Abstract
This paper extends the long-run risks (LRR) models coined by Bansal and Yaron (2004) and Bansal, Kiku, and Yaron (2012) with Epstein and Zin’s (1989) recursive preferences into the LRR models with Chen and Epstein’s (2002) recursive multiple-priors utility, which captures the investor’s ambiguity on uncertainties. Our empirical analysis shows that considering ambiguity on uncertainties, in particular, large ambiguity on the long-run risks, in the LRR models can significantly improve the fitting performance of the equity premium and the risk-free rate when using a low relative risk aversion coefficient.