A two-step indirect inference approach to estimate the long-run risk asset pricing model

Abstract

The long-run consumption risk model provides a theoretically appealing explanation for prominent asset pricing puzzles, but its intricate structure presents a challenge for econometric analysis. This paper proposes a two-step indirect inference approach that disentangles the estimation of the model's macroeconomic dynamics and the investor's preference parameters. A Monte Carlo study explores the feasibility and efficiency of the estimation strategy. We apply the method to recent U.S. data and provide a critical re-assessment of the long-run risk model's ability to reconcile the real economy and financial markets. This two-step indirect inference approach is potentially useful for the econometric analysis of other prominent consumption-based asset pricing models that are equally difficult to estimate.

Empirical Asset Pricing with Multi-Period Disaster Risk: A Simulation-Based Approach

“…There are variants of these methodologies with similarities to the present study. For example, Grammig and Küchlin (2018) propose a two-step II approach to estimate the parameters of economic growth processes in a first stage, then use the results to estimate the parameters of the EZW pricing kernel in a second stage. Their simulation of data based on preliminary parameter estimates (and not only calibrated values) is analogous to our approach, but it is also quite uncommon.…”

“…For example,Grammig and Küchlin (2018) adopt a strategy outlined byNewey and McFadden (1994) to provide the asymptotic distribution theory for a simulation-based sequential estimation ofBansal and Yaron's (2004) long-run risk (LRR) model. The efficient use of samples from different periods for GMM estimation is addressed bySingleton (2006, Ch.…”