WPS8405 Policy Research Working Paper 8405 Capital Inflows, Equity Issuance Activity, and Corporate Investment Charles W. Calomiris Mauricio Larrain Sergio L. Schmukler Development Research Group Macroeconomics and Growth Team April 2018 Policy Research Working Paper 8405 Abstract This paper uses issuance-level data to study how equity cap- other countries’ attractiveness to foreign investors. Shifts ital inflows that enter emerging market economies affect in the supply of foreign capital are important drivers of equity issuance and corporate investment. It shows that increased equity inflows. Instrumented contemporaneous foreign inflows are strongly correlated with country-level and lagged capital inflows lead large firms to raise new equity, issuance. The relation especially reflects the behavior of which they use to fund investment. The results indicate large firms, defined as those with large market value of that inflows imply more than a transfer of equity owner- equity. To identify supply-side shocks, capital inflows into ship from domestic to foreign investors. Foreign purchases each country are instrumented with exogenous changes in of equity have financial and real consequences for firms. This paper is a product of the Macroeconomics and Growth Team, Development Research Group. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/research. The authors may be contacted at cc374@columbia.edu, mauricio.larrain@uc.cl, and sschmukler@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Capital Inflows, Equity Issuance Activity, and Corporate Investment Charles W. Calomiris Mauricio Larrain Sergio L. Schmukler * JEL Classification Codes: F23, F32, F65, G11, G15, G31 Keywords: capital flows, corporate financing, emerging markets, domestic investors, foreign investors, use of funds * Calomiris: Columbia Business School, Hoover Institution, and NBER, email: cc374@columbia.edu; Larrain: Universidad Catolica de Chile School of Management and Financial Market Commission of Chile, email: mauricio.larrain@uc.cl; Schmukler: World Bank Research Department, email: sschmukler@worldbank.org. We thank Laura Alfaro, Gustavo Araujo (discussant), Nathan Converse (discussant), Erik Gilje (discussant), Aart Kraay, Jeanne Lafortune, Atif Mian, Gabriel Natividad, Amit Seru, Luis Serven, Jose Tessada, Shang-Jin Wei, Ilknur Zer (discussant), and participants at presentations held at the ASSA Annual Meetings, Central Bank of Brazil’s Conference on Financial Stability and Banking, Finance UC Conference, Fordham University, Hebrew University, HEC Paris, IMF, Kansas City Fed, LACEA-LAMES Annual Meetings, Southern Economic Association Annual Meetings, Telfer Annual Conference on Accounting and Finance, University of Chile, University of Michigan, University of Santiago Chile, and Vienna Graduate School of Finance for useful comments. We are grateful to Soha Ismail for superb research assistance, to Facundo Abraham, Juan Cortina, Marta Guasch Rusiñol, and Ruth Llovet for their valuable help at different stages of the project, and to Tatiana Didier and Tomas Williams for facilitating access to data. Larrain acknowledges funding from Proyecto Fondecyt Iniciación #11160879. Schmukler thanks the financial support of the World Bank Knowledge for Change Program (KCP) and Strategic Research Program (SRP). The views expressed here do not necessarily represent those of the World Bank. I. Introduction Capital inflows are prevalent in emerging market countries. In 2016, foreign investors invested around 64 billion U.S. dollars into emerging countries in the form of portfolio equity, i.e., foreign investors’ purchases of stocks of publicly traded emerging market firms. Despite this large magnitude, one can question whether capital inflows are economically important in spurring economic development in emerging economies. It is conceivable that large flows of equity associated with significant cross-border diversification of investors’ equity holdings might not translate into large effects on firms raising capital in equity markets. A central question is whether capital inflows benefit recipient countries by encouraging firms to issue new equity to fund investment and growth. Most of the evidence that has been assembled to address that question has been indirect, based on aggregate relations rather than firm-level behavior (Kose et al., 2009). In this paper, we contribute to the debate about the impact of equity capital inflows by providing direct micro evidence at the firm level on the channels of transmission. We study how equity capital inflows affect the economies that receive them by analyzing their connection to equity financing (through firm-level equity issuance) and to real economic activity (through changes in corporate balance sheets). We investigate whether publicly traded firms in emerging countries issue more equity when their countries experience increases in foreign equity purchases and what firms do with the equity capital they raise. In particular, we ask whether the firms that raise new equity use those funds to finance corporate investment.1 1 We focus on equity financing, which is especially important for expanding the growth opportunities of firms. Most of the literature on financial globalization and financial liberalization has focused on equity markets. Moreover, a larger array of firms from emerging economies issue equity than issue public debt, which allows us to consider heterogeneous effects across a wider range of firms. Furthermore, well-known indexes for equity markets exist, which guide foreign investors to invest in those economies. In addition, financial statements data on firms across countries are mostly available for those firms listed in equity markets. These data are important to understand the real effects of capital inflows. 1 We also analyze whether firms of different size differ in the degree to which their issuance of new equity responds to increased funding by foreign equity investors. Ours is the first study, of which we are aware, that examines the links between capital inflows and investment using issuance-level data. To do so, we assemble a large granular dataset containing information on 20,306 seasoned equity issuances (SEOs) and financial statements for 12,723 firms in 25 emerging market countries, in addition to capital inflows during the 26-year period 1991-2016. The total number of firm-year observations is 330,798. Figure 1 presents the evolution of the aggregate amount of equity investing by foreign investors into our sample countries alongside the total value of seasoned equity raised by firms in those countries. The figure shows that periods of large capital inflows coincide with periods of active equity issuance activity. This shows that inflows are associated with more than a simple transfer of equity ownership from domestic to foreign investors, and suggests that issuances are not purchased solely by domestic investors. In terms of causality, this correlation could reflect supply-side variation. For example, foreign equity inflows resulting from greater global liquidity increase the supply of funds available to firms.2 Such an increase in the supply of foreign participation in equity markets implies a higher investor demand for equity and lower required equity returns, which in turn encourage destination countries’ firms to issue shares. Alternatively, the positive correlation between inflows and issuances could reflect shifts in the demand for capital by firms (resulting, for example, from improvements in investment opportunities at the country level). According to that view, better firms’ prospects induce foreign investors to send more funds to the country. 2Chari et el. (2017) document significant effects of U.S. monetary policy surprises around FOMC meetings on capital flows from the U.S. to a range of emerging markets as well as on the associated emerging market valuations. They find that equity positions and valuations are more sensitive to monetary policy shocks than debt positions and valuations. 2 By using firm-level data together with a novel set of instrumental variables, we distinguish between supply-side and demand-side influences. We first estimate the country- level relation between capital inflows and equity issuance. We regress each country’s equity issuances on equity inflows, taking into account country and year fixed effects. We find a strong association between the two. For every million U.S. dollars foreign investors purchase of emerging market equity, the value of seasoned issuance proceeds increases by 160,000 U.S. dollars. We then move to the micro-level analysis to better understand the mechanisms that drive these aggregate issuance patterns. We first regress firm-level issuance proceeds on equity capital inflows, controlling for firm and year fixed effects. We find only a weakly significant correlation between capital inflows and the value of issuance proceeds for the typical emerging market firm. When we divide firms into groups according to size, we find a large and statistically significant effect of equity inflows on the issuances of large firms. We define large firms as those in the top decile of size, measured by market value of equity within a country in the previous year. This result is consistent with the well-known fact that international investors are particularly interested in investing in the shares of large, well-established emerging market firms. The higher responsiveness of large firms’ issuances to equity capital inflows remains even after controlling for country-year fixed effects, which allows us to control for all time- varying country shocks. Moreover, we find that the relation between equity capital inflows and equity issuance is monotonically increasing in firm size. Given that an increase in issuance activity could reflect either an increase in foreign equity supply by investors or greater domestic equity demand by firms, we develop an instrumental variable approach that identifies supply-side influences. We identify supply 3 shocks to equity inflows by measuring shifts in foreign investor interest unrelated to changes in domestic firms’ prospects. In particular, we instrument equity inflows using various measures of the attractiveness of other countries’ equity markets to foreign investors, which we argue are plausibly exogenous to demand-side shocks to the subject country. The idea is that for a given amount of total capital inflows to emerging markets as a whole, positive shocks to other countries’ attractiveness to foreign investors constitute negative shocks to the subject country’s supply of funds.3 We provide regression results using three alternative instruments, all of which are strongly correlated with equity capital inflows. As our first instrument, we use the lagged weight of a country in the MSCI Emerging Markets stock index. The MSCI weight of a country is a function of both the relative market capitalization of that country’s stock market (in relation to the global market capitalization of 25 emerging economies) and sporadic decisions by MSCI on the index constitution. When institutional investors receive funds from their underlying investors, they tend to invest those funds into emerging economies’ equity markets according to the weights of those economies in the MSCI index (Raddatz et al., 2017). The time variation of each country’s MSCI weight should primarily reflect shocks to the market values of the other 24 countries’ stock markets.4 3 As we discuss in Section 3, other facts about the association between equity inflows and issuances are suggestive of supply-driven causality. The fact that the response of issuances to equity inflows is greater among large firms suggests a supply-side channel. If inflows were simply responding to improved economic conditions in the country, one might expect all firms, not just large ones, to issue more equity at times of large capital inflows (our country-year fixed effects control for all shocks that affect all firms equally in a country). The finding that inflows tend to prompt issuances particularly by large firms suggests that those firms may be especially attractive to foreign equity investors. However, one cannot rule out that demand-side shocks are heterogeneous across firms. The fact that lagged inflows predict equity issuances is also suggestive of supply-driven causality, but here too, it is conceivable that serially correlated demand shocks could account for the correlation between lagged equity inflows and equity issuances. 4 Our results remain unchanged when we exclude the largest countries in the MSCI index from our analysis. 4 Because the MSCI weights are partially affected by domestic shocks that change a country’s relative market value, we also employ as alternative instruments the sum of other countries’ total equity value and other countries’ volume of equity issuances. Those two instruments depend exclusively on foreign shocks and are therefore plausibly exogenous to demand shocks of the subject country. To deal with the possibility that equity inflows to emerging markets can be correlated across countries, our specification includes year fixed effects, which control for common shocks across them. To remove any further common shocks, we also employ orthogonalized versions of those same instruments, where we orthogonalize the sum of equity values (the sum of issuances) with respect to the market value (the issuances volume) in the subject country. In all cases, we find that instrumented inflows, contemporaneous and lagged, lead large firms to raise significantly more equity. Our data on capital inflows are measured annually and the precise timing lag between the stimulus of instrumented capital inflows and the response of issuances is not known a priori. Therefore, when modeling the timing of the connection between instrumented capital inflows and equity issuances, we consider two alternative dynamic formulations. In one formulation, we regress issuances on contemporaneous inflows (where it is possible that issuances could occur at the same time as the capital inflows, or slightly before or after the inflows but within the same year). In the second formulation, we regress issuances on lagged inflows. Our results are similar under either of the two formulations.5 In the last section of the paper, we analyze the real effects of equity capital inflows. First, we estimate the effect of instrumented equity capital inflows on a variety of potential 5 Because we employ instrumented inflows, we do not rely on lagging inflows for identification. However, by using lagged values of instrumented inflows, we reduce the likelihood that the relation between inflows and issuances reflects the response of inflows to increased demand for capital by issuing firms. 5 uses of funds: capital expenditures (CAPEX), corporate acquisitions, research and development expenses (R&D), inventory accumulation, cash and short-term investments, and long-term debt reduction.6 We show that instrumented capital inflows lead large emerging market firms to significantly increase CAPEX. They also tend to make more acquisitions, increase R&D, accumulate inventories, hoard cash and short-term investments, and reduce long-term debt. Second, we study more precisely how large firms use the funds raised in their equity offerings over time, controlling also for other sources of funding. We measure the increases in each use of funds over a four-year period. Our estimates indicate that the largest use of funds after the first year is CAPEX. For every million U.S. dollar raised in an offering, large firms as a group spend on average 700,000 U.S. dollars on investment four years after the issuance. This effect takes place as firms use the additional cash and short-term investments accumulated during the issuance year. Using a back-of-the-envelope calculation, our analysis indicates that every million U.S. dollar of foreign equity capital results in seasoned equity offerings that fund an increase of about 110,000 U.S. dollars of corporate investment. The 110,000-dollar estimate is the result of 160,000 U.S. dollars of additional secondary equity issuance (in response to every million U.S. dollar of inflows) and 700,000 U.S. dollars of average additional spending on investment (for every million U.S. dollar of capital raised). Capital inflows appear to reduce the cost of equity finance, allowing emerging market firms to finance new investments. In all, equity issuance seems to be an important channel through which capital inflows affect real economic activity. Our paper is related to several strands of the literature. First, there is a literature on how aggregate economic activity is affected by the liberalization and flow of equity capital 6 These are the six uses of funds analyzed by Kim and Weisbach (2008) and Erel et al. (2011). 6 (Henry, 2000a; Henry, 2000b; Alfaro et al., 2004; Bekaert et al., 2005; Kose et al., 2010). These papers show that equity inflows are associated with a boom in aggregate investment and higher economic growth of the recipient countries. However, we know relatively little about the channels through which equity inflows affect real economic activity.7 Our paper adds to this literature by studying for the first time the effects of capital inflows using issuance-level data. We show that supply-side changes in capital inflows allow firms to raise new financing and expand investment, which might be behind the patterns documented in this literature. Moreover, our paper shows that the effects are not uniform across types of firms. Capital inflows seem to reduce the cost of equity finance for large firms more than for other firms, which likely reflects the fact that differences in firm size is a proxy for differences in global investors’ knowledge and interest in firms. The literature has found it challenging to disentangle supply and demand influences when gauging the effects of capital inflows on financial and real economic activity in a multi- country setting. Examining the case of one country, using detailed firm-level data, Baskaya et al. (2017a) and Baskaya et al. (2017b) isolate supply-side influences on capital inflows. In this paper, we propose a novel set of instruments in a multi-country context to distinguish between supply-side and demand-side effects on capital inflows, and we find that the supply side is important. Our paper also contributes to another literature that asks why firms issue equity and bonds geared toward foreign investors. Part of this literature has studied firms’ issuance activity in international markets, characterizing which firms issue abroad and why. Foreign markets can offer benefits compared to domestic ones in terms of access to better financing 7 Mitton (2006), Gupta and Yuan (2009), Levchenko et al. (2009), and Igan et al. (2016) use industry- and firm- level data to study the effects of liberalizing equity markets on industry growth and firms’ operating performance. 7 conditions, greater visibility, and enhanced corporate governance, among others (Pagano et al., 2002; Benos and Weisbach, 2004; Doidge, 2004; Karolyi, 2006; Schmukler and Vesperoni, 2006; Claessens and Schmukler, 2007; Forbes, 2007; Doidge et al., 2009).8 Although this literature assumes that issuances abroad target foreign investors, it has not shown that facilitating foreign investor participation actually influences issuances. 9 In practice, it is hard to track the influence of foreign investors on firm behavior because there are no data identifying the nationality of who buys each security. Data are available, however, on the change in net purchases by foreigners of each country’s publicly traded firms’ equity, which is our measure of equity capital inflows. Our approach to identification allows us to use those data to link foreign participation in equity markets with consequences for each country’s equity issuances in domestic and foreign markets. A separate literature (Pagano et al., 1998; Kim and Weisbach, 2008; Brown et al., 2009; Erel et al., 2011; Didier et al., 2015) analyzes how firms use new capital market financing from various sources. We complement this strand of the literature by linking the use of funds with inflows of foreign capital. In particular, we study how shifts in the supply of equity financing affect the use of funds by the emerging market firms that tend to raise capital when their country receives capital inflows. We also find that firms use the proceeds primarily to expand investment, aside from any activity geared toward retiring debt and accumulating cash. The rest of this paper is organized as follows. Section II discusses the data sources. Sections III explains our empirical strategy. Section IV reports country- and firm-level results linking capital inflows and issuance activity. Section V reports instrumented results for the 8 Other papers argue that, as liquidity became more abundant in the aftermath of the global financial crisis of 2008-09, firms issued more foreign currency bonds to take advantage of carry-trade opportunities (Chui et al., 2014; Powell, 2014; Caballero et al., 2016; Bruno and Shin, 2017; Moreno and Serena Garralda, 2018). 9 Forbes (2007) studies the effects of the “encaje” controls on capital inflows in Chile from 1991 to 1998. She finds evidence that imposing the encaje on equity inflows reduced aggregate equity issuances. 8 responses of issuances to supply-side factors. Section VI reports the use-of-funds analysis. Section VII concludes. II. Data We collect data on capital inflows using balance of payments information from the International Monetary Fund (IMF). The IMF provides data on annual private gross capital inflows and outflows by category: foreign direct investment (FDI), portfolio equity, portfolio debt, bank credit, and others. We focus on portfolio equity inflows, defined as the difference between foreign purchases of domestic shares and foreign sales of domestic shares. Equity inflows are positive (negative) when foreign investors purchase more (less) domestic securities than what they sell. Foreign retail investors and foreign institutional investors (such as mutual funds, pension funds, hedge funds, and sovereign wealth funds) are often behind the foreign purchases and sales of domestic shares. Those investors purchase both existing and newly issued shares. To verify that the relation we observe between equity capital inflows and equity issuances reflects influences that are specific to equity markets, in additional analysis reported in the Appendix, we also use data on non-equity capital inflows, including other categories of capital inflows to the private sector (FDI, portfolio debt, and other debt investments in banks and other sectors, excluding general government and central banks), which are in fact positively correlated with equity inflows. Our sample consists of the 25 emerging market countries included in the MSCI Emerging Markets index (explained below) during the 26-year period 1991-2016. The countries are: the Arab Republic of Egypt, Argentina, Brazil, Chile, China, Colombia, Czech Republic, Hungary, India, Indonesia, Israel, Jordan, Malaysia, Mexico, Morocco, Pakistan, 9 Peru, the Philippines, Poland, the Republic of Korea (or South Korea), the República Bolivariana de Venezuela, the Russian Federation, South Africa, Thailand, and Turkey. We focus mainly on positive equity inflows, which represent more than 84% of all inflow observations. Our focus on positive inflows reflects our goal to analyze whether firms issue more equity after foreign capital arrives to their country, and how those equity proceeds are employed. Negative capital inflows, on the other hand, represent a departure of foreign capital. Although it is conceivable that firms might repurchase equity when foreign capital departs their country (a negative issuance), existing empirical evidence finds no connection between outflows and investment behavior by publicly traded firms, which suggests that negative issuance is not a common response to outflows.10 Our issuance data (explained below) do not provide information on stock repurchases, so we focus on the positive issuance implications of positive capital inflows. However, for robustness, we analyze the effects of negative inflows on equity issuance. Our findings confirm the view that negative equity inflows have no significant effects on equity issuance.11 The data on equity issuance activity come from the Thomson Reuters Security Data Corporation Platinum database (SDC Platinum). This database provides transaction-level information on new issuances of common equity by publicly traded firms. The transactions in the database include 20,306 SEOs. The database also includes transactions related to initial public offerings (IPOs), which we do not use except when explicitly indicated. Because the database covers the universe of issuance transactions, we assume zero issuance activity for the 10 Tong and Wei (2010) and Claessens et al. (2012) investigate stock price reactions and real investment changes associated with the large capital outflows produced by the global financial crisis. They find a significant negative effect on stock prices, but no effect on investment. That finding is consistent with firms not responding to capital outflows and lower stock prices with significant repurchases of their shares. 11 Specifically, we find that negative inflows tend to produce an asymmetric response in equity issuance. Because negative inflows reduce issuances only slightly, even large negative inflows are still associated with large positive issuances. 10 firm-year observations when no positive transactions are recorded for a given publicly listed firm in a given year (as is common in the finance and trade literatures that work with transaction-level data). A firm enters the sample since its IPO year. We obtain the aggregate issuance data by adding all the transactions for firms in each country-year. In both our aggregate and firm-level regression analysis we focus on SEOs for two reasons. First, our firm-level regression analysis employs lagged information about firms, including the lagged market value of equity, which is only available for firms that were publicly traded in the prior period. To ensure comparability with our firm-level estimates, we exclude IPOs from our aggregate analysis. Second, we wish to understand how publicly traded firms’ issuance decisions respond to foreign investor interest. Including initial offerings would mix two different phenomena: the responses of private firms (which choose to become public) with the responses of preexisting publicly traded firms. In the Appendix, we show that our aggregate results remain unchanged if we include IPOs in our sample. The data include issuances in international and domestic equity markets.12 Equity issuances are sold to a combination of domestic and foreign investors. We have data on a total of 12,723 firms and a total of 330,798 firm-year observations. We include both financial and non-financial firms. Each group has a significant share of the issuance activity. The issuance activity of financial firms is relevant for the financing of investment by non-financial corporations, although financial firms do not directly engage in capital investment. In the Appendix, we show that the results on issuance activity are robust to excluding financial firms from our sample. 12 An issuance is classified as international if the firm’s country of origin is different than the country where the equity is raised. SDC classifies the majority of newly issued shares that are destined to become depository receipts (including American Depositary Receipts and Global Depositary Receipts) as international issuances. 11 The sample of firms used is a result of merging the SDC data with Thomson Reuters Worldscope data, which provide information on firms’ financial statements (balance sheets, income statements, and cash flow statements) for publicly listed firms. We need this information both to classify firms as large and for the use-of-funds analysis. After matching the two sources, Worldscope data are available for publicly listed firms and for 70% of the equity issuers contained in the SDC database of the 25 emerging markets under study. The information on market value of equity, which we use to classify firms by size, is available from the IPO year onward. Table 1 reports summary statistics of issuance activity by country. Column (1) reports the number of firms included in the sample. Columns (2) and (3) show the average annual value of equity issuance proceeds for all firms in a country and the value of proceeds per firm, respectively. In a typical year, the average firm in the sample (including issuers and non-issuers) issues equity worth 6 million U.S. dollars. Because most firms are not actively issuing equity every year, the value of proceeds per issuing firm is much larger. For the first of our three instrumental variables, we collect data from the MSCI Emerging Markets Index for the period 1996-2016.13 The MSCI index is a stock market index covering 25 emerging market countries representing 10% of global stock market capitalization. The index covers approximately 85% of the free float-adjusted market capitalization in each country. The index is maintained by MSCI Inc., formerly Morgan Stanley Capital International, and is used as a common benchmark for international equity mutual funds. Appendix Figure 2 plots the average weights of the 25 countries in the MSCI Index. 13The instrumental variable analysis that uses the lagged MSCI weights is restricted to the sample 1997-2016 because the MSCI weights are available only for 1996-2015. An advantage of the other two instruments (other countries’ market value and issuance volume) is that they can be estimated over the whole sample period, 1991- 2016. The results using the second and third instruments are robust to restricting the sample to the period covered by the MSCI data. 12 III. Empirical Strategy and Identification Our presentation of empirical findings begins with ordinary least squares (OLS) results at the country level. The results show a strong empirical relation between country-level equity inflows and seasoned equity issuances. These results do not provide a causal interpretation of the links between issuances and equity inflows, but they do document an important new fact: increases in equity inflows are associated with increases in equity issuance, and that is true after controlling for country and time fixed effects. The country-level results do not permit a causal interpretation because they do not distinguish between supply-side and demand-side influences. Supply-side factors include increased global liquidity or global appetite for risk (depending on each country’s sensitivity to those global shocks), or idiosyncratic changes in foreign appetite for investing in particular countries, which could reflect changes in constraints on international investments, improvements in a destination country’s property rights, or legal institutional improvements (Stulz, 2005; Karolyi, 2015). Demand-side factors are any changes that affect investment opportunities, such as changes in productivity, technology, or local economic conditions. For example, improvements in firm productivity within the subject country might drive both equity inflows and issuances. In that case, although foreign investors’ willingness to provide equity inflows could facilitate adjustment to demand-side shocks (by reducing the cost of issuances), changes in foreigners’ interest in investing might not be an important source of change in either inflows or issuances. Several papers document that supply-side factors have been more important than demand-side factors in explaining capital inflow episodes in emerging economies (Forbes and Warnock, 2012; Fratzscher, 2012; Avdjiev et al., 2018). 13 In this paper, we employ instrumental variable estimation to disentangle demand-side from supply-side effects in explaining the relation between capital inflows and equity issuances. We make use of the fact that, for a given amount of capital inflows to emerging markets as a whole, positive shocks to other countries’ attractiveness to foreign investors constitute negative shocks to the subject country’s supply of funds. Our instruments capture changes in the attractiveness to foreign investors of other emerging market countries (for a given total amount of inflows, which we capture by a time fixed effect). Valid instruments should be strongly correlated with capital inflows and should also satisfy the exclusion restriction that they are not correlated with demand-side influences within the subject country. We employ three alternative measures as instruments. To start, we instrument the equity inflows received by a country with the lagged weight of that country in the MSCI Emerging Markets stock index. The MSCI weight of a country depends primarily on the market capitalization of that country’s stock market, relative to the global market capitalization of 25 emerging economies. The weight also depends, to a lesser extent, on MSCI’s adjustments to country weights for factors that they regard as relevant to foreign investors. Because all country weights sum to 100%, variation in a country’s weight in the MSCI index should primarily reflect shocks to the market values of the other 24 countries’ stock markets (and to MSCI’s adjustments to country weights), which are plausibly exogenous to subject country demand-side shocks. That should be especially true for small countries. Changes in MSCI weights should affect capital inflows not only as an indicator of market value changes in other countries, but also because some foreign investors, such as emerging market mutual funds, follow closely the MSCI index when setting their portfolio holdings (Raddatz et al., 2017). When those investors receive funds from their ultimate fund suppliers, they invest those funds into emerging economies’ equity markets according to the 14 proportion of those economies in the MSCI index. For illustrative purposes, Panel A of Figure 2 shows the positive relation between Mexico’s MSCI weight and its equity inflows. In robustness tests, we run the same instrumental variable regressions for a sample of emerging markets that excludes large countries (which should have greater effects on their own MSCI weights by virtue of their size). Although the primary source of variation in MSCI weights is external to each country, MSCI weights are still partially affected by domestic shocks (including demand-side shocks to firms’ productivity). For that reason, we also employ alternative instruments that do not suffer from that problem. Our alternative measures of the attractiveness of other countries are the aggregate value of equity in other emerging market countries, and the volume of equity issuances in other emerging markets. Both of these instrumental variables will also affect the MSCI weights of a country, but only as the result of variation coming from outside the country. Whereas the market value of equity, or issuances, in other countries reflects a mix of supply- and demand-side influences within those other countries, from the standpoint of the subject country, they are plausibly exogenous influences on the supply of funding. If increases in the value or volume of issuances in other countries are associated with capital inflows into those other countries (as our aggregate results suggest), then from the standpoint of the subject country, the diversion of capital inflows into other countries is a negative capital supply shock. For illustrative purposes, Panel B of Figure 2 plots Mexico’s equity inflows against the equity issuances of other countries, which are negatively correlated. Lastly, it is conceivable that market values of equity, or equity issuances, in other countries could be correlated with local economic conditions in the subject country in the case of common shocks (Forbes and Warnock, 2012; Ahmed and Zlate, 2014). Such a correlation 15 would violate the exclusion restriction. In our first-stage regression, we account for year fixed effects, which control for common shocks. That said, to be sure that we get rid of common- shock influences, we orthogonalize other countries’ equity value, or issuances, by removing any correlation between other countries’ equity value or issuances with the equity value or issuances in the subject country. Even before employing our instrumenting, one could argue that firm-level differences in the magnitude of the connection between capital inflows and issuances are indicative of a supply-side story. If demand-side influences are largely common across firms of different sizes, then greater issuance responses to capital inflows for large firms likely reflect supply-side differences resulting from differential access to international investors. From that perspective, even our firm-level OLS analysis, which controls for firm and year fixed effects, and which displays a stronger association between equity inflows and issuance for large firms, is suggestive of supply-side influences.14 Furthermore, the relation between inflows and issuances also applies to lagged inflows. A positive relation between lagged inflows and issuances suggests a supply-side influence whereby equity inflows put upward pressure on stock prices, which reduces the cost of equity capital, encouraging destination countries’ firms to issue equity, a decision that occurs with a lag. Although differential effects for large and small firms and observed connections between lagged inflows and subsequent issuances are both suggestive of supply-side causation, these facts by themselves are not conclusive. It is conceivable that demand-side influences differ across firms of different sizes. And serial correlation in demand shocks could produce 14In a prior version of this paper (available as NBER Working Paper 24433), we sorted firms by size according to the average size of their issuances over the sample period and obtained similar results. 16 a relation between lagged inflows and current issuances. For these reasons, although we believe that all of the various pieces of evidence point toward a supply-driven story, we do not rely only on firm-level differences or lags for identification, but emphasize our instrumenting of equity inflows. IV. Equity Inflows and Issuances A. Capital Inflows and Issuance Activity in the Aggregate As explained in the Introduction, Figure 1 displays the relation between global capital inflows and global equity issuance values. These two worldwide time series are significantly positively correlated: the correlation coefficient is 0.35.15 In Figure 3, we alternatively plot the time series of global equity inflows scaled by GDP and global equity issuances scaled by GDP. To control for country and year effects, we estimate the following country-level panel regression: log⁡ (1 + ) = + + log() + , (1) where denotes the value of seasoned equity issuance proceeds (in million U.S. dollars) by all firms of country c in year t and refers to equity capital inflows (in million U.S. dollars) received by country c in year t. We use the log of issuance plus one (million U.S. dollars) to account for country-year observations with zero issuances (15% of the total). and capture country and year fixed effects, respectively. We cluster standard errors of this regression, and all other analogous regressions reported below, by country and year.16 15 Appendix Figure 1 reports aggregate patterns that include both IPOs and SEOs. The fluctuations over time are similar to those that use only SEOs, but the size of the value of issuance activity is substantially greater when IPOs are included. 16 Our results remain unchanged if we cluster standard errors at the country level. 17 Table 2 shows a highly significant positive relation between capital inflows and country-level issuance proceeds in emerging markets. Column (1) shows that the elasticity of issuances to inflows is 0.52.17 This indicates that inflows imply more than a simple transfer of equity ownership from domestic to foreign investors. The result implies that for the typical country in a typical year, every million U.S. dollars of equity capital received from foreign investors is associated with an increase in the value of seasoned equity proceeds of 160,000 U.S. dollars.18 To make sure that our results are not affected by the log specification, which excludes negative inflows, we re-estimate Equation (1) scaling country issuances and all equity inflows (positive and negative) by GDP. Column (2) of Table 2 reports the results of this alternative specification. Increases in capital inflows, relative to GDP, are strongly correlated with greater equity issuances, relative to GDP. We also report results separately for positive and negative equity inflows relative to GDP in Columns (3) and (4) of Table 2. Interestingly, the coefficient magnitude for negative inflows is much smaller and statistically insignificant, suggesting only a small reduction in issuances, which remain positive, even when capital inflows are highly negative. Given this asymmetry in the relation between equity inflows and aggregate issuances, we focus on positive 17 In Appendix Table 1, Column (2), we re-estimate Equation (1) using as dependent variable the sum of SEOs and IPOs. The results remain unchanged, while the dollar effect increases to 300,000 U.S. dollars raised for every million U.S. dollar of inflows. In Appendix Table 1, Column (3), we re-estimate Equation (1) excluding financial sector firms from our sample. Again, the results do not change. 18 To calculate the dollar effects, we first calculate the predicted equity issued for each country-year pair by replacing the corresponding equity inflows into Equation (1) and using the estimated coefficients from the regression results. As fixed effects, we use the coefficients for each year and country for the corresponding country-year pair. We then increase equity inflows by one million U.S. dollars and repeat the procedure, which yields the new predicted issuance. Next, we compute the difference between the two predicted values. For each country, we take the median of the differences across all years and report the value for the median country. 18 equity inflows in our empirical analysis of the effects of equity inflows on issuance decisions at the firm level.19 B. Capital Inflows and Firms’ Issuance Activity To analyze the impact of equity capital inflows on firms’ issuance activity, we estimate a firm- level panel regression accounting for firm and year fixed effects: log⁡ (1 + ) = + + log⁡ () + , (2) where is the value of seasoned equity raised (in million U.S. dollars) by firm i in country c in year t. Firms issue equity sporadically, so firm issuances exhibit lumpy behavior. As in the previous section, we add a one (million U.S. dollars) to the log of issuances to account for firm-year observations with zero issuances. ⁡ and denote firm and year fixed effects, respectively. Table 3 reports the results. Interestingly, Column (1) shows that the effect of capital inflows on firm-level issuance is only weakly statistically different from zero on average for the whole sample of firms. To explore the heterogeneity that could be driving the aggregate results, we divide the sample of equity issuers into two groups: large firms (those in the top decile of size, as measured by the prior period’s market value of equity within a country and year) and other firms.20 We lag firms’ market value of equity by one year to ensure that our measure of firm size is unaffected by current-year capital inflows. In Column (2) of Table 3, 19 One can explain this finding from the perspective of corporate capital structure decisions: firms in emerging markets have strong incentives to issue equity when the cost of doing so is low, but they do not have to reduce outstanding equity when foreign withdrawals of equity cause prices to fall. Given the high costs of external finance in emerging markets, firms in these economies tend to have highly productive unrealized investment opportunities (from a Tobin’s q perspective), which explains why issuances tend to be positive even when inflows are small or negative, and why repurchases of equity are rare. 20 We estimate the regression: log⁡ () + log⁡ (1 + ) = + + log⁡ () × ⁡−1 + . 19 we find that while the average effect for all firms is not statistically significant, large firms display a highly significant positive differential response to equity inflows. In Column (3), we replace firm and year fixed effects with firm and country-year fixed effects. 21 This specification allows us to control for all time-varying country shocks. The coefficient of interest is identified purely from the within-country variation between large and small firms. The differential response of large firms remains positive, statistically significant, and of nearly identical value as in Column (2). Columns (4) and (5) of Table 3 report results that are the same as those in Columns (2) and (3), except that lagged inflows are used in place of contemporaneous inflows. Results are similar in magnitude and statistical significance.22 In results not reported here, we also checked to see whether the responses of firms issuing in foreign markets are different from those of large firms issuing in domestic markets and we found that they are rather similar. We also perform a test to be sure that it is truly equity capital inflows, rather than capital inflows per se, that are driving the differential issuance behavior of large firms. In particular, we are concerned that large firms’ issuance of equity may be affected by non-equity forms of capital inflows that are correlated with equity flows, and that non-equity flows may have different consequences for large and small firms. In Appendix Table 2, we estimate regressions identical in structure to those in Columns (3) and (5) of Table 3 (also reported in Columns (1) and (2) of Appendix Table 2 for ease of comparison), but where we also include non-equity capital inflows (measured as the sum of all the non-equity flows mentioned earlier). 21 The effect of capital inflows, which varies at the country-year level, is absorbed by the country-year fixed effects. The equation for the new specification with interacted country-year fixed effects is: log⁡ (1 + () × ⁡−1 +⁡ . ) = + + log⁡ 22 In unreported results, we run specifications including both contemporaneous and lagged inflows. The point estimate for contemporaneous inflows remains unchanged, however due to the autocorrelation of inflows it is difficult to infer from these estimates the relative importance of the effects of inflows at different times. 20 Appendix Table 2 shows that equity inflows remain positively related to equity issuances and statistically significant, whereas non-equity inflows are not related or negatively related to equity issuances. For the contemporaneous specification, the coefficient on equity issuances is statistically larger in Column (3) than in Column (1). This reflects the significant negative coefficient on non-equity inflows. Equity and non-equity inflows are positively correlated (with a correlation coefficient of 0.25). To the extent that non-equity inflows (i.e., bond purchases) encourage large firms to issue bonds, they might actually reduce their incentive to issue equity. In other words, it could be that once the correlation between equity and bond inflows is controlled for, large firms’ responsiveness to equity inflows is even greater than our baseline estimate. For our purposes here, it is not necessary to orthogonalize equity inflows so long as it is clear that our baseline estimate reflects the effect of non-orthogonalized equity inflows. However, this finding implies that a full accounting of the consequences of equity inflows’ effects on investment must take account of other sources of funding that could arrive simultaneously with equity inflows. We return to this topic in Section VI below. For the lagged inflow specification in Column (4), the coefficient on equity inflows is not statistically different to that of Column (5) in Table 3. In Table 4, we examine how our coefficient estimates for large firms’ equity issuances change as the definition of a large firm varies. In Panel A of Table 4, we report specifications identical to Column (3) of Table 3, where we vary the percentile cutoff that defines large firms. Panel B is identical in structure except that it uses lagged equity inflows instead of contemporaneous. In both Panels A and B, as the percentile cutoff becomes less selective (more firms are included), the coefficient declines monotonically in value. From the first to the fourth decile, the declines indicate that the incremental decile of firms displays a smaller, but still positive, response than firms in more selective deciles. After the fourth decile, the 21 response of incremental firms (those in the fifth decile) is zero. Nonetheless, our purpose is not to model the responsiveness of each size category, but rather to explore important differences in behavior between large and small firms. For that purpose, a cutoff that distinguishes large from small is useful, and the results in Table 4 show that it does not matter much whether one picks the first, second, third, or fourth decile when defining large firms. The precision of estimates, however, is highest at the first decile of largest size, and partly for that reason, we prefer to use that definition. Furthermore, as we show in Appendix Table 3, there are a priori reasons to prefer the first decile as the definition of large firms. On average, the firms that choose to list issues abroad tend to be within the largest 12th percentile of firms. Those listed in the MSCI Index tend to be within the largest sixth percentile. These are two a priori indicators of firms that have access to foreign investors, one more selective than the other. Thus, we conclude that the firms in the top decile of firm size, which issue either abroad or in the domestic market, are large enough to be of interest to foreign investors. V. Instrumental Variables Approach As discussed in the Introduction and in Section III, an increase in issuance activity could reflect an increase in foreign equity funding supply or domestic equity funding demand, or some combination of the two. This section analyzes the importance of the supply-side channel and whether it can explain the response, documented in the previous section, of the issuance activity of a large firms to equity inflows. To do so, it presents our instrumental variable (IV) regressions, which identify supply-side shocks affecting capital inflows.23 We report our IV 23 The structure of our model combines aggregate country-level data in the first-stage regressions and firm-level data in the second-stage regressions. To account for most of the variation in the data, we use country and year fixed effects in the first-stage regressions and firm and country-year fixed effects in the second-stage regressions. 22 results for issuances in Tables 5-7. Panel A of each of these tables employs contemporaneous equity inflows, whereas Panel B employs lagged equity inflows. Standard errors are bootstrapped and clustered both by country and by year.24 In our first IV regression, we instrument equity inflows with the lagged MSCI Emerging Market Index country weights. For the second-stage regression, the relevant regressor is the interaction between equity inflows and the large firm dummy, which we instrument with the interaction between the lagged MSCI weights and the large firm dummy, which measures large firms based on their size in the previous year. Table 5 reports the results of the first-stage and second-stage regressions.25 Column (1) reports the first-stage regression. It shows that the instrument is positively and highly correlated with equity capital inflows. The F-statistic is 29 in Panel A and 27 in Panel B, indicating a powerful first-stage influence of the instrument.26 Column (2) of Table 5 shows the results of the second-stage regression.27 Consistent with the OLS results reported in the previous section, we find that when a country receives a supply-driven capital inflow, large firms issue more equity. As a robustness test, in Appendix However, the main results remain unchanged if we use the same firm-level structure of the data and the same set of fixed effects in the first- and second-stage regressions. 24 To account for the fact that we use an estimated regressor in the second stage, we bootstrap the standard errors. Our approach follows the methods outlined in Cameron et al. (2006), Cameron and Trivedi (2009), and Cameron et al. (2015) and adapts them to our data structure, clustering separately at the country and year level and then computing standard errors that take into account the two-way (country and year) clustering. The bootstrapped standard errors are obtained jointly for the two stages, for each clustering level. We obtained similar results when we estimated the model by bootstrapping the standard errors drawing independent samples in each stage. We report results with 1,000 sample draws for each clustering level. 25 The first-stage regression using the MSCI-weight instrument is: ⁡log⁡ () = + + log(⁡ℎ−1 ) + . 26 We also tried using two-year and three-year lagged weights of the MSCI, which should be even less related to contemporary demand shocks. Although the effect remains significant, the power of the instrument decreases with more lags, as one would expect. For this reason, we focus on the one-year-lag specification, which provides the strongest first-stage relation. 27 ̂ × ⁡−1 + , where We estimate: (1 + ) = + + ̂ denotes the fitted values of the first-stage regression. 23 Table 4, we exclude from the sample emerging market countries with the largest MSCI weights (Brazil, China, and the Republic of Korea) because the variation in their weights could be large enough that their own country’s demand-side changes could produce much of the variation in their own country weights. The average MSCI weights for these three countries are 11.7%, 10.8%, and 13.6%, respectively. We find that the results are not statistically different to those in Table 5.28 As discussed in Section III, the market value of other countries’ equity, and the equity issuances of other countries, offer alternative measures of the attractiveness of the subject country’s equity market to foreign investors. The advantage of those alternative instruments is that both of them affect a country’s MSCI weight exclusively through external influences. We report IV results using these measures in Tables 6 and 7, respectively. The instruments both are powerful negative predictors of equity capital inflows in the first-stage regression reported in Column (1). The results of the second-stage regressions in Column (2) are similar to those reported in Table 5. Columns (3) and (4) of Tables 6 and 7 report the IV results using as instruments the orthogonalized market value of equity and orthogonalized equity issuances in other countries. We derive the orthogonalization by removing the covariance, that is, by regressing the total market value of equity (equity issuances) of all emerging markets on the subject country’s 28 We also considered specifications in which lagged MSCI weights were substituted for contemporaneous weights, or were added to contemporaneous weights as an additional instrument. The results were similar to specifications with only the lagged MSCI weight. In another robustness test, we substituted the log of the equity index value for the log of the MSCI weight. In this alternative specification, the instrument captures only the role of market value changes of other countries’ indexes as an indicator of the subject country’s attractiveness to foreign investors. Note that this specification does not capture any causal effect from investors’ desires to track country weights. Results remained unchanged. In a final robustness test of this approach, the log of the market value of other countries’ stock indexes was orthogonalized (regressed on the subject country’s index to remove covariance). The results remained again unchanged. 24 market value of equity (equity issuances), and then using the residuals as our instrument.29 The first-stage regressions indicate equally powerful instruments than the non-orthogonalized instruments. The second-stage regressions in Columns (4) are similar in value and not statistically significantly different from the non-orthogonalized specifications reported in Columns (2). The IV coefficients in Tables 5-7 are slightly smaller or equal in magnitude, but not statistically different, than the comparable OLS coefficients reported in Table 3. In the absence of measurement error of capital inflows, the OLS coefficient should be greater than or equal to the IV coefficient because OLS captures supply and demand effects and those effects are additive. Our results are consistent with that expectation. The equality of the magnitude of the OLS and IV coefficient suggest that supply-side influences account for most of the variation in equity capital inflows. However, there is reason to believe that equity inflows are measured with error, which biases the OLS coefficient downwards, implying that the true OLS coefficient is probably larger than the estimated value.30 In the absence of measurement error, a larger OLS coefficient would suggest that demand-side influences also are present. Overall, we find that whether one measures the attractiveness of other countries’ equity markets to foreign investors using the MSCI weights, other countries’ market value or other countries’ equity issuance volume, the results are similar: supply-side effects of 29 The orthogonalized market value _̂ for country c in year t is constructed as the residual of the following () + , where is the sum of all 25 emerging markets’ value regression: () = log⁡ of market capitalization in year t, is the value of own market capitalization for country c in year t, and _ is the residual to be predicted. The orthogonalized issuance volume is calculated in the same way. For each instrument, 25 time-series regressions are estimated separately for each country. Country values of market capitalization are downloaded from the World Bank’s WDIs. 30 As Lane and Milesi-Ferretti (2017, p. 21) note in their discussion of the capital inflow data: “One concern … is the increasing difficulty in properly assessing external exposures … particularly in light of the size of cross - border asset trade intermediated by financial centers [which complicates the measurement of inflows into a particular country] … This difficulty affects virtually all categories of cross-border holdings …” As a result, under the assumption that our identification is correct, the OLS coefficient in our setting could be larger or smaller than the IV estimate. 25 instrumented equity inflows are large and statistically significant. Results are robust to using lagged or contemporaneous values of instruments, or to using orthogonalized or non- orthogonalized measures. We conclude that supply-side shocks are an important driver of equity capital flows, and that plausibly exogenous changes in the supply of foreign equity inflows have important consequences for equity issuances by large firms. VI. Capital Inflows and Uses of Funds Having established a connection between equity capital inflows and equity issuances, we now study the real effects of capital equity inflows. Following the approach of Kim and Weisbach (2008) and Erel et al. (2011), we focus on six uses of funds: CAPEX, acquisitions, R&D, inventory accumulation, cash accumulation, and long-term debt reduction.31 We conduct two set of tests. First, we estimate the effect of capital inflows on each of these potential uses of funds. Second, we study how large firms use, on average, the funds raised in their equity offerings. We report results for all firms in Tables 8 and 9, and we obtain nearly identical results for a subsample that is restricted to non-financial firms (Appendix Tables 5 and 6). This reflects the fact that non-financial firms comprise most of our sample (representing almost 90% of our observations). Table 8 reports the results of estimating the effects of equity capital inflows separately on each use of funds, for both contemporaneous and lagged inflows. We report IV results, using lagged MSCI weight as an instrument for equity inflows. We obtain similar results using the other two instruments (other countries’ market value and other countries’ issuance volume). We report those alternative results in Appendix Tables 7 and 8. Column (1) of Table 31 We obtain the variables CAPEX, acquisitions, R&D, and long-term debt reduction from the income and cash flow statements and the variables inventory accumulation and cash accumulation from the balance sheets. 26 8 shows that equity inflows lead to a significant increase in capital expenditures by large firms. Columns (2) and (3) show that, after the arrival of equity inflows, large firms also tend to undertake more corporate acquisitions and invest more in R&D. The final three columns of Table 8 show that increased equity inflows lead to inventory accumulation, cash accumulation, and a reduction in long-term debt. The previous results reveal the connections between capital inflows and different uses of funds, but the analysis does not link the uses of funds to the actual equity issuances. Given that equity inflows are positively correlated with debt inflows, and that debt inflows are negatively correlated with equity issuances (as shown in Appendix Table 2), the coefficients in Table 8 should be seen as lower bound estimates of the effect of exogenous equity capital inflows. Note that the amount of debt inflows is not held constant in Table 8. To better analyze the linkages among issuances and uses of funds, taking into account the presence of other sources of funding, in our second test we adopt the methodology of Kim and Weisbach (2008) and Erel et al. (2011). We focus on the six uses of funds described above, measuring the change in each use of funds over a variety of time intervals, ranging from one year to four years. Following those authors, we begin by calculating the use of funds after each firm’s equity offering (whether caused by capital inflows or something else) by estimating the following regression for the equity offerings of large firms: = + + ⁡log [1 + ( ) ] (5) ℎ + log [1 + ( ) ] + log[ ] + , where = ⁡log[(∑ =1 /) + 1] for the income- and cash flow-statement items ( = CAPEX, acquisitions, R&D, long-term debt reduction), and = ⁡log[(( − 0 )/) + 1] for the balance-sheet items ( = inventory, cash holdings). N=1,2,3,4 denotes the years 27 following the issuance. denotes total assets in the year just prior to the equity issuance ∑ −1(⁡ −) (n=0). ℎ⁡ = ⁡log [( ) + 1],⁡where total sources of funds represent the total funds generated by the firm internally and externally during a given year.32 Table 9 reports the results of estimating Equation (5) separately for each use of funds, for each time interval considered. We report the estimated elasticities and also the dollar effects, for the average firm of the typical country in a typical year.33 The table shows that for every million U.S. dollars raised in an offering, large firms increase CAPEX on average by 180,000 U.S. dollars in the year after the offering. The effect on CAPEX increases to 700,000 U.S. dollars when the equation is estimated over a four-year period. After four years, issuers spend about 150,000 U.S. dollars in acquisitions. The effect on R&D is not statistically significant when we include all firms but becomes significant when we exclude financial sector firms (Appendix Table 5). Overall, the largest use of funds is CAPEX. Firms also spend some of the equity proceeds in accumulating cash and short-term investments. This effect is largest in the first year that the firm issues, then (similar to the findings in Kim and Weisbach, 2008) it diminishes markedly and becomes statistically not significant after four years of the issuance activity. The effect on long-term debt reduction is not significant at any horizon. The fact that 32 There are advantages and disadvantages to the alternative estimation approaches of Tables 8 and 9. As noted above, Table 8 does not control for debt inflows. On the other hand, Table 9’s estimates do control for the presence of other funding sources, and by following the Kim and Weisbach (2008) method, we facilitate comparisons with prior studies. However, the estimates in Table 9 employ all issuances, not issuances that specifically result from exogenous equity capital inflows. We regard the two approaches as complementary. The results in Table 8 may understate the true effects, but they link responses of issuances to exogenous inflows. The results in Table 9 control for other funding sources, but are not computed to be conditional on capital inflows. 33 To calculate the dollar effects, we first calculate the predicted values of the dependent variables for each firm- year observation by plugging the actual values of firm issuances, other sources of funds, and total assets into Equation (5). For the fixed effects, we use the coefficients for each year and country of the corresponding country-year pair. We then re-calculate the predicted values of the dependent variables after adding one million U.S. dollars to the issuance value. Next, we calculate the difference of the two predicted values for each firm- year observation. To aggregate the differences, we first take the time-average of the differences per firm, we then take the median firm-average per country and subsequently the median country in our sample. 28 firms spend most of the issuance proceeds to fund corporate investment suggests that capital inflows reduce equity financing costs. The result provides evidence against alternative explanations for equity issuance, such as market timing (Baker and Wurgler, 2000). Overall, our results indicate that equity issuance is an important channel through which capital inflows can affect real economic activity. In the aggregate analysis, we documented that one million U.S. dollars of equity inflows was associated with an increase of 160,000 U.S. dollars of country-level seasoned equity issuances. On the other hand, in this section we have shown that large firms invest on average 700,000 U.S. dollars of each million raised in an equity offering. Combining both results, a back-of-the-envelope calculation indicates that for every million U.S. dollars of equity capital received from foreign investors, emerging market firms use the proceeds of seasoned equity offerings to increase corporate investment by at least 110,000 U.S. dollars (=0.16x0.7 for every million U.S. dollar received). VII. Conclusions There is a growing literature documenting that greater capital inflows are associated with important increases in aggregate investment and higher economic growth. A separate large literature studies the issuance activity of firms. This paper is the first study to use issuance- level data to document a new channel, firm-level equity issuance, through which capital inflows affect real economic activity. We seek to determine whether increases in equity capital inflows into emerging market countries are associated with increases in equity issuance and corporate investment by publicly traded firms, and whether any observed association can be attributed to supply-side influences from exogenous changes in international investors’ interest in investing in particular countries. 29 We find that increases in equity inflows into emerging markets are associated with higher values of country-level equity issuance proceeds. This indicates that inflows imply more than a simple transfer of equity ownership from domestic to foreign investors. Using firm- level data, we show that large firms disproportionately drive this relation. Instrumenting equity inflows (both contemporaneous and lagged) with various alternative measures that capture the exogenous variation in other countries’ attractiveness to foreign investors, we show that our results are driven by variation in foreign equity capital supply. Lastly, we find that equity capital inflows lead large firms to increase corporate investment. We also show that large firms invest, on average, a substantial fraction of the funds raised in equity offerings. Our evidence is consistent with capital inflows lowering equity financing costs, which allows firms to raise funds to finance new investments. More generally, our results indicate capital inflows affect real economic activity through equity issuance activity, among other possible channels. Our work shows how micro data can provide unique insights into how subsets of firms drive aggregate relations. Our findings suggest that the issuance and investment behavior of large firms in emerging markets is highly responsive to equity inflows. But apparently, many other emerging market firms are not the target of global market investors’ share purchases. For those smaller firms, large flows of funds connecting their countries to global markets have little direct effect on their propensity to issue equity. This suggests that it can be useful to divide firms in emerging economies into two categories: those for which equity capital inflows have significant and important direct effects on the cost of issuing capital, and those for which they have little effect. To the extent that equity inflows lower the cost of finance for large firms, that may create a competitive advantage for those firms. At the same time, it is possible that large firms 30 may share some of the benefits of their access to international investors with other firms. Other firms could benefit indirectly from more abundant trade credit, or increased demand for their products and services. Also, if equity issuances reduce issuers’ demands for local bank debt, that could make it easier for non-issuers to borrow locally. Furthermore, financial firms might use their new equity issuance proceeds in support of greater lending to local firms. These two influences could be particularly beneficial for small and medium-sized firms (de la Torre et al., 2010). More broadly, future work could examine the extent to which the selective reductions in the cost of equity either promote greater efficiency in the economy (i.e., by reducing financing constraints for relatively productive firms, and by providing indirect benefits for other firms), or result in inefficiencies by increasing the market power of a small number of large firms. 31 References Ahmed, S. and A. Zlate (2014). “Capital Flows to Emerging Market Economies: A Brave New World? Journal of International Money and Finance 48 (B), 221-248. Alfaro, L., A. Chanda, S. Kalemli-Ozcan, and S. Sayek (2004). “FDI and Economic Growth: The Role of Local Financial Markets.” Journal of International Economics 64 (1), 89-112. Avdjiev, S., B. Hardy, S. Kalemli-Ozcan, and L. Serven (2018). “Gross Capital Inflows to Banks, Corporates, and Sovereigns.” NBER Working Paper No. 23116. Baskaya, Y., J. di Giovanni, S. Kalemli-Ozcan, and M. Ulu (2017a). “International Spillovers and Local Credit Cycles.” NBER Working Paper No. 23149. Baskaya, Y., J. di Giovanni, S¸. Kalemli-Ozcan, J.L. Peydro, and M. Ulu (2017b). “Capital Flows and the International Credit Channel.” Journal of International Economics 108 (1), 15-22. Baker, M. and J. Wurgler (2000). “The Equity Share in New Issues and Aggregate Stock Returns.” Journal of Finance 55, 2219-57. Bekaert, G., C. Harvey, and C. Lundblad (2005). “Does Financial Liberalization Spur Growth?” Journal of Financial Economics 77 (1), 3-55. Benos, E. and Weisbach, M. (2004). “Private Benefits and Cross-listings in the United States.” Emerging Markets Review 5 (2), 217-240. Brown, J., S. Fazzari, and B. Petersen (2009). “Financing Innovation and Growth: Cash Flow, External Equity and the 1990s R&D Boom.” Journal of Finance 64 (1), 151-185. Bruno, V. and H.S. Shin (2017). “Global Dollar Credit and Carry Trades: A Firm-level Analysis”. Review of Financial Studies 30(3), 703-749. Caballero, J., U. Panizza and A. Powell (2016). “The Second Wave of Global Liquidity: Why are Firms Acting Like Financial Intermediaries?” Working Paper IDB-WP-641, Inter- American Development Bank. Cameron, A.C., J. Gelbach, and D. Miller (2006). “Robust Inference with Multi-way Clustering.” NBER Technical Working Paper 327. Cameron, A.C., and D. Miller (2015). “A Practitioner’s Guide to Cluster-Robust Inference.” Journal of Human Resources 50, 317-372. Cameron, A.C., and P. Trivedi (2009). “Microeconometrics Using Stata.” Chapter 13 of Bootstrap Methods. Stata Press. Chari, A., K.S. Stedman, and C. Lundblad (2017) “Taper Tantrums: QE, Its Aftermath, and Emerging Market Capital Flows.” NBER Working Paper No. 23474. Chui, M., I. Fender, and V. Sushko (2014). “Risks Related to EME Corporate Balance Sheets: The Role of Leverage and Currency Mismatch.” BIS Quarterly Review September 2014, Bank for International Settlements. Claessens, S., H. Tong, and S.-J. Wei (2012). “From the Financial Crisis to the Real Economy: Using Firm-level Data to Identify Transmission Channels.” Journal of International Economics 88 (2), 375-387. Claessens, S. and S. Schmukler (2007). “International Financial Integration through Equity Markets: Which Firms from which Countries Go Global?” Journal of International Money and Finance 26 (5), 788-813. de la Torre, A., M.S. Martinez Peria, and S. Schmukler (2010). “Bank Involvement with SMEs: Beyond Relationship Lending.” Journal of Banking and Finance 34 (9), 2280-2293. Didier, T., R. Levine, and S. Schmukler (2015). “Capital Market Financing, Firm Growth, Firm Size Distribution.” NBER Working Paper 20336 and World Bank Policy Research Paper 7353. 32 Doidge, C. (2004). “U.S. Cross-listings and the Private Benefits of Control: Evidence from Dual-Class Firms.” Journal of Financial Economics 72 (3), 519-553. Doidge, C., A. Karolyi, K. Lins, D. Miller, and R. Stulz (2009). “Private Benefits of Control, Ownership, and the Cross-Listing Decision.” Journal of Finance 64 (1), 425-466. Erel, I., B. Julio, W. Kim, and M. Weisbach, (2011). “Macroeconomic Conditions and Capital Raising.” Review of Financial Studies 25 (2), 341-376. Forbes, K. (2007). “One Cost of the Chilean Capital Controls: Increased Financial Constraints for Smaller Traded Firms.” Journal of International Economics 71 (2), 294-323. Forbes, K. and F. Warnock (2012). “Capital Flow Waves: Surges, Stops, Flight and Retrenchment.” Journal of International Economics 88 (2), 235-251. Fratzscher, M. (2012). “Capital Flows, Push versus Pull Factors and the Global Financial Crisis.” Journal of International Economics, 88 (2), 341-356. Gupta, N. and K. Yuan (2009). “On the Growth Effect of Stock Market Liberalizations.” Review of Financial Studies 22(11), 4715-4752. Henry, P. (2000a). “Do Stock Market Liberalizations Cause Investment Booms?” Journal of Financial Economics 58, 301–334. Henry, P. (2000b). “Stock Market Liberalization, Economic Reform, and Emerging Market Equity Prices.” Journal of Finance 55 (1-2), 529-564. Igan, D., A. Kutan, and A. Mirzae (2016). “Real Effects of Capital Inflows in Emerging Markets.” IMF Working Paper 16/235, International Monetary Fund. Karolyi, G. (2006). “The World of Cross-Listings and Cross-Listings of the World: Challenging Conventional Wisdom.” Review of Finance 10 (1), 99-152. Karolyi, G.A. (2015). Cracking the Emerging Markets Enigma. New York: Oxford University Press. Kim, W. and M. Weisbach (2008). “Motivations for Public Equity Offers.” Journal of Financial Economics 87(2), 281-307. Kose, A., K. Rogoff, E. Prasad, and S.-J. Wei (2009). “Financial Globalization: A Reappraisal,” IMF Staff Papers, Vol. 16 (1), 8-62. Kose, A., K. Rogoff, E. Prasad, and S.-J. Wei (2010). “Financial Globalization and Economic Policies,” Handbook of Development Economics, Vol. 5, edited by Dani Rodrik and Mark Rosenzweig, North-Holland, 2010, pp. 4283-4362. Lane, P. R., and G.G.M Milesi-Ferretti (2017). “International Financial Integration in the Aftermath of the Global Financial Crisis”. IMF Working Paper No. 17/115. Levchenko, A., R. Ranciere, and M. Thoenig (2009). “Growth and Risk at the Industry Level: The Real Effects of Financial Liberalization.” Journal of Development Economics 89 (2), 210-222. Mitton, T. (2006). “Stock Market Liberalization and Operating Performances at the Firm Level.” Journal of Financial Economics 81(3), 625-47. Moreno, R., and J.M. Serena Garralda (2018). “Firms’ Credit Risk and the Onshore Transmission of the Global Financial Cycle,” BIS Working Paper 712. Pagano, M., F. Panetta, and L. Zingales (1998). “Why Do Companies Go Public?” Journal of Finance 53 (1), 27-64. Pagano, M., A. Röell, and J. Zechner (2002). “The Geography of Equity Listing: Why Do Companies List Abroad?” Journal of Finance 57 (6), 2651-2694. Powell, A. (2014). Global Recovery and Monetary Normalization: Escaping a Chronicle Foretold? Washington, DC: Inter-American Development Bank. Raddatz, C., S. Schmukler, and T. Williams (2017). “International Asset Allocations and Capital flows: The Benchmark Effect.” Journal of International Economics, 108, 413-430. 33 Schmukler, S. and E. Vesperoni (2006). “Financial Globalization and Debt Maturity in Emerging Economies.” Journal of Development Economics 79 (1), 183-207. Stulz, Rene (2005). “The Limits of Financial Globalization.” Journal of Finance 60(4), 1595-1638. Tong, H., and S.-J. Wei (2010). “The Composition Matters: Capital Flows and Liquidity Constraint during a Global Economic Crisis,” Review of Financial Studies, 2010, 24(6), 2023-2052. 34 Figure 1 Emerging Market Equity Issuances and Equity Capital Inflows This figure plots the total value of equity issued by firms in 25 emerging market countries (right axis) and total portfolio equity inflows to those emerging markets (left axis) during the 1991-2016 period. All values are reported in billions of 2011 U.S. dollars (USD). 200 300 150 250 (Billion 2011 USD) Equity Inflows Equity Issuances (Billion 2011 USD) 100 200 50 150 0 100 -50 50 -100 0 Equity Inflows Equity Issuances 35 Figure 2 MSCI Weights, Other Countries' Issuance Volume, and Equity Capital Inflows for Mexico Panel A of this figure plots the time series of MSCI Emerging Market index weights (left axis) and portfolio equity inflows (right axis) for Mexico during the 1991-2016 period. Panel B plots the time series of the sum of the issuance volume of the countries in the MSCI Emerging Market index except Mexico (left axis) and portfolio equity inflows for Mexico (right axis) during the 1991-2016 period. Equity inflows and issuances are reported in billions of 2011 U.S. dollars (USD). Panel A. MSCI Weights and Equity Inflows for Mexico 13% 10 12% 8 11% 6 (Billion 2011 USD) Equity Inflows MSCI Weight 10% 4 9% 2 8% 0 7% -2 6% -4 5% -6 4% -8 MSCI Weight Equity Inflows Panel B. Other Countries' Issuance Volume and Equity Inflows for Mexico 350 22 18 300 Other Countries' Issuance Volume 14 250 (Billion 2011 USD) 10 Equity Inflows (Billion 2011 USD) 200 6 150 2 -2 100 -6 50 -10 0 -14 Other Countries' Issuances Equity Inflows 36 Figure 3 Emerging Market Equity Issuances and Equity Capital Inflows, Scaled by GDP This figure plots the average value of equity issued by firms in 25 emerging market countries over GDP (right axis) and the average value of portfolio equity inflows to those emerging markets over GDP (left axis) during the 1991-2016 period. Both equity issuances and inflows are scaled by each country's GDP and then averaged across countries. 1.20% 0.90% 1.00% 0.80% 0.80% 0.70% Equity Issuances/GDP Equity Inflows/GDP 0.60% 0.60% 0.40% 0.50% 0.20% 0.40% 0.00% 0.30% -0.20% 0.20% -0.40% 0.10% -0.60% 0.00% Equity Inflows/GDP Equity Issuances/GDP 37 Table 1 Summary Statistics of Number of Firms and Equity Issuance Activity This table reports summary statistics of firms' equity issuance activity for each of the 25 emerging market countries in our sample during the 1991-2016 period. All issuance values are in millions of 2011 U.S. dollars (USD). Number Average Annual Average Annual Country Issuance Issuance Value of Firms Value / Number of Firms (Million USD) (Million USD) (1) (2) (3) Argentina 85 626 7 Brazil 323 5,572 17 Chile 152 1,156 8 China 3,660 24,863 7 Colombia 33 524 16 Czech Republic 7 66 9 Egypt, Arab Rep. 140 461 3 Hungary 21 93 4 India 2,105 6,388 3 Indonesia 499 3,280 7 Israel 226 1,009 4 Jordan 127 149 1 Korea, Rep. 1,136 6,851 4 Malaysia 146 2,687 2 Mexico 52 1,522 10 Morocco 129 96 2 Pakistan 48 100 1 Peru 231 105 2 Philippines 423 1,519 7 Poland 196 1,139 3 Russian Federation 231 3,745 19 South Africa 1,825 2,310 10 Thailand 651 2,162 3 Turkey 237 966 4 Venezuela, R.B. 40 104 3 Total (Column 1) and 12,723 2,700 6 Average (Columns 2 and 3) 38 Table 2 Equity Inflows and Equity Issuance Activity: Aggregate Evidence This table reports country-level panel OLS regressions of aggregate equity issuance on equity inflows for 25 emerging market countries during the 1991-2016 period. Column (1) reports the regression for the log of one plus aggregate equity issuance on the log of portfolio equity inflows. Column (2) reports the regression for aggregate equity issuance/GDP on equity inflows/GDP. Columns (3) and (4) report the regressions for aggregate equity issuance/GDP on equity inflows/GDP, restricting the sample to observations with positive and negative inflows, respectively. All variables are winsorized at the 1% level. All regressions include country and year fixed effects. Standard errors are double clustered at the country and year levels. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Equity Equity Log(1+Equity Equity Issuance / GDP Issuance / GDP Issuance) Issuance / GDP (Inflows > 0) (Inflows < 0) (1) (2) (3) (4) Log(Equity Inflows) 0.5240 *** (0.093) Equity Inflows / GDP 0.0652 ** 0.1527 *** 0.0260 (0.029) (0.046) (0.037) Country FE Yes Yes Yes Yes Year FE Yes Yes Yes Yes Number of Observations 428 581 427 154 39 Table 3 Equity Inflows and Firms' Equity Issuance Activity This table reports firm-level panel OLS regressions of the log of one plus equity issuance on the log of portfolio equity inflows and its interaction with the large firm dummy variable for 25 emerging market countries during the 1991-2016 period. A firm is classified as large if its prior period's market value of equity is in the top decile of the market value distribution within a country and year. Columns (1)-(3) and (4)-(5) report the analysis for contemporaneous and lagged equity inflows, respectively. All variables are winsorized at the 1% level. Regressions in columns (1), (2), and (4) include firm and year fixed effects. Regressions in columns (3) and (5) include firm and country- year fixed effects. Standard errors are double clustered at the country and year levels. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Log(1+Equity Issuance) (1) (2) (3) (4) (5) Log(Equity Inflows) 0.0395 * 0.0360 (0.022) (0.022) Log(Equity Inflows) * Large Firm 0.0223 *** 0.0210 *** (0.002) (0.001) Log(Lagged Equity Inflows) 0.0260 (0.025) Log(Lagged Equity Inflows) * Large Firm 0.0183 *** 0.0173 *** (0.003) (0.003) Firm FE Yes Yes Yes Yes Yes Year FE Yes Yes No Yes No Country-Year FE No No Yes No Yes Number of Observations 88,314 88,314 88,314 88,112 88,112 40 Table 4 Equity Inflows and Firms' Equity Issuance Activity Effects for Firms of Different Sizes This table reports firm-level panel OLS regressions of the log of one plus equity issuance on the log of contemporaneous or lagged portfolio equity inflows and its interaction with the large firm dummy variable in Panels A and B, respectively, for 25 emerging market countries during the 1991-2016 period. A firm is classified as large if its prior period's market value of equity is in the top 10th (column 1), 20th (column 2), 30th (column 3), 40th (column 4), or 50th (column 5) percentile of the market value distribution within a country and year. All variables are winsorized at the 1% level. All regressions include firm and country-year fixed effects. Standard errors are double clustered at the country and year levels. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Panel A. Contemporaneous Inflows Log(1+Equity Issuance) Top 10 Top 20 Top 30 Top 40 Top 50 Log(Equity Inflows) * Large Firm (1) (2) (3) (4) (5) 0.0210 *** 0.0175 *** 0.0133 ** 0.0124 ** 0.0095 ** (0.001) (0.004) (0.006) (0.005) (0.004) Firm FE Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Number of Observations 88,314 88,314 88,314 88,314 88,314 Panel B. Lagged Inflows Log(1+Equity Issuance) Top 10 Top 20 Top 30 Top 40 Top 50 Log(Lagged Equity Inflows) * Large Firm (1) (2) (3) (4) (5) 0.0173 *** 0.0139 ** 0.0115 * 0.0112 ** 0.0084 (0.003) (0.005) (0.007) (0.005) (0.005) Firm FE Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Number of Observations 88,112 88,112 88,112 88,112 88,112 41 Table 5 Equity Inflows and Firms' Equity Issuance Activity, Instrumental Variable Approach Instrument: Lagged MSCI Weight This table reports the first- and second-stage regressions for the instrumental variable approach, using the lagged MSCI emerging market weight as instrument, for 25 emerging market countries during the 1997-2016 period. In Panel A, column (1) reports a country-level panel OLS regression of the log of contemporaneous portfolio equity inflows on the log of the one-year lag of the MSCI emerging market weight. Column (2) reports the firm-level panel second-stage regression of the log of one plus equity issuance on the interaction of the log of contemporaneous portfolio equity inflows with the large firm dummy (top decile of the prior period's equity market value distribution within a country and year). Panel B reports the same regressions using lagged portfolio equity inflows and the two-year lag of the MSCI emerging market weight. Regressions in column (1) include country and year fixed effects. Regressions in column (2) include firm and country-year fixed effects. All variables are winsorized at the 1% level. First-stage standard errors are double clustered at the country and year levels. Second-stage standard errors are block bootstrapped with 1,000 repetitions, double clustering at the country and year levels. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Panel A. Contemporaneous Inflows First Stage Second Stage Log(Equity Inflows) Log(1+Equity Issuance) (1) (2) Log(Lagged MSCI Weight) 0.5182 *** (0.096) Log(Equity Inflows) * Large Firm 0.0250 *** (0.004) Firm FE No Yes Country FE Yes No Year FE Yes No Country-Year No Yes Number of Observations 297 72,888 Kleibergen-Paap Wald F-Stat 28.97 .. Panel B. Lagged Inflows First Stage Second Stage Log(Lagged Equity Inflows) Log(1+Equity Issuance) (1) (2) Log(Two-Year Lagged MSCI Weight) 0.5109 *** (0.098) Log(Lagged Equity Inflows) * Large Firm 0.0195 *** (0.005) Firm FE No Yes Country FE Yes No Year FE Yes No Country-Year No Yes Number of Observations 283 72,997 Kleibergen-Paap Wald F-Stat 27.32 .. 42 Table 6 Equity Inflows and Firms' Equity Issuance Activity, Instrumental Variable Approach Instrument: Other Countries' Equity Value This table reports the first- and second-stage regressions for the instrumental variable approach, using the sum of other countries' equity value and other countries' orthogonalized value as instruments, for 25 emerging market countries during the 1991-2016 period. We compute the orthogonalized equity value for each country as the residual of a time-series regression of the log of market vale of emerging markets on the log of own-country market value. In Panel A, columns (1) and (3) report country-level panel OLS regressions of the log of portfolio equity inflows on the log of the sum of other countries' equity value and orthogonalized equity value, respectively. Columns (2) and (4) report firm-level panel second-stage regressions of the log of one plus equity issuance on the interaction of the log of portfolio equity inflows with the large firm dummy (top decile of the prior period's equity market value distribution within a country and year). Panel B reports the same regressions using lagged portfolio equity inflows and the one-year lag of other countries' equity value or orthogonalized equity value. Regressions in columns (1) and (3) include country and year fixed effects. Regressions in columns (2) and (4) include firm and country-year fixed effects. All variables are winsorized at the 1% level. First-stage standard errors are double clustered at the country and year levels. Second-stage standard errors are block bootstrapped with 1,000 repetitions, double clustering at the country and year levels. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Panel A. Contemporaneous Inflows Other Countries' Equity Value, Other Countries' Equity Value Orthogonalized First Stage Second Stage First Stage Second Stage Log(Equity Log(1+Equity Log(Equity Log(1+Equity Inflows) Issuance) Inflows) Issuance) (1) (2) (3) (4) Log(Other Countries’ Equity Value) -2.8308 *** -0.5652 *** (0.810) (0.165) Log(Equity Inflows) * Large Firm 0.0208 *** 0.0204 *** (0.003) (0.004) Firm FE No Yes No Yes Country FE Yes No Yes No Year FE Yes No Yes No Country-Year No Yes No Yes Number of Observations 428 88,314 329 84,963 Kleibergen-Paap Wald F-Stat 12.20 .. 11.69 .. Panel B. Lagged Inflows Other Countries' Equity Value, Other Countries' Equity Value Orthogonalized First Stage Second Stage First Stage Second Stage Log(Lagged Log(1+Equity Log(Lagged Log(1+Equity Equity Inflows) Issuance) Equity Inflows) Issuance) (1) (2) (3) (4) Log(Lagged Other Countries’ Equity Value) -3.0342 *** -0.6199 *** (0.951) (0.172) Log(Lagged Equity Inflows) * Large Firm 0.0174 *** 0.0174 *** (0.004) (0.005) Firm FE No Yes No Yes Country FE Yes No Yes No Year FE Yes No Yes No Country-Year No Yes No Yes Number of Observations 413 88,112 315 83,160 Kleibergen-Paap Wald F-Stat 10.19 .. 12.98 .. 43 Table 7 Equity Inflows and Firms' Equity Issuance Activity, Instrumental Variable Approach Instrument: Other Countries' Equity Issuance Volume This table reports the first- and second-stage regressions for the instrumental variable approach, using the sum of other countries' equity issuance and other countries' orthogonalized issuance as instruments, for 25 emerging market countries during the 1991-2016 period. We compute the orthogonalized equity issuance for each country as the residual of a time-series regression of the log of equity issuance of emerging markets on the log of one plus own-country equity issuance. In Panel A, columns (1) and (3) report country-level panel OLS regressions of the log of portfolio equity inflows on the log of the sum of other countries' equity issuance and orthogonalized equity issuance, respectively. Columns (2) and (4) report firm-level panel second-stage regressions of the log of one plus equity issuance on the interaction of the log of portfolio equity inflows with the large firm dummy (top decile of the prior period's equity market value distribution within a country and year). Panel B reports the same regressions using lagged portfolio equity inflows and the one-year lag of other countries' equity issuance or orthogonalized equity issuance. Regressions in columns (1) and (3) include country and year fixed effects. Regressions in columns (2) and (4) include firm and country-year fixed effects. All variables are winsorized at the 1% level. First-stage standard errors are double clustered at the country and year levels. Second-stage standard errors are block bootstrapped with 1,000 repetitions, double clustering at the country and year levels. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Panel A. Contemporaneous Inflows Other Countries' Equity Issuance Other Countries' Equity Issuance Volume Volume, Orthogonalized First Stage Second Stage First Stage Second Stage Log(Equity Log(1+Equity Log(Equity Log(1+Equity Inflows) Issuance) Inflows) Issuance) (1) (2) (3) (4) Log(Other Countries’ Equity Issuance Volume) -2.7897 ** -0.7548 *** (1.161) (0.187) Log(Equity Inflows) * Large Firm 0.0200 *** 0.0215 *** (0.003) (0.003) Firm FE No Yes No Yes Country FE Yes No Yes No Year FE Yes No Yes No Country-Year No Yes No Yes Number of Observations 428 88,314 428 88,314 Kleibergen-Paap Wald F-Stat 5.78 .. 16.25 .. Panel B. Lagged Inflows Other Countries' Equity Issuance Other Countries' Equity Issuance Volume Volume, Orthogonalized First Stage Second Stage First Stage Second Stage Log(Lagged Log(1+Equity Log(Lagged Log(1+Equity Equity Inflows) Issuance) Equity Inflows) Issuance) (1) (2) (3) (4) Log(Lagged Other Countries’ Equity Issuance Volume) -3.4375 ** -0.7976 *** (1.383) (0.192) Log(Lagged Equity Inflows) * Large Firm 0.0166 *** 0.0180 *** (0.003) (0.004) Firm FE No Yes No Yes Country FE Yes No Yes No Year FE Yes No Yes No Country-Year No Yes No Yes Number of Observations 413 88,112 413 88,112 Kleibergen-Paap Wald F-Stat 6.18 .. 17.25 .. 44 Table 8 Real Economic Effects, Instrumental Variable Approach This table reports the second stage of firm-level panel instrumental variable regressions of the log of one plus firm real and financial outcomes on the interaction of the log of contemporaneous (Panel A) and lagged (Panel B) portfolio equity inflows with the large firm dummy (top decile of the prior period's equity market value distribution within a country and year) for 25 emerging market countries during the 1997-2016 period. We use the one-year and two-year lags of MSCI emerging market weights as instruments for the contemporaneous and lagged portfolio equity inflows, respectively. The dependent variables are capital expenditures, acquisitions, research and development expenditures, inventory accumulation, cash and short-term investments accumulation, and reduction of long-term debt. All variables are winsorized at the 1% level. Second-stage standard errors are block bootstrapped with 1,000 repetitions, double clustering at the country and year levels. All Regressions include firm and country- year fixed effects. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Panel A. Contemporaneous Inflows Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ CAPEX) Acquisitions) R&D) Inventory) Cash&ST Inv.) LT Debt Red.) (1) (2) (3) (4) (5) (6) Log(Equity Inflows) * Large Firm 0.0629 *** 0.0264 *** 0.0347 *** 0.0375 *** 0.0682 *** 0.0185 * (0.005) (0.004) (0.006) (0.004) (0.008) (0.011) Firm FE Yes Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Yes Number of Observations 69,864 50,603 23,129 67,591 68,354 50,940 Panel B. Lagged Inflows Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ CAPEX) Acquisitions) R&D) Inventory) Cash&ST Inv.) LT Debt Red.) (1) (2) (3) (4) (5) (6) Log(Lagged Equity Inflows) * Large Firm 0.0627 *** 0.0251 *** 0.0349 *** 0.0390 *** 0.0683 *** 0.0192 * (0.006) (0.003) (0.007) (0.004) (0.008) (0.010) Firm FE Yes Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Yes Number of Observations 69,951 50,640 23,046 67,741 68,488 51,049 45 Table 9 Equity Issuances and Subsequent Use of Funds by Large Firms This table reports firm-level panel OLS regressions for the use-of-funds analysis for equity issuers classified as large firms (top decile of the prior period's equity market value distribution within a country and year) for 25 emerging market countries during the 1991-2016 period. The analysis follows the specification of Kim and Weisbach (2008). The dependent variable for balance-sheet variables (inventory or cash and short-term investment) is Y = log[((Vi - V0)/Assets) + 1]. The dependent variable for cash-flow statement and income statement variables (capital expenditures, acquisitions, research and development expenditures, or reduction of long-term debt) is Y = log[(∑iVi/Assets) + 1]. Independent variables are equity issuance value and other sources of funds, both normalized by total assets, in addition to the log of total assets. Total assets are measured at the value of the year just before the issuance. Dollar changes capture the change in the dependent variable resulting from a one-million-dollar increase in a firm’s equity issuance. All variables are winsorized at the 1% level. All regressions include country and year fixed effects. Standard errors are clustered at the industry (two-digit SIC) level. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Years After Issuance N Log 1 $ Change R2 Issuance Assets (Issuance at t=1) β1 t-stat ∑CAPEX 1 1,579 0.1797 *** 3.538 0.1812 0.258 2 1,529 0.3848 *** 5.736 0.4065 0.334 3 1,361 0.4883 *** 5.379 0.5409 0.361 4 1,179 0.5950 *** 5.334 0.6921 0.392 ∑Acquisitions 1 1,464 0.2058 *** 2.733 0.2005 0.174 2 1,350 0.1743 *** 2.709 0.1702 0.182 3 1,126 0.2237 *** 2.785 0.2172 0.214 4 937 0.1570 1.636 0.1529 0.188 ∑R&D 1 464 0.0090 1.414 0.0083 0.239 2 407 0.0132 0.810 0.0123 0.220 3 343 0.0866 1.508 0.0815 0.308 4 266 0.1528 1.471 0.1439 0.326 Δ Inventory 1 1,183 0.0673 * 1.762 0.0636 0.129 2 1,160 0.0979 ** 2.464 0.0930 0.150 3 1,040 0.1282 *** 3.626 0.1218 0.172 4 899 0.1256 *** 2.796 0.1201 0.184 Δ Cash & ST Inv. 1 1,207 0.5089 *** 7.470 0.4872 0.273 2 1,188 0.2767 *** 3.735 0.2642 0.172 3 1,068 0.2572 ** 2.521 0.2429 0.189 4 927 0.1604 1.572 0.1554 0.216 ∑ LT Debt 1 1,495 -0.0204 -0.201 -0.0208 0.319 Reduction 2 1,418 -0.1184 -1.281 -0.1234 0.358 3 1,237 -0.0873 -0.824 -0.0952 0.392 4 1,052 -0.0570 -0.505 -0.0656 0.408 46 Appendix Figure 1 Emerging Market Equity Issuances Including IPOs and Equity Capital Inflows This figure plots the total value of equity issued, including initial public offerings (IPOs), by firms in 25 emerging market countries (right axis) and total portfolio equity inflows to those emerging markets (left axis) during the 1991-2016 period. All values are reported in billions of 2011 U.S. dollars (USD). 250 350 200 300 150 250 (Billion 2011 USD) Equity Inflows Equity Issuances (Billion 2011 USD) 100 200 50 150 0 100 -50 50 -100 0 Equity Inflows Equity Issuances 47 Appendix Figure 2 Average MSCI Emerging Market Index Weights by Country This figure plots the average weights of the 25 countries included in the MSCI Emerging Market index during the 1997-2016 period. 14% 13% 12% 11% Average MSCI Weight 10% 9% 8% 7% 6% 5% 4% 3% 2% 1% 0% 48 Appendix Table 1 Equity Inflows and Equity Issuance Activity: Aggregate Evidence Robustness Checks This table reports country-level panel OLS regressions of the log of one plus aggregate equity issuance on the log of portfolio equity inflows, for 25 emerging market countries during the 1991-2016 period. Issuances in column (1) include only seasoned equity offerings (SEOs). Issuances in column (2) include both SEOs and IPOs. Issuances in column (3) include only SEOs and excludes issuances by finance, insurance, and real estate firms. All variables are winsorized at the 1% level. All regressions include country and year fixed effects. Standard errors are double clustered at the country and year levels. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Log(1+Equity Issuance) All Sectors, All Sectors, Non-Financial Sectors, Excluding IPOs Including IPOs Excluding IPOs (1) (2) (3) Log(Equity Inflows) 0.5240 *** 0.5305 *** 0.5130 *** (0.093) (0.100) (0.099) Country FE Yes Yes Yes Year FE Yes Yes Yes Number of Observations 428 428 428 49 Appendix Table 2 Equity Inflows, Non-Equity Inflows, and Firms' Equity Issuance Activity This table reports firm-level panel OLS regressions of the log of one plus equity issuance on the log of portfolio equity inflows, the log of non-equity inflows, and their interaction with the large firm dummy variable (top decile of the prior period's equity market value distribution within a country and year) for 25 emerging market countries during the 1991-2016 period. Non-equity inflows are computed as the sum of inflows of FDI in debt securities, portfolio debt, and other debt investments in banks and other sectors, excluding general government and central banks. Columns (1) and (3) report the analysis for contemporaneous inflows. Columns (2) and (4) report the analysis for lagged inflows. All variables are winsorized at the 1% level. All regressions include firm and country- year fixed effects. Standard errors are double clustered at the country and year levels. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Log(1+Equity Issuance) (1) (2) (3) (4) Log(Equity Inflows) * Large Firm 0.0210 *** 0.0522 *** (0.001) (0.016) Log(Lagged Equity Inflows) * Large Firm 0.0173 *** 0.0285 ** (0.003) (0.011) Log(Non-Equity Inflows) * Large Firm -0.0303 * (0.015) Log(Lagged Non-Equity Inflows) * Large Firm -0.0082 (0.011) Firm FE Yes Yes Yes Yes Country-Year FE Yes Yes No Yes Number of Observations 88,314 88,112 74,575 75,723 50 Appendix Table 3 Market Value of Foreign Equity Issuers and Firms Included in the MSCI Index Percentiles This table reports the percentiles associated with foreign equity issuers and firms included in the MSCI index in the total firm size distribution for each country. The country size is computed by taking the mean of the firm’s market value of equity within country. The mean percentile across countries is computed by taking the average of the countries’ percentile. Columns (1) and (2) report the percentiles for foreign equity issuers during the 1991-2016 period and for the firms included in the MSCI Index during the 2006-2016 period, respectively. Percentile of Foreign Equity Issuers Percentile of Firms in the MSCI Country (%) Index (%) (1) (2) Argentina 95 84 Brazil 95 91 Chile 90 98 China 97 98 Colombia 99 91 Czech Republic 46 76 Egypt, Arab Rep. 98 97 Hungary 52 95 India 91 99 Indonesia 97 98 Israel 82 97 Jordan 91 98 Korea, Rep. 98 98 Malaysia 92 99 Mexico 93 93 Morocco N/A 96 Pakistan 86 98 Peru 64 N/A Philippines 95 97 Poland 95 98 Russian Federation 89 90 South Africa 86 93 Thailand 98 99 Turkey 94 96 Venezuela, R.B. 96 88 Average 88 94 51 Appendix Table 4 Equity Inflows and Firms' Equity Issuance Activity, Instrumental Variable Approach Instrument: Lagged MSCI Weight Excluding the Largest MSCI-Weight Countries This table reports the first- and second-stage regressions for the instrumental variable approach using the lagged MSCI emerging market weights as instrument for 24 emerging market countries, excluding China (columns 1-2), Brazil (columns 3-4), or Republic of Korea (columns 5-6), during the 1997-2016 period. In Panel A, columns (1), (3), and (5) report country-level panel OLS regressions of the log of portfolio equity inflows on the log of the one-year lag of the MSCI emerging market weight. Columns (2), (4), and (6) report firm-level panel second-stage regressions of the log of one plus equity issuance on the interaction of the log of contemporaneous portfolio equity inflows with the large firm dummy (top decile of the prior period's equity market value distribution within a country and year). Panel B reports the same regressions using lagged portfolio equity inflows and the two-year lag of the the MSCI Emerging Market weight. Regressions in columns (1), (3), and (5) include country and year fixed effects. Regressions in columns (2), (4), and (6) include firm and country-year fixed effects. All variables are winsorized at the 1% level. First-stage standard errors are double clustered at the country and year levels. Second-stage standard errors are block bootstrapped with 1,000 repetitions, double clustering at the country and year levels. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Panel A. Contemporaneous Inflows Excluding China Excluding Brazil Excluding Republic of Korea First Stage Second Stage First Stage Second Stage First Stage Second Stage Log(Equity Log(1+Equity Log(Equity Log(1+Equity Log(Equity Log(1+Equity Inflows) Issuance) Inflows) Issuance) Inflows) Issuance) (1) (2) (3) (4) (5) (6) Log(Lagged MSCI Weight) 0.5926 *** 0.5126 *** 0.5507 *** (0.177) (0.095) (0.115) Log(Equity Inflows) * Large Firm 0.0276 *** 0.0264 *** 0.0248 *** (0.006) (0.005) (0.006) Firm FE No Yes No Yes No Yes Country FE Yes No Yes No Yes No Year FE Yes No Yes No Yes No Country-Year No Yes No Yes No Yes Number of Observations 277 45,620 279 70,311 282 62,480 Kleibergen-Paap Wald F-Stat 11.26 .. 28.96 .. 23.01 .. Panel B. Lagged Inflows Excluding China Excluding Brazil Excluding South Korea First Stage Second Stage First Stage Second Stage First Stage Second Stage Log(Lagged Log(1+Equity Log(Lagged Log(1+Equity Log(Lagged Log(1+Equity Equity Inflows) Issuance) Equity Inflows) Issuance) Equity Inflows) Issuance) (1) (2) (3) (4) (5) (6) Log(Two-year Lagged MSCI Weight) 0.5940 *** 0.5036 *** 0.5448 *** (0.182) (0.097) (0.117) Log(Lagged Equity Inflows) * Large Firm 0.0166 ** 0.0203 *** 0.0184 *** (0.008) (0.005) (0.006) Firm FE No Yes No Yes No Yes Country FE Yes No Yes No Yes No Year FE Yes No Yes No Yes No Country-Year No Yes No Yes No Yes Number of Observations 264 45,729 266 70,439 269 62,930 Kleibergen-Paap Wald F-Stat 10.70 .. 27.03 .. 21.65 .. 52 Appendix Table 5 Real Economic Effects, Instrumental Variable Approach Excluding Financial Firms This table reports the second stage of firm-level panel instrumental variable regressions of the log of one plus firm real and financial outcomes on the interaction of the log of contemporaneous (Panel A) and lagged (Panel B) portfolio equity inflows with the large firm dummy (top decile of the prior period's equity market value distribution within a country and year) for 25 emerging market countries during the 1997-2016 period, excluding from the sample all finance, insurance, and real estate firms. We use the one-year and two-year lags of the MSCI emerging market weight as instruments for the contemporaneous and lagged portfolio equity inflows, respectively. The dependent variables are capital expenditures, acquisitions, research and development expenditures, inventory accumulation, cash and short-term investments accumulation, and reduction of long-term debt. All variables are winsorized at the 1% level. Second-stage standard errors are block bootstrapped with 1,000 repetitions, double clustering at the country and year levels. All Regressions include firm and country-year fixed effects. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Panel A. Contemporaneous Inflows Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ CAPEX) Acquisitions) R&D) Inventory) Cash&ST Inv.) LT Debt Red.) (1) (2) (3) (4) (5) (6) Log(Equity Inflows) * Large Firm 0.0659 *** 0.0283 *** 0.0355 *** 0.0426 *** 0.0625 *** 0.0169 ** (0.005) (0.005) (0.006) (0.004) (0.006) (0.007) Firm FE Yes Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Yes Number of Observations 62,223 44,521 22,846 63,052 63,259 45,141 Panel B. Lagged Inflows Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ CAPEX) Acquisitions) R&D) Inventory) Cash&ST Inv.) LT Debt Red.) (1) (2) (3) (4) (5) (6) Log(Lagged Equity Inflows) * Large Firm 0.0663 *** 0.0270 *** 0.0337 *** 0.0431 *** 0.0621 *** 0.0158 ** (0.005) (0.004) (0.006) (0.004) (0.006) (0.007) Firm FE Yes Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Yes Number of Observations 62,304 44,587 22,751 63,187 63,387 45,259 53 Appendix Table 6 Equity Issuances and Subsequent Use of Funds by Large Firms Excluding Financial Firms This table reports firm-level panel OLS regressions for the use-of-funds analysis for equity issuers classified as large firms (top decile of the prior period's equity market value distribution within a country and year) for 25 emerging market countries during the 1991-2016 period, excluding from the sample all finance, insurance, and real estate firms. The analysis follows the specification of Kim and Weisbach (2008). The dependent variable for balance-sheet variables (inventory or cash and short-term investment) is Y = log[((Vi - V0)/Assets) + 1]. The dependent variable for cash-flow statement and income statement variables (capital expenditures, acquisitions, research and development expenditures, or reduction of long-term debt) is Y = log[(∑iVi/Assets) + 1]. Independent variables are equity issuance value and other sources of funds, both normalized by total assets, in addition to the log of total assets. Total assets are measured at the value of the year just before the issuance. Dollar changes capture the change in the dependent variable resulting from a one-million-dollar increase in a firm’s equity issuance. All variables are winsorized at the 1% level. All regressions include country and year fixed effects. Standard errors are clustered at the industry (two-digit SIC) level. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Years After Issuance N Log 1 $ Change R2 Issuance Assets (Issuance at t=1) β1 t-stat ∑CAPEX 1 1,209 0.1321 *** 2.757 0.1351 0.216 2 1,189 0.3181 *** 5.292 0.3461 0.310 3 1,066 0.3972 *** 5.241 0.4604 0.370 4 933 0.4586 *** 4.656 0.5590 0.421 ∑Acquisitions 1 1,114 0.1797 ** 2.579 0.1746 0.214 2 1,046 0.1341 ** 2.465 0.1293 0.234 3 880 0.1763 ** 2.596 0.1688 0.262 4 744 0.1105 1.212 0.1058 0.240 ∑R&D 1 494 0.0095 1.386 0.0088 0.247 2 432 0.0225 1.156 0.0211 0.219 3 361 0.0788 * 2.006 0.0749 0.355 4 281 0.1513 ** 2.069 0.1434 0.376 Δ Inventory 1 1,207 0.0629 * 1.746 0.0593 0.134 2 1,190 0.0903 ** 2.324 0.0855 0.158 3 1,067 0.1226 *** 3.288 0.1155 0.184 4 932 0.1171 ** 2.644 0.1101 0.195 Δ Cash & ST Inv. 1 1,208 0.5025 *** 7.654 0.4796 0.269 2 1,190 0.2879 *** 3.919 0.2752 0.166 3 1,068 0.2753 ** 2.652 0.2629 0.205 4 933 0.2286 ** 2.351 0.2180 0.228 ∑ LT Debt 1 1,183 0.0279 0.296 0.0282 0.415 Reduction 2 1,159 0.0057 0.073 0.0060 0.484 3 1,028 0.0826 0.964 0.0938 0.515 4 887 0.0677 0.766 0.0814 0.527 54 Appendix Table 7 Real Economic Effects, Instrumental Variable Approach This table reports the second stage of firm-level panel instrumental variable regressions of the log of one plus firm real and financial outcomes on the interaction of the log of contemporaneous (Panel A) and lagged (Panel B) portfolio equity inflows with the large firm dummy (top decile of the prior period's equity market value distribution within a country and year) for 25 emerging market countries during the 1991-2016 period. We use the contemporaneous and one-year lag of the sum of other countries' equity value as instruments for the contemporaneous and lagged portfolio equity inflows, respectively. The dependent variables are capital expenditures, acquisitions, research and development expenditures, inventory accumulation, cash and short-term investments accumulation, and reduction of long-term debt. All variables are winsorized at the 1% level. Second-stage standard errors are block bootstrapped with 1,000 repetitions, double clustering at the country and year levels. All Regressions include firm and country-year fixed effects. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Panel A. Contemporaneous Inflows Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ CAPEX) Acquisitions) R&D) Inventory) Cash&ST Inv.) LT Debt Red.) (1) (2) (3) (4) (5) (6) Log(Equity Inflows) * Large Firm 0.0696 *** 0.0275 *** 0.0355 *** 0.0442 *** 0.0714 *** 0.0299 ** (0.004) (0.003) (0.007) (0.005) (0.006) (0.012) Firm FE Yes Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Yes Number of Observations 84,121 62,343 26,479 81,328 82,364 62,151 Panel B. Lagged Inflows Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ CAPEX) Acquisitions) R&D) Inventory) Cash&ST Inv.) LT Debt Red.) (1) (2) (3) (4) (5) (6) Log(Lagged Equity Inflows) * Large Firm 0.0693 *** 0.0260 *** 0.0353 *** 0.0444 *** 0.0714 *** 0.0276 *** (0.004) (0.004) (0.006) (0.005) (0.006) (0.010) Firm FE Yes Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Yes Number of Observations 83,930 62,226 26,308 81,192 82,202 62,066 55 Appendix Table 8 Real Economic Effects, Instrumental Variable Approach This table reports the second stage of firm-level panel instrumental variable regressions of the log of one plus firm real and financial outcomes on the interaction of the log of contemporaneous (Panel A) and lagged (Panel B) portfolio equity inflows with the large firm dummy (top decile of the prior period's equity market value distribution within a country and year) for 25 emerging market countries during the 1991-2016 period. We use the contemporaneous and one-year lag of the sum of other countries' equity issuance volume as instruments for the contemporaneous and lagged portfolio equity inflows, respectively. The dependent variables are capital expenditures, acquisitions, research and development expenditures, inventory accumulation, cash and short-term investments accumulation, and reduction of long-term debt. All variables are winsorized at the 1% level. Second-stage standard errors are block bootstrapped with 1,000 repetitions, double clustering at the country and year levels. All Regressions include firm and country-year fixed effects. *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. Panel A. Contemporaneous Inflows Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ CAPEX) Acquisitions) R&D) Inventory) Cash&ST Inv.) LT Debt Red.) (1) (2) (3) (4) (5) (6) Log(Equity Inflows) * Large Firm 0.0698 *** 0.0274 *** 0.0350 *** 0.0443 *** 0.0713 *** 0.0305 *** (0.004) (0.003) (0.007) (0.005) (0.007) (0.011) Firm FE Yes Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Yes Number of Observations 84,121 62,343 26,479 81,328 82,364 62,151 Panel B. Lagged Inflows Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ Log(1+ CAPEX) Acquisitions) R&D) Inventory) Cash&ST Inv.) LT Debt Red.) (1) (2) (3) (4) (5) (6) Log(Lagged Equity Inflows) * Large Firm 0.0694 *** 0.0257 *** 0.0350 *** 0.0444 *** 0.0714 *** 0.0278 *** (0.004) (0.003) (0.006) (0.005) (0.006) (0.009) Firm FE Yes Yes Yes Yes Yes Yes Country-Year FE Yes Yes Yes Yes Yes Yes Number of Observations 83,930 62,226 26,308 81,192 82,202 62,066 56