My research focuses on how financial markets process information, why predictable return patterns arise, and how investor incentives, learning, and systematic risk affect asset prices.
Summary. We revisit Pástor and Veronesi's rational explanation for stock-price bubbles during technological revolutions. The bubble depends on two assumptions: a single adoption decision at a prespecified date and simultaneous adoption by all firms. Relaxing either assumption removes the bubble. Strong coordination and preemption incentives can restore it, but only through a socially inefficient, value-destroying rush to adopt.
Recent studies show that using firm-level earnings to systematically learn about a common factor induces a conditional risk premium on announcement days. We find that only the first weeks of quarterly earnings seasons earn the well-documented announcement return premium, as these are the only weeks having a measurable signal about aggregate earnings.
As a delegated information monitor, The Wall Street Journal (WSJ) makes coverage decisions to provide value to its subscribers. We use those choices to reveal how the marginal value of earnings information changes over time, which tracks time-series variation in the informativeness of small-stock prices, because small stocks live at the margins of WSJ coverage.
We document a long-lasting continuation in weekly stock returns that arises after the initial well-known short-horizon reversal. The subsequent momentum is strong enough to offset the initial reversal over the full year after portfolio formation.
We identify two types of momentum — relative-return momentum and momentum in firm-specific abnormal returns — and show that their long-run behaviors differ sharply. We relate both patterns to agency incentives in the money-management industry.
We investigate whether a real-time investor could have used book-to-market equity, firm size, and lagged returns to generate portfolio profits. Recursive out-of-sample tests reveal a marked distinction between ex post and ex ante predictability.
We show that momentum profits depend strongly on the prior state of the market. Momentum following up markets is large and later reverses, while momentum following down markets is weak, consistent with overreaction-based theories.