Asset Pricing with Recursive Utility and Heterogenous Investors

Topic & Objectives

Fundamental research in asset pricing is concerned with the determinants of expected returns on risky assets as a function of fundamental dynamics and investor characteristics. In this project, the focus was on recursive utility which significantly generalizes the standard time-additive preferences mostly of the type featuring constant relative risk aversion (CRRA). Furthermore, the assumption of only one type of investors being present in the economy was also relaxed. The form of heterogeneity we were interested in related to investors’ perceptions of the dynamics of fundamental sources of risk like expected consumption growth.

As an important first step, we still looked at the CRRA casewith two investors who differ in their beliefs but both being potentially wrong in their respective assessments. The interesting question is then which types of errors (e.g., with respect to the uncertainty around the expected growth rate or with respect to the speed with which the growth rate will revert back to its long-term mean) lead to more significant wealth losses over the long term.

In yet ongoing work we build on the insights from this analysis and examine a case where two investors equipped with recursive utility have different views on the likelihood of sudden drops in expected consumption (optimistic versus pessimistic). This model can also shed light on the question whether the presence of just a few optimists would already be enough to make the equity premium disappear as it is claimed in parts of the literature. Furthermore, we can explicitly investigate the relevance of market incompleteness.

Key Findings

• When heterogenous investors commit different kinds of filtering errors, there may indeed be errors by one investor of a type which cannot be compensated by error on the part of the other investor. These are considered the worst types of errors.
• In a model with recursive utility the presence of optimists does not make the equity premium vanish. Market incompleteness becomes relevant as soon as the non-traded risk factors are actually important in the sense of representing major fundamental risks for the economy (and not just minor variations in consumption growth).
• Our results highlight the importance of the specification of preferences in equilibrium asset pricing models and thus represent an important contribution to fundamental research in finance.