Essays in empirical asset pricing

Cesare Robotti, Boston College

Abstract

My dissertation consists of three essays in empirical asset pricing. My first essay, titled "Dynamic Strategies, Asset Pricing Models, and the Out-of-Sample Performance of the Tangency Portfolio", analyzes the behavior of an investor with unit risk aversion who maximizes a utility function defined over the mean and the variance of a portfolio's return. The main findings of the paper are summarized as follows: (i) The estimates of the tangency portfolio weights are extremely volatile and imprecise. Using an asset pricing model to constrain mean asset returns eliminates extreme short positions in the underlying securities, and improves the precision of the estimates of the weights. (ii) Fully restricting mean asset returns according to single-index and multi-index asset pricing models worsens the in-sample and out-of-sample performance of the tangency portfolio. (iii) Passive investment strategies, i.e., strategies that disregard the role played by conditioning information in investment decisions, dominate dynamic investment strategies that incorporate relevant conditioning information when individual stocks are considered. On the contrary, the use of dynamic investment strategies is effective in improving the out-of-sample performance of the tangency portfolio when aggregate assets are used. My second essay, titled "The Price of Inflation and Foreign Exchange Risk in International Equity Markets", shows how the premia on inflation and foreign exchange risks are given by minus the cross moments between an admissible nominal pricing kernel scaled by the risk-free rate, and the innovations in inflation and exchange rates. Of the many admissible kernels, I choose the one with the minimum variance. Only global market risk is significant both conditionally and unconditionally as indicated by the high positive and significant relative Sharpe ratio. In addition, I test several international asset pricing models using a variety of diagnostics. All models are rejected by the data even if their pricing implications are significantly different. My third essay, titled "Minimum-Variance Kernels and Economic Risk Premia", analyzes risk premia associated with economic risk variables, i.e., risk variables which are not traded, based on the minimum-variance kernel of Hansen and Jagannathan (1991). This approach is appealing because it allows the estimation of the risk premia without the need to make assumptions on a pricing model. We find that several economic variables affecting the risk-return trade-off command significant risk premia, and that the signs of the premia are mostly consistent with the effects on the distribution of returns.

Recommended Citation

Cesare Robotti, "Essays in empirical asset pricing" (January 1, 2001). Boston College Dissertations and Theses. Paper AAI3021591.
http://escholarship.bc.edu/dissertations/AAI3021591