and Kouretas (2011) found a significant increase in dynamic correlation between the U.S. and German stock markets and Central and Eastern European stock markets, especially during the 2007-2009 crisis. More recently, Horvath and Petrovski (2013) and Gjika and Horvath (2013) confirmed the increasing comovements between Central European and Western European stock markets.
On the other hand, the previous literature shows that emerging European equity can still offer diversification benefits to developed Europe and U.S. investors. Gilmore et al. (2005) found that investment into three Central European equity markets (the Czech Republic, Hungary, and Poland) provided significant diversification benefits for U.S. and German investors. In the same spirit, Egert and Kocenda (2011) noted a trivial intraday correlation between equity market returns of three developed Europe (France, Germany, and the United Kingdom) and two emerging (the Czech Republic and Poland) European stock markets. However, the author notes that the diversification opportunity may change over time as emerging European countries integrate towards the eurozone. Finally, Avdulaj and Barunik (2013) used the intraday highfrequency price data of the Czech and German stock markets from 2008 to 2013 and concluded that there are still time-varying international diversification benefits. Consistent with the second strand of literature, we find that the correlation between the equities of emerging European and developed countries is smaller than the correlation between equities of other emerging and developed countries. European emerging equity ETFs thus may still serve as a vehicle for investors from developed countries to enter these markets and improve their diversification. However, the correlation between emerging European equity and developed countries is the highest during the 2005-2009 crisis period, similar to the findings of Syllignakis and Kouretas (2011).
Second, our paper relates to the literature on the consequences of ETFs' growth for the financial markets of emerging economies. Converse et al. (2020) showed from a sample of 33,019 mutual funds and 6,431 ETFs that returns in countries where ETFs hold a large share of the equity were significantly more sensitive to global financial ETFs and 31 emerging Europe ETFs, covering around 30% of the fund universe. We further require that ETFs have complete data for all other variables (TNA, monthly net returns, month-end closing price, month-end NAV, yearly expense ratio, the monthly number of stock holdings, and primary prospectus benchmark monthly returns). We also exclude actively managed ETFs in developed Europe since no actively managed ETFs are investing in emerging European equity. Our final sample consists of 375 equity ETFs: 364 ETFs in developed and 11 emerging European countries. Our research sample covers, on average, 89% and 62% of total net assets of ETFs investing in Developed Europe and Emerging Europe, respectively. Therefore, even though we do not have complete data of all ETFs that invest in the markets of interest, our sample is a good representative sample of the universe of ETFs in developed and emerging Europe.     This replication method is used primarily by large funds and funds tracking indices with a smaller number of constituents. Partial replication refers to replication that uses only a sample of securities of the underlying index. Although this method results www.ieeca.org/journal 264 in higher tracking errors, it decreases costs, especially for smaller funds or funds trading in markets with higher transaction costs.   Not surprisingly, correlations among developed markets are higher than correlations among developed and emerging markets. However, correlations of developed markets with emerging European markets are lower (0.62) than their correlations with other emerging markets, especially since 2015. The lowest correlation is between the U.S. markets and Emerging Europe; in the time period 2015 to 2019, this correlation is only 0.53. This low correlation between equity returns of the U.S. and emerging European markets suggests diversification benefits between these markets. As a result, U.S. investors may benefit more from international Diversification to Emerging European markets than from Diversification to other emerging markets.

TRACKING ERRORS
This section examines whether tracking errors of emerging European equity ETFs are significantly different from tracking errors of developed European ETFs. Tracking errors depict the deviation of fund returns from returns of the underlying benchmark. Our descriptive statistics document significantly higher tracking errors for emerging European ETFs. Previous literature, however, suggests that tracking errors are related to fund characteristics, such as fund size, age, expense ratio, fund return volatility, and the number of assets in the fund's holdings (Vardharaj, Fabozzi, and Jones, 2004). To address this issue, we create a dummy variable Emerging that takes a value of one if the ETF invests in emerging European equity. Then we regress the tracking error on this dummy variable and other fund characteristics: Where: T.E. is a tracking error, Log(Age), Log(TNA), Expense, Volatility, and Holdings are lagged control variables depicting the age, size (as total net assets), volatility of the benchmark Log(Age), Log (TNA), Expense, Volatility, and Holdings are lagged control variables depicting the size, age, volatility of the benchmark returns, and the number of stocks in the Fund's holdings. The regressions include month fixed effects, and standard errors are clustered at the fund level.
The results of the regression are shown in Table   5, columns 1 and 2. The dummy variable Emerging is not significantly different from zero, suggesting that after controlling for Fund's characteristics, tracking errors of European emerging equity ETFs are not significantly different from tracking errors of developed Europe ETFs. Larger and younger ETFs tend to have smaller tracking errors; the number of holdings does not have a significant effect on tracking error.
Because ETFs in emerging Europe tend to use synthetic replication, we create a dummy variable Synthetic replication that takes the value of one for funds that use synthetic replication. The coefficient on this dummy variable is negative, suggesting that synthetic replication lowers tracking error, but the effect is not significant ( Table 5, column 3).

DETERMINANTS OF FLOWS
This section examines the determinants of flows to ETFs investing in emerging European equity and their counterparts investing in developed Europe. We are especially interested in whether previous returns attract investors into emerging European equity. Following previous research that documents the dependence of Fund flows on fund characteristics and previous fund flows (Clifford, Fulkerson, and Jordan, 2014), we estimate the following regression: Where Flow is the monthly ETF net flow. Tracking error and fund volatility are based on a 24-month rolling standard deviation of monthly benchmark adjusted returns, monthly fund returns, and monthly benchmark returns. Return is the Fund's monthly net return. All variables are lagged by one month. The regressions include month fixed effects, and standard errors are clustered at the fund level.
The results are presented in Table 6. Flows into the ETFs investing in developed Europe are significantly higher for funds with higher previous returns. Surprisingly, this is not true for ETFs investing in emerging Europe, whose flows are not significantly related to previous returns.
Looking at other variables, in the developed Europe category, ETFs that are smaller, younger, and those with lower fees and higher volatility enjoy significantly higher flows. For emerging European ETFs, the only variables that have a significant effect on flows are size and tracking errors. Within this category, the smaller ETFs and ETFs with lower tracking errors attract higher flows.
Emerging markets ETFs represent a low-cost vehicle for foreign investments to enter local markets. Filippou et al. (2019) showed that the flows into emerging market ETFs depend on economic shocks in developed countries, particularly the U.S. As the U.S. investors react to the domestic economic conditions, they also adjust their foreign investments, including those in emerging markets. Changes in net flows to emerging market ETFs then contribute to the transmission of U.S. shocks into those markets, limiting the diversification benefits of emerging market strategies.
To examine if shocks in developed markets affect flows into ETFs of emerging European countries, we add market volatility in the U.S. (rolling 24month standard deviation of S&P 500 Index return) and developed Europe (rolling 24-month standard deviation of FTSE Developed Europe All Cap Index return) to our regression. Results are reported in Table 6, columns 5 to 8.
We find that net flows into equity ETFs in developed Europe are negatively affected by market volatility in the U.S. However, surprisingly, market volatilities in the U.S. and developed Europe do not affect the equity flows into European emerging equity ETFs; the finding is contrary to Filippou et al. (2019) for other emerging markets.

CONCLUSIONS
The emergence of international markets ETFs in the last two decades has significantly improved investors' ability to diversify globally, including diversification into emerging markets. While most earlier studies focus on the emerging markets of Latin American, Asian, and Indonesian countries, markets of emerging European countries remain understudied. In this paper, we investigate ETFs that invest in emerging European equity and examine their tracking errors and factors that affect flows into these ETFs.
We discovered that emerging European equity ETFs are significantly smaller and younger than ETFs of developed Europe. They utilize mainly synthetic replication, probably to avoid low liquidity in these markets. However, after adjusting for fund characteristics, their tracking errors are not significantly different from the tracking errors of their counterparts in developed Europe.
Emerging European equity ETFs may represent diversification benefits for investors from developed countries because the correlation of their returns with equity returns of developed countries is smaller than the correlation of other emerging countries. In addition, the net flows into these funds are not affected by the shocks in developed equity markets. This may be because of the low participation of investors from developed countries in these funds or the lower sensitivity of the underlying emerging European equity to these shocks. Further work should examine in more detail the diversification benefits of emerging European equity, the factors that affect flows into these ETFs, and why these ETFs are not experiencing growth similar to other emerging markets ETFs.