| In this dissertation we propose a theoretical model for conditional betas. Within a rational expectation equilibrium model, we provide a precise characterization of the dynamics of betas and the price of beta risk in terms of the model's primitive parameters and state variables. The expressions reveal that during periods of higher uncertainty, the investor requires a higher market premium. Likewise, the conditional betas also respond to levels of uncertainty; depending on the cash-flow properties of the asset, the asset's beta can increase or decrease on higher uncertainty. Because of the connection with uncertainty, conditional betas derive the stochastic properties from investor beliefs. One of such properties is the asymmetric response to positive and negative news.;We also provide empirical evidence of the model's predictions about the dynamics of betas. For this empirical investigation, we propose an econometric specification that provides time-varying estimates of betas and relates them, non-linearly, to investor beliefs. As a by-product, we suggest proxies for investor beliefs and uncertainty that can be extracted from stock returns. The dynamics implied by the estimated parameters confirms the model's prediction that value and growth betas have opposing sensitivity to the levels of uncertainty. |
