Based on a comprehensive visual deception score constructed from China’s mandatorily disclosed IPO roadshow videos using computer vision and machine learning, we find that visual cues embedded in these videos predict IPO financial fraud. The predictive power is incremental to financial, textual, and audio indicators. Further analyses show that these cues are well grounded in psychology research and focusing on the psychology-based deception-cue-area features improves fraud predicting performance. Furthermore, decomposing each video into five topics, we find that only the deception score of the topic on financial performance is positively associated with financial fraud. We further show that insiders of firms with higher visual deception scores are more likely to sell shares shortly after lock-up periods expire, and these firms experience significantly lower subsequent stock returns. Our study demonstrates the feasibility and potential value of using video information for fraud detection in accounting and finance.

We provide initial evidence that stock exchange procedures around closing auctions advantage speed traders at the expense of auction participants. We show that, on Nasdaq and NYSE Arca, 4:00 pm earnings releases result in informed trading in the continuous regular-hour session in the short window between 4:00 pm and the closing auction; this trading subsequently moves closing prices in the direction of the earnings news. The ability of speed traders to submit 4:00-pm-news orders to the auction through the continuous session earns them up to 1.5 percent profit and creates an unlevel playing field because most auction participants are not allowed to cancel their orders. When stock exchanges recommended that firms delay disclosures until after the market closes, those with higher institutional ownership were more likely to voluntarily do so. Our study has implications regarding the timing of information releases and the design of the closing process.

Practitioners have long criticized risk-factor disclosures in the 10-K as generic and boilerplate. In response, regulators emphasize the importance of being specific. By using a computing algorithm, this paper establishes a new measure (Specificity) to quantify the level of specificity of firms’ qualitative risk-factor disclosures. We first examine determinants of variations in Specificity, and document that firms with high proprietary costs provide less specific risk-factor disclosures. More importantly, we find that, controlling for numerous determinants, the market reaction to the 10-K filing is positively and significantly associated with Specificity. In addition, our results suggest that analysts are better able to assess fundamental risk when firms’ risk-factor disclosures are more specific. Together, these findings suggest that more specific risk-factor disclosures benefit users of financial statements.

This study examines how short-selling threats affect auditors. During 2005–2007, the SEC ordered a pilot program in which one-third of the Russell 3000 index firms were arbitrarily chosen to be exempted from short-sale price tests. As a result, these stocks faced significantly higher short-selling threats. We implement a difference-in-differences test with firm fixed effects to show that auditors react to the increased threats and charge higher audit fees to the pilot firms. Further, we find that the impact only exists when auditors are concerned with the bankruptcy risk or when managers are less likely disciplined by short sellers.

Using 2015–2019 intraday short sale data from CBOE, we show that shorting flows near the open, middle, and close all negatively predict future returns, but the shorting flows near the open and middle have stronger predictive power than shorting flows near the close. We relate our findings to three informed trading models with different predictions on the timing of the trades. The long term predictive power of shorting flows near the open and midday is consistent with Kyle’s (1985) model of steady trading; the intraday variation in shorting flows’ predictive power is more consistent with Holden and Subrahmanyam’s (1992) aggressive trading model, in the sense that predictive power of shorting flows is stronger when there is greater urgency to trade at open and when the securities lending market is more competitive; and the liquidity timing hypothesis from Collin-Dufresne and Fos (2016) is also supported by the finding that opening shorting flows increase for firms with better liquidity conditions.