Dividend Payments And Stock Price Crash Risk
University of Waterloo
School of Management, Huazhong University of Science and Technology
University of Kentucky – Von Allmen School of Accountancy
March 9, 2016
Abstract:
Using a large sample of U.S. firms over the period of 1991-2011, this study examines whether and how cash dividend payments affect stock price crash risk. We predict and find that dividend payments are negatively associated with stock price crash risk. In addition, we find that the effect of dividend payments on lowering crash risk is more pronounced for firms with higher information asymmetry (proxied by higher market-to-book ratios or positive R&D expenditures) and for those with larger free cash flows. Moreover, we document that dividend payments reduce bad news hoarding (overinvestment) whereas bad news hoarding (overinvestment) leads to crash risk, suggesting that dividend payments (1) enhance financial reporting quality and investment efficiency and (2) mitigate stock price crash risk through two channels: curbing bad news hoarding and curtailing overinvestment. Our main results are robust to various sensitivity checks including controls for endogeneity concerns. Our findings are important to investors, policy makers, and academics, because they (1) suggest that dividend payments are an easily observable indicator that investors can use to assess firms’ financial reporting quality and crash risk and (2) support some policymakers’ proposition that dividends mitigate earnings manipulation and fraud.
Dividend Payments And Stock Price Crash Risk – Introduction
Stock price crashes, defined as the occurrence of extreme negative stock returns, cause significant losses in investors’ wealth and in their confidence in the capital markets. The question of what causes stock price crashes has attracted much research. Prior studies identify two fundamental causes for crashes. First, due to agency conflicts, managers have incentives to hoard bad news for extended periods (Ball, 2009; Kothari et al., 2009; Graham et al., 2005). When the accumulated bad news reaches a tipping point, it is likely to be released to outside investors all at once, leading to an abrupt, large-scale decline in a firm’s stock price, or a stock price crash (Jin and Myers, 2006; Hutton et al., 2009; Kim et al, 2011a; Kim and Zhang, 2015). Second, due to agency conflicts, managers have incentives to overinvest in privately beneficial but negative net present value (NPV) projects and to hide poor performance of these unprofitable projects, which hinders owners and the board of directors from taking corrective actions at an early stage. As a result, bad projects are kept alive for too long and poor performance accumulates over time, which eventually leads to a crash in asset prices (Bleck and Liu, 2007; Benmelech et al., 2010).
The purpose of this study is to examine whether and how a firm’s dividend payout policies affect stock price crash risk. We expect dividend payments to reduce crash risk based on findings from prior research. First, Skinner and Soltes (2011) show that earnings quality of dividend-paying firms is higher than that of non-dividend-paying firms. Caskey and Hanlon (2013) find that paying dividends reduces a firm’s probability of committing accounting fraud, suggesting that dividend payments indicate higher financial reporting quality. Lawson and Wang (2015) document that dividend-paying firms pay lower audit fees than non-dividend-paying firms after controlling for common determinants of audit fees, suggesting that auditors perceive financial reporting quality of dividend-paying firms as higher. Intuitively, high quality financial reporting reduces managers’ leeway to hoard bad news. Second, Easterbrook (1984) and Jensen (1986) point out that dividend payments increase the possibility that firms need to access external capital markets. Dividend payments, thus, can subject managers to external monitoring and disciplinary actions by outside stakeholders such as financial analysts, institutional investors, commercial and investment banks, and credit rating agencies. This possibility of being subject to close scrutiny reduces managers’ incentives and abilities to hide bad news. In sum, to the extent that dividend payments enhance financial reporting quality and subject managers to external scrutiny, they curb managerial bad news hoarding, thereby lowering stock price crash risk.
Third, Jensen (1986) identifies free cash flows, i.e., cash flows in excess of what are needed to fund positive NPV projects, as a major source that exacerbates agency conflicts between managers and shareholders. The presence of free cash flows enables managers to overinvest in unprofitable projects (e.g., Harford, 1999), pursue “empire-building” activities, or extract perquisite consumption (Yermack, 2006). Therefore free cash flows associate negatively with shareholder value. Paying dividends reduces free cash flows available to managers and thus constrains managers’ ability to squander shareholder wealth via overinvestment or excessive spending.1 Existing evidence is, by and large, consistent with the view that dividend payments reduce agency costs of free cash flows (e.g., Lang and Litzenberger, 1989; Lie, 2000; Officer, 2011) and deter managers from engaging in value-destroying overinvestment. In reviewing the dividend literature in the past two decades, Farre-Mensa et al. (2014) conclude that “[t]he accumulated evidence on payout and agency indicates that firms use payouts to reduce potential overinvestment by management.” To the extent that dividend payments reduce free cash flows available to managers and thus overinvestment, they contribute to lowering stock price crash risk.
We first test our main hypothesis that dividend payments are negatively associated with stock price crash risk. Following Chen et al. (2001), Hutton et al. (2009), and Kim et al. (2011a, 2011b), we adopt two measures of firm-specific stock price crash risk: (1) the likelihood of the occurrence of extreme negative firm-specific weekly returns and (2) the negative conditional skewness of firm-specific weekly returns. Using a large sample of 63,795 firm-year observations during 1991-2011 from the intersection of Compustat and CRSP, we find that crash risk is significantly and negatively associated with an indicator variable for cash dividends, suggesting that dividend-paying firms are less susceptible to crashes than non-dividend-paying firms. When we limit our sample to firm-years with positive cash dividends, we find that crash risk is significantly and negatively associated with dividend yield. This suggests that for dividend-paying firms, those that pay more dividends (i.e., higher dividend yield) are less prone to crashes than those that pay fewer dividends (lower, but non-zero, dividend yield).
Second, we investigate the potential channels through which cash dividends reduce crash risk by examining firm-specific factors that explain cross-sectional variations in the negative association between crash risk and dividends. If dividends reduce crash risk by curbing bad news hoarding due to dividends enhancing financial reporting quality (Skinner and Soltes, 2011; Caskey and Hanlon, 2013; Lawson and Wang, 2015) and subjecting managers to external monitoring (Easterbrook, 1984; Jensen, 1986), then we expect that the mitigating effect of cash dividends on crash risk is more pronounced for firms with higher information asymmetry, because such firms tend to offer greater managerial leeway for hoarding bad news. If dividends reduce crash risk by curtailing free cash flows (Jensen, 1986), then we expect that the effect of dividend payments on mitigating crash risk is greater for firms with larger free cash flows.
See full PDF below.
