In recent years, there have been significant changes in the information environment facing firms. In particular, the explosive growth in computing power and the reduction in the costs of disseminating economically relevant information due to the widespread use of the internet have significantly enhanced the ability of investors to produce and transmit information useful for valuing firms’ equity and for evaluating firm’s financing and investment policies. To give one example, Several internet sites allow employees to rate a firm’s management, work culture, compensation schemes, and overall prospects, e.g., Glassdoor Employee Ratings. While it is difficult to argue that such ratings directly reduce the information asymmetry facing firms in the equity market (since the employees giving the ratings may not be senior officers of these firms), such ratings may be taken into account by equity market investors when valuing firms’ equity.
These ratings by employees can therefore be viewed as “soft information” affecting firms’ equity values and potentially their external financing decisions. However, there has been no empirical analysis in the literature so far analyzing how the existence of such soft information affects the external financing and investment policies of firms. The objective of our paper is to fill this gap in the literature by analyzing the effect of soft information (such as that contained in the online employee ratings of firms) on firms’ external financing and investment policies.
We view online employee ratings as soft information available to equity market investors (over and above any publicly available information available to them about the firm), which they can use to value a firm’s equity and evaluate its financing policies such as its decision whether to issue equity to finance its projects; its choice of external financing between equity and debt; and its investment policy. We use a sample of around 1.1 million employee ratings from the Glassdoor website covering a sample of 2842 public firms during the period 2008 to 2017 to construct our proxy for soft information. We analyze the relation between the above soft information proxy and the external financing policies of firms, their stock returns upon equity issue announcements, their propensity to have a positive stock return upon an equity issue announcement, and their choice between issuing debt and equity. We also analyze the relation between the above soft information proxy and firms’ investment policies, the participation of institutional investors in their seasoned equity issues, and their post-issue operating performance.
We develop a theoretical framework to develop testable hypotheses for our empirical analyses based on the theoretical model of Chemmanur and Jiao (2005). The intuition there is that expectations about soft information affect firms’ external financing decisions. Their model builds on Myers and Majluf (1984), but differentiates itself by having two sources of information observable by outsiders: the equity issue or no issue decision and the realization of soft information. Firm insiders have some private information, while the outsiders can only observe the soft information realization about the firm, e.g., employee ratings. The soft information is imperfectly correlated with the quality of the firm. Firms have expectations about the soft information realization received by investors and base their equity issue decisions on it. In contrast to the Myers and Majluf (1984), there is a non-monotonic relation between firm quality and equity issuance. Further, seasoned equity offerings (SEOs) may have a positive announcement effect in equilibrium, again in contrast to Myers and Majluf (1984), where SEOs have only a negative announcement effect on average.
We use the above theoretical framework (with some additional assumptions) to develop testable hypotheses on the relation between the realizations of the online employee ratings of a firm (prior to an equity issue) and the following: the algebraic value of the abnormal stock return upon the announcement of its equity issue (i.e., the announcement effect); the propensity of the above announcement return to be positive; the propensity of the firm to issue equity rather than debt; firm investment; institutional investor participation in the firm’s equity issue; and its long-term post-SEO operating performance.
We use various employee ratings (overall ratings, management ratings, and other rating measures) available on the Glassdoor website to construct the main independent variables for our empirical analyses. All these rating variables are correlated and capture different aspects of online employee ratings. Since our focus is on capturing soft information contained in employee reviews, and given that different employee rating measures capture different pieces of information, we also use common factor analysis to create a “common rating factor” that captures the information content common to all the above rating measures.
We show, first, that firms with higher average online employee ratings prior to equity issues are associated with algebraically greater abnormal stock returns upon announcement of an equity issue (three-day cumulative abnormal stock return (CAR)), i.e., with higher announcement effects. Our results are also economically significant: A one standard deviation increase in overall rating is associated with a 1.7 percentage point increase in the abnormal stock return.
Second, we show that firms with higher average online employee ratings are associated with a greater propensity to have positive announcement effects. Again, the above results are economically significant. A one standard deviation increase in overall rating is associated with a 5.5 percentage point increase in the probability of a firm having a positive announcement effect.
Third, we show that firms with higher prior average online employee rating realizations are associated with a greater propensity to issue equity rather than debt to raise external financing. A one standard deviation increase in overall rating is associated with a 1.6 percentage point increase in the probability of issuing equity over debt.
Fourth, we show that firms with higher prior average online employee rating realizations are associated with a greater level of annual investment expenditures. A one standard deviation increase in overall rating is associated with an increase of 7.5 percentage point in annual investment expenditures made by a firm.
Fifth, we show that firms with higher prior average online employee rating realizations are associated with greater participation by institutional investors in their SEOs. A one standard deviation increase in overall rating is associated with an increase of five institutional investors investing in the firm’s SEO.
Lastly, we show that firms with higher prior average online employee rating realizations are associated with better long-run post-SEO operating performance. A one standard deviation increase in the common rating factor is associated with an increase of 6.3 percentage points in the operating performance of a firm over the three years subsequent to the SEO.
Our baseline (OLS) analyses may suffer from some endogeneity concerns. It may be possible that higher quality firms have higher announcement returns and have higher online employee ratings. Further, such firms may have greater probability of positive announcement returns and may have cheaper access to external financing from the equity market.
To address the above endogeneity concern, we use the staggered adoption of Anti-SLAPP laws, which protect the First Amendment rights of U.S. citizens, across the U.S. states. SLAPP (strategic lawsuit against public participation) is a retaliatory lawsuit filed against someone who has spoken against the plaintiff in a public forum. SLAPP lawsuits are filed against online reviewers as well. SLAPP lawsuits may have chilling effects on online reviewers due to the threat of litigation and may reduce the information content of those reviews. The provision of anti-SLAPP laws in some states may therefore potentially affect the propensity of individuals to provide online reviews (and ratings) of firms and to provide more informative reviews. This is because anti-SLAPP laws provide for recovery of attorney fees of defendants and swift dismissal of meritless lawsuits and put the burden of proof on plaintiffs.
We make use of the staggered adoption (and in some cases rejection) of anti-SLAPP laws in U.S. states using a difference-in-differences (DID) approach. We expect that online employee ratings given for firms headquartered in states with anti-SLAPP laws to have more information. Hence, outside investors may put greater weight on the information contained in online employee ratings for firms headquartered in states with anti-SLAPP laws. The passage of anti-SLAPP laws is clearly exogenous to the issuance of equity by firms so that these laws provide an appropriate economic setting for our identification tests.
Our DID analysis shows that external financing of firms headquartered in states having anti-SLAPP laws are affected to a greater extent by the corresponding online employee ratings. We find that firms with higher online employee ratings headquartered in states with anti-SLAPP laws have greater announcement effects; a greater probability of a positive announcement effect; and a greater probability of issuing equity rather than debt, compared with similar firms headquartered in states without anti-SLAPP laws.
This post comes to us from Professor Thomas Chemmanur at Boston College’s Carroll School of Management; Harshit Rajaiya, a PhD candidate at Boston College’s Carroll School of Management; and Professor Jinfei Sheng at the University of California-Irvine. It is based on their recent article, “How Does Soft Information Affect External Firm Financing? Evidence from Online Employee Ratings,” available here.