Research

Monetary Factor Momentum , with Adam Farago, Erik Hjalmarsson, Hongtao Qiu, coming soon

We document strong persistence in factor returns following Federal Open Market Committee (FOMC) announcements, as factors tend to sustain their performance afterward. This monetary factor momentum is both economically sizable and statistically significant, and it is distinct from previously documented time-series and cross-sectional factor momentum. The profitability stems from much stronger autocorrelation of factor returns in the FOMC cycle. Consistent with a slow diffusion of the information released at FOMC announcements, the strategy’s returns accumulate gradually over several weeks after the announcements. The effect is stronger after FOMC announcements that reveal negative central bank information. The monetary factor momentum effect is also concentrated to FOMC announcements with high monetary policy uncertainty, consistent with investors underreacting to new information under uncertainty.

Earnings Persistence Cliff and Post-Earnings-Announcement Drift , with Liyao Wang

We document a sharp persistence cliff in firm-level earnings surprises at the fiscal-year boundary. Earnings-surprise persistence is stable across quarter-to-quarter transitions within a fiscal year but declines markedly from FQ4 to the subsequent FQ1. We develop a regime-switching framework in which investors infer the persistence of earnings news from past surprises. Because the weak relation between FQ4 and FQ1 surprises leads investors to underestimate the persistence of FQ1 news, prices underreact to its implications for future earnings, generating stronger postearnings-announcement drift (PEAD). Consistent with this prediction, we find that the PEAD following FQ1 announcements is approximately 83% larger than following announcements in other fiscal quarters. The effect is robust across return horizons and sectors. Moreover, the FQ1 PEAD remains economically and statistically significant after the mid-2000s, even as the overall PEAD largely disappears. We also document similar persistence cliffs in other firm characteristics associated with investor underreaction. Our findings are not attributable to microcap firms, prior earnings-surprise streaks, calendar-year seasonality, or earnings management, nor are they subsumed by existing explanations for PEAD.

Presidential Cycles in PEAD, with Zhi Da and Liyao Wang

Post-earnings announcement drift (PEAD) displays presidential cycles: it earns 4.1% per year during Democratic presidencies but its profitability increases significantly to 14.9% during Republican presidencies. Survey-based evidence also indicates substantial underreaction to earnings news when the US president is Republican. The tax component of firm earnings exhibits significantly higher volatility during Republican periods, likely reflecting higher tax policy uncertainty. Consistently, we find that the PEAD is much stronger for firms with larger exposure to tax policy changes during Republican presidencies. This explanation accounts for the observed presidential cycles in PEAD, whereas existing explanations for PEAD cannot. The cycles are more pronounced among non-microcap firms.

Analysts Are Good at Ranking Stocks , with Adam Farago and Erik Hjalmarsson, R&R at  Journal of Finance

Sell-side analysts' forecasts of stock returns are biased and the consensus forecast is a poor cross-sectional predictor. In sharp contrast, the implicit ranking of stocks by each analyst is highly informative of subsequent returns. Long-short portfolios sorted on these rankings result in large and highly significant excess returns that cannot be explained by previous anomaly characteristics. The strong performance is most easily understood by noting the similarity between rankings and within-analyst demeaned forecasts. The latter are equivalent to removing each analyst's fixed effect and thus controlling for unobservable analyst-specific biases.

Inflation Risk, Ambiguity, and the Cross-Section of Stock Returns, with Guihai Zhao, submitted

Inflation premia among individual stocks are small on average but vary substantially over time. We identify a novel driver of this variation: the correlation between inflation and inflation ambiguity, which we term the nominal-ambiguity correlation (NAC). A higher NAC makes high-inflation-beta stocks better hedges against ambiguity, thereby lowering their required returns and the cross-sectional inflation premium. Empirically, the monthly inflation premium falls from 0.45% when NAC is negative to -1.57% when it is positive. The industry-level analysis provides similar evidence. An equilibrium model with ambiguity and worst-case beliefs rationalizes these findings, highlighting ambiguity as a key driver of inflation premia.

Factor MAX and Predictable Factor Returns, with Liyao Wang, R&R at  Review of Asset Pricing Studies

Factors with the highest maximum daily returns in the previous month outperform those with the lowest by 0.32% per month (t= 5.89). This factor MAX strategy cannot be explained by factor momentum. An asset pricing model with dispersed private information shows that limited investor attention and underreaction to private information are likely the sources of predictability. Consistently, factor MAX is stronger among low-attention factors, during low-attention periods, and among factors with greater investor disagreement. The factor MAX is related to the well-known stock MAX effect, as both share an underreaction component arising from dispersed information about common fundamentals.