How ETFs Amplify the Global Financial Cycle in Emerging Markets

K.7 How ETFs Amplify the Global Financial Cycle in Emerging Markets Converse, Nathan, Eduardo Levy-Yeyati, and Tomas Williams Please cite paper as: Converse, Nathan, Eduardo Levy-Yeyati, and Tomas Williams (2020). How ETFs Amplify the Global Financial Cycle in Emerging Markets. International Finance Discussion Papers 1268. https://doi.org/10.17016/IFDP.2020.1268 International Finance Discussion Papers Board of Governors of the Federal Reserve System Number 1268 January 2020

Board of Governors of the Federal Reserve System International Finance Discussion Papers Number 1268 January 2020 How ETFs Amplify the Global Financial Cycle in Emerging Markets Nathan Converse, Eduardo Levy-Yeyati, and Tomas Williams NOTE: International Finance Discussion Papers (IFDPs) are preliminary materials circulated to stimulate discussion and critical comment. The analysis and conclusions set forth are those of the authors and do not indicate concurrence by other members of the research staff or the Board of Governors. References in publications to the International Finance Discussion Papers Series (other than acknowledgement) should be cleared with the author(s) to protect the tentative character of these papers. Recent IFDPs are available on the Web at www.federalreserve.gov/pubs/ifdp/. This paper can be downloaded without charge from the Social Science Research Network electronic library at www.ssrn.com.

How ETFs Amplify the Global Financial Cycle in Emerging Markets∗ Nathan Converse† Eduardo Levy-Yeyati ‡ Tomas Williams § Abstract This paper examines how the growth of exchange-traded funds (ETFs) has affected the sensitivity of international capital flows to global financial conditions. Using data on individual emerging market funds worldwide, we employ a novel identification strategy that controls for unobservable time-varying economic conditions at the investment destination. We find that the sensitivity of flows to global financial conditions for equity (bond) ETFs is 2.5 (2.25) times higher than for equity (bond) mutual funds. We then show that our findings have macroeconomic implications. In countries where ETFs hold a larger share of the equity market, total cross-border equity flows and returns are significantly more sensitive to global financial conditions. Our results imply that the increasing role of ETFs as a channel for international capital flows has amplified the global financial cycle in emerging markets. JEL Classification: F32, G11, G15, G23 Keywords: exchange-traded funds; mutual funds; global financial cycle; global risk; push and pull factors; capital flows; emerging markets ∗WewouldliketothankMichaelBauer(discussant),FernandoBroner,TitoCordella,JeffreyFrankel,LindaGoldberg, Maurizio Habib (discussant), Graciela Kaminsky, Stefan Nagel, Dongwon Lee (discussant) Lorenzo Pandolfi, Claudio Raddatz, Tara Sinclair, Jay Shambaugh, Bryan Stuart, and Daniel Beltran for their insightful comments. WealsoreceivedvaluablefeedbackfromseminarparticipantsattheCentralBankofIreland,FederalReserveBoard, and the GWU International Finance Seminar, as well as participants in the 3rd CEPR Symposium in Financial Economics, 7th Annual West Coast Workshop in International Finance, AEA/ASSA Annual Meeting, EMG-ECB Workshop on International Capital Flows, IMF Capital Flows Group, the 2018 Federal Reserve System Committee on International Economic Analysis, the European Economic Association Meetings, LACEA Annual Meetings, and the RIDGE International Macroeconomics Workshop for helpful comments and discussions. †Division of International Finance, Federal Reserve Board. Nathan.L.Converse@frb.gov ‡School of Government, Universidad Torcuato di Tella. ely@utdt.edu §Department of Economics, George Washington University. tomaswilliams@gwu.edu 1

1 Introduction Recent work has documented how changes in US financial conditions are transmitted to other countries in a so-called global financial cycle (Rey, 2015). While much of this research has focused on bank flows (Bruno and Shin, 2015b,a), changes in US monetary policy and risk appetite are also transmitted via portfolio flows (Forbes and Warnock, 2012; Fratzscher, 2012; Avdjiev et al., 2017). Atthesametime,itisclearthatexposuretotheglobalfinancialcyclevariesacrosscountries (Ceruttietal.,2017;Choietal.,2017)andovertime(AhmedandZlate,2014). Indeed,asshownby the green line in Figure 1, portfolio capital flows to emerging markets have become more sensitive to changes in global financial stress over the last 15 years. This increased sensitivity has coincided with another shift over the same period: the growing importance of exchange traded funds (ETFs) in the financial markets of emerging economies, illustrated by the blue line in Figure 1. Are these twotrendsrelated? Andifso, how? InthispaperweshowthatriseofETFsasaconduitforcapital flows to emerging markets has indeed amplified the transmission of global financial shocks to those economies, and present evidence suggesting that this is due to the particular pool of investors who favor ETFs. The growing importance of ETFs is not limited to emerging markets. Even as the mutual fund industry has expanded rapidly in recent years, accounting for US$20 trillion in assets worldwide (Khorana et al., 2005; ICI, 2017) and acting as an important channel for cross-border portfolio capital flows (Didier et al., 2013), the assets of ETFs have grown even faster. The share of fund assetsheldbyETFshasgonefromonly3.5percentin2005to14percentin2017(Figure2). Indeed, the rising popularity of ETFs has been one of the most notable developments in the fund industry over the past decade (Cremers et al., 2016). Nonetheless the rise of ETFs has been particularly striking for emerging markets (EM) funds, where the ETF asset share reached 20 percent in 2017.1 We explore the relationship between the growth of ETFs and the sensitivity of EM capital flows to global factors—also referred to as push factors—in two steps. First, we present robust evidence that fund-level investor flows to ETFs respond more to changes in global financial conditions than flows to traditional mutual funds. By contrast ETF flows respond less, if at all, to changing economic conditions in the countries in which the ETFs invest. Second, we show that where ETFs 1ETFs account for an even larger share (25 percent) of the assets of equity funds dedicated to investing in emerging markets. While the share of EM bond funds assets held by ETFs is much lower, it has been growing very rapidly. Prior to 2006 there were essentially no EM bond ETFs. 2

hold a larger share of the host country’s market capitalization, both aggregate portfolio flows and equity returns are more sensitive to global factors. These findings indicate that the rise of ETFs as a vehicle for international capital flows has amplified the effects of the global financial cycle in emerging markets. Our analysis uses comprehensive data from EPFR Global on monthly investor flows to mutual funds and ETFs over the period 1997 to 2017.2 The dataset contains more than 33,000 mutual fundsandmorethan6,000ETFs, withmorethanUS$29trillioninassetsundermanagementatthe end of June 2017. Beyond its extensive coverage, this database has several appealing features. In additiontodataonnetfundflowsandmonth-endassets,EPFRprovidesinformationoneachfund’s investment scope, indicating the country or set of countries where the fund invests. Importantly, the coverage of the dataset is sufficiently broad that the investment scope varies for ETFs and mutual funds, so that both categories include global, regional, and dedicated country funds. EPFR also provides information on each fund’s location, allowing us to control for domicile-specific push factors and thus focus on the effects of truly global factors. Ournovelempiricalapproachexploitsthesefeaturesofthedata. Weexaminehowinvestorflows into funds respond to global push factors and test whether the response differs for mutual funds and ETFs. Throughout the paper, we control for economic conditions in each fund’s investment destination in two different ways. First we use a specific variable: growth in industrial production, averaged across countries for multi-country funds.3 Second, we use investment-scope-time fixed effects to absorb any time-varying, investment-scope-specific factors that might affect fund flows. ThisallowsustocleanlyidentifyhowglobalfactorsdifferentiallyaffectflowsgoingintoETFsversus mutual funds. Consistent with previous research on fund flows, we find that increases in global financial stress are negatively related to investor flows into dedicated emerging market funds, both mutual funds and ETFs. However, we go on to show that the sensitivity of ETF flows to these push factors is significantly larger than that of flows into mutual funds, a fact previously undocumented in the literature.4 Quantitatively, the sensitivity of EM fund flows to push factors is almost 2.5 times 2Throughout the paper we use the term investor flows and fund flows interchangeably to refer to end investors’ purchases and redemptions of shares in mutual funds and ETFs. 3In robustness checks, we demonstrate that our findings arerobust to alternative ways of aggregating pull factors in multi-countryfunds. Wealsoexperimentwithotherpullfactorvariables,suchasdestinationcountryinterestrates. 4Importantly,becausewearedoingouranalysisatamonthlyfrequency,thisfindingisnotamechanicalresultofthe fact that ETFs are continuously traded while mutual funds are not. 3

bigger for equity ETFs, and 2.25 times larger for bond ETFs, than it is for mutual funds. In order to understand better why ETF flows exhibit greater sensitivity to global financial conditions, we also present evidence consistent with the view that ETFs appeal to a different clientele than traditional mutual funds. Specifically, our findings suggest ETF investors place particular value on liquidity and are relatively inattentive to local economic conditions in the countries where the funds invest. The key features that define ETFs are their low cost, their passive management, and their liquidity (ETFs are continuously traded while mutual funds are not). Becausesomemutualfundsofferlowfeesorarepassivelymanged, weareabletotestwhether these two characteristics are associated with greater sensitivity to global financial conditions. And we find that they are not. By process of elimination, our results therefore suggest that the liquidity of ETFs attracts investors who behave differently. At the same time we show that at the monthly frequency, mutual fund investors respond to changes in economic conditions in the countries where these funds invest, but ETF investors do not, indicating that they are relatively inattentive. Having used fund-level data to clearly identify the greater sensitivity of ETF flows to global financial conditions, we then show that our findings have economically significant implications at the aggregate level. Analyzing a panel of 43 emerging markets, we regress total portfolio equity flows from abroad on our measure of global financial stress, allowing the coefficient to vary with the share of each country’s equity market held by foreign ETFs. We find that in countries where ETFs hold a larger share of the equity market, aggregate portfolio equity inflows are more sensitive to global financial conditions. We repeat the exercise using aggregate equity market returns as the dependent variable and find similar results. Quantitatively, a one- standard-deviation increase in the share of local equity held by ETFs is associated with an a response of portfolio equity inflows that is 2.9 times higher. For stock prices, a similar increase in the ETF share is associated with an exposure to global factors almost 1.2 times larger. It follows that while ETFs may attract new investors to the EM asset class, the benefits of a broader investor base for EM issuers may be partially offset by the fact that the greater sensitivity of ETF flows deepens exposure to the global financial cycle, raising the volatility of financing conditions in recipient economies.5 Throughout the paper we explicitly address concerns about endogeneity and omitted variable bias in order to ensure that the results do in fact reflect a causal effect of ETFs on the sensitivity 5See Converse (2018) for a detailed exploration of the negative effects of capital flow volatility on the real economy in emerging markets. 4

of capital flows to global financial conditions. Endogeneity is a concern in our analysis because financial institutions may create ETFs to cater to investors seeking exposure to volatile or highbeta markets. To deal with this, we include investment scope-time fixed effects in our fund-level analysis, so that we are effectively comparing ETF flows with the flows into mutual funds which have the same investment destination. This ensures that our fund-level results are not driven by ETFs tending to invest in more volatile markets. In our country-level regressions, we tackle the potential endogeneity of our ETF share variable in two ways. First, we construct a narrower measure the ETF share which excludes the single-country ETFs that may have been created to provide access to high-beta markets. And second, we use a proxy for the ETF share which is not subject to concerns about endogeneity. In both cases, our core result survives: ETF participation is significantly related to the sensitivity of EM capital flows and equity prices to global shocks. In our aggregate-level regressions, omitted variable bias is a also concern because ETFs are likely to own a larger share of the equity market in economies which are more financially integrate, and financial integration is in turn associated with greater sensitivity to global financial shocks. We therefore re-run our aggregate-level regressions with an exhaustive set of de facto and de jure financial integration measures, and interact those integration measures with our measure of global financial stress. The estimate of the coefficient measuring the relationship between ETF participation and sensitivity to global shocks remains virtually unchanged when we do this, indicating that our country-level results are not driven by omitted variable bias. Ourpaperrelatestofourstrandsoftheliterature. Inadditiontothealreadymentionedbodyof workontheglobalfinancialcycle,wecontributetothelargeliteratureonthedriversofcapitalflows to emerging markets (Ahmed and Zlate, 2014) and the relative importance of global push factors and local pull factors (Forbes and Warnock, 2012; Cerutti et al., 2015), in particular work using mutual fund data to explore the issue (Fratzscher, 2012). In this context, Jotikasthira et al. (2012) also study withdrawals and redemptions by end investors and how they affect the transmission of shocksacrosscountriesbutdonotdifferentiatebetweentypesoffundsaswedo.6 WhereasRaddatz and Schmukler (2012) and Miyajima and Shim (2014) study whether the portfolio decisions of fund managers differ from those of end investors, we analyze the differences in the behavior of end investors in two different types of funds—ETFs and traditional mutual funds.7 In a closely related 6IntheirpaperJotikasthiraetal.(2012)buildempiricalevidenceattheinternationallevelbasedonalargeliterature both theoretical (Shleifer and Vishny (1997)) and empirical (Coval and Stafford (2007)) on asset fire sales. 7In one of the first papers making use of mutual fund data, Borensztein and Gelos (2003) compare capital flows via 5

paper, Brandao-Marques et al. (2015) do compare the sensitivity of ETFs and mutual funds in the EPFR data, but study flows to individual countries rather than fund flows, so that they capture the combined responses of fund managers and end-investors. Additionally, Brandao-Marques et al. (2015) restrict their analysis to fund-level data while we also provide evidence on aggregate macro financial variables such as capital inflows and country asset prices.8 Third, our paper relates to the rapidly growing literature studying the consequences of the growth of ETFs for financial markets and economic activity. Broadly speaking, our paper contributes that literature, nearly all of which has focused on US financial markets, by examining the consequences of ETFs for emerging markets. To our knowledge, only two previous papers analyze the effects of ETFs in an international context. Baltussen et al. (2019) show that ETF ownership is associated with greater negative serial correlation in returns, a phenomenon closely related to the volatility we study in this paper. Analyzing 41 funds provided by iShares, Filippou et al. (2019) show that US investors’ demand for foreign country ETFs comoves significantly with the VIX, but is uncorrelated with analogous implied volatility measures in destination countries. We confirm this result using data from a larger sample of funds, but focus our attention on differences in the behavior of ETF investors relative to mutual fund investors, rather than differences in ETF investors’ sensitivity to global as opposed to local shocks. OurfindingscomplementintwowaysworkanalyzingtheeffectsofETFsonUSequitymarkets, which has showed that ETF ownership increases both comovement (Da and Shive, 2017; Israeli et al., 2017) and volatility (Ben-David et al., 2018). First, our results confirm that equity flows and stock prices are more volatile in international markets with greater ETF ownership. Second, we highlightaspecificmechanism throughwhichETFsboostvolatilityandcomovement: byincreasing sensitivity to global financial shocks. In addition, the evidence we present that ETFs attract investors seeking liquid assets and who are inattentive to local economic conditions in the funds’ investment destination is consistent with the findings of several papers analyzing the role of ETFs US corporate bond markets (Dannhauser, 2017; Dannhauser and Hoseinzade, 2017; Nam, 2017). AndtheamplificationmechanismsmodeledinBhattacharyaandO’Hara(2017)mayhelptoexplain the greater sensitivity of ETF flows that we identify. open-ended funds with those via closed-ended funds 8More broadly, this study is related to a large literature studying international mutual funds and how these institutional investors affect international financial markets and asset prices. See among others Kaminsky et al. (2004); Gelos and Wei (2005); Broner et al. (2006); Gelos (2011); Shek et al. (2015); Forbes et al. (2016). 6

Fourth and finally, our paper relates to the literature on the drivers of investor flows into managed funds (for a survey see Christoffersen et al., 2014), which has explored in depth the relationshipbetweenfundflowsandperformance. Wetakeonboardtheinsightsfromthisliterature by controlling for the past performance of funds in our main specifications, but study how another set of variables—global financial conditions and local economic conditions in the countries where the funds invest— affect flows to different types of funds. Therestofthepaperisstructuredasfollows. Section2presentsinformationontheinstitutional details and the mechanics of ETFs. Section 3 details the data we use. In Section 4, we outline our empirical strategy, present our main results concerning the sensitivity of fund flows to global factors, and then assess whether something other than ETFs can explain our findings. Section 5 analyzestheaggregateimplications, particularlythelinkbetweenETFparticipationandtheglobal financial cycle. Section 6 concludes. 2 ETFs and Institutional Details This section presents a brief description of the structure and functioning of exchange traded funds (ETFs), focusing on the ways in which they differ from traditional mutual funds.9 Like a mutual fund, an ETF is an investment vehicle which owns a basket of underlying assets, usually stocks or bonds.10 Often the basket is constructed to track the performance of a particular index. Although actively managed ETFs do exist, they are rare– of more than 700 ETFs in our dataset which focus on emerging markets, only 7 are actively managed. When open-ended mutual fund investors buy or sell shares, they enter into a transaction with the fund, and the price at which the transaction happens is determined by the fund’s net asset value (NAV) at the end of the trading day on which the buy or sell request is made. By contrast, ETF shares are continuously traded on equity exchanges, allowing investors to buy or sell shares at any time at the current market price. In this sense ETFs are like closed-end mutual funds, which also have exchange-traded shares. The continuous trading of ETF shares not only makes them easy 9This section is informed by the concise and insightful institutional detail in Ben-David et al. (2018) and Da and Shive (2017), as well as the comprehensive chapter by Deville (2008). 10TherearealsoothertypesofETFs,forexample,commodityETFs. BecausetheEPFRdatacontainonlyequityand bondETFs, here we limitour discussion to these ETF types. In markets outsidethe U.S., there arealso synthetic ETFswhichreplicatetheperformanceofadesignatedbasketofsecuritiesthroughthetradingofderivatives. While flows in and out of synthetic ETFs do not directly generate capital flows, they nonetheless affect asset prices. 7

for investors to buy and sell at low cost, but also greatly reduces the need for the fund to hold a cash allocation to satisfy redemptions, eliminating the cash drag that is an implicit cost mutual fund investing. Whereas closed-end mutual funds have a fixed number of shares, set at the fund’s IPO, ETF shares can be created or redeemed. Indeed, the creation and redemption of ETF shares ensures that the value of the ETF’s shares outstanding closely tracks the basket of underlying assets. The ETF has a number of so-called authorized participants (APs), large financial institutions that can create or redeem shares in the fund. To create new ETF shares, an AP buys up the underlying assets and exchanges them for fund shares. When an AP redeems shares, it returns shares to the fund administrators and receives the corresponding quantity of underlying assets. If the value of ETF shares differs from the value of the underlying basket, there is an arbitrage opportunity for the fund’s APs. For example, when an ETF’s outstanding shares are more valuable than the underlying, an AP can buy up the underlying, exchange it for fund shares, then sell the fund shares at a profit. These sales will cause the price of the ETF shares to fall until the ETF and the underlying are equal in value. Of course, if the underlying assets are relatively illiquid, there is scope for the price of the ETF to diverge from the underlying since arbitrage will not always be possible. Importantly, although shares in the emerging market ETFs in our sample are generally traded on exchanges in developed markets, the creation and redemption process nonetheless means that investor flows into these funds generate cross border capital flows. For example, should end investors’purchasesofETFsharespushtheirpriceabovethatoftheunderlyingasset,theETF’sAPs will buy the underlying assets and redeem them to make an arbitrage profit. Because the APs are generally large financial institutions in countries with developed financial markets, their purchase of the underlying asset represents a foreign purchase of an emerging market asset, classified as a gross portfolio capital inflow in the balance of payments. 8

3 Data 3.1 Fund Flows Data We obtain monthly fund-level data on mutual funds and ETFs from the commercial data provider EPFR Global.11 The dataset includes both equity and bond funds, with the data on equity funds covering the period January 1997 to August 2017 and the bond fund data running from January 2002 to August 2017. The data are an unbalanced panel with funds both entering and leaving the sample, so that the data do not suffer from survivorship bias. The full EPFR database contains 33,019 mutual funds (of which roughly 65 percent are equity funds) and 6,431 ETFs (of which 80 percent are equity funds). At the end of June 2017, EPFR funds held US$26.4 trillion in assets undermanagement,accountingforapproximately66percentofthetotalworldwideassetsofmutual funds and ETFs.12 Official data on US holdings of foreign assets show that US-domiciled mutual funds held around US$1.7 trillion in emerging market assets, and US funds tracked by EPFR hold roughly 50 percent of these (TIC, 2017).13 Our primary variable of interest is investor flows (F ), defined as the US dollar value of the net it purchases or redemptions of shares in fund i in month t.14 Throughout our analysis, we normalize flows into each fund by its assets under management at the end of the previous month (A ) so it−1 (cid:16) (cid:17) that our measure of fund flows is f = Fit . Importantly, the dataset includes a field classifying it Ait−1 each fund according to what we refer to as its investment scope, meaning the country or group of countries where the fund invests. Example of multi-country investment scope categories include “Global Emerging Markets” and “Latin America Regional.” See Appendix Table A2 for a list of the investment scope categories in the dataset and how many funds and observations are assigned to each. In addition, EPFR also provides data on each fund’s performance, meaning the month-onmonth percent change in the fund’s net asset value (NAV). Throughout our analysis, we control for 11For detailed variable definitions and sources see Table A1. 12According to ICI (2017) the total assets of the fund industry are roughly US$40 trillion. 13HerewecomparetheholdingsofUS-domiciledfundswithUSdataonoverseasholdingsbecausemostcountriesdo not yet report the institutional sector of asset holders. 14WeusethefundflowsvariablegeneratedbyEPFR,whichiscalculatedbysubtractingthechangeinthefund’snet assetvalue(NAV)fromthechangeinthefund’stotalassets: F =(A −A )−A ×(%∆NAV ). Forfunds it it it−1 it−1 t that report assets and returns in currencies other than the US dollar EPFR converts flows using the average spot exchangerateforthemonth,sothatF =F∗×ER ,whereF∗ isflowsmeasuredinthefund’sreportingcurrency it it it it and ER is the average exchange rate for the month (measured as fund’s reporting currency per US dollar). it 9

the lagged performance of each fund relative to the average performance of funds with the same investment scope. EPFR also provides a host of other fund characteristics which we use in our analysis, such as each fund’s domicile and it’s declared benchmark. We clean the EPFR dataset using procedures standard in research using fund-level data, dropping funds with less than one year of data and funds with average assets lower than US$10 million. In addition, we drop funds with extreme values of performance and inflows (measured as a share of lagged assets), specifically funds with observations in the top and bottom one percent for these variables. Because our analysis is focused on the role of mutual funds and ETFs in international capital flows, we exclude from the dataset domestic funds, which investing only in the country in which they are domiciled. We also exclude funds domiciled in a country that is included in the fund’s investment scope (e.g. a Latin America regional fund domiciled in Brazil). See Table A3 for the number of funds and observations in each domicile in our cleaned dataset. This procedure leaves us with 12,852 mutual funds and 2,525 ETFs in our dataset. Table 1 presents summary statistics and provides a first glimpse of our main result. The volatility of fund flows normalized by assets is much larger for ETFs than for mutual funds.15 The greater volatility of ETF investor flows can be seen even more clearly in Figure 3, where we plot the aggregate fund flowsnormalizedbyaggregateinitialassetsforthetwotypesoffunds. Evenaftertheglobalfinancial crisis, fund flows for ETFs appear to be much more volatile and less persistent than investor flows for mutual funds. 3.2 Additional Variables We analyze the drivers of fund flows using data on pull and push factors. Our main measure of global push factors is the St. Louis Fed Financial Stress Index, which is the first principal component of 18 mostly US financial variables including interest rates, spreads, and equity and bond market implied volatility. Putting changes in the financial stress index in context, the index jumped by 1.5 standard deviations following the September 11, 2001 terrorist attacks and during the 2013 Taper Tantrum. During the 2011 peak of the Eurozone crisis and after the 2015 surprise devaluation of the Chinese currency, the index increased by roughly two standard deviations. In robustness checks, we use a variety of other commonly used measures of risk sentiment 15Table A4 contains summary statistics for the assets under management of funds. 10

and liquidity conditions. As indicators or risk, we employ the Chicago Board Options Exchange Market Volatility Index (VIX), the effective yield of the Bank of America Merrill Lynch US High Yield Master II Index (US HY), and the spread between 3-month LIBOR and 3-month Treasury Bill (TED spread). Following the literature, we also run our analysis using the effective federal funds rate (FF Rate) to measure global financial conditions. Since the US policy rate was at the zero lower bound for a substantial portion of our sample period, we also make use of the shadow federal funds rate developed by Wu and Xia (2016) (FF Shadow Rate). With the exception of the shadow fed funds rate, which is made available by the Atlanta Fed, our risk and monetary policy variables were obtained from the Federal Reserve Economic Data (FRED) system at the end of each month.16 Our analysis also takes into account push factors specific to each fund’s home country. Specifically, we use monthly stock market returns measured in dollars from MSCI for the domicile country reported by EPFR. For funds domiciled in financial centers, we assign the major stock market most closely associated with the financial center as its home market.17 To capture pull factors for fund investors we use the month-on-month change in countryspecific seasonally adjusted industrial production (IP) indexes from the IMF’s International Financial Statistics (IFS) database.18 For multi-country funds, we construct investment scope-level aggregate pull factors by taking the cross-country median value for IP growth for the countries within the fund’s scope.19 Our results are not sensitive to the method used to aggregate across countries in each investment scope; using the mean value of IP growth or taking a weighted average produced quantitatively similar results. In robustness checks, we also include monthly one year ahead forecasts of short-term interest rates in the economies included in each fund scope, obtained from Consensus Economics. We avoid using market interest rates or equity returns as push factors because of the potential for reverse causality, as these variables are themselves affected by fund flows. 16For summary statistics on these global factors see Table A5. 17Funds domiciled in Ireland, the British Virgin Islands, and the Channel Islands were matched with UK stock market returns. Funds domiciled in other Caribbean financial centers were matched with US stock returns. Funds domiciled in Luxembourg were assigned German equity returns. 18IP data were seasonally adjusted using the X12-ARIMA method developed by the U.S. Census Bureau. For summary statistics on IP growth see Table A6. 19Funds to which EPFR has assigned the same investment scope classification may invest in a slightly different set of countries (e.g. not all EM Asia funds invest in Taiwan). In constructing our aggregates, we use the set of countrieswhichMSCIassignstoeachcountrygroupeachperiod. Asaresult,thesetofcountriesincludedineach category varies over time. For example, we include Greece in “Emerging Europe” after November 2013, when it was downgraded from MSCI’s developed markets index. 11

4 Empirical Strategy and Results 4.1 Empirical Strategy The dependent variable in our fund-level regressions is investor flows into each fund, rather than flows to individual countries. We thus avoid constructing estimates of capital flows at the fundcountry-time level, which has been common in the literature but introduces measurement error.20 We model fund flows as a function of global factors, local factors, and lagged fund returns. More specifically, we use the following baseline specification: 3 X f = θ +βGF +γ(GF ∗ETF )+λLF +η(LF ∗ETF )+ δ R +ε (1) it i t t i it it i k it−k it k=1 where f is investor flows into fund i during month t, normalized by the fund’s assets at the start it of month t. The variable GF (“Global Factor”) is a measure of global financial conditions, LF t it (“LocalFactor”)capturespullfactorsinthefund’sinvestmentdestination, ETF isadummyequal i to one if the fund is an ETF, and ε is an error term.21 This baseline specification includes fixed it effectsatthefundlevelθ . Sincealargebodyofworkhasshownthatpastperformanceaffectsfund i flows, we include three lags of the fund’s returns relative to other funds with the same investment scope (R ). it Throughout the paper we try to keep the specification parsimonious and therefore include generally only one pull and one push factor in each regression. For GF our main variable is the St. t Louis Fed Financial Stress Index, a broad measure of global financial conditions.22 Fund flows f it represent an adjustment in end investors’ holdings of fund i, which could be due reallocation across funds or to a change in the size of the portfolio of investors who hold fund i. We therefore include the global factor variable in differences, so that β represents the change in investors’ holdings of fund i in response to a change in global financial conditions at time t.23 The sum β +γ captures the sensitivity of ETF investor flows to push factors, and the main parameter of interest to us is 20Measurement error occurs because of the need to approximate each fund’s country-level returns using a publicly available price index. 21TheETFdummydoesnotentertheregressiononitsownbecauseitisnottimevarying. Fundsdonotswitchfrom being a mutual fund to being an ETF, or vice versa. 22In Section 4.4 we show that using narrower various narrower measures of global risk sentiment and liquidity conditions does not alter our results. 23More specifically we take log differences of the St Louis Fed Financial Stress Index. 12

γ, the difference in sensitivity between ETF flows and mutual fund flows. The focus of this paper is the difference in the responses of ETF and mutual fund investor flows to global financial shocks, but we do include a local factor in our regression and allow its coefficient to differ for ETFs for two reasons. First it allows us to verify that our results are in line with other research on the drivers of fund flows. And second, knowing η—the differential response of ETF investor flows to local factors—may help us better understand our results regarding γ. Our main measure of local factors, often referred to as pull factors in the literature on the drivers of capital flows, is month-on-month growth in industrial production (as described in Section 3). We use IP because measures of local returns (equity returns or interest rates) would raise serious concerns about endogeneity bias, since large fund flows can generate price changes (as documented in, for example,Jotikasthiraetal.,2012).24 AsdiscussedinSection3ourmainspecificationsmeasurepull factors using the median industrial production growth for the group of countries included in the fund’s investment scope, but our results are robust to using either the simple or the GDP-weighted mean of IP growth. The response of ETF investor flows to pull factors is given by λ+η. Beyondthisbaselinespecification,weuseanalternativeapproachexploitinghigherdimensional fixed effects as follows: 3 X f = θ +θ +γ(GF ∗ETF )+η(LF ∗ETF )+ δ R +e , (2) it i st t i it i k it−k it k=1 where θ are fixed effects at the investment scope-time level. This set of fixed effects absorbs all st time-varying shocks non-parametrically at the investment scope level. Thus, we can more cleanly identify the difference in sensitivities coming from the difference in the type of fund. For instance, if financial institutions create ETFs to service country or regions with higher sensitivity to push factors, this would generate a high γ in Equation 1 even if ETF flows per se were not more sensitive. The use of scope-time fixed effects addresses this concern because it allows us to compare the sensitivities of ETFs and mutual funds with the same investment scope, controlling for any time-varying factors specific to the investment scope.25 24While fund flows can also affect the cost of capital in the countries where the fund invests, in turn affecting real investment and thus industrial production, this effect is unlikely to work within a single month. 25In principle the structure of our database allows us to use fixed effects at finer levels, such as the fund domicileinvestment scope-time level, or the benchmark-time level. However, especially for bond funds, there are too few ETFs within these more granular sub-categories. We therefore favor the investment scope-time fixed effects for most of the paper. 13

4.2 Main Results We begin by estimating equation (1) for the dedicated emerging market funds in our sample (Table 2).26 Consistent with previous work, the results show that an increase in global financial stress is associated with a reduction in investor flows to both EM equity (Column 1) and EM bond funds (Column 4). Columns 2 and 5 present this paper’s main result: flows to dedicated EM ETFs are significantly more sensitive to the global push factor than EM mutual fund flows, for both equity and bond funds.27 Indeed, ETF flows’ exposure to our global factor is almost 2.5 times bigger for equity funds and 2.25 times larger for bond funds.28 Toensurethatourmainparameterofinterest, γ inequations1and2, iswellidentified, wenext estimate equation 2, which includes investment scope-time fixed effects. Including this set of fixed effects allows us to compare ETFs with mutual funds that have the same investment scope and also control for any time-varying determinants specific to that investment scope. Again, this strategy helpsuscontrolforthefactthatfinancialinstitutionsmaychoosetocreateETFsspecificallytocater to investment scope categories which, for other reasons, exhibit more volatility in fund flows. The resulting point estimates for γ (found in columns 3 and 5) are somewhat smaller in magnitude than those in column 2, which is consistent with endogenous ETF creation generating an upward bias in our estimates of equation 2. Nonetheless, the coefficients on the global factor-ETF interaction term do not change dramatically and remain significant. Our baseline regression results imply that the greater sensitivity of ETF flows has an economicallysignificanteffectonthesizeofflowstodedicatedemergingmarketfunds. Toillustratethis, we analyze the 2013 Taper Tantrum in light of our results. Following Fed Chairman Ben Bernanke’s May 22, 2013 comments regarding the possibility of scaling back the Federal Reserve’s asset purchase program, investors withdrew US$38 billion (2.8 percent of fund assets) from dedicated EM 26While all our regressions contain fund performance controls, we do not report the estimated coefficients for compactness. Full results including our estimates for δ in equations 1 and 2 are presented in Table A7. it−k 27In Appendix Table A8 we show that for developed market (DM) funds the sensitivity of ETF flows to both push andpullfactorsisnotsignificantlydifferentfromthatoftraditionalmutualfunds. Wetheninvestigatethebehavior of flows to developed market ETFs in detail in appendix table A9 and find that DM ETF flows do appear more sensitivetoglobalfinancialconditionsthanDMmutualfundflowsoncewemodifyourdatasetintwoways. First, we re-include funds investing in the country where they are domiciled. We do this because in developed markets thesefundscatertoforeignaswellasdomesticinvestors,unlikeinEMswheretheirinvestorbaseislargelydomestic. Second, we exclude DM funds investing exclusively in German, Japanese, and U.S. government bonds, which are widelyconsideredsafe-havenassets. However,sincethefocusofthispaperisonflowstodedicatedemergingmarket funds we leave further exploration of the behavior of investors in developed market funds to future work. 28This is calculated as β+γ where the numerator is the sensitivity of ETFs flows to the global factor, while the β denominator is the sensitivity of mutual fund flows to the global factor. 14

funds in June 2013, of which US$11.5 billion came out of ETFs. Concurrently, the St. Louis Fed Financial Stress index that is our main measure of global financial conditions increased by 1.5 standard deviations. The coefficient estimates in column 5 of Table 2 imply that approximately US$26.3 billion (1.9 percent of assets) of the total outflow can be attributed to the increase in financial stress that followed Bernanke’s speech. If we impose on ETF flows the same sensitivity thatweestimateformutualfundflows, theoutflowduetofinancialstresswouldhavebeenUS$20.8 billion(1.5percentoffundassets). Thisbackoftheenvelopecalculationthusimpliesthattheextra sensitivity of ETF flows boosted outflows by US$5.5 billion (0.5 percent of fund assets), meaning that outflows would have been roughly 15 percent smaller if ETF flows had the same sensitivity as mutual fund flows. Thus our results suggest that the extra sensitivity of ETF flows is economically important, but at the same time do not imply implausibly large effects. The difference in sensitivity to global financial shocks between ETF and mutual fund flows that we have found implies that the growth of ETFs as a conduit for international capital flows has contributed to the strengthening of the relationship between global financial conditions and portfolio flows depicted in Figure 1. However, we also find evidence investor flows to ETFs have become more sensitive to our global push factor over time. Figure 4 plots the 36-month rolling slope of a regression of aggregate fund flows on our chosen measure of global financial conditions. Except for a brief period after the 2008 global financial crisis the sensitivity of ETF flows to push factors is greater (in absolute terms) than for traditional mutual funds. Moreover, the sensitivity of investorflowsintoETFshasbeenincreasingsteadilysince2012whilethesensitivityofmutualfund flows has essentially remained constant over the period. This suggests that the rising sensitivity of aggregate flows to dedicated EM funds that we highlighted in the introduction is due not only to the growing use of ETFs as a channel for cross-border investment but also to the increase in the sensitivity of ETF flows. The focus of our analysis is on how investor flows to ETFs and mutual funds respond differently to global shocks; however, the results presented in Table 2 also reveal differences in how flows to the two types of funds comove with what are often called pull factors in the capital flows literature: economicconditionsinthecountrieswherethefundsinvest. Incolumn2ofTable2weseethatflows to dedicated EM mutual funds are positively and significantly related to economic conditions in the funds’ investment destination. For bond funds, the coefficient is of a similar magnitude but not statistically significant, perhaps because our data contain many fewer observations for bond funds. 15

The coefficient on the interaction of the local factor and the ETF dummy (η in equation 1) is never significant, so we cannot reject the null that ETF and mutual fund flows respond to local factors in the same way. At the same time, it is noteworthy that the sum of the local factor coefficient (again, λ) and the coefficient on its interaction with the ETF dummy (η) is never statistically different from zero. This can be seen in the the row labeled “Local Factor ETF” in the bottom section of Table 2, which gives the ETF-specific coefficient λ+η and the row below, which gives the p-value from a test of the null that the sum is equal to zero. Because we calculate pull factors for multi-country funds by averaging across the countries in the funds’ investment scope, there is a concern that lack of a statistically significant response by ETF investors could be the result of attrition bias generated by measurement error in our local factor. But in Table A10 we confirm thatETFflowsdonotrespondtothelocalfactorevenwhenwelimitoursampletocountry-specific funds, for which such measurement error is not a concern. It therefore appears that while ETF investors respond strongly to changes in global financial conditions, they respond little if at all to changes in local economic conditions in the particular countries where the ETF invests.29 This somewhat surprising finding is in fact consistent with the hypothesis put forward by Nam (2017) and Dannhauser (2017) that ETFs attract investors who are relatively uninformed about the fundamentals of the assets they trade. Moreover, our results regarding ETF investors’ insensitivity to destination-specific shocks offers a potential explanation for the finding of Israeli et al. (2017) that US stock prices respond less to firm-specific information about future earnings when ETFs hold a larger share of the stock. 4.3 Alternative Hypotheses We now explore several alternative explanations for our main finding that ETF flows respond more to changes in global financial conditions. First, we consider the possibility that our global shock measureiscorrelatedwithfinancialconditionsinsomecountrieswhicharealsohometomoreETFs. Second, we show that our results are not merely a reflection of ETFs having grown in popularity faster in countries where fund investors are more responsive to shocks. Third, we assess whether ourfindingssimplyreflectdifferentspeedsofportfolioadjustmentbytestingwhethermutualfunds’ 29WealsotestwhetherflowstomutualfundsandETFsresponddifferentlytolaggedfundperformanceandfindthey donot(theseresultsareavailableonrequest). Onceagain,theinclusionoftheseadditionalinteractiontermsdoes not substantially change our coefficient estimates for global factors. 16

sensitivity to global financial shocks is the same as those of ETFs at longer time horizons. Having shown that these alternative hypotheses cannot explain our findings, we explore which specific features of ETFs are associated with greater sensitivity to global conditions. The first alternative explanation for our findings concerns, on the one, hand the correlation between global financial shocks and country-specific financial shocks in countries where ETF investors reside and, on the other hand, the fact that the popularity of ETFs is different across different domicile countries. If our measure of global shocks is more highly correlated with countryspecific shocks in places where ETFs are more popular, this could explain our result that ETF investor flows are more sensitive to our measure of global shocks. To determine whether this is the case, we now run a set of regressions in which we control for push factors that are specific to the country where each fund is domiciled. In particular, we add to our baseline specification the stock market returns in each fund’s domicile country in order to capture financial conditions at home for the fund’s investors. The resulting coefficient estimates are presented Table 3. Higher stock market returns at home are associated with larger investor flows to EM funds (Column 1), presumably reflecting portfolio rebalancing to maintain a desired weight on EM assets as the size of investors’ total portfolios grows. Despite its importance, the inclusion of this variable does not alter our main conclusions. ETF flows are still significantly more exposed to the global factor than mutual fund flows (Columns 1 and 2). Furthermore, the effect is now larger for both equity (Panel A) and bond funds (Panel B). In column 3, we introduce domicile-investment scope-time fixed effects to our estimation. In doing so, we go beyond just controlling for stock market performance and control for all time-varying unobservables that might be affecting fund flows at the fund domicile or investment scope level. For example, one might think that central bank policy rates at the fund domicile might play a role on top of stock market returns. This specification controls for such factors, and the results remain very similar to our baseline. Nonetheless, stock market returns in the domicile of the fund do seems to be an important explanatory variable, and we therefore include this variable as a fund-level control in the rest of the estimations presented in this section. Thesecondalternativeexplanationforourmainfindinghastodowithdifferencesinthegrowth of ETFs relative to traditional mutual funds across domicile countries. If ETFs have simply grown more quickly in domicile countries where investor flows tend to be more sensitive to global financial shocks, this could explain why we find ETF flows are more sensitive. We therefore run a set of regression where we control for the fact that investor flows to ETFs exhibit different long-term 17

trends than flows to mutual funds. As noted in the introduction, the share of fund assets held by ETFshasrisensteadilyoverthelast15years. ThisgrowthislargelyduetoETFsreceivingsteadily growing inflows during the period. To ensure that our results are not an artifact of the upward trend in ETF flows over the period we analyze, we run a set of regressions that include fixed effects at the fund domicile-year-ETF level. In other words, we interact a set of dummy variables for the funds’ country of domicile with our ETF dummy and a full set of year fixed effects. We include the domicile fixed effects in this interaction to account for the fact that the rise in the popularity of ETFs has been more pronounced in some countries than in others. The results of this regression (colums4and5ofTable3)alsoverifythatourresultsarenotdrivenby, forexample, U.S.investors being more sensitive to global shocks and also more eager to shift to using ETFs as an investment vehicle. The point estimates for coefficient on the global factor-ETF dummy interaction (gamma) is smaller than our baseline estimate, but still implies that ETF flows are more than twice as sensitive to global financial conditions than are mutual fund flows. Third, we consider the possibility that mutual fund investors are in fact equally as sensitive to global financial shocks as ETF investors, but simply respond to such shocks more slowly. To test this hypothesis, in Table 4 we re-estimate our baseline specifications at quarterly frequency (columns 1 and 2) and again at the 6-month frequency (columns 3 and 4). Once again, we confirm our baseline result that ETF fund flows are significantly more sensitive to push factors than flows to mutual funds. Indeed, the excess sensitivity of ETF flows appears even larger for bond funds than in the monthly regressions. Thus, our baseline results do not merely reflect differences in the timing of ETFs’ versus mutual funds responses to global shocks. Rather, the lower frequency analysis in Table 4 provides further evidence that country capital flows channeled via ETFs are much more sensitive to push factors than regular mutual funds. With respect to pull factors, equity ETFs flows do behave differently at a lower frequency. Whereas flows into equity ETFs were not sensitive to local conditions in our monthly frequency analysis, at the quarterly and 6-month frequency equity ETF flows are positively associated with IP growth in the investment destination, as is the case for mutual funds (in Table 4 this can be seen by looking at the p-values associated the “local factor ETF” coefficient). By contrast, investor flows to bond ETFs remain uncorrelated with pull factors over longer horizons,. So it appears that equity ETF investors are not totally inattentive to economic conditions in the funds’ investment destinations, but rather acquire such information or act on it only with a lag. This 18

finding regarding the timing of equity ETF investors’ reaction to local shocks is consistent with Giglio et al. (2019), who present evidence that changes in retail investors’ beliefs do not affect the likelihood that they adjust their portfolios, but that when portfolio adjustment occurs, it does reflect changes in beliefs. Once again, the focus of this paper is differences in fund flows’ sensitivity to global factors; nonetheless, our results regarding funds’ sensitivity to local factors do help us paint a fuller picture of how ETF investors differ from traditional mutual fund investors. Having ruled out three alternative explanations for our findings, we now determine whether the greater sensitivity that we have identified is related to fund characteristics which in turn are positivelycorrelatedwithETFstatus,butwhichcanalsobefeaturesofmutualfunds. Inparticular, we examine how fund size, passive investment strategy, and multi-country scope are related to the sensitivity of fund flows to changes in global financial conditions. We find that none of these fund characteristics can explain the excess sensitivity result we obtained for ETF flows in Table 2, suggesting that it is some other feature of ETFs—perhaps the liquidity they offer—that draws investors who behave differently from mutual fund investors. The average emerging market ETF in our sample is around 50 percent larger than the average mutual fund, so we test whether it is in fact large funds that are more sensitive to changes in global financial conditions (Table 5, columns 1 and 2). We allow the coefficient on the global factor to vary according to the size of the mutual fund by interacting the global factor with a dummy variable equal to one if the mutual fund is large, defined as having more than $250 million in assets.30 Large equity mutual funds do not seem to have a significantly higher sensitivity to global factors (Table 5, Panel A). While flows into large bond mutual funds do have a higher exposure to our measure of global push factors than flows to small bond mutual funds, the large funds are nonetheless significantly less sensitive than ETFs (Panel B). Because larger mutual funds tend to have lower fees, our results regarding large funds also suggest that it is not the relatively low cost of investing via ETFs that motivates investors to behave differently Another key characteristic of ETFs is their passive management strategy.31 We therefore next examine whether passively managed mutual funds are more sensitive to changes in global financial conditionsthanareactivelymanagedmutualfunds. Theresultsincolumns3and4ofTable5show 30In Table A11 we experiment with alternative thresholds for what constitutes a large fund, and obtain very similar results. 31While active ETFs do exist, there are very few. Our dataset includes more than 500 ETFs investing in emerging markets, of which only seven are actively managed. 19

that passive equity and bond mutual funds are not significantly different from other mutual funds, and further that the change in specification does not alter the estimated coefficients on the global factor-ETF interaction. We therefore conclude that it is not ETFs’ passive management that sets them apart from mutual funds in terms of their sensitivity to global shocks. Becausecountry-specificETFsaremuchlesscommonthancountry-specificmutualfunds, there is a concern that our results may reflect differences in the sensitivity of flows to multi-country (global and regional) funds relative to that of single-country funds, rather than any feature specific to ETFs. This is a particularly important concern given that global and regional funds may cater to less specialized, possibly less sophisticated, investors who are more sensitive to changes global financial conditions. However, these concerns are dispelled in columns 5 and 6 of Table 5, where we test whether the sensitivity of flows to country-specific mutual finds differs significantly from that of multi-country funds.32 We fail to reject the null that the sensitivity of flows to single-country mutualfundsdiffersfromthatofflowstomulti-countrymutualfunds,andalsofailtorejectthenull that flows to county-specific ETFs have a different sensitivity than flows to multi-country ETFs. Overall then the results presented in the final two columns of Table 5 demonstrate that the greater sensitivity of ETF flows to global financial conditions which we have identified is not merely a reflection of the prevalence of multi-country ETFs, but rather some other feature of this type of fund. Wehavenowruledoutanumberofcompetingexplanationsforourbaselineresults. Neitherthe size, nor the passive management strategy, nor the multi-country investment scope of ETFs explain their greater sensitivity to changes in global financial conditions relative to traditional mutual funds. Taken together, the findings in this section also suggest that it is not the case that ETFs’ low fees explain their greater sensitivity. This is because the lowest cost mutual funds are large, passively managed, or both, and we have confirmed that these characteristics on their own do not induce higher responsiveness to global shocks. By process of elimination, our results suggest that the distinctive characteristic of ETFs which generates excess sensitvity is the enhanced liquidity they provide, due to the fact that ETF shares can be traded intra-day while mutual fund shares 32In Table A10 we also estimate equations 1 and 2 separately for, on the one hand, global and regional funds and on the other hand country funds. For equity funds, results for the two groups are qualitatively and quantitatively similar(PanelA).Inthecaseofbondfunds,wecannotrejectthenullthatflowstocountry-specificETFshavethe same sensitivity to global financial conditions as country-specific mutual funds, but this is likely due to the small number of country-specific bond funds in our sample. Our dataset contains 98 country-specific EM bond funds, of which only eight are ETFs. 20

cannot.33 This result is consistent with various pieces of evidence presented by Ben-David et al. (2018) that US equity ETFs attract clientele who trade frequently and thus value liquidity more highly than other investors. 4.4 Robustness In this section, we demonstrate the robustness of the results reported above along three different dimensions: varying the time period covered by our analysis, using different measures of global financial shocks, and using different measures of local pull factors. We discuss the results of each exercise here but present the full estimation results in the Appendix (Tables A12 through A15) to conserve space. We begin our robustness checks by considering only the period after the global financial crisis of 2007/2008. We do this to address the concern that our results reflect that fact that ETFs captured substantial market share only after the financial crisis, a period when sensitivity of fund flows to global financial conditions was high for reasons unrelated to the existence of ETFs. This is a reasonable worry, given that pre-crisis, there are periods in our sample with few or no ETFs. By restricting our sample to the post-crisis period, we ensure that a substantial number of ETFs as well as mutual funds are present in each month. And indeed, the results in Table A12 are both qualitatively and quantitatively similar to our baseline estimation. Thus our results are not a reflection of ETFs happening to be come popular at a time when sensitivity to global financial conditions was generally high. Next, we exclude the period of the global financial crisis of 2007/2008, dropping the months between March 2007 until March 2009 from our estimations (Table A12). We do this because crisis periods tend to disproportionately increase the cross-country correlation of financial variables (Forbes and Rigobon, 2002). However the results in Table A12 confirm that when we exclude the global financial crisis, the sensitivity of fund flows to ETFs is still significantly larger than for mutual funds, both for equity and bond funds. Similarly, our finding that investor flows to ETFs are not significantly associated with local pull factors remains once we remove the crisis period from our sample. 33At the same time, note that because we are studying flows at a monthly frequency, the greater sensitivity we encounter is not a mechanical result of the fact that ETF shares can be traded intra-day. 21

Third, weshowthatourresultsarerobusttousingdifferentmeasuresofglobalfinancialshocks. We begin by using three common measures of global risk appetite: the VIX, the TED spread, and the US high yield spread (Table A13). For equity funds, our findings are very similar to the baseline: ETF flows always exhibit higher sensitivity to the global variable than mutual fund flows, while ETFs flows response to local factors is not significantly different from zero. Flows to EM bond funds do not exhibit elevated sensitivity to the VIX or the TED spread, but do respond significantly more than EM mutual fund flows to changes in the US high yield spread. In the first two columns of Table A14 we verify that when we measure global financial conditions using the first principal component of these variables (PCA1) to measure global push factors, the results are very similar to our baseline for both equity and bond funds. In columns 3 to 6 of Table A14 we capture global financial shocks using two different measures of U.S. monetary policy. Because the U.S. policy rate was at the zero lower bound for a substantial part of the period weanalyze, we not only re-run our specification using the fed funds rate, but also with the so-called shadow fed funds rate developed by Wu and Xia (2016). Here again we see that ETFs fund flows respond much more strongly to changes in financial conditions than do mutual fund flows. All in all, our findings are robust to the use of different variables to capture changes in global financial conditions. Fourth, we demonstrate that our core results do not change when we use alternative measures of local economic conditions—the local factor in specifications 1 and 2. Recall that because we are analyzing flows to funds, many of which invest in multiple countries, we need to average any measure of local conditions across the countries within the fund’s investment scope. In columns (1) and (2) of Table A15 we average using the unweighted mean of month-on-month growth across countries, while in columns 3 and 4 we take the GDP-weighted average. The results are very similar to those in Table 2. In columns 5 to 8 of Table A15 we include a measure of local financial conditions, the Consensus forecast for the country’s short-term interest rate in the following year. Whether we include this variable on its own (columns 5 and 6) or alongside our measure of real conditions (columns 7 and 8), our finding that ETF flows are more sensitive to changes in global financial conditions remains. Moreover, the response of ETF flows to changes in local economic conditions is never significantly different from zero. 22

5 Aggregate Implications: ETFs and the Global Financial Cycle in Emerging Markets Having presented evidence that investor flows into dedicated emerging market ETFs are more sensitive to changes in global conditions than flows into EM mutual funds, in this section we ask whether this greater sensitivity affects countries’ exposure to the global financial cycle at the aggregate level. After all, ETFs account for less that half of EM mutual fund assets, and mutual funds are only a subset of cross-border investors. We address this question in two steps. First, we present graphical evidence suggesting that ETFs have boosted the sensitivity of aggregate EM fund flows to global financial conditions in the period since the global financial crisis. Second, we provide a quantitative assessment of this enhanced sensitivity for capital flows and equity prices at the country level.34 We begin our discussion of the aggregate implications of our fund-level findings by looking again at Figure 4, which plots the 36-month rolling slope coefficient for aggregate flows into all EM funds (the green line), for aggregate EM mutual fund flows (the red line), and for total EM ETF flows (the blue dashed line) with respect to our measure of global financial stress. The sensitivity of flows to both mutual funds and ETFs, and thus total flows, spiked during the financial crisis, fell back to its pre-crisis value relatively quickly, and jumped again during the 2011 Eurozone crisis. Since then, the sensitivity of traditional mutual funds flows to global financial conditions has trended back towards its pre-crisis average. But the sensitivity of traditional mutual funds flows to global financial conditions has not reverted to its pre-crisis level. Rather, it has remained more than 30 percent higher than it’s pre-crisis average. Thus, Figure 4 demonstrates that the growing importance of ETFs in the fund industry combined with the rise in ETF flows’ sensitivity over the last several years has resulted in fund flows overall becoming more sensitive to changes in global financial conditions. To formally explore the macro-level implications of our fund-level results, we construct a measure of ETFs’ market penetration in each country, defined as the share of the country’s equity 34Throughout this section we focus on portfolio equity flows and equity prices. We do this because both portfolio capital flows and bonds prices are much more diverse and more difficult to aggregate. For instance, portfolio debt liability flows in the balance of payments include purchases of both sovereign and corporate securities, both of whichmaybedenominatedineitherdomesticorforeigncurrency. Accordingly,thereareseparatepriceindexesfor sovereign and corporate debt in domestic and foreign currency. We therefore restrict our analysis to the aggregate implications for equity. 23

market capitalization held by ETFs: P w A ETF Share = i∈ETF ict it (3) ct Mcap ct where w is the share of ETF i’s assets invested in country c at time t, and A is the ETF’s total ict it assets under management measured in U.S. dollars. In using share of outstanding held by ETFs to measure their importance, we follow previous work analyzing the effects of ETF ownership on the behavior of US stock returns (Glosten et al., 2016; Israeli et al., 2017; Ben-David et al., 2018). Both w and A are obtained from EPFR. The numerator thus captures the dollar value of ETFs’ ict it assets in country c at time t, while the denominator is the stock market capitalization of country c (Mcap , also measured in U.S. dollars). ct We test whether capital flows and asset prices are more exposed to global factors in countries with a greater ETF presence using the following specification: y = α +βGF +µ(GF ∗Share ETF )+δ(Share ETF )+θ +ν (4) ct c t t ct−1 ct−1 t ct where y is the aggregate variable of interest, either quarterly portfolio equity liability flows from ct the balance of payments (normalized by GDP) or monthly MSCI country stock market returns. The global factor GF is defined as before. We lag the ETF share variable one period to avoid t reverse causality, since large capital inflows in period t could mechanically boost the ETF share for the same period. We also include a set of country fixed effects (α ) and in some specification add c time fixed effects (θ ) as well. In equation 4, µ captures how the sensitivity of capital flows and t prices to global factors varies with the presence of ETFs.35 The results of our macro-level regressions, presented in Table 6, suggest that a greater ETF share is associated with a higher aggregate exposure to global financial shocks for both equity flows (Panel A) and stock returns (Panel B). We first confirm that, as one would expect, portfolio equity inflows and local equity returns are negatively related to increases the global financial stress index that we use to measure global financial conditions (column 1). Then in column 2 we interact our measure of global conditions with the ETF share in order to test whether the greater sensitivity we found at the fund level generates macro-level effects. The negative and significant coefficient on the 35Summary statistics for the relevant macro variables are presented in Table A16. 24

interactoin term indicates that the association between global shocks and equity flows and returns is indeed larger (in absolute value) when the ETF share of the local equity market is greater. The result holds even when we add time fixed effects, which strip out among other things any time trend in the ETF share (Column 3). The coefficient estimates in columns 2 and 3 thus constitute the core result of our macro-level regressions, implying that our findings at the micro level have implications for aggregate financial variables. How large is the effect? With the ETF share of equity assets at its mean (ETF Share = 0.43 ct percentagepoints),thecountry’sinflowsbetawithrespecttotheglobalfinancialconditionsis-0.27; for a country with an ETF share one standard deviation higher (ETF Share = 1.16 percentage ct points), this beta increases to -0.79, which implies an exposure 2.9 times higher. The conclusions are qualitatively similar when looking at aggregate stock market returns (Panel B). Increasing the ETF share by one standard deviation relative to the average ETF share, the beta associated with the global factor is 1.2 times higher. Thus the effects are economically as well as statistically significant. One potential concern about these estimates is that of omitted variable bias. For instance, greater financial integration may lead to an increase in both the ETF share and the equity market co-movement with global factors. Our regressions already include country fixed effects that could controlforthecross-countrydifferencesintheleveloffinancialintegration,whichhasbeenshownto varylittleovertime. Nonetheless, weperformtwoadditionalexercisesinordertodemonstratethat our core results are not driven by a correlation between the ETF share and financial integration. First, in columns 4 and 5 of Table 6, we verify that it is not holdings of equity by investment funds more generally that is associated with higher sensitivity to changes in global financial conditions. To do this, we include alongside the ETF share the share of assets held by mutual funds (Mutual Fund Share) and interact this variable with the global factor. If our results on the ETF share were merely reflecting the fact that ETFs own more stocks in countries that are more financially integrated, and that, in turn, more financially integrated countries are more sensitive, then we would expect to find the same result for the mutual fund share that we do for ETFs. But we do not—the interaction with the mutual fund share is not statistically significant, with a point estimate dramatically smaller than that for the ETF share interaction. Furthermore, when we add the variables capturing the mutual fund share, the coefficient on the ETF share interacted with the 25

global factor is not affected at all, suggesting that financial integration as measured by the mutual fund share is not a relevant omitted variable. The second exercise we perform in order to verify that our result are not driven by omitted variable bias is to explicitly control for countries’ general degree of international financial integration. To this end, in columns 5 and 7 of Table 7 we include as a control the ratio of each country’s gross financial assets and liabilities measured as a share of its GDP.36 This is a measure of de facto international financial integration widely used in the literature. In Table A17 we experiment with alternative measures of financial integration, including de jure and other defacto measures, such as equity and FDI de facto financial integration. We also interact this broad measure of financial integration with our global financial stress measure in order to assess whether it is in fact general international financial integration that renders capital flows to emerging markets more sensitive to global conditions.37 As one would expect, in the flows regressions in Panel A of Table 7 financial integrationispositivelyandsignificantlyassociatedwiththelevelofcapitalflows—moreintegrated countries tend to receive higher capital flows. At the same time, we find that the estimated coefficient on the interaction between the ETF share and global financial conditions changes very little when we control for general financial integration, both for portfolio equity flows (Panel A) and equity returns (Panel B). And this is true whether we include only country fixed effects (column 6) or country and time fixed effects (column 7). These results confirm that the relationship that we have found between, on the one hand, ETF ownership of local equities and, on the other hand, the sensitivity of capital flows and equity prices to global financial conditions is not simply driven by a correlation between ETF participation and international financial integration. Reverse causality is also a concern in our country-level regressions. Financial institutions do not set up new ETFs at random. Rather, they are usually created in response to demand from investors. For example, if there are investors who would like to quickly move in and out of risky assets,assetmanagerswilllikelysetupETFsthatprovideexposuretothoseassets. Weaddressthis concern about endogeneity in two ways. We first note that for endogenous ETF creation to drive our results, two very specific and arguably implausible conditions must hold. We then conduct a robustness test in which we examine a subset of cross-border ETF holdings which previous work 36We obtain data on cross-border assets and liabilities from (Lane and Milesi-Ferretti, 2018). 37We drop the mutual fund share and its interaction from these regressions in order to make clear that the lack of significance of the financial integration variable is not due to colinearity. However, results are basically unchanged if the mutual fund share variables are included. 26

suggests are exogenously determined. Under what conditions would reverse to causality rather than a causal relationship explain our finding that total portfolio equity flows and stock prices are more sensitive to global factors where ETFs hold a larger share of the market? First, it must be the case that the launch of an ETF in a volatile market does not attract new investors to that market. If the launch of ETFs investing in a volatile market does draw in new investors, the responsiveness of total portfolio equity capital flows (measured as a share of GDP) will increase, and the introduction of the ETF will have had a causal effect onthatresponsiveness. Second, for reversecausalityto explainour resultsit mustalso be the case that the ETF launch did not lead those investors who had exposure to that market to change their behavior by reacting more to global financial shocks. Both of these assumptions seem implausible since the appeal of ETFs relative to other investment vehicles is that they are low cost, and thus attract to new investors, and more liquid, which likely prompts a change the behavior of investors. Nonetheless, to confirm that the creation of ETFs to provide access to already volatile or highbeta markets does not drive our results, in Table 7 we focus the relationship between ETF holdings which are exogenously determined and sensitivity to global financial shocks. We isolate exogenous differences in the ETF share in two different ways. First, in columns 1 and 2 we construct an ETF sharemeasurewhichincludesonlytheemergingmarketassetsheldbyglobalandregionalETFs(as opposed to country specific funds). Because these funds’ holdings are diversified across countries, it is less likely that their holdings are endogenously determined by a desire to access high-beta emerging markets. More importantly, these funds’ holdings of any particular country’s stocks are largely determined by benchmark weights, as documented by Raddatz et al. (2017). As a result, we can be confident that the share of a given country’s market capitalization held by this subset of ETFs is not endogenously determined. When when redo our analysis using this exogenous measure of ETF holdings in Table 7, we find results very similar to those in Table 6. Once again, both flows and returns respond more to changes in global financial conditions in markets where ETFs own a larger share of the equity market capitalization. We further address concerns about heterogeneity bias in columns 3 and 4 of Table 7, where we replace the ETF share with a dummy variable (MSCI EM) equal to one for countries that 27

MSCI classifies as emerging markets and zero for countries MSCI deems frontier markets.38 Many ETFs track MSCI indexes, and more ETFs track MSCI’s emerging market index than its frontier index. Thus the degree to which ETFs own the local market is correlated with the country’s MSCI classification. However as demonstrated in Raddatz et al. (2017), the timing of changes in MSCI classification is largely exogenous. Thus our MSCI variable can be regarded as an arguably exogenousproxyfortheETFsharevariable, whichweinteractwiththeglobalfactorinaregression that also includes country fixed effects. Re-estimating equation 4 using the MSCI classification as a proxyforETFshare(again, column1of7), wefindthattheinteractiontermisonceagainnegative and significant. Since we include country fixed effects in the regression and focus on within-country variation, this implies that MSCI upgrades of a country are associated with an increase in the exposure of capital flows and equity returns to global financial conditions. Ourfindingsregardingthemacro-levelimplicationsofETFs’growingroleincross-bordercapital flows are summarized in Figure 5. We plot the relationship between, on the one hand, the crosssectional betas for portfolio equity inflows (the left panel) and stock market returns (the right panel) for the 2000-2017 period and, on the other hand, the average share of assets held by ETFs for a given country and period. Furthermore, we find that the inclusion (exclusion) from important benchmark indexes tracked by ETF investors raises (reduces) the country’s betas, even when we look exclusively at the cross-section of countries, which is again consistent with the hypothesis that ETFs amplify the incidence of global factors on local markets. 6 Conclusion Since the early 2000s, the asset management industry has undergone a significant change as the assets under management of ETFs have expanded rapidly. In this paper, we present evidence that the growing role of ETFs as a channel for cross-border capital flows has increased the exposure of emerging markets to the global financial cycle. We use detailed monthly micro data at the fund levelfrom1997until2017todocumentthatinvestorflowsintodedicatedemergingmarketETFsare moresensitivetoglobalpushfactorsthanflowsintoemergingmarketmutualfunds. Thisdifference is economically large, with betas to global factors almost 2.5 times bigger for equity ETFs, and 38Our sample also includes one country, Israel, that shifted from being classified as emerging to developed markets by MSCI. For simplicity, we drop Israel from our sample for this exercise. 28

2.25 time bigger for bond ETFs, relative to non-ETFs. By contrast, while flows into mutual funds respond to changes in local economic conditions, ETF flows do not. Our findings are robust to the inclusion of fund and investment scope-time fixed effects, time-varying fund controls such as past performance and economic conditions in the domicile of the fund. We explicitly consider and rule out a host of alternative explanations for the patterns we uncover in the data, confirming that it is specifically ETF flows which are particularly sensitive to changes in global financial conditions. In addition, we demonstrate that our findings have important implications for aggregate crossborder capital flows: we find that greater holdings of equity by foreign ETFs is associated with a higher exposure to global financial conditions both for aggregate portfolio equity flows and stock market returns. These results are not only statistically significant, but of economic importance. A one standard deviation increase in the percentage of local assets held by ETFs implies a sensitivity to global financial shocks that is 2.5 times in terms of portfolio equity flows and almost 1.4 times larger for prices. Overall, our results suggest that greater use of ETFs as a conduit for capital flows to emerging markets has increased the exposure of these economies to the global financial cycle. Our findings also present one example of how the rising popularity of passively managed, benchmarked instruments contributes to market co-movement and capital flows synchronicity at the expense of local fundamentals. Finally, the results presented here raise the question of why ETF flows respond differently to global and local factors, whether this is due to the perceived liquidity of ETFs shares or differences in the investor base of ETFs. This is a natural line for future research. 29

References Ahmed, Shaghil and Andrei Zlate, “Capitalflowstoemergingmarketeconomies: Abravenew world?,” Journal of International Money and Finance, 2014, 48 (Part B), 221 – 248. Current account imbalances and international financial integration. Avdjiev, Stefan, Bryan Hardy, Sebnem Kalemli-Ozcan, and Luis Servén, “Gross Capital Inflows to Banks, Corporates and Sovereigns,” NBER Working Papers 23116, National Bureau of Economic Research, Inc January 2017. Baltussen, Guido, Sjoerd van Bekkum, and Zhi Da, “Indexing and stock market serial dependence around the world,” Journal of Financial Economics, 2019, 132 (1), 26 – 48. Ben-David, Itzhak, Francesco Franzoni, and Rabih Moussawi, “Do ETFs Increase Volatility?,” The Journal of Finance, 2018, 73 (6), 2471–2535. Bhattacharya, Ayan and Maureen O’Hara, “Can ETFs Increase Market Fragility? Effect of Information Linkages in ETF Markets,” 2017. Mimeo. Borensztein, Eduardo R. and R.Gaston Gelos, “Leaders and followers: emerging market fund behavior during tranquil and turbulent times,” Emerging Markets Review, 2003, 4 (1), 25 – 38. Brandao-Marques, Luis, R. G Gelos, Hibiki Ichiue, and Hiroko Oura, “Changes in the Global Investor Base and the Stability of Portfolio Flows to Emerging Markets,” IMF Working Papers 15/277, International Monetary Fund December 2015. Broner, Fernando, Gaston Gelos, and Carmen Reinhart, “When in peril, retrench: Testing the portfolio channel of contagion,” Journal of International Economics, June 2006, 69 (1), 203– 230. Bruno, Valentina and Hyun Song Shin,“Capitalflowsandtherisk-takingchannelofmonetary policy,” Journal of Monetary Economics, 2015, 71 (Supplement C), 119 – 132. and , “Cross-Border Banking and Global Liquidity,” The Review of Economic Studies, 2015, 82 (2), 535–564. 30

Cerutti, Eugenio M, Stijn Claessens, and Damien Puy, “Push Factors and Capital Flows to Emerging Markets; Why Knowing Your Lender Matters More Than Fundamentals,” IMF Working Papers 15/127, International Monetary Fund June 2015. Cerutti, Eugenio, Stijn Claessens, and Andrew K. Rose, “How Important is the Global Financial Cycle? Evidence from Capital Flows,” NBER Working Papers 23699, National Bureau of Economic Research, Inc August 2017. Choi, Woon Gyu, Taesu Kang, Geun-Young Kim, and Byongju Lee, “Global liquidity transmissiontoemergingmarketeconomies,andtheirpolicyresponses,”Journal of International Economics, 2017, 109 (Supplement C), 153 – 166. Christoffersen, Susan E.K., David K. Musto, and Russ Wermers, “Investor Flows to Asset Managers: Causes and Consequences,” Annual Review of Financial Economics, 2014, 6 (1), 289–310. Converse, Nathan,“Uncertainty,capitalflows,andmaturitymismatch,”JournalofInternational Money and Finance, 2018, 88, 260 – 275. Coval, Joshua and Erik Stafford, “Asset fire sales (and purchases) in equity markets,” Journal of Financial Economics, November 2007, 86 (2), 479–512. Cremers, Martijn, Miguel A. Ferreira, Pedro Matos, and Laura Starks, “Indexing and active fund management: International evidence,” Journal of Financial Economics, 2016, 120 (3), 539–560. Da, Zhi and Sophie Shive, “Exchange Traded Funds and Asset Return Correlations,” European Financial Management, 2017. Forthcoming. Dannhauser, Caitlin and Saeid Hoseinzade, “The Transformation of Corporate Bond Investors and Fragility: Evidence on Mutual Funds and ETFs,” 2017. Unpublished. Dannhauser, Caitlin D., “The impact of innovation: Evidence from corporate bond exchangetraded funds (ETFs),” Journal of Financial Economics, 2017, 125 (3), 537 – 560. Deville, Laurent, “Exchange Traded Funds: History, Trading, and Research,” Handbook of Financial Engineering, 2008, pp. 67–98. 31

Didier, Tatiana, Roberto Rigobon, and Sergio L. Schmukler, “Unexploited Gains From International Diversification: Patterns Of Portfolio Holdings Around The World,” The Review of Economics and Statistics, December 2013, 95 (5), 1562–1583. Filippou, Ilias, Arie E. Gozluklu, and Hari Rozental, “ETF Arbitrage and International Diversification,” 2019. Mimeo. Forbes, Kristin J. and Francis E. Warnock, “Capital flow waves: Surges, stops, flight, and retrenchment,” Journal of International Economics, 2012, 88 (2), 235–251. and Roberto Rigobon, “No Contagion, Only Interdependence: Measuring Stock Market Comovements,” The Journal of Finance, 2002, 57 (5), 2223–2261. Forbes, Kristin, Marcel Fratzscher, Thomas Kostka, and Roland Straub, “Bubble thy neighbour: Portfolio effects and externalities from capital controls,” Journal of International Economics, 2016, 99 (C), 85–104. Fratzscher, Marcel, “Capital flows, push versus pull factors and the global financial crisis,” Journal of International Economics, 2012, 88 (2), 341 – 356. {NBER} Global. Gelos, R. G, “International Mutual Funds, Capital Flow Volatility, and Contagion - A Survey,” IMF Working Papers 11/92, International Monetary Fund April 2011. Gelos, R. Gaston and Shang-Jin Wei, “Transparency and International Portfolio Holdings,” Journal of Finance, December 2005, 60 (6), 2987–3020. Giglio, Stefano, Matteo Maggiori, Johannes Stroebel, and Stephen Utkus, “Five Facts about Beliefs and Portfolios,” March 2019. Unpublished. Glosten, Lawrence R., Suresh Nallareddy, and Yuan Zou,“ETFActivityandInformational Efficiency of Underlying Securities,” Technical Report 16, Columbia Business School December 2016. ICI, Fact Book, Investment Company Institute, 2017. Israeli, Doron, Charles M. C. Lee, and Suhas A. Sridharan, “Is there a dark side to exchange traded funds? An information perspective,” Review of Accounting Studies, Sep 2017, 22 (3), 1048–1083. 32

Jotikasthira, Chotibhak, Christian Lundblad, and Tarun Ramadorai, “Asset Fire Sales and Purchases and the International Transmission of Funding Shocks,” Journal of Finance, December 2012, 67 (6), 2015–2050. Kaminsky, Graciela, Richard K. Lyons, and Sergio L. Schmukler, “Managers, investors, and crises: mutual fund strategies in emerging markets,” Journal of International Economics, October 2004, 64 (1), 113–134. Khorana, Ajay, Henri Servaes, and Peter Tufano, “Explaining the size of the mutual fund industry around the world,” Journal of Financial Economics, October 2005, 78 (1), 145–185. Lane, Philip R. and Gian Maria Milesi-Ferretti, “The External Wealth of Nations Revisited: International Financial Integration in the Aftermath of the Global Financial Crisis,” IMF Economic Review, Mar 2018, 66 (1), 189–222. Miyajima, Ken and Ilhyock Shim, “Asset managers in emerging market economies,” BIS Quarterly Review, 2014. Nam, Jayoung, “Market Accessibility, Corporate Bond ETFs and Liquidity,” November 2017. Unpublished. Raddatz, Claudio and Sergio Schmukler,“Ontheinternationaltransmissionofshocks: Microevidence from mutual fund portfolios,” Journal of International Economics, 2012, 88 (2), 357– 374. , , and Tomas Williams, “International Asset Allocations and Capital Flows: The Benchmark Effect,” Journal of International Economics, 2017, 108, 413–430. Rey, Hélène, “Dilemma not Trilemma: The Global Financial Cycle and Monetary Policy Independence,” CEPR Discussion Papers 10591, C.E.P.R. Discussion Papers May 2015. Shek, Jimmy, Ilhyock Shim, and Hyun Song Shin, “Investor redemptions and fund manager sales of emerging market bonds: how are they related?,” BIS Working Papers 509, Bank for International Settlements August 2015. Shleifer, Andrei and Robert W Vishny, “ The Limits of Arbitrage,” Journal of Finance, March 1997, 52 (1), 35–55. 33

TIC, “Annual Report on U.S. Portfolio Holdings of Foreign Securities as of December 31, 2016,” TechnicalReport, DepartmentoftheTreasuryandFederalReserveBankofNewYorkandBoard of Governors of the Federal Reserve System 2017. Wu, Jing Cynthia and Fan Dora Xia, “Measuring the Macroeconomic Impact of Monetary PolicyattheZeroLowerBound,”JournalofMoney, CreditandBanking,2016,48(2-3),253–291. 34

Tables and Figures Figure 1: ETF Market Share and Emerging Markets’ Exposure to Global Financial Shocks Note: ThisfigureshowstheportfolioequityliabilityflowstoemergingmarketsasashareofGDP.Rollingbetaistheslopeofa 36-monthrollingregressionoftheportfolioequityliabilityflowsoverGDPversusthefirstdifferenceintheSt. LouisFinancial StressIndex. ETFMarketShare(rightaxis)representstheassetsundermanagementheldbyequityETFdividedbythetotal assetsundermanagementofallemergingmarketfundsinpercentage. Figure 2: The Growth of ETFs Note: ThisfigureshowstheassetsundermanagementofETFandnon-ETFintheEPFRdata. Thedataisatplottedatthe end of each year. Share ETF (right axis) represents the assets under management held by ETF divided by the total assets undermanagementofallfundsinpercentage. 35

Figure 3: Fund Flow Volatility, ETFs vs Mutual Funds Note: This figure depicts the timeevolution ofinvestor flowsover initial assets forETFsand non-ETF funds. InvestorFlows arethesumofinjectionsandredemptionsateachpointintime. AUMaretheinitialassetsundermanagementaggregatedat eachpointintime. Figure 4: Comparing Sensitivity to Global Factors Over Time Note: Thisfigurepresentsthesensitivityofinvestorflowstoglobalfactorsfortotalflowstoallfunds,aggregateflowstomutual funds,andaggregateflowstoETFs. Thebetaflows/AUMtoGlobalFactoristheslopeofa36-monthrollingregressionofthe aggregateinvestorflowsoverinitialassetsversusthedifferenceintheSt. LouisFinancialStressIndex. 36

Figure 5: Country Betas and ETF Share of Market Capitalization Note: ThisfiguredepictstheexposuretoglobalfactorsandtherelationshipwiththepresenceofETFsineachemergingcountry. TheleftpanelshowsthecoefficientofaregressionofBalanceofPaymentsPortfolioEquityInflowstothedifferenceintheSt. LouisFinancialStressIndexintheverticalaxis. Therightpanelpresentsintheverticalaxisthecoefficientofaregressionof MSCIstockmarketreturnsforeachcountrytothedifferenceintheSt. LouisFinancialStressIndex. Theseregressionsarefor theperiod2010-2017. ThehorizontalaxisforbothpanelsindicatestheequityassetsheldbyETFsineachcountrydividedby thetotalstockmarketcapitalization. SlopeandR-squaredreferstothecorrespondingstatisticsforthelinearfitofthescatter plot. 37

Table 1: Summary Statistics, Fund Flows over Initial Assets Panel A: Equity Funds FullSample DevelopedMarkets EmergingMarkets (1) (2) (3) (4) (5) (6) ETFs MutualFunds ETFs MutualFunds ETFs MutualFunds Mean 0.85 -0.14 0.93 -0.19 0.64 -0.05 StandardDeviation 9.20 5.97 9.43 5.86 8.54 6.23 10thPercentile -7.07 -4.48 -7.24 -4.38 -6.72 -4.73 25thPercentile -0.77 -1.74 -0.77 -1.71 -0.77 -1.82 Median 0.00 -0.30 0.00 -0.36 0.00 -0.16 75thPercentile 2.39 0.98 2.68 0.95 1.45 1.05 90thPercentile 10.10 4.48 10.48 4.27 9.03 4.96 NumberofFunds 1858 9150 1380 6621 479 2551 Observations 109888 657800 81050 457014 28838 200786 Panel B: Bond Funds FullSample DevelopedMarkets EmergingMarkets (1) (2) (3) (4) (5) (6) ETFs MutualFunds ETFs MutualFunds ETFs MutualFunds Mean 1.30 0.02 1.26 -0.05 1.58 0.23 StandardDeviation 9.85 6.45 9.88 6.19 9.70 7.18 10thPercentile -7.14 -5.01 -7.14 -4.84 -7.13 -5.56 25thPercentile -0.53 -1.93 -0.51 -1.88 -0.63 -2.10 Median 0.00 -0.18 0.00 -0.24 0.00 -0.03 75thPercentile 3.66 1.43 3.62 1.34 4.15 1.73 90thPercentile 11.82 5.50 11.62 5.11 12.85 6.81 NumberofFunds 406 3595 353 2738 53 859 Observations 20447 202285 17732 151399 2715 50886 Note: This table reports the summary statistics for fund flows over initial assets (in percentage) for the sample used in the main analysisfortheallthesample,developedandemergingmarketfunds. ThesampleisdividedintoETFandnon-ETF.PanelAshows statisticsforequityfundsandPanelBforbondfunds. Fundflowsoverinitialassetsarewinsorizedatthe1and99percentlevel. 38

Table 2: Baseline Results: Sensitivity of Mutual Fund and ETF Flows to Global Shocks DependentVariable: FundFlowsoverInitialAssets EquityFunds BondFunds (1) (2) (3) (4) (5) (6) GlobalFactor -2.118*** -1.857*** -3.294*** -3.169*** (0.344) (0.305) (0.475) (0.460) LocalFactor 0.170*** 0.187*** 0.099 0.116 (0.047) (0.045) (0.127) (0.123) GlobalFactor*ETF -2.733*** -2.256*** -3.948** -3.030* (0.607) (0.519) (1.951) (1.823) LocalFactor*ETF -0.133 0.030 -0.352 -0.255 (0.087) (0.073) (0.332) (0.359) FundPerformanceControls Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes InvestmentScope-TimeFE No No Yes No No Yes LocalFactorETF 0.054 -0.236 P-value 0.584 0.524 Observations 210,392 210,392 209,696 50,510 50,510 50,029 N.ofFunds 2,908 2,908 2,899 910 910 901 R2 0.064 0.064 0.138 0.092 0.092 0.177 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand differentsetsoffixedeffects. Theleftthreecolumnsshowtheresultsforequityfundsandthreeright-handcolumnsshowresultsfor bondfunds. LocalFactoristhemedianmonthlyindustrialproductiongrowthfortheinvestmentscopeofeachfund. GlobalFactoris thedifferenceintheSt. LouisFinancialStressIndex. ETFisadummyindicatingwhetherafundisanETFornot. FundPerformance Controls indicates whether the regression includes three lags of the portfolio returns of the fund minus the average fund returns at theinvestmentscopelevel. LocalFactorETFindicatesthesumofthecoefficientsforLocalFactorsandLocalFactors*ETF.P-value shows the significance test for LocalFactorETF = 0. Fund flows over initial assets are winsorized at the 1 and 99 percent level. Driscoll-Kraayrobuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 39

Table 3: Domicile-Specific Controls Panel A: Equity Funds DependentVariable: FundFlowsoverInitialAssets Domicilefinancialconditions DomicilespecifictimeFE (1) (2) (3) (4) (5) GlobalFactor -1.248*** -1.266*** (0.271) (0.213) GlobalFactor*ETF -2.945*** -2.265*** -2.789*** -1.699*** -1.652*** (0.740) (0.646) (0.762) (0.470) (0.417) LocalFactor 0.193*** 0.044 (0.043) (0.035) LocalFactor*ETF -0.136 0.028 0.172* -0.177** -0.082 (0.086) (0.073) (0.098) (0.073) (0.072) StkMktatFundDomicile 5.164*** 1.573* 4.558*** 1.925** (1.011) (0.834) (0.856) (0.754) StkMktatFundDomicile*ETF -2.393 0.078 0.165 (2.639) (2.254) (2.946) FundPerformanceControls Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No No Yes Domicile*Inv. Scope*TimeFE No No Yes No No FundDomicile*ETF*YearFE No No No Yes Yes LocalFactorETF 0.057 -0.133 P-value 0.563 0.098 Observations 210,194 209,498 195,690 210,189 209,493 N.ofFunds 2,906 2,897 2,750 2,906 2,897 R2 0.066 0.138 0.216 0.091 0.148 Panel B: Bond Funds DependentVariable: FundFlowsoverInitialAssets Domicilefinancialconditions DomicilespecifictimeFE (1) (2) (3) (4) (5) GlobalFactor -2.391*** -2.198*** (0.465) (0.416) GlobalFactor*ETF -5.970*** -4.768** -7.038*** -2.625** -2.240** (2.276) (2.132) (2.363) (1.206) (1.130) LocalFactor 0.137 0.049 (0.119) (0.091) LocalFactor*ETF -0.374 -0.267 -0.119 -0.237 -0.212 (0.322) (0.354) (0.396) (0.287) (0.313) StkMktatFundDomicile 5.943*** 4.173** 5.574*** 3.771*** (2.056) (1.699) (1.838) (1.224) StkMktatFundDomicile*ETF -14.464** -11.459* -13.875** (5.843) (6.000) (6.766) FundPerformanceControls Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No No Yes Domicile*Inv. Scope*TimeFE No No Yes No No FundDomicile*ETF*YearFE No No No Yes Yes LocalFactorETF -0.237 -0.188 P-value 0.517 0.571 Observations 50,510 50,029 48,254 50,509 50,028 N.ofFunds 910 901 870 910 901 R2 0.094 0.177 0.226 0.129 0.190 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand different sets of fixed effects for emerging market funds. Local Factor is the median monthly industrial production growth for the investment scope of each fund. Global Factor is the difference in the St. Louis Financial Stress Index. ETF is a dummy indicating whetherafundisanETFornot. FundControlsindicateswhethertheregressionincludesfundcontrolvariables. Thesevariablesare thethreelagsoftheportfolioreturnsofthefundminustheaveragefundreturnsattheinvestmentscopelevelandthedifferenceinlogs oftheMSCIstockmarketindexinthedomicileofeachfund. LocalFactorETFindicatesthesumofthecoefficientsforLocalFactors andLocalFactors*ETF.P-valueshowsthesignificancetestforLocalFactorETF=0. Fundflowsoverinitialassetsarewinsorizedat the1and99percentlevel. Driscoll-Kraayrobuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 40

Table 4: Analysis at Lower Frequencies Panel A: Equity Funds DependentVariable: FundFlowsoverInitialAssets Frequency Quarterly 6-MonthPeriods (1) (2) (3) (4) GlobalFactor -0.853 -0.063 (0.601) (0.974) GlobalFactor*ETF -5.018*** -4.343*** -7.545** -5.688** (1.726) (1.470) (3.077) (2.726) LocalFactor 0.649*** 1.050*** (0.107) (0.219) LocalFactor*ETF 0.205 0.463** 0.428 0.677 (0.227) (0.200) (0.512) (0.528) FundControls Yes Yes Yes Yes FundFE Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes LocalFactorETF 0.854 1.478 P-value 0.001 0.029 Observations 208,134 200,756 199,787 199,091 N.ofFunds 2,906 2,888 2,895 2,885 R2 0.112 0.203 0.154 0.252 Panel B: Bond Funds DependentVariable: FundFlowsoverInitialAssets Frequency Quarterly 6-MonthPeriods (1) (2) (3) (4) GlobalFactor -2.004 -2.607 (1.275) (1.794) GlobalFactor*ETF -12.249*** -9.996*** -19.000*** -14.767*** (3.262) (3.373) (4.518) (4.097) LocalFactor 0.597** 1.373*** (0.299) (0.454) LocalFactor*ETF -0.716 -0.666 -1.913** -1.345 (0.648) (0.702) (0.879) (0.856) FundControls Yes Yes Yes Yes FundFE Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes LocalFactorETF -0.119 -0.540 P-value 0.883 0.655 Observations 49,877 49,401 46,881 46,422 N.ofFunds 910 901 906 898 R2 0.150 0.253 0.198 0.302 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand different sets of fixed effects for emerging market funds. Fund flows are cumulative during an horizon of 3 and 6 months and are divided by the initial assets. Panel A shows the results for equity funds and Panel B for bond funds. Local Factor is the median monthly industrial production growth for the investment scope of each fund. Global Factor is the difference in logs of the variable indicatedatthetopofeachcolumn. ETFisadummyindicatingwhetherafundisanETFornot. FundControlsindicateswhether theregressionincludesfundcontrolvariables. Thesevariablesarethethreelagsoftheportfolioreturnsofthefundminustheaverage fund returns at the investment scope level and the difference in logs of the MSCI stock market index in the domicile of each fund. LocalFactorETFindicatesthesumofthecoefficientsforLocalFactorsandLocalFactors*ETF.P-valueshowsthesignificancetest forLocalFactorETF=0. Fundflowsoverinitialassetsarewinsorizedatthe1and99percentlevel. Driscoll-Kraayrobuststandard errorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 41

Table 5: Exploring Alternate Hypotheses Panel A: Equity Funds DependentVariable: FundFlowsoverInitialAssets ETFsvsLargeFunds ETFsvsPassive ETFsvsGlobalFunds (1) (2) (3) (4) (5) (6) GlobalFactor -1.174*** -1.271*** -1.135*** (0.264) (0.277) (0.252) GlobalFactor*ETF -2.705*** -2.341*** -2.608*** -2.274*** -3.018*** -2.655*** (0.652) (0.534) (0.620) (0.522) (0.825) (0.737) LocalFactor 0.192*** 0.194*** 0.192*** (0.043) (0.043) (0.043) LocalFactor*ETF -0.134 0.028 -0.135 0.030 -0.133 0.029 (0.086) (0.073) (0.086) (0.073) (0.085) (0.073) GlobalFactor*>250M -0.260 -0.175 (0.185) (0.158) GlobalFactor*Passive -0.177 0.082 (0.667) (0.692) LocalFactor*Passive -0.080 0.075 (0.152) (0.153) GlobalFactor*CountryFund -0.496 (0.333) GlobalFactor*CountryFund*ETF 1.022 0.824 (0.831) (0.854) FundControls Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes No Yes LocalFactorETF 0.058 0.058 0.059 P-value 0.547 0.547 0.541 Observations 210,194 209,498 210,194 209,498 210,194 209,498 N.ofFunds 2,906 2,897 2,906 2,897 2,906 2,897 R2 0.066 0.138 0.066 0.138 0.066 0.138 Panel B: Bond Funds DependentVariable: FundFlowsoverInitialAssets ETFsvsLargeFunds ETFsvsPassive ETFsvsGlobalFunds (1) (2) (3) (4) (5) (6) GlobalFactor -2.020*** -2.453*** -2.481*** (0.467) (0.476) (0.488) GlobalFactor*ETF -4.300** -3.571* -3.877** -3.116* -4.109* -3.109* (2.013) (1.915) (1.948) (1.836) (2.157) (1.834) LocalFactor 0.135 0.138 0.136 (0.119) (0.119) (0.119) LocalFactor*ETF -0.372 -0.264 -0.375 -0.266 -0.375 -0.265 (0.331) (0.360) (0.332) (0.360) (0.332) (0.360) GlobalFactor*>250M -0.886*** -0.991*** (0.328) (0.296) GlobalFactor*Passive -3.304 -2.357 (2.354) (2.429) LocalFactor*Passive -0.453* -0.112 (0.240) (0.243) GlobalFactor*CountryFund 0.591 0.000 (0.874) (.) GlobalFactor*CountryFund*ETF 2.893 (5.124) FundControls Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes No Yes LocalFactorETF -0.237 -0.237 -0.239 P-value 0.526 0.526 0.521 Observations 50,510 50,029 50,510 50,029 50,510 50,029 N.ofFunds 910 901 910 901 910 901 R2 0.094 0.177 0.093 0.177 0.093 0.177 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand differentsetsoffixedeffectsforEMfunds. PanelAshowstheresultsforequityfundsandPanelBforbondfunds. LocalFactoristhe medianmonthlyindustrialproductiongrowthfortheinvestmentscopeofeachfund. GlobalFactoristhedifferenceintheSt. Louis Financial Stress Index. ETF is a dummy indicating whether a fund is an ETF or not. >250M is a dummy variable that is 1 when a fund’s AUM is larger than $250 million. Fund Controls indicates whether the regression includes three lags of the fund’s returns (measuredasthedeviationfromtheaveragereturnoffundswiththesameinvestmentscope)levelandthelogdifferencedMSCIstock marketindexinthefund’sdomicile. Fundflowsoverinitialassetsarewinsorizedatthe1and99percentlevel. Driscoll-Kraayrobust standarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 42

Table 6: Aggregate Economic Significance: ETF Assets and Country Betas Panel A Dependent Variable: Balance of Payments Gross Portfolio Equity Inflows (% of GDP) (1) (2) (3) (4) (5) (6) (7) GlobalFactor -0.201** 0.031 0.031 0.028 (0.090) (0.110) (0.109) (0.128) GlobalFactor*ETFShare -0.712*** -0.724*** -0.711*** -0.730*** -0.722*** -0.715*** (0.097) (0.071) (0.108) (0.081) (0.101) (0.072) ETFShare -0.127** -0.027 -0.127** -0.026 -0.137** 0.001 (0.055) (0.051) (0.054) (0.050) (0.055) (0.051) GlobalFactor*MutualFundShare -0.000 0.004 (0.013) (0.012) MutualFundShare -0.008 -0.015 (0.014) (0.015) GlobalFactor*Fin. Integration 0.009 0.016 (0.024) (0.023) FinancialIntegration(L-MF) 0.051*** 0.059*** (0.008) (0.008) CountryFE Yes Yes Yes Yes Yes Yes Yes TimeFE No No Yes No Yes No Yes Observations 2,035 2,035 2,032 2,035 2,032 2,035 2,032 N.ofCountries 43 43 43 43 43 43 43 R2 0.136 0.155 0.222 0.155 0.222 0.183 0.257 Panel B Dependent Variable: MSCI Country Stock Markets Returns (basis points) (1) (2) (3) (4) (5) (6) (7) GlobalFactor -9.09*** -7.49*** -7.00*** -7.42*** (1.57) (1.53) (1.40) (1.55) GlobalFactor*ETFShare -5.13*** -4.28*** -4.36*** -3.11*** -5.14*** -4.16*** (0.88) (0.77) (0.96) (0.72) (0.88) (0.73) ETFShare -0.73*** -0.02 -0.73*** -0.36** -0.72** -0.38** (0.28) (0.10) (0.28) (0.15) (0.28) (0.15) GlobalFactor*MutualFundShare -0.41* -0.51** (0.24) (0.24) MutualFundShare 0.11* 0.05 (0.06) (0.05) GlobalFactor*Fin. Integration -0.03 0.02 (0.04) (0.04) FinancialIntegration(L-MF) -0.03* -0.02 (0.02) (0.01) CountryFE Yes Yes No Yes Yes Yes Yes TimeFE No No Yes No Yes No Yes Observations 7,613 7,613 7,606 7,613 7,606 7,613 7,606 N.ofCountries 49 49 49 49 49 49 49 R2 0.112 0.124 0.354 0.125 0.361 0.124 0.359 Note: This table reports the OLS coefficients from a regression of Balance of Payments Portfolio Equity Liability Flows over GDP (Panel A) or MSCI Country Stock Market Returns (Panel B) on different explanatory variables and different sets of fixed effects for emerging markets at the quarterly frequency. Global Factor is the difference in the St. Louis Financial Stress Index. ETF Share is theassetsundermanagementofequityETFsdividedbythetotalequitymarketcapitalization. MutualFundShareistheassetsunder management of equity of funds that are not ETFs divided by the total equity market capitalization. Financial Integration is total financialassetsplusliabilitiesdividedbyGDPfromLaneandMilesi-Ferretti(2018). Alltheestimationsarefortheperiod2000-2017. PanelAestimationsareatthequarterlyfrequencyandPanelBatthemonthlyfrequency. Thedependentvariableiswinsorizedatthe 1and99percentlevel. Driscoll-Kraayrobuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 43

Table 7: Aggregate Economic Significance: Robustness Panel A: Dependent Variable: Balance of Payments Gross Portfolio Equity Inflows (% of GDP) (1) (2) (3) (4) GlobalFactor 0.012 -0.021 (0.119) (0.090) GlobalFactor*ETFShare(excl. CountryFunds) -0.634*** -0.657*** (0.129) (0.116) ETFShare(excl. CountryFunds) -0.295*** -0.104 (0.092) (0.136) GlobalFactor*MFShare(excl. CountryFunds) -0.050*** -0.046** (0.019) (0.020) MutualFundShare(excl. CountryFunds) 0.005 -0.008 (0.013) (0.013) GlobalFactor*MSCIEM -0.337*** -0.395*** (0.093) (0.097) MSCIEM 0.473*** 0.273 (0.153) (0.175) CountryFE Yes Yes Yes Yes TimeFE No Yes No Yes Observations 2,007 2,004 1,845 1,842 N.ofCountries 43 43 40 40 R2 0.157 0.223 0.149 0.221 Panel B: Dependent Variable: MSCI Country Stock Markets Returns (basis points) (1) (2) (3) (4) GlobalFactor -6.519*** -6.961*** (1.373) (1.627) GlobalFactor*ETFShare(excl. CountryFunds) -6.069*** -4.373*** (2.236) (1.490) ETFShare(excl. CountryFunds) -1.631** -1.396*** (0.668) (0.418) GlobalFactor*MFShare(excl. CountryFunds) -1.101** -1.223*** (0.465) (0.378) MutualFundShare(excl. CountryFunds) 0.335*** 0.122 (0.126) (0.083) GlobalFactor*MSCIEM -3.968*** -4.489*** (1.084) (1.260) MSCIEM 0.458 -0.351 (0.519) (0.443) CountryFE Yes Yes Yes Yes TimeFE No Yes No Yes Observations 7,525 7,518 7,525 7,518 N.ofCountries 49 49 49 49 R2 0.129 0.365 0.117 0.361 Note: This table reports the OLS coefficients from a regression of Balance of Payments Portfolio Equity Liability Flows over GDP (Panel A) or MSCI Country Stock Market Returns (Panel B) on different explanatory variables and different sets of fixed effects for emerging markets at the quarterly frequency. Global Factor is the difference in the St. Louis Financial Stress Index. ETF Share is theassetsundermanagementofequityETFsdividedbythetotalequitymarketcapitalization. MutualFundShareistheassetsunder managementofequityoffundsthatarenotETFsdividedbythetotalequitymarketcapitalization. ETFandMutualFundShareare computed by using only global or regional funds (i.e. excluding country funds). MSCI EM is a variable that is 0 when a country is a frontier markets or standalone market under, 1 when it is an emerging market, and 2 when it is a developed market under MSCI classificationscheme. Alltheestimationsarefortheperiod2000-2017. PanelAestimationsareatthequarterlyfrequencyandPanel Batthemonthlyfrequency. Thedependentvariableiswinsorizedatthe1and99percentlevel. Driscoll-Kraayrobuststandarderrors inparenthesis. ***p<0.01,**p<0.05,*p<0.10. 44

Appendix Table A1: Variable Definition and Source Variable Definition Source Fit Injections/RedemptionstofundiattimetinUSdollars EPFR Ait AssetsundermanagementtofundiattimetinUSdollars EPFR FundPerformance Portfolioreturnofeachfundminustheaveragereturnattheinvestmentscopelevel EPFR STLFSI St. LouisFinancialStressIndex FRED VIX ChicagoBoardOptionsExchangeMarketVolatilityIndex FRED USHY EffectiveyieldoftheBankofAmericaMerrillLynchUSHighYieldMasterIIIndex FRED TEDRate Spreadbetween3-monthLIBORand3-monthTreasuryBill FRED PCA1 PrincipalComponentofthemonthlygrowth(inlogs)ofVIX,USHYandTEDRate Own FFRate EffectiveFederalFundsRate FRED FFShadowRate Wu-XiaFederalFundsRate AtlantaFed MedianIPGrowth MedianoftheMonthlyIndustrialProductionGrowthattheInvestmentScopeLevel IMFIFS MeanIPGrowth MeanoftheMonthlyIndustrialProductionGrowthattheInvestmentScopeLevel IMFIFS GDPWeightedIPGrowth GDPweightedMonthlyIndustrialProductionGrowthattheInvestmentScopeLevel IMFIFS StkMktatDomicile MonthlyGrowthoftheMSCIStockMarketIndexattheDomicileoftheFund MSCI 45

Table A2: Funds by Investment Scope Panel A: Developed Markets EquityFunds BondFunds (1) (2) (3) (4) Observations Funds Observations Funds Europe 84007 1221 17180 355 Europeex-UK 45518 697 39421 847 Global 189738 2928 82413 1397 Globalex-US 67734 868 3820 54 Japan 35554 515 496 9 Pacific 13223 185 148 5 UnitedStates 65646 1089 16032 306 Other 36644 530 9621 219 Total 538064 7999 169131 3091 Panel B: Emerging Markets EquityFunds BondFunds (1) (2) (3) (4) Observations Funds Observations Funds Asiaex-Japan 51377 648 5395 110 China 16299 275 1212 31 EmergingEurope 15610 162 2024 26 GlobalEmergingMarkets 68527 924 40469 652 GreaterChina 12641 123 68 2 India 10072 156 308 9 LatinAmerica 12691 145 1291 21 Other 42407 617 2834 66 Total 229624 3029 53601 912 Note: Thistableshowsthestatisticsfortheinvestmentscopeofthefunds. PanelAreportsthedevelopedmarketfunds,andPanelB theemergingmarketfunds. Otherisaresidualcategoryindicatingallotherdomiciles. 46

Table A3: Funds by Domicile Panel A: Number of Funds EquityFunds BondFunds (1) (2) (3) (4) (5) (6) FullSample DM EM FullSample DM EM Canada 522 455 67 78 70 8 France 714 607 109 267 256 11 Germany 296 280 16 118 113 5 Ireland 1188 852 339 440 349 92 Japan 245 125 120 166 91 75 Luxembourg 3303 2366 944 1752 1327 425 Switzerland 270 222 49 181 173 8 UnitedKingdom 733 521 215 131 118 13 UnitedStates 2347 1764 589 474 311 163 Other 1409 827 583 399 286 114 Total 11005 7999 3029 4001 3091 912 Panel B: Number of Observations EquityFunds BondFunds (1) (2) (3) (4) (5) (6) FullSample DM EM FullSample DM EM Canada 36028 29972 6056 4144 3767 377 France 47004 39530 7474 13647 13033 614 Germany 21715 20496 1219 5952 5639 313 Ireland 73366 48846 24520 26658 20735 5923 Japan 10374 5451 4923 7241 3968 3273 Luxembourg 235145 161760 73385 94549 69651 24898 Switzerland 18518 14610 3908 9250 9096 154 UnitedKingdom 57137 37372 19765 9302 8293 1009 UnitedStates 182008 132528 49480 32289 21478 10811 Other 86393 47499 38894 19700 13471 6229 Total 767688 538064 229624 222732 169131 53601 Note: Thistableshowsthestatisticsforthedomicileofthefunds. PanelAreportsthenumberoffunds,andPanelBthenumberof observationsinthesampleforeachdomicile. Fundsaredividedintodevelopedoremergingmarketfunds. Otherisaresidualcategory indicatingallotherdomiciles. 47

Table A4: Summary Statistics, Assets Under Management Panel A: Equity Funds AllSample DevelopedMarkets EmergingMarkets (1) (2) (3) (4) (5) (6) ETFs MutualFunds ETFs MutualFunds ETFs MutualFunds Mean 609.40 714.52 569.05 825.99 722.62 460.95 StandardDeviation 2782.76 3649.96 2475.34 4270.73 3501.88 1430.28 p10 12.91 20.00 13.41 23.14 11.72 15.56 p25 27.56 49.32 29.24 57.09 23.27 36.42 Median 86.78 151.16 90.40 171.55 76.68 112.37 p75 315.93 462.10 321.36 518.28 303.65 349.00 p90 1031.81 1296.78 1020.59 1458.55 1065.87 976.13 NumberofFunds 1858 9150 1380 6621 479 2551 Observations 110435 658470 81422 457385 29013 201085 Panel B: Bond Funds AllSample DevelopedMarkets EmergingMarkets (1) (2) (3) (4) (5) (6) ETFs MutualFunds ETFs MutualFunds ETFs MutualFunds Mean 489.09 784.45 457.45 878.80 694.65 504.17 StandardDeviation 1105.49 2356.44 1055.87 2655.26 1367.77 999.38 p10 13.67 28.71 13.60 33.70 14.30 20.25 p25 32.95 76.11 32.85 87.57 34.69 50.83 Median 123.22 221.78 125.90 248.70 105.40 156.59 p75 443.83 670.71 430.55 752.21 569.13 456.71 p90 1273.78 1736.33 1181.92 1877.18 2397.69 1262.66 NumberofFunds 406 3595 353 2738 53 859 Observations 20573 202587 17829 151567 2744 51020 Note: Thistablereportsthesummarystatisticsfortheassetsundermanagement(inmillionsUSD)forthesampleusedinthemain analysisfortheallthesample,developedandemergingmarketfunds. ThesampleisdividedintoETFandnon-ETF.PanelAshows statisticsforequityfundsandPanelBforbondfunds. 48

Table A5: Summary Statistics, Global Factors Summary Statistics Global Variables (1) (2) (3) (4) (5) (6) St. LouisFSI VIX TEDRate USHY FFRate FFShadowRate Mean -0.01 -0.00 -0.00 0.00 -0.02 -0.02 StandardDeviation 0.26 0.16 0.22 0.09 0.17 0.20 p10 -0.22 -0.18 -0.29 -0.09 -0.21 -0.25 p25 -0.12 -0.09 -0.14 -0.06 -0.02 -0.08 Median -0.02 -0.01 -0.00 -0.01 0.00 0.00 p75 0.07 0.07 0.14 0.04 0.02 0.09 p90 0.18 0.18 0.25 0.11 0.15 0.20 Observations 248 248 248 248 248 248 Note: Thistablereportsthesummarystatisticsforthevariablesusedasglobalfactors. TheSt. LouisFSI,theFFRateandtheFF ShadowRateareindifferences. TheVIX,TEDRate,USHYareindifferencesoflogs. Table A6: Summary Statistics, Local Factors Industrial Production Growth in Investment Scope AllSample DevelopedMarkets EmergingMarkets (1) (2) (3) (4) (5) (6) (7) (8) (9) Median Mean GDPWeighted Median Mean GDPWeighted Median Mean GDPWeighted mean 0.18 0.18 0.19 0.05 0.05 0.05 0.27 0.26 0.28 sd 2.61 2.63 2.60 2.44 2.46 2.42 2.71 2.74 2.72 p10 -2.23 -2.27 -2.22 -2.22 -2.22 -2.20 -2.25 -2.30 -2.25 p25 -0.76 -0.81 -0.72 -0.81 -0.84 -0.77 -0.72 -0.79 -0.69 p50 0.25 0.25 0.26 0.10 0.10 0.10 0.37 0.35 0.38 p75 1.19 1.24 1.17 0.93 0.98 0.91 1.32 1.38 1.32 p90 2.45 2.50 2.46 2.21 2.25 2.15 2.59 2.65 2.63 N 9424 9424 9424 3699 3699 3699 5725 5725 5725 Note: Thistablereportsthesummarystatisticsforthevariablesusedaslocalfactors. Median,meanandGDPweightedindicateshow themonthlygrowthinindustrialproductionwasaggregatedattheinvestmentscopelevel. Allthevariablesareinpercentages. 49

Table A7: Lagged Performance Coefficients With Lagged Performance Controls FundFlowsoverInitialAssets (1) (2) (3) (4) Equity Equity Bond Bond Lagged(1)FundPerformance 11.941*** 9.072*** -10.187*** 19.799*** (1.631) (1.351) (3.413) (3.186) Lagged(2)FundPerformance 7.729*** 6.925*** -3.959 13.096*** (1.048) (0.951) (3.344) (3.472) Lagged(3)FundPerformance 5.788*** 5.158*** -1.975 11.361*** (1.111) (1.065) (2.611) (2.498) LocalFactor 0.187*** 0.116 (0.045) (0.123) LocalFactor*ETF -0.133 0.030 -0.352 -0.255 (0.087) (0.073) (0.332) (0.359) GlobalFactor -1.857*** -3.169*** (0.305) (0.460) GlobalFactor*ETF -2.733*** -2.256*** -3.948** -3.030* (0.607) (0.519) (1.951) (1.823) FundFE Yes Yes Yes Yes InvestmentScope-TimeFE No Yes No Yes LocalFactorETF 0.054 -0.236 P-value 0.584 0.524 Observations 210392 209696 50510 50029 N.ofFunds 2908 2899 910 901 R2 0.064 0.138 0.092 0.177 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand different sets of fixed effects for emerging market funds. Local Factor is the median monthly industrial production growth for the investment scope of each fund. Global Factor is the difference in logs of the variable indicated at the top of each column. ETF is a dummyindicatingwhetherafundisanETFornot. Lagged(n)FundPerformanceisthenthlagoftheportfolioreturnsofthefund minustheaveragefundreturnsattheinvestmentscopelevel. LocalFactorETFindicatesthesumofthecoefficientsforLocalFactors andLocalFactors*ETF.P-valueshowsthesignificancetestforLocalFactorETF=0. Fundflowsoverinitialassetsarewinsorizedat the1and99percentlevel. Driscoll-Kraayrobuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 50

Table A8: Developed Market Funds - Baseline Results DependentVariable: FundFlowsoverInitialAssets EquityFunds BondFunds (1) (2) (3) (4) (5) (6) GlobalFactor -0.997*** -0.941*** -1.367*** -1.380*** (0.198) (0.158) (0.258) (0.253) LocalFactor 0.058 0.073* 0.158* 0.153 (0.044) (0.043) (0.095) (0.104) GlobalFactor*ETF -0.485 -0.438 0.242 -0.148 (0.552) (0.472) (0.829) (0.779) LocalFactor*ETF -0.097 -0.065 0.047 0.261** (0.108) (0.103) (0.141) (0.125) FundPerformanceControls Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes InvestmentScope-TimeFE No No Yes No No Yes LocalFactorETF -0.025 0.200 P-value 0.824 0.070 Observations 467,681 467,681 467,263 142,806 142,806 142,600 N.ofFunds 7,840 7,840 7,840 3,046 3,046 3,042 R2 0.077 0.077 0.104 0.088 0.088 0.115 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand differentsetsoffixedeffects. Theleftthreecolumnsshowtheresultsforequityfundsandthreeright-handcolumnsshowresultsfor bondfunds. LocalFactoristhemedianmonthlyindustrialproductiongrowthfortheinvestmentscopeofeachfund. GlobalFactoris thedifferenceintheSt. LouisFinancialStressIndex. ETFisadummyindicatingwhetherafundisanETFornot. FundPerformance Controls indicates whether the regression includes three lags of the portfolio returns of the fund minus the average fund returns at theinvestmentscopelevel. LocalFactorETFindicatesthesumofthecoefficientsforLocalFactorsandLocalFactors*ETF.P-value shows the significance test for LocalFactorETF = 0. Fund flows over initial assets are winsorized at the 1 and 99 percent level. Driscoll-Kraayrobuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. Table A9: Developed Market Funds - Additional Tests Developed Market Funds FundFlowsoverInitialAssets (1) (2) (3) (4) (5) (6) Equity Equity Equity Bond Bond Bond LocalFactor 0.017 0.026 0.179** 0.187** (0.032) (0.030) (0.074) (0.080) GlobalFactor -1.112*** -0.968*** -1.315*** -1.208*** (0.172) (0.145) (0.319) (0.312) LocalFactor*ETF -0.055 -0.028 -0.116 0.023 (0.084) (0.082) (0.232) (0.229) GlobalFactor*ETF -1.065** -0.967** -1.762** -1.914** (0.496) (0.473) (0.884) (0.813) FundPerformanceControls Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes InvestmentScope-TimeFE No No Yes No No Yes LocalFactorETF -0.029 0.071 P-value 0.745 0.727 Observations 818356 818356 818049 287285 287285 287074 N.ofFunds 13107 13107 13107 5387 5387 5382 R2 0.088 0.088 0.108 0.114 0.114 0.138 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand different sets of fixed effects for developed market funds. Coefficients were estimated using data that included domestic funds but excludedfundsinvestingexclusivelyinJapanese,German,orU.S.governmentbonds. LocalFactoristhemedianmonthlyindustrial productiongrowthfortheinvestmentscopeofeachfund. GlobalFactoristhedifferenceintheSt. LouisFinancialStressIndex. ETF isadummyindicatingwhetherafundisanETFornot. FundPerformanceControlsindicateswhethertheregressionincludesthree lagsoftheportfolioreturnsofthefundminustheaveragefundreturnsattheinvestmentscopelevel. LocalFactorETFindicatesthe sumofthecoefficientsforLocalFactorsandLocalFactors*ETF.P-valueshowsthesignificancetestforLocalFactorETF=0. The estimationsforbondfundsdonothavefundsinvestingingovernmentdebtofsafeheavencountries(Germany,Japan,UnitedStates). Theestimationsdonotcontaintheheightenoftheglobalfinancialcrisis(August2007,SeptemberandOctober2008)andcontainboth domestic and international mutual funds. Fund flows over initial assets are winsorized at the 1 and 99 percent level. Driscoll-Kraay robuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 51

Table A10: Global and Country Funds - Additional Tests Panel A: Equity Funds DependentVariable: FundFlowsoverInitialAssets GlobalFunds CountryFunds (1) (2) (3) (4) GlobalFactor -1.205*** -1.500*** (0.249) (0.441) GlobalFactor*ETF -3.019*** -2.640*** -1.996*** -1.872*** (0.809) (0.727) (0.661) (0.591) LocalFactor 0.167*** 0.237*** (0.053) (0.044) LocalFactor*ETF 0.154 0.253** -0.334*** -0.172** (0.142) (0.124) (0.082) (0.077) LocalFactorETF 0.321 -0.097 P-value 0.046 0.277 Observations 150,851 150,832 59,342 58,666 N.ofFunds 1,987 1,987 921 912 R2 0.070 0.121 0.056 0.176 Panel B: Bond Funds DependentVariable: FundFlowsoverInitialAssets GlobalFunds CountryFunds (1) (2) (3) (4) GlobalFactor -2.479*** -2.030** (0.499) (1.024) GlobalFactor*ETF -4.094* -3.424* -1.758 2.562 (2.156) (1.963) (4.169) (4.189) LocalFactor 0.154 0.016 (0.133) (0.131) LocalFactor*ETF -0.502 -0.291 0.488 0.479 (0.384) (0.360) (0.411) (0.954) FundControls Yes Yes Yes Yes FundFE Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes LocalFactorETF -0.348 0.504 P-value 0.418 0.164 Observations 47,060 46,964 3,450 3,065 N.ofFunds 812 809 98 92 R2 0.092 0.167 0.090 0.299 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand differentsetsoffixedeffectsforemergingmarketfunds. PanelAshowstheresultsforequityfundsandPanelBforbondfunds. Local Factor is the median monthly industrial production growth for the investment scope of each fund. Global Factor is the difference in theSt. LouisFinancialStressIndex. ETFisadummyindicatingwhetherafundisanETFornot. FundControlsindicateswhether theregressionincludesfundcontrolvariables. Thesevariablesarethethreelagsoftheportfolioreturnsofthefundminustheaverage fund returns at the investment scope level and the difference in logs of the MSCI stock market index in the domicile of each fund. LocalFactorETFindicatesthesumofthecoefficientsforLocalFactorsandLocalFactors*ETF.P-valueshowsthesignificancetest forLocalFactorETF=0. Fundflowsoverinitialassetsarewinsorizedatthe1and99percentlevel. Driscoll-Kraayrobuststandard errorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 52

Table A11: ETFs vs Large Mutual Funds - Alternative Definitions of “Large” Panel A: Equity Funds DependentVariable: FundFlowsoverInitialAssets 100M Top25thPercentile (1) (2) (3) (4) GlobalFactor -1.076*** -1.164*** (0.272) (0.257) GlobalFactor*ETF -2.802*** -2.411*** -2.714*** -2.270*** (0.645) (0.523) (0.666) (0.544) LocalFactor 0.192*** 0.192*** (0.043) (0.043) LocalFactor*ETF -0.134 0.028 -0.134 0.028 (0.086) (0.073) (0.086) (0.073) GlobalFactor*>100M -0.319* -0.225 (0.183) (0.189) GlobalFactor*>TopQuartileAv.Assets -0.149 0.008 (0.233) (0.221) LocalFactorETF 0.058 0.058 P-value 0.547 0.547 Observations 210,194 209,498 210,194 209,498 N.ofFunds 2,906 2,897 2,906 2,897 R2 0.066 0.138 0.066 0.138 Panel B: Bond Funds DependentVariable: FundFlowsoverInitialAssets 100M Top25thPercentile (1) (2) (3) (4) GlobalFactor -1.926*** -1.926*** (0.443) (0.454) GlobalFactor*ETF -4.401** -3.715* -4.402** -3.696* (2.030) (1.905) (2.051) (1.925) LocalFactor 0.136 0.136 (0.119) (0.119) LocalFactor*ETF -0.372 -0.265 -0.372 -0.265 (0.331) (0.360) (0.331) (0.360) GlobalFactor*>100M -0.761* -0.898** (0.398) (0.399) FundControls Yes Yes Yes Yes FundFE Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes LocalFactorETF -0.237 -0.237 P-value 0.526 0.526 Observations 50,510 50,029 50,510 50,029 N.ofFunds 910 901 910 901 R2 0.094 0.177 0.094 0.177 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand differentsetsoffixedeffectsforemergingmarketfunds. PanelAshowstheresultsforequityfundsandPanelBforbondfunds. Local Factoristhemedianmonthlyindustrialproductiongrowthfortheinvestmentscopeofeachfund. GlobalFactoristhedifferenceinthe St. LouisFinancialStressIndex. ETFisadummyindicatingwhetherafundisanETFornot. (>TopQuartileAv.Assets)isadummy variablethatis1whentheaverageassetsundermanagementinafundarelargerthanthetopquartileaverageassetsinoursample for each equity and bond funds separately. Fund Controls indicates whether the regression includes fund control variables. These variablesarethethreelagsoftheportfolioreturnsofthefundminustheaveragefundreturnsattheinvestmentscopelevelandthe differenceinlogsoftheMSCIstockmarketindexinthedomicileofeachfund. LocalFactorETFindicatesthesumofthecoefficients forLocalFactorsandLocalFactors*ETF.P-valueshowsthesignificancetestforLocalFactorETF=0. Fundflowsoverinitialassets arewinsorizedatthe1and99percentlevel. Driscoll-Kraayrobuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 53

Table A12: Robustness - Alternate Time Periods Panel A: Equity Funds DependentVariable: FundFlowsoverInitialAssets PostGFC ExcludingGFC (1) (2) (3) (4) GlobalFactor -2.495*** -1.746*** (0.443) (0.321) GlobalFactor*ETF -2.457*** -2.499*** -2.817*** -2.607*** (0.595) (0.594) (0.731) (0.592) LocalFactor 0.122* 0.195*** (0.062) (0.047) LocalFactor*ETF -0.122* -0.015 -0.147 0.029 (0.072) (0.077) (0.089) (0.076) LocalFactorETF -0.001 0.049 P-value 0.994 0.635 Observations 136,893 136,562 194,390 193,744 N.ofFunds 2,498 2,492 2,885 2,875 R2 0.087 0.136 0.069 0.139 Panel B: Bond Funds DependentVariable: FundFlowsoverInitialAssets PostGFC ExcludingGFC (1) (2) (3) (4) GlobalFactor -3.963*** -3.480*** (0.897) (0.860) GlobalFactor*ETF -5.087** -4.954** -5.301** -5.033** (1.973) (1.984) (2.062) (2.004) LocalFactor 0.106 0.091 (0.141) (0.125) LocalFactor*ETF -0.261 -0.182 -0.247 -0.180 (0.326) (0.371) (0.332) (0.372) FundControls Yes Yes Yes Yes FundFE Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes LocalFactorETF -0.155 -0.157 P-value 0.687 0.683 Observations 42,144 41,804 46,632 46,196 N.ofFunds 871 864 905 898 R2 0.106 0.186 0.100 0.182 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand differentsetsoffixedeffectsforemergingmarketfunds. Columns1and2excludethemonthsbetweenMarch2007andMarch2009. Columns 3 and 4 exclude the months before April 2009. Local Factor is the median monthly industrial production growth for the investment scope of each fund. Global Factor is the difference in the St. Louis Financial Stress Index. ETF is a dummy indicating whetherafundisanETFornot. FundControlsindicateswhethertheregressionincludesfundcontrolvariables. Thesevariablesare thethreelagsoftheportfolioreturnsofthefundminustheaveragefundreturnsattheinvestmentscopelevelandthedifferenceinlogs oftheMSCIstockmarketindexinthedomicileofeachfund. LocalFactorETFindicatesthesumofthecoefficientsforLocalFactors andLocalFactors*ETF.P-valueshowsthesignificancetestforLocalFactorETF=0. Fundflowsoverinitialassetsarewinsorizedat the1and99percentlevel. Driscoll-Kraayrobuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 54

Table A13: Robustness - Global Factors, Part 1 Panel A: Equity Funds DependentVariable: FundFlowsoverInitialAssets GlobalFactorVariable: VIX TEDSpread USHighYieldSpread (1) (2) (3) (4) (5) (6) GlobalFactor -1.637*** -0.961*** -3.806*** (0.378) (0.256) (1.018) GlobalFactor*ETF -2.464*** -2.662*** -1.289* -1.609*** -6.270*** -7.157*** (0.655) (0.628) (0.740) (0.575) (2.073) (1.781) LocalFactor 0.191*** 0.188*** 0.184*** (0.045) (0.044) (0.042) LocalFactor*ETF -0.135 0.027 -0.149* 0.010 -0.136* 0.022 (0.083) (0.071) (0.083) (0.071) (0.082) (0.069) FundControls Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes No Yes LocalFactorETF 0.056 0.039 0.048 P-value 0.562 0.682 0.601 Observations 210,194 209,498 210,194 209,498 210,194 209,498 N.ofFunds 2,906 2,897 2,906 2,897 2,906 2,897 R2 0.066 0.138 0.065 0.138 0.067 0.138 Panel B: Bond Funds DependentVariable: FundFlowsoverInitialAssets GlobalFactorVariable: VIX TEDSpread USHighYieldSpread (1) (2) (3) (4) (5) (6) GlobalFactor -2.603*** -1.391** -7.031*** (0.780) (0.648) (1.649) GlobalFactor*ETF -1.173 -1.129 -1.067 -1.803 -10.441*** -10.445*** (1.625) (1.469) (2.311) (1.977) (3.972) (3.307) LocalFactor 0.145 0.131 0.164 (0.122) (0.118) (0.117) LocalFactor*ETF -0.413 -0.280 -0.425 -0.301 -0.331 -0.211 (0.337) (0.363) (0.336) (0.363) (0.328) (0.352) FundControls Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes No Yes LocalFactorETF -0.269 -0.295 -0.167 P-value 0.489 0.440 0.652 Observations 50,510 50,029 50,510 50,029 50,510 50,029 N.ofFunds 910 901 910 901 910 901 R2 0.091 0.177 0.090 0.177 0.094 0.177 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand differentsetsoffixedeffectsforemergingmarketfunds. PanelAshowstheresultsforequityfundsandPanelBforbondfunds. Local Factor is the median monthly industrial production growth for the investment scope of each fund. Global Factor is the difference in logsofthevariableindicatedatthetopofeachcolumn. ETFisadummyindicatingwhetherafundisanETFornot. FundControls indicateswhethertheregressionincludesfundcontrolvariables. Thesevariablesarethethreelagsoftheportfolioreturnsofthefund minustheaveragefundreturnsattheinvestmentscopelevelandthedifferenceinlogsoftheMSCIstockmarketindexinthedomicile of each fund. Local Factor ETF indicates the sum of the coefficients for Local Factors and Local Factors*ETF. P-value shows the significancetestforLocalFactorETF=0. Fundflowsoverinitialassetsarewinsorizedatthe1and99percentlevel. Driscoll-Kraay robuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 55

Table A14: Robustness - Global Factors, Part 2 Panel A: Equity Funds DependentVariable: FundFlowsoverInitialAssets GlobalFactorVariable: PCA1 FedFundsRate ShadowFedFundsRate (1) (2) (3) (4) (5) (6) GlobalFactor -0.270*** 0.291 -0.188 (0.056) (0.459) (0.386) GlobalFactor*ETF -0.362*** -0.421*** -4.704*** -2.994** -2.402*** -1.834** (0.099) (0.088) (1.553) (1.350) (0.894) (0.718) LocalFactor 0.182*** 0.194*** 0.196*** (0.044) (0.044) (0.044) LocalFactor*ETF -0.141* 0.016 -0.147* 0.017 -0.167* 0.004 (0.079) (0.066) (0.088) (0.074) (0.091) (0.077) FundControls Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes No Yes LocalFactorETF 0.041 0.047 0.030 P-value 0.643 0.644 0.774 Observations 204,181 203,567 210,194 209,498 210,194 209,498 N.ofFunds 2,906 2,897 2,906 2,897 2,906 2,897 R2 0.069 0.139 0.064 0.138 0.064 0.138 Panel B: Bond Funds DependentVariable: FundFlowsoverInitialAssets GlobalFactorVariable: PCA1 FedFundsRate ShadowFedFundsRate (1) (2) (3) (4) (5) (6) GlobalFactor -0.479*** 1.834 -0.563 (0.094) (1.342) (1.009) GlobalFactor*ETF -0.411** -0.430** -14.136*** -10.676** -4.651** -3.830** (0.206) (0.174) (5.340) (4.869) (1.961) (1.556) LocalFactor 0.142 0.156 0.149 (0.115) (0.118) (0.127) LocalFactor*ETF -0.368 -0.255 -0.549 -0.398 -0.524 -0.368 (0.339) (0.362) (0.336) (0.351) (0.331) (0.354) FundControls Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes No Yes LocalFactorETF -0.226 -0.393 -0.375 P-value 0.555 0.312 0.328 Observations 50,510 50,029 50,510 50,029 50,510 50,029 N.ofFunds 910 901 910 901 910 901 R2 0.096 0.177 0.090 0.177 0.090 0.177 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand differentsetsoffixedeffectsforemergingmarketfunds. PanelAshowstheresultsforequityfundsandPanelBforbondfunds. Local Factoristhemedianmonthlyindustrialproductiongrowthfortheinvestmentscopeofeachfund. GlobalFactoristhefirstprincipal componentofthedifferenceinlogsfortheVIX,TEDRateandUSHYforthefirsttwocolumns. ForColumns(3)-(6)isthedifference of the variable indicated at the top of each column. ETF is a dummy indicating whether a fund is an ETF or not. Fund Controls indicateswhethertheregressionincludesfundcontrolvariables. Thesevariablesarethethreelagsoftheportfolioreturnsofthefund minustheaveragefundreturnsattheinvestmentscopelevelandthedifferenceinlogsoftheMSCIstockmarketindexinthedomicile of each fund. Local Factor ETF indicates the sum of the coefficients for Local Factors and Local Factors*ETF. P-value shows the significancetestforLocalFactorETF=0. Fundflowsoverinitialassetsarewinsorizedatthe1and99percentlevel. Driscoll-Kraay robuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 56

Table A15: Robustness - Local Factors Panel A: Equity Funds DependentVariable: FundFlowsoverInitialAssets AggregationMethod: Mean GDP-WeightedMean (1) (2) (3) (4) (5) (6) (7) (8) GlobalFactor -1.295*** -1.267*** -1.206*** -1.181*** (0.276) (0.270) (0.277) (0.270) GlobalFactor*ETF -2.603*** -2.280*** -2.616*** -2.275*** -3.362*** -2.845*** -3.341*** -2.796*** (0.618) (0.521) (0.620) (0.518) (0.669) (0.618) (0.685) (0.622) IPGrowth 0.125*** 0.256*** 0.236*** (0.031) (0.045) (0.046) IPGrowth*ETF -0.051 0.045 -0.125 0.086 -0.080 0.108 (0.068) (0.061) (0.102) (0.090) (0.122) (0.118) ExpectedShort-TermInterestRate -0.204*** -0.203*** (0.055) (0.053) ExpectedShortRate*ETF 0.062 -0.225* 0.070 -0.224 (0.127) (0.131) (0.137) (0.138) IPGrowthETF 0.074 0.130 0.156 P-value 0.341 0.237 0.237 ShortRateETF -0.142 -0.133 P-value 0.385 Observations 210,194 209,498 210,194 209,498 170,490 170,326 168,327 168,164 N.ofFunds 2,906 2,897 2,906 2,897 2,394 2,391 2,392 2,389 R2 0.066 0.138 0.066 0.138 0.071 0.133 0.073 0.132 Panel B: Bond Funds DependentVariable: FundFlowsoverInitialAssets AggregationMethod: Mean GDP-WeightedMean (1) (2) (3) (4) (5) (6) (7) (8) GlobalFactor -2.481*** -2.432*** -2.464*** -2.419*** (0.477) (0.474) (0.499) (0.492) GlobalFactor*ETF -3.848** -3.086* -3.936** -3.127* -4.283** -3.740** -4.322** -3.776** (1.936) (1.828) (1.951) (1.843) (1.978) (1.833) (1.988) (1.852) IPGrowth 0.089 0.249* 0.240* (0.080) (0.137) (0.139) IPGrowth*ETF -0.244 -0.192 -0.120 -0.075 -0.122 -0.030 (0.237) (0.248) (0.416) (0.465) (0.437) (0.479) ExpectedShort-TermInterestRate -0.270*** -0.249** (0.100) (0.104) ExpectedShortRate*ETF 0.950 1.304* 0.982 1.354* (0.906) (0.764) (0.909) (0.773) FundControls Yes Yes Yes Yes Yes Yes Yes Yes FundFE Yes Yes Yes Yes Yes Yes Yes Yes InvestmentScope*TimeFE No Yes No Yes No Yes No Yes IPGrowthETF -0.155 0.130 0.118 P-value 0.586 0.771 0.802 ShortRateETF 0.680 0.734 P-value 0.471 0.439 Observations 50,510 50,029 50,510 50,029 49,056 48,777 48,721 48,456 N.ofFunds 910 901 910 901 871 864 871 864 R2 0.093 0.177 0.094 0.177 0.094 0.174 0.094 0.173 Note: ThistablereportstheOLScoefficientsfromaregressionoffundflowsoverinitialassetsondifferentexplanatoryvariablesand differentsetsoffixedeffectsforemergingmarketfunds. PanelAshowstheresultsforequityfundsandPanelBforbondfunds. Local Factor is the either the mean or GDP weighted monthly industrial production growth for the investment scope of each fund. The variableusedisindicatedatthetopofeachcolumn. GlobalFactoristhedifferenceintheSt. LouisFinancialStressIndex. ETFisa dummyindicatingwhetherafundisanETFornot. FundControlsindicateswhethertheregressionincludesfundcontrolvariables. These variables are the three lags of the portfolio returns of the fund minus the average fund returns at the investment scope level and the difference in logs of the MSCI stock market index in the domicile of each fund. Local Factor ETF indicates the sum of the coefficientsforLocalFactorsandLocalFactors*ETF.P-valueshowsthesignificancetestforLocalFactorETF=0. Fundflowsover initial assets are winsorized at the 1 and 99 percent level. Driscoll-Kraay robust standard errors in parenthesis. *** p< 0.01, ** p<0.05,*p<0.10. 57

Table A16: Summary Statistics - Macro Variables, ETF Share, and Mutual Fund Share Panel A: Quarterly Sample PortfolioEquityInflows(%ofGDP) ETFShare MutualFundShare (1) (2) (3) Mean 0.365 0.440 1.964 StandardDeviation 1.079 0.734 2.413 10thPercentile -0.524 0.000 0.001 25thPercentile -0.084 0.000 0.253 Median 0.126 0.081 1.414 75thPercentile 0.617 0.685 2.904 90thPercentile 1.626 1.313 4.172 NumberofCountries 43 43 43 Observations 2035 2276 2276 Panel A: Monthly Sample StockMarketsReturns(basispoints) ETFShare MutualFundShare (1) (2) (3) Mean 0.003 0.434 1.840 StandardDeviation 0.082 0.727 2.288 10thPercentile -0.087 0.000 0.048 25thPercentile -0.035 0.000 0.271 Median 0.007 0.081 1.328 75thPercentile 0.048 0.657 2.658 90thPercentile 0.092 1.306 3.990 NumberofCountries 49 49 49 Observations 7755 8134 8134 Note: This table reports the summary statistics for macro variables for the sample used in the macro-level regressions. The sample is divided into the quarterly (Portfolio Equity Inflows) and monthly sample (Stock Market Returns). Panel A shows statistics for quarterlysampleandPanelBthemonthlysample. PortfolioEquityInflowsandStockMarketReturnsarewinsorizedatthe1and99 percentlevel. 58

Table A17: Robustness - Aggregate-level Regressions and Alternate Financial Integration Measures Panel A Dependent Variable: Balance of Payments Gross Portfolio Equity Inflows (% of GDP) (1) (2) (3) (4) (5) (6) (7) (8) GlobalFactor 0.038 0.020 0.044 -0.169 (0.110) (0.130) (0.093) (0.193) GlobalFactor*ETFShare -0.716*** -0.729*** -0.727*** -0.714*** -0.687*** -0.718*** -0.708*** -0.711*** (0.097) (0.069) (0.099) (0.070) (0.101) (0.077) (0.101) (0.073) ETFShare -0.129** -0.026 -0.132** -0.006 -0.112** -0.038 -0.120** -0.018 (0.056) (0.050) (0.055) (0.051) (0.055) (0.050) (0.057) (0.053) GlobalFactor*EquityIntegration -0.015 -0.014 (0.051) (0.052) EquityIntegration(L-MF) 0.050 0.066 (0.077) (0.081) GlobalFactor*FDIIntegration 0.054 0.068 (0.090) (0.085) FDIIntegration(L-MF) 0.133*** 0.148*** (0.021) (0.021) GlobalFactor*EquityInflowsControls -0.087 -0.058 (0.173) (0.195) CapitalControlsEq. Inflows -0.309** -0.231 (0.130) (0.168) GlobalFactor*Chinn-ItoFO 0.334* 0.334* (0.183) (0.182) Chinn-ItoFO -0.279 -0.283 (0.280) (0.276) Observations 2,035 2,035 2,035 2,035 1,745 1,745 2,030 2,030 N.ofCountries 43 43 43 43 37 37 43 43 R2 0.156 0.223 0.186 0.258 0.165 0.235 0.155 0.220 Panel B Dependent Variable: MSCI Country Stock Markets Returns (basis points) (1) (2) (3) (4) (5) (6) (7) (8) GlobalFactor -10.68*** -10.50*** -12.19*** -10.54*** (1.74) (1.76) (2.06) (1.80) GlobalFactor*ETFShare -4.17*** -3.70*** -4.22*** -3.70*** -4.21*** -3.40*** -4.51*** -4.07*** (1.05) (0.71) (1.05) (0.71) (1.13) (0.74) (1.03) (0.69) ETFShare -0.67** -0.29* -0.66** -0.30* -0.68** -0.34** -0.71** -0.39** (0.32) (0.16) (0.32) (0.16) (0.31) (0.17) (0.33) (0.16) GlobalFactor*EquityIntegration 0.16 0.15 0.18* (0.11) (0.10) (0.10) EquityIntegration(L-MF) -0.01 0.11 (0.16) (0.10) GlobalFactor*FDIIntegration -0.17 -0.13 (0.16) (0.15) FDIIntegration(L-MF) -0.10* -0.03 (0.05) (0.04) GlobalFactor*EquityInflowsControls 3.08*** 3.44*** (0.85) (0.73) CapitalControlsEq. Inflows -0.69 0.04 (0.86) (0.64) GlobalFactor*Chinn-ItoFO -0.04 -0.56 (1.61) (1.59) Chinn-ItoFO 0.79 1.39* (1.47) (0.77) Observations 7,755 7,755 7,755 7,755 6,756 6,756 7,584 7,584 N.ofCountries 49 49 49 49 42 42 48 48 R2 0.188 0.417 0.188 0.417 0.190 0.431 0.187 0.416 Note: This table reports the OLS coefficients from a regression of Balance of Payments Portfolio Equity Liability Flows over GDP (Panel A) or MSCI Country Stock Market Returns (Panel B) on different explanatory variables and different sets of fixed effects for emergingmarketsatthequarterlyfrequency. GlobalFactoristhedifferenceintheSt. LouisFinancialStressIndex. ETFShareisthe assets under management of equity ETFs divided by the total equity market capitalization. Equity (FDI) Integration is total equity (FDI)assetsplusequity(FDI)liabilitiesdividedbyGDPfromLaneandMilesi-Ferretti(2018). EquityInflowsControlsistheequity infowscapitalcontrolsmeasurein?. Chinn-ItoFOisthefinancialopennessfrom?. Alltheestimationsarefortheperiod2000-2017. PanelAestimationsareatthequarterlyfrequencyandPanelBatthemonthlyfrequency. Thedependentvariableiswinsorizedatthe 1and99percentlevel. Driscoll-Kraayrobuststandarderrorsinparenthesis. ***p<0.01,**p<0.05,*p<0.10. 59