Abstract Prior research has shown that information diffuses gradually across stocks that are economically linked at the industry level. I document a similar pattern when stock portfolios are formed based on characteristics that are used in the anomaly literature (e.g., size, value, asset growth). Specifically, characteristics are useful to identify economic links, and earnings surprises contain information about future returns of other firms that share similar characteristics (i.e., “similar-style” firms). Such style-based earnings surprises can be used to predict style returns in the time series. For the cross-section of stocks, I create a composite style-based earnings surprise measure (SESM), which generates an equal-weighted (value-weighted) long–short strategy return of 167 (101) basis points per month. I do not find that industry spillovers, the traditional post-earnings announcement drift, unconditional abnormal style returns or risk can explain the return predictability. My findings suggest a further channel of gradual information diffusion in security markets.

We analyze whether the diversification discount is driven by the book value bias of corporate debt. Book values of debt may be a more downward biased proxy of the market value of debt for diversified firms, relative to undiversified firms, as diversification leads to lower firm risk. Thus, measures of firm value based on book values of debt undervalue diversified firms relative to focused firms. Our paper complements recent literature which uses market values to test the risk reduction hypothesis for a subsample of firms for which debt is traded. Alternatively, we employ market value of debt estimates for the whole firm universe. Consistent with the above hypothesis, we show that the use of book values of debt underestimates the value of diversified firms. There is no discount for mainly equity financed firms and lower distress risk and equity volatility for diversified firms. More concentrated ownership increases firm valuation.

We leverage the capabilities of GPT-3 to generate historical business descriptions for over 63,000 global firms . Utilizing these descriptions and advanced embedding models from OpenAI, we construct time-varying business networks that represent business links across the globe. We showcase the performance of these networks by studying the lead–lag effect for global stocks and predicting target firms in M&A deals. We demonstrate how masking firm-specific details can mitigate look-ahead bias concerns that may arise from the use of embedding models with a recent knowledge cutoff, and how to differentiate between competitor, supplier, and customer links by fine-tuning an open-source language model .

Motivated by McLean and Pontiff (2016), we study the pre- and post-publication return predictability of 241 cross-sectional anomalies in 39 stock markets. We find, based on more than two million anomaly country-months, that the United States is the only country with a reliable post-publication decline in long-short returns. Collectively, our meta-analysis of return predictors suggests that barriers to arbitrage trading can create segmented markets and that anomalies tend to represent mispricing instead of data mining.

We evaluate the relation between 133 anomalies/factors and market states using a sample of 57 countries from 1980 to 2019. The vast majority (96 of 133; 50 significant at the 5% level) performs better in bad times; 9 anomalies perform significantly better in good times, including market, size, value, and momentum. The value-weighted four-factor alpha amounts to 47.7 (31.8) bps in bad (good) times. 92.9% of the performance gain in bad times is driven by the anomaly short side. Findings are robust to controlling for sentiment or recession indicators and highlight the importance of mispricing in explaining anomaly returns.

We examine the value of analysts’ recommendations using a dataset of 45 countries, 3.8 million firm-month observations, and 222 return anomalies from 1994 to 2019. Unlike U.S.-based evidence, recommendations lead to subsequent highly significant abnormal returns in international markets. Furthermore, analysts do not seem to strengthen mispricing in international markets, as they give more favorable recommendations to (anomaly-ranked) underpriced stocks, and inconsistencies between recommendations and composite anomaly ranks lead to lower, not higher, abnormal returns. Recommendations are more valuable in less developed and less individualistic markets. Our results suggest that analysts’ recommendations provide more value to investors than previously thought.

This study provides new insight into the recent debate on profitability and investment patterns in the cross-section of expected returns. Relying on implied risk premia of U.S. corporate bonds, we document a strong negative relation between exposure to the profitability factor and cost of debt. We do not observe a robust relation between exposure to the investment factor and cost of debt. Our findings are consistent with profitability being a risk factor, but suggest that high profitability implies lower (and not higher) risk. Because the market portfolio consists of all risky assets including corporate bonds, our findings challenge a risk-based explanation for the profitability and investment patterns in stock returns.

Relying on 2.2 million articles from forty-five national and local U.S. newspapers between 1989 and 2010, we find that firms particularly covered by the media exhibit, ceteris paribus, significantly stronger momentum. The effect depends on article tone, reverses in the long run, is more pronounced for stocks with high uncertainty, and is stronger in states with high investor individualism. Our findings suggest that media coverage can exacerbate investor biases, leading return predictability to be strongest for firms in the spotlight of public attention. These results collectively lend credibility to an overreaction-based explanation for the momentum effect.

We evaluate the relative and absolute performance of competing factor-based asset pricing models in international regions and globally. Our holistic analysis controls for model transaction costs and incorporates both right-hand-side tests (based on maximum squared Sharpe ratios) and left-hand-side tests (individual return predictors, composite mispricing proxies). The overall view of the tests shows that recently proposed models tend to perform better than classical models, but otherwise perform comparably. This finding, the performance of the models in some of the LHS tests as well as further results collectively suggest the need for new powerful asset pricing models for global equity markets.