This paper explores strategies for estimating the prices of risks in Merton’s intertemporal capital asset pricing model (ICAPM). Our preferrred constrained ordinary least squares (OLS) method is tractable, scales to allow multiple test assets, and produces unbiased estimates and standard errors in our sample sizes.
When including additional test assets to the conditional CAPM, the authors find a positive and significant coefficient of relative risk aversion, which controls the intertemporal risk-return tradeoff. The estimates are robust across different test assets, and they remain significant when we include additional state variables.
Whereas the previous literature has mostly found insignificant estimates of the risk-return tradeoff when using only the market return, the authors conclude that including test assets in addition to the market return consistently produces positive and significant estimates.