Essays on corporate finance

Resumen

This thesis consists of three essays. The first, titled “Do institutional investors care about corporate social responsibility and irresponsibility?" investigates how Corporate Social Responsibility (CSR) and Corporate Social Irresponsibility (CSiR) change when the degree of monitoring by institutional investors varies. CSR is defined as the set of proactive policies that a firm voluntarily adopts to improve the wellbeing of various stakeholders and society at large, including charitable contributions and voluntary programs for employees (Di Giuli and Kostovetsky, 2014). Conversely, CSiR captures social and environmental misbehavior linked to future economic penalties or civil fines (Di Giuli and Kostovetsky, 2014; Ioannou and Serafeim, 2015; Kruger, 2015). Companies worldwide are taking social and environmental objectives into account when designing their corporate strategies. Institutional investors are the largest type of shareholder in public corporations in the United States, and the empirical evidence shows that they influence corporate policies and governance. Institutional investors are generally a key monitoring agent for corporate governance. However, their role on CSR and CSiR have received less attention. In this study, we show that tighter monitoring by this type of shareholder reduces both CSR and CSiR. While the reduction in the former is driven by firms prone to agency conflicts, the drop in the latter is cause by firms with a larger need for external advising. The literature on CSR proposes on the one hand, that these activities enhance shareholder value by developing a competitive advantage through efficient contracting with different stakeholders of the corporation (Deng, Kang, and Low, 2013) or providing an insurance Godfrey, Merrill, and Hansen (2009); Hong and Liskovich (2014) in case of corporate misbehavior. On the other hand, the literature also suggests that these activities can destroy shareholders' value when the cost of these activities do not compensate its benefits (Friedman, 1970; Jensen and Meckling, 1976; Benabou and Tirole, 2010). Meanwhile, the literature on CSiR suggests that reducing CSiR might lead to lower future costs (Chatterji, Levine, and Toffel, 2009; Dimson, Karakas, and Xi, 2015) and lower stock return volatility (Bansal and Clelland, 2004; Godfrey et al., 2009; Hoepner, Oikonomou, Sautner, Starks, and Zhou, 2016). However, the benefits of corporate misbehavior could outweigh its costs, through higher chances of obtaining potentially corrupt arrangements with policy makers (Borisov, Goldman, and Gupta, 2015), or a lower hazard rate of being discovered, and being less likely to be detected by regulators (Yu and Yu, 2011). Monitoring efforts by institutional investors are not randomly assigned. Therefore, we use a plausible source of exogenous variation in monitoring intensity by this type of shareholder (Kempf, Manconi, and Spalt, 2017; Ben-Rephael, Da, and Israelsen, 2017a; Liu, Low, Masulis, and Zhang, 2017). We analyze how CSR and CSiR change when institutional investors deviate their attention away from the firm for reasons unrelated to that firm. Following Kempf et al. (2017), we exploit the fact that when stock returns in one industry are extremely high or low, shareholders shift their attention to the firms in their portfolio that belong to those industries, paying less attention to the rest of their investments. This loosening of monitoring activities not only creates an opportunity for managers to extract rents from shareholders, but also blocks managers access to external advising. We find that CSR increases when institutional investors are distracted, consistent with the literature suggesting that this investors deter those activities that destroy value (Hong, Kubik, and Scheinkman, 2012; Di Giuli and Kostovetsky, 2014; Adhikari, 2016; Masulis and Reza, 2015). We also find that distraction increases CSiR, consistent with institutional investors promoting activities that are more likely to lead to increases in shareholder value (Chatterji et al., 2009; Dimson et al., 2015; Bansal and Clelland, 2004; Godfrey et al., 2009; Hoepner et al., 2016). Institutional investors often do not content themselves with the pure monitoring of managerial decision making (e.g., Jensen and Meckling, 1976; Jensen, 1986; Almazan, Hartzell, and Starks, 2005) but offer expertise and external advising to aid the top management team (e.g., Lerner, 1995; Re-pullo and Suarez, 2004; Tirole, 2010). We explore the effect of institutional investor monitoring in settings prone to agency conflicts and those where advising is more valuable or necessary. We find that institutional investor distraction leads to larger levels of CSR only in settings that are prone to present agency conflicts. Whereas, institutional investor attention leads to decreases in CSiR in settings where external advising is more necessary. Our results are robust to alternative definitions of institutional investor monitoring, CSR, and the timing of the effect. First, we assume that the larger the fraction of shares owned by institutional investors is, the tighter the monitoring. Thus, we exploit variation in institutional ownership due to assignment to the Russell 1000/2000 Indexes (Appel, Gormley, and Keim, 2016; Crane, Michenaud, and Weston, 2016). This mechanical increase in institutional investor ownership exogenously increases monitoring intensity in treated firms, which allows for a causal interpretation of the results. The results from this alternative setting match our main findings. In addition, we show robustness tests for alternative measures of CSR provided by Thomson Reuters' ESG Scores (formerly known as ASSET4 Equally Weighted Ratings) and employed in previous literature (Dyck, Lins, Roth, and Wagner, 2018) and our results remain unchanged. Finally, we address concerns regarding the timing of CSR and CSiR, because previous literature suggests that changes in social and environmental policies arise with some lag (e.g., Adhikari, 2016). We show that our results are statistically more significant when we measure CSR and CSiR in the following year. There is ample evidence that institutional investors use shareholders' proposals as a mechanism to change governance policies (see, for instance, Appel et al., 2016; McCahery, Sautner, and Starks, 2016). We show dis-traction does not seem to affect the number of social and environmental proposals submitted at shareholder meetings, which, together with recent empirical evidence suggests that institutional investors prefer to engage via constructive dialogue through various engagement actions such as meetings, calls, emails or letters (McCahery et al., 2016; Hoepner et al., 2016). Various studies attempt to empirically assess the relationship between institutional ownership and CSR (see Coffey and Fryxell, 1991; Graves and Waddock, 1994; Johnson and Greening, 1999; Neubaum and Zahra, 2006; Dyck et al., 2018). Coffey and Fryxell (1991) find contradictory results using different measures of CSR. The closest paper to ours is the article by Dyck et al. (2018), which studies the impact of institutional ownership on the environmental and social performance of an international sample, and finds that institutional ownership has a positive impact on firms' social and environmental performance. Our results differ from Dyck et al. (2018) in the scope of our samples, our measurement of CSR and the different sources of exogenous variation employed. Our paper is also related to the literature on the effect of institutional investors and monitoring on managerial compensation (Almazan et al., 2005), market for corporate control (Chen, Harford, and Li, 2007; Fich, Harford, and Tran, 2015; Kempf et al., 2017), innovation (Aghion, Van Reenen, and Zingales, 2013) and dividend payout (Crane et al., 2016). We add to these sets of papers by showing that institutional investors play a pure monitoring role with respect to CSR and an advisory role regarding CSiR. In the second chapter, “The bright side of stock repurchases," I provide evidence of actual stock repurchases increasing future investment in firms subject to strong asymmetric information that are reliant on external capital markets. American firms spend several millions of dollars in stock repurchases each quarter, drawing criticism from politicians and academics. Almeida, Fos, and Kronlund (2016) conclude that managers are willing to shift away investment and employment towards stock repurchases that allow them to meet analysts' earnings per share (EPS) targets. In this paper I present evidence of actual repurchases increasing investment in firms that suffer the most from asymmetric information and rely on external finance. This positive effect of stock repurchases on investment is consistent with a theory of firms using this payout form as a signal that allows them to access capital markets later in better terms. It is well established that in the presence of informational asymmetries and no information transfer, market value must reflect the average firm quality (Akerlof, 1970; Leland and Pyle, 1977; Ofer and Thakor, 1987), meaning that good (bad) firms will be undervalued (overvalued). The credibility of actual repurchases as a signal comes from the fact that for bad firms repurchasing shares is too costly, as long-term investors would be diluted if manager engage in that kind of transaction. To the contrary, good- firm insiders benefit from actual repurchases when the firm is undervalued (Fried, 2014, 2015), what allows for a separating equilibrium. The existence of informed shareholders with a long-term horizon seems a reasonable assumption in the average American publicly traded firm, for which Perez-Gonzalez (2002) and Holderness (2009) document high insider ownership. In this signaling context, stock repurchases are expected to increase investment in firms suffering the most from asymmetric information, as that would allow them to convey their type to capital providers. The empirical results are consistent with this argument. In particular, the estimates indicate that for a typical increase in market-based asymmetric information, a one-standard-deviation increase in stock repurchases boosts investment by 10%. However, stock repurchases are only expected to affect investment in those firms that rely on external capital markets to finance investment and, therefore, need to signal their type. Indeed, when I sort firms into external finance dependent (EFD) and those that are not reliant on external finance (No EFD), I find that the results are driven by the former. To further support the causal interpretation of the results, I instrument repurchases using the exogenous price pressure created by distressed mutual funds (Coval and Stafford, 2007; Khan, Kogan, and Serafeim, 2012; Edmans, Goldstein, and Jiang, 2012). Coval and Stafford (2007), Edmans et al. (2012) and Khan et al. (2012) analyze the effect of forced sales and purchases by distressed mutual funds on equity markets and show that trades concentrated in a limited number of securities significantly affect stock prices, resulting in transaction prices that deviate from fundamental values. The authors document ex post price reversals, which are inconsistent with information-based trading. These papers show that this type of misvaluation is unrelated to firms' prospects, providing support for the exclusion restriction required by the instrumental variables approach. The first-stage estimation shows that firms suffering from exogenous price pressures from distressed mutual funds are more likely to repurchase shares and, on average, increase repurchases relative to shares outstanding by 13.6%. In the second-stage estimation, I first show that stock repurchases motivated by mutual funds-fire sales neither increase nor decrease investment on average. I further instrument the interaction between stock repurchases and information asymmetry, and the results confirm the panel fixed-effect results: stock repurchases increase investment, particularly in firms with higher information asymmetry, and the results are driven by EFD rms. In additional tests, I replace the firm-level variable for asymmetric information with the VIX index, which proxies for aggregate volatility in firms' market value (Kim and Kung, 2017). The results hold, and in general are stronger, under the alternative proxy for asymmetric information, and this suggests that the market finds the signal more useful when overall uncertainty over firms' value is higher. I also analyze the effect of stock repurchases on financing policies. On the one hand, if EFD firms repurchase shares to signal their type, they should be raising external financing in the following period. I find that stock repurchases are followed by debt issuance in the EFD subsample (but not for cash-rich firms). Moreover, I do not find changes in cash holding, thus reducing concerns that the firms with excess cash are the ones repurchasing shares and increasing investment. Regarding equity issuance, the results are inconclusive, but in general, I find no significant changes in this source of financing. However, the results should be interpreted with caution because the instrument is very likely to violate the exclusion restriction for financing policies and, in particular, for equity issuance. Finally, stock repurchases should allow good firms to borrow at a lower cost of capital if they can actually convey information to borrowers and investors. Using data on new debt issuance I document a negative association between stock repurchases and the at-issue yield spread on new debt issuance. Moreover, the association is increasing in asymmetric information. In addition, I find that firms that repurchase shares before seasoned equity offerings (SEOs) experience lower discounts in the [0,1] window around the announcement, which is consistent with Billett and Xue (2007) and Bond and Zhong (2016). The effect is stronger for firms that suffer the most from adverse selection, which is in line with the mechanism proposed in this paper. In additional tests I show that the results are stronger when I drop firms with negligible managerial ownership, or when firms are less likely to have informed (and long-term oriented) investors, as expected. Moreover, I the results are economically and statistically stronger when I drop firm-quarter in which firms with short-term incentives (close to the zero EPS threshold). This paper contributes to the existing literature in several directions. First, I contribute to the payout policy literature providing a new motive for actual repurchases that was not previously explored in the literature (see Allen and Michaely, 2003; Farre-Mensa, Michaely, and Schmalz, 2014, for a comprehensive review), with distinct implications for investment. I contribute to this vast literature arguing that repurchase may play an important signaling role that results in improved firms investment. On the other hand, this paper contributes to the current debate on the real effect of stock repurchases. The extensive literature on payout policy in general and stock buybacks in particular, has generally left the real effects unaddressed, and if anything, it provides evidence of a negative association between stock repurchases and real outcomes (Grullon and Michaely, 2004; Bens, Nagar, and Wong, 2002), or a negative causal impact (Almeida et al., 2016). To the contrary, I provide causal evidence in the opposite direction. Finally, this paper also adds to the recent literature suggesting that financing, payout, and investing decisions have to be analyzed in a unified framework (Bond and Zhong, 2016; Farre-Mensa, Michaely, and Schmalz, 2017). Similar to Bond and Zhong (2016), this paper provides a rationale for why firms with financial deficit might engage in stock repurchases, violating the pecking order hypothesis. The third chapter, “The role of accounting quality during mutual fund fire sales," explores the role of accounting quality in mitigating firm's undervaluation in the stock market generated by mutual funds liquidity needs. Lee and So (2015) argue that “Market prices are buffeted by a continuous ow of information, or rumours and innuendos disguised as information" (pp. 64). The process of incorporating relevant information into prices takes time and e ort, and as a consequence firm stock prices might suffer deviations from the fundamental value. The objective of financial reporting is to provide information about the firm that is useful for current and potential investors decision-making. Higher financial reporting quality should mitigate information asymmetries, providing a better estimation of firms' fundamental, reducing adverse selection in the trading of securities, and promoting the efficient allocation of capital (e.g. Biddle, Hilary, and Verdi, 2009; Dechow, Ge, and Schrand, 2010). Previous studies show that reporting quality matters in broad samples and over long periods (Francis, LaFond, Olsson, and Schipper, 2005; Jin and Myers, 2006; Biddle et al., 2009). However, some scholars cast doubts on the usefulness of accounting information due to the use of noisy estimates and judgment calls, the lack of comparability of financial statements, and managerial incentives to misreport (Sherman and Young, 2016). If that were the case, investors might decide to rely on alternative (arguably, more timely and reliable) sources of information, such as analysts recommendations (Sulaeman and Wei, 2018) or management earning forecasts (Kadach, 2017), to assess the value of a rm. That is, whether accounting quality matters, particularly during periods of severe underpricing, is ultimately an empirical question. While previous studies have mainly analyzed the role of accounting quality in incorporating fundamental information into stock prices, its role during the arrival of non-fundamental information has been generally left unaddressed. Recent studies documented that correlated mutual funds liquidity needs are an important source of price pressure for the stocks that these mutual funds hold in their portfolios, causing temporary mispricing in these securities (e.g. Coval and Stafford, 2007; Ali, Wei, and Zhou, 2011; Khan et al., 2012; Sulaeman and Wei, 2018). Analyzing the role of accounting quality in this setting is motivated by (i) the exogeneity of the shock, which allow us to better identify the impact of accounting quality, and (ii) the strong impact that this type of shock has on prices, making it an ideal setting to analyze the role of accounting quality. The proxy for price pressure is constructed following Coval and Stafford (2007) (as in Chapter 2). Fire sales due to mutual funds liquidity needs increase uncertainty regarding the value of firms suffering price pressures, and therefore, increases the need for reliable information to accurately price stocks. In this situation, if financial reports fulfill their role of reducing information asymmetries between the firm and market participants, firms with better accounting quality should experience less severe mispricings. We construct our main proxy for accounting quality based on the afore-mentioned argument that financial reporting should provide information about the firm's operations, especially its cash flows. In particular, the accruals component of earnings are valuable if they can be linked to cash flows and provide valuable information to estimate the firm's stock price. Building on this idea, Dechow and Dichev (2002) derive a measure of accrual quality as the residuals from the regression of changes in working capital on past, present, and future operating cash flows. We use the augmented version suggested by McNichols (2002) as our main proxy for accounting quality. Consistent with our prediction, we find evidence that firms with higher financial reporting quality have lower deviations from fundamental value compared to their lower quality counterparts. In particular, we find that an increase of one decile of accounting quality increases the abnormal return by 0.255% in the quarter of the shock, 0.603% when including the previous quarter, and 0.708% when considering also the following quarter. These figures imply a 17% to 20% reduction in the mispricing during those event windows. These results hold after controlling for firm and stock characteristics previously considered in the literature, and are robust to a wide set of additional tests, such as alternative benchmarks to estimate abnormal returns, the inclusion of additional controls, the use of median regression that is less sensitive to the effect of outliers, or the exclusion of financial and regulated firms and alternative earnings quality measures. These findings are consistent the value relevance of accounting information (Francis et al., 2005), and suggests that having better accounting quality help investors dis-entangling noise from fundamental information in equity prices, and causing smaller deviations from fundamental value. We rule out some alternative explanations for our results, such as governance, or firms complexity including controls for institutional ownership and analysts coverage, two widely used proxies for governance, and firm economic fundamentals that previous studies find to affect accruals quality (Francis et al., 2005). Moreover, our result are robust to alternative definitions of accounting quality, in particular, conditional conservatism (LaFond and Watts, 2008; Khan and Watts, 2009; Garcia-Lara, Garcia-Osma, and Penalva, 2016). We find that more conservative firms return to fundamental value faster, but we do not find consistent evidence of conservatism reducing the mispricing during or before the event quarter. Finally, we exclude multiple shocks to a firm in the same year to rule out the possibility of the results being driven by firms suffering continuous shocks. All the aforementioned conclusions remain unchanged in this reduced sample. This paper contributes to the literature exploring how internal and external stakeholders actions can mitigate misvaluations driven by mutual fund liquidity needs (Ali et al., 2011; Kadach, 2017; Sulaeman and Wei, 2018). On the other hand, we add to the literature on the usefulness of high quality financial reporting reducing information asymmetries when the firm faces exogenous market shocks (see also, Hilary, 2008). Finally, our paper is also related to the extensive body of research that shows that stock prices underreact to major corporate events (e.g., Ball and Brown, 1968; Womack, 1996). Some explanations for the price drifts include methodological issues (Fama, 1998), behavioral explanations (Frazzini, 2006), the lack of institutional investors attention to these news (Ben-Rephael, Da, and Israelsen, 2017b), or firm complexity (Cohen and Lou, 2012; Barinov, Park, and Yildizhan, 2016). We contribute to this debate proposing that the quality of the accounting information might partly explain the slow incorporation of news to prices.