Using brokerage mergers and closures as two sources of exogenous shock to analyst coverage, this study explores the causal effect of analyst coverage on ex ante expected crash risk as captured by the options implied volatility smirk. We find a significant increase in a firm's ex ante expected crash risk subsequent to an exogenous drop in analyst coverage; this positive effect is stronger for firms initially receiving less coverage. Further, we find analysts' ability matters to investors' assessment of future crash risk. Specifically, we find the impact is more pronounced for the coverage terminations of analysts with more firm-specific or general experience, with greater access to resources, or whose prior forecasts are more accurate than those of their peers. Overall, our results suggest that investors in the options market do recognize analysts as important information intermediaries and monitors and, thus, that analyst coverage influences the underlying stock's expected crash risk.

Standard setters contend that fair value accounting yields the most relevant measurement for financial instruments. We examine this claim by comparing the value relevance of banks' financial statements under fair value accounting with that under current GAAP, which is largely based on historical costs. We find that the combined value relevance of book value of equity and income under fair value is less than that under GAAP. We also find that fair value income is less value-relevant than GAAP income because of the inclusion of transitory unrealized gains and losses in fair value income. More surprisingly, we find that book value of equity under fair value is not more value-relevant than under GAAP, due both to divergence between exit value and value-in-use and to measurement error in fair value estimates. Overall, our results suggest that financial statements under fair value accounting provide less relevant information for bank valuation than financial statements under current GAAP.

We investigate whether CEO power influences a firm's decision to change its compensation system in response to regulatory and public pressure. In particular, we assess whether CEO power influences the choice of performance measures as a form of camouflage to minimize the impact of these reforms on their wealth. We examine one component of CEO pay, namely, the use of performance-vested stock option (PVSO) plans, and find that firms with powerful CEOs attach less challenging targets in the initial PVSOs granted to their CEOs. Such firms also appear to adopt PVSO plans early, and are more likely to do so when faced with public outrage over executive compensation. Our results suggest that powerful CEOs attempt to appease public outrage by quickly adopting PVSOs, but that adopting PVSOs early does not appear to be an optimal strategy for increasing shareholder value. Regulators intended that implementation of PVSOs would be beneficial to shareholders by improving the link between CEO pay and firm performance. However, our results indicate that powerful CEOs can negate some of the beneficial effect of PVSOs through their influence on adoption and choice of performance targets. Data Availability: All data used in this study are publicly available from the sources indicated in the paper.

We analyze the effects of limited investor attention on the stock market reaction to innovation announcements and develop a new measure of patents’ economic value. We hypothesize that, when some investors pay delayed attention to innovation announcements, there will be a post-announcement drift in addition to the announcement effect, with the former decreasing and the latter increasing in investor attention. Using media coverage and abnormal Google search volume as investor attention proxies, we find consistent evidence. Our new attention-weighted measure of patents’ economic value has greater predictive power for future firm performance than measures based on the announcement effect alone.

This research proposes a novel firm-based model for pricing cyber insurance. Our model considers two types of cyber risk: virus attacks and data breaches. Virus attacks deliver adverse shocks to the firm’s productivity, while data breaches cause premium customer departures that worsen the prospect of the firm’s product demand. We derive the endogenous structural form of cyber losses in firms and utilize it to solve the formula for cyber insurance premiums. Our quantitative results show that the consensus prediction about a strictly positive premium-risk nexus is no longer valid. Asymmetries in the sub-premium’s sensitivity to cyber risks from different sources and the premium customer loss rates jointly shape the complexity of the relation between cyber insurance premiums and cyber risks. Improvements in the product demand conditions enhance firms’ incentives to hedge cyber losses and push premiums higher. Lastly, we discuss the influence of product price competition on premiums.

Previous research has revealed that return spreads between stocks with high and low characteristics-based factor beta remain insignificant. This study investigates the time variation in the pricing of various characteristics-based factors, uncovering a notable two-regime pattern: high-beta portfolios yield higher returns than low-beta portfolios after high-sentiment periods, while the opposite occurs after low-sentiment periods. Remarkably, this two-regime pattern is completely reversed for macro factors. Mutual fund and hedge fund returns corroborate these findings. Our results suggest that exposure to characteristics-based factors likely represents mispricing levels, particularly during high-sentiment periods, whereas exposure to macro factors likely represents risk, particularly during low-sentiment periods.

This paper explores implications of weak identification in common ‘long memory’ and recent ‘rough’ approaches to modeling volatility dynamics of financial assets. We unveil an asymptotic near-observational equivalence between a long memory model with weak autoregressive dynamics and a rough model with a near-unit autoregressive root. Standard methods struggle to distinguish them, and conventional asymptotics are invalid. We propose an identification-robust approach to construct confidence sets that reveal the uncertainty and aid inference. Empirical studies based on realized volatility and trading volume often fail to statistically reject either model, thereby providing evidence of their potential coexistence.

Social media has become a popular venue for individuals to share the results of their own analysis on financial securities. This paper investigates the extent to which investor opinions transmitted through social media predict future stock returns and earnings surprises. We conduct textual analysis of articles published on one of the most popular social media platforms for investors in the United States. We also consider the readers' perspective as inferred via commentaries written in response to these articles. We find that the views expressed in both articles and commentaries predict future stock returns and earnings surprises.

Social ties between mutual funds and the companies in which they invest (investees) can both facilitate information transfers and encourage favoritism. Using the investment choices of mutual funds in China, we compare investment performance of holdings in companies that are socially connected to mutual funds versus those that are not. We find that funds allocate more investment to connected investees' stocks, especially when a fund is weakly monitored. This overweighting is greater in times of poor investee performance, when the benefits of additional investment to the connected investees are high. Weakly monitored funds' preference for connected stocks hurts the returns of these funds, yielding a 6.6 percent lower annualized risk-adjusted return, relative to closely monitored funds. These results suggest that, absent sufficient monitoring, agency conflicts generated by social networks can dominate the information advantages of these networks.

In the presence of litigation-facing suppliers, the supply chain relationship is at risk. Suppliers with principal customers (dependent suppliers) have a higher concentration of sales to customers, and they are more at risk relative to suppliers without principal customers (non-dependent suppliers). As a result, we predict and find that litigation disclosure patterns differ for the two supplier types: dependent suppliers are more likely to delay bad news and accelerate good news related to litigation outcomes, compared to non-dependent suppliers. Such strategic disclosure patterns in our end-game setting are opposite to those documented in the existing supply chain literature for the repeated-game setting (for example, Hui, Klasa, and Yeung 2012).