[We study the effect of the Global Analyst Research Settlement and related regulations on sell-side research. These regulations attempted to mitigate the interdependence between research and investment banking. We document that following the regulations many brokerage houses have migrated from the traditional five-tier rating system to a three-tier system. Optimistic recommendations have become less frequent and more informative, whereas neutral and pessimistic recommendations have become more frequent and less informative. Importantly, the overall informativeness of recommendations has declined. The likelihood of issuing optimistic recommendations no longer depends on affiliation with the covered firm, although affiliated analysts are still reluctant to issue pessimistic recommendations.]

We study the dynamics of information demand and supply in capital markets, focusing on how firms’ disclosures align with analysts’ information needs. Using a novel dataset from Visible Alpha, we analyze granular data from analysts’ forecasting models to understand the breadth of information they seek and how firms meet these demands through mandatory and voluntary disclosures. We document significant variation in the complexity of analysts’ models and the extent of firms’ disclosures, leading to some items in analysts’ models remaining undisclosed. This unmet information demand gives rise to a novel concept we term “uncertainty about the past” ( UP ). We investigate its implications for key capital market outcomes, including analyst forecast dispersion, market reactions to earnings announcements, and stock market liquidity. Our results demonstrate that UP plays a significant role in shaping the information environment, challenging the assumption that earnings announcements fully resolve uncertainty about past performance.

Predictability of future returns using ex ante information (e.g., analyst forecasts) violates market efficiency. We show that predictability can be due to non-random data deletion, especially in skewed distributions of long-horizon security returns. Passive deletion arises because some firms do not survive the post-event long horizon. Active deletion arises when extreme observations are truncated by the researcher. Simulations demonstrate that data deletion induces a negative relation between future returns and ex ante information variables. Analysis of actual data suggests a 30–50% bias in the estimated relations. We recommend specific robustness checks when testing return predictability using ex ante information.

Financial analysts closely follow a firm’s operations and assess the risks that it faces. In this paper, we examine whether analysts incorporate ESG risks into their stock recommendations and target prices. Specifically, we use a unique firm-day level dataset on negative ESG risk incidents to proxy for unobservable risk assessments of analysts. We find that analyst outputs predict future ESG incidents, suggesting that analysts incorporate ESG risks into their models. Our results are robust to controlling for ESG incidents that firms experienced in the past, and are stronger in more transparent information environments, and in the presence of more guidance on ESG issues from the Sustainability Accounting Standards Board. Importantly, we find that analysts incorporate ESG risks through adjusting discount rates rather than cash flow estimates. Overall, our results highlight the ability of financial analysts to synthesize and integrate ESG risks into their research.

Using a large sample of debt contracts, we study the determinants of excluding nonrecurring items from covenant calculations. We investigate this choice across firms, across items, and through time. We find that nonrecurring items are more likely to be excluded when the agency costs of debt are higher and less likely to be excluded when they predict borrowers' performance. Our evidence further suggests that the interplay between agency costs and nonrecurring items' predictive ability affects the decision to exclude these items from covenant computations. Finally, when examining the exclusion by different nonrecurring item types, we find confirmatory evidence that the probability of exclusion decreases with the predictive ability for borrowers' future performance of major nonrecurring item types. Overall, our research extends the literature on the determinants of contract design and improves understanding of the usefulness of accounting information in debt contracting.

Industry expertise is an important aspect of sell-side research. We explore this aspect using a novel dataset of industry recommendations, which are often issued by strategy analysts. We study sell-side analysts' ability to rank industries relative to each other (across-industry expertise), and how it relates to analysts' ability to rank firms in a particular industry (within-industry expertise). We find that analysts express more optimism towards industries with higher levels of investment, past profitability, and past returns. Analysts exhibit across-industry expertise, as portfolios based on industry recommendations generate abnormal returns over both short and long horizons, beyond what would be explained by industry momentum. Additionally, industry recommendations contain information, which is orthogonal to the information revealed in firm recommendations, and more so for brokers who benchmark their firm recommendations to industry peers. Consequently, the investment value of sell-side analysts' recommendations is enhanced when both dimensions of industry expertise are utilized by considering industry and firm recommendations in combination.

To test hypotheses about earnings management, many studies investigate managers' manipulation of real activities (real earnings management, REM). Tests using measures of abnormal REM hinge critically on the measurement of normal real activities. Yet, there is no systematic evidence on the statistical properties of commonly used REM measures. We provide such evidence by documenting the Type I error rates and power of the test of the REM measures commonly used in the literature. We find these measures are often misspecified with Type I error rates that deviate from the nominal significance level of the test, especially in samples of firms with extreme performance or firm characteristics. We also compare the specification and power of traditional REM measures with performance‐matched REM measures to see if the latter provide better specified and more powerful tests. While performance‐matched REM measures are not immune from misspecification in all settings, in general they are better specified under the null hypothesis (i.e., in terms of Type I errors) than are traditional REM measures. Comparisons of the power to detect abnormal REM reveal that neither approach, traditional or performance‐matched, is consistently more powerful than the other in terms of detecting abnormal REM ranging from 1 to 10 percent of (lagged) total assets. The absence of a dominant approach to measure abnormal REM leads us to recommend that future researchers report results using both traditional and performance‐matched measures, so that readers are able to clearly assess the reliability of the inferences drawn about the magnitude and significance of the abnormal REM documented in a given study.