We investigate, in a theoretical framework, the joint role played by investors' attention to news and learning uncertainty in determining asset prices. The model provides two main predictions. First, stock return variance and risk premia increase with both attention and uncertainty. Second, this increasing relationship is quadratic. We empirically test these two predictions, and we show that the data lend support to the increasing relationship. The evidence for a quadratic relationship is mixed. Overall, our study shows theoretically and empirically that both attention and uncertainty are key determinants of asset prices.

This paper highlights a positive and significant beta-return relationship in high expected market return states, as suggested by the ICAPM. The ICAPM has strong out-of-sample predictive power for equity returns. As a result, timing strategies exploiting this predictive power have Sharpe ratios about double those of the buy-and-hold strategies, alphas of about 5% per annum, and average returns increasing sharply with unconditional betas. Our findings relate to the positive beta-return relation uncovered overnight, on macroeconomic announcement days, and in low inflation times because these periods share an important common feature: high market returns.

We build an equilibrium model to explain why stock return predictability concentrates in bad times. The key feature is that investors use different forecasting models, and hence assess uncertainty differently. As economic conditions deteriorate, uncertainty rises and investors' opinions polarize. Disagreement thus spikes in bad times, causing returns to react to past news. This phenomenon creates a positive relation between disagreement and future returns. It also generates time‐series momentum, which strengthens in bad times, increases with disagreement, and crashes after sharp market rebounds. We provide empirical support for these new predictions.

We model how investor learning about monetary-policy transmission impacts asset prices. In an asset-pricing model, investors learn from realized inflation surprises how effectively monetary policy steers future inflation. Downward revisions in perceived effectiveness raise expected inflation persistence, increasing return volatility and risk premia. These effects intensify when policy deviates significantly from neutral or monetary-transmission uncertainty is high. We estimate the model using U.S. macro and policy data from 1954 to 2023. The resulting dynamics align with observed patterns in equity returns and volatility. Empirical tests support the model’s core prediction: investor learning turns central-bank credibility into a priced risk factor.

Recent empirical findings document downward-sloping term structures of equity return volatility and risk premia. An equilibrium model with rare disasters followed by recoveries helps reconcile theory with empirical observations. Indeed, recoveries outweigh the upward-sloping effect of time-varying disaster intensity and expected growth, generating downward-sloping term structures of dividend growth risk, equity return volatility, and equity risk premia. In addition, the term structure of interest rates is upward-sloping when accounting for recoveries and downward-sloping otherwise. The model quantitatively reconciles high risk premia and a low risk-free rate with the shape of the term structures, which are at odds in other models.

We investigate, in a theoretical framework, the joint role played by investors' attention to news and learning uncertainty in determining asset prices. The model provides two main predictions. First, stock return variance and risk premia increase with both attention and uncertainty. Second, this increasing relationship is quadratic. We empirically test these two predictions, and we show that the data lend support to the increasing relationship. The evidence for a quadratic relationship is mixed. Overall, our study shows theoretically and empirically that both attention and uncertainty are key determinants of asset prices.

The term structure of equity risk has been shown to be downward sloping. We capture this feature using return dynamics driven by both a transitory and a permanent component. We study the asset allocation and portfolio performance when transitory and permanent components cannot be observed and therefore need to be estimated. Strategies that account for the observed timing of equity risk outperform those that do not, particularly so out of sample. Indeed, the mean (median) certainty equivalent return increases from about 13% (12%) to about 21% (15%) because properly modeling the timing of equity risk implies surges in portfolio returns.

Persistence risk is an endogenous source of risk that arises when a rational agent learns about the length of business cycles. Persistence risk is positive during recessions and negative during expansions. This asymmetry, which solely results from learning about persistence, causes expected returns, return volatility, and the price of risk to rise during recessions. Persistence risk predicts future excess returns, particularly at 3- to 7-year horizons. Its predictability is strongest around business-cycle peaks and troughs. We confirm the model’s predictions in the data and provide evidence that persistence risk is priced in financial markets.Received October 13, 2017; editorial decision September 19, 2018 by Editor Stijn Van Nieuwerburgh.