Risks for the Long Run: Estimation with Time Aggregation

Abstract

The long-run risks (LRR) asset pricing model emphasizes the role of low-frequency movements in expected growth and economic uncertainty, along with investor preferences for early resolution of uncertainty, as an important economic-channel that determines asset prices. In this paper, we estimate the LRR model. To accomplish this we develop a method that allows us to estimate models with recursive preferences, latent state variables, and time-aggregated data. Time-aggregation makes the decision interval of the agent an important parameter to estimate. We find that time-aggregation can significantly affect parameter estimates and statistical inference. Imposing the pricing restrictions and explicitly accounting for time-aggregation, we show that the estimated LRR model can account for the joint dynamics of aggregate consumption, asset cash flows and prices, including the equity premia, risk-free rate and volatility puzzles.