My research examines how financial information is disclosed, processed, and used in capital markets. I study three related questions: how the design and timing of corporate disclosures shape the information environments of investors and competitors; how individual investors become aware of, acquire, and act on financial information — and the role of platform design in that process; and how prediction markets can be used to measure market expectations about managerial disclosure and to study market efficiency in novel institutional settings.
When Are Concurrent Quarterly Reports Useful for Investors? Evidence from ASC 606
With Nikki Skinner and Andrew Stephan
Review of Accounting Studies, 29(2), 1360–1406, 2024
Prior research suggests that quarterly reports released concurrently with earnings depress trading due to information overload. In this study, we predict that concurrent reports help investors trade when they face uncertainty about how to interpret earnings news. We rely on the implementation of ASC 606 as a quasi-exogenous increase in uncertainty about how to trade in response to earnings news. Specifically, we find that when uncertainty is high in the first quarter of ASC 606 implementation, 10-Qs released concurrently with earnings are associated with increased trading around the earnings announcement. We find the relation is more pronounced when investors face greater uncertainty about earnings and when firms increase disclosure in their revenue recognition footnote. We also find some evidence that our results hold in a broader sample of accounting standard changes and generalize to other proxies for investor uncertainty. Our results suggest that concurrent quarterly reports are informative to investors when uncertainty about earnings is especially high.
With Joe López-Vilaró (University of Cincinatti) and Andrew Stephan (Indiana University East)
The Accounting Review Early-Stage Research Track · 2026 AAA Global Connect Meeting
Discussant: Greg Miller (University of Michigan)
We exploit prediction markets to measure narrative disclosure surprises during earnings calls and assess whether surprises affect equity market activity. These markets facilitate trading on whether specific words will be spoken during calls, allowing us to recover implied probabilities of prediction market outcomes and calculate a direct measure of expected disclosure. We compare expectations with realized disclosure to construct word-level disclosure surprises. We find that disclosure expectations are accurate but incomplete: words with higher implied probabilities are more likely to be spoken, yet meaningful deviations from expectations occur. When managers speak words with low ex ante probabilities, equity market trading volume declines and spreads widen during the call. However, we do not observe significant associations when highly expected words are not spoken. Introducing a novel approach to the study of qualitative disclosure, our results suggest that narrative disclosure is most informative when managers communicate unanticipated content relative to expectations.
Better Together: Information Processing When Firms File Their 10-K with Their Earnings Announcement
With Jared Flake (Boise State University) and Benjamin Whipple (University of Georgia)
Arif et al. (2019) document a growing practice in which firms file their 10-K concurrently with the earnings announcement (concurrent EA/10-Ks), and they read the accompanying muted market response as consistent with information overload. The practice has since spread to a majority of firms—a trend difficult to reconcile with a disclosure that persistently impairs investors' information environments. We revisit the market consequences and show that the muted response is also consistent with a different mechanism: receiving a more complete information set at once, investors interpret firm performance more similarly. Consistent with this convergence rather than overload, we find that investors engage more with the 10-K and less with the earnings announcement; that retail order flow and social-media opinion are more aligned; that analyst forecast dispersion is lower and the consensus forecast more accurate; and that prices adjust with less trading per unit of return, at narrower spreads and lower price impact. Prices also incorporate the period's information more quickly. The evidence indicates that concurrent EA/10-Ks improve information processing by delivering a more complete information set when investor attention is highest, narrowing rather than widening differences among investors and aligning with regulators' goal of reducing information disparities.
Data Visualization and Retail Investor Trading on Earnings
With Austin Moss (University of Colorado Boulder)
Using three complementary research analyses centered on the Robinhood trading platform, which displays earnings visually, we provide convergent evidence that visualization facilitates retail investors’ trading on earnings information. First, a framed field experiment with retail investors confirms that visualization improves their ability to correctly perceive the direction of the earnings surprise—that is, to form a usable signal from the earnings data. Second, we document that Robinhood investors’ trading around earnings announcements is more sensitive to visual earnings surprises than that of non-Robinhood retail and institutional investors. Third, a difference-in-differences analysis around Robinhood’s 2017 introduction of visual earnings reveals an increased trading reaction to visual earnings for high-Robinhood-ownership stocks. However, our analyses also show that the buy-sell decisions these investors make in response to earnings do not increase future returns. Net buying increases with the magnitude of negative visual earnings surprises (a signal that predicts lower future returns) but is not associated with the magnitude of positive surprises (a signal that predicts higher future returns). Taken together, visualization lowers the cost of forming a usable signal from earnings but does not change how investors act on it; the benefit of cheaper processing is therefore conditional on the quality of the resulting trades, and here it produces more trading without better trades.
Price as an Information Monitoring Signal
With Austin Moss (University of Colorado Boulder)
We investigate how investors react to price alerts and whether they learn about the existence of new firm information from prices. Leveraging institutional details regarding Robinhood’s price alert push notifications, we identify the moment a set of investors gain awareness of a stock’s price change and study these investors' subsequent trading decisions. Comparing Robinhood investors to non-Robinhood investors in a difference-in-differences event study design, we document an immediate and substantial increase in trading after price alerts. To examine whether investors learn about new information from price, we compare trading activities prompted by price alerts that are clearly tied to information events to those that are not. We find a higher trading response to price alerts clearly driven by new information. Since the alerts only contain price information, these results indicate that investors use price as a signal for when to search for firm information. The ability for investors to learn from prices is vital to many financial economic theories but there is limited empirical evidence evaluating whether and what investors learn. We document that investors learn about the existence of new public information from price.
Sophisticated Investors on a Retail Platform: Short Selling and Social Media Equity Analysis
With Wei Sun (Saginaw Valley State University) and James Anderson (Saginaw Valley State University)
Presented: AAA Southwestern Region Meeting (2024)
Do short sellers have an informational advantage in the era of financial social media? We investigate this question using intraday trading data around analyst articles on SeekingAlpha.com, a setting that allows us to observe how short sellers and retail investors differentially respond to user-generated equity analysis. We find that short selling increases in the 30 minutes following article publication, concentrated in bearish and uncertain articles and attenuated in bullish ones, patterns that are consistent with informed trading rather than positioning against retail order flow. Post-article short selling is also associated with subsequent price corrections. We then ask whether short sellers' processing advantages persist on a platform designed for retail audiences. Retail investors react more on average, but when articles impose higher awareness, acquisition, or integration costs, short sellers react more relative to retail investors. The informational edge traditionally attributed to short sellers extends to social media, and accessibility alone does not equalize processing ability.
Investor Information Demand Following Politician Trade Disclosures
With Gabriel Brull (University of Colorado Boulder) and Scott Robinson (University of Oregon)
Presented: AAA Global Connect (2026)
We examine whether disclosures of U.S. politicians’ stock trades prompt investors to demand firm-specific information. Using Google search activity as a revealed-preference measure of information demand, we find that investors significantly increase firm-level search following the public disclosure of a politician’s trade. This increase is concentrated in the days immediately after disclosure and varies systematically with the perceived informativeness of the trade. Investor information demand is stronger following purchase disclosures as compared to sale disclosures, but this asymmetry disappears during the early COVID-19 pandemic when politician sales were widely perceived to be privately informed. In addition, information demand is greater following trades disclosed by politicians who trade more frequently. Collectively, these findings suggest that politician trade disclosures increase investors’ incentives to gather firm-specific information.