In the first chapter, A Unified Theory of the Term Structure and the Beta Anomaly'', I propose a novel generalized framework which allows for disentangling agent's risk aversion, elasticity of intertemporal substitution, and the agent's preference for early or late resolution of uncertainty. I apply this framework to a consumption-based asset pricing model in which the representative agent's consumption process is subject to rare but large disasters. The calibrated model matches major asset pricing moments, while higher exposure to systematic risks may lead to lower risk premia. This is consistent with empirical finding, while existing consumption-based asset pricing models fail to deliver. The second chapter, A Model of Two Days: Discrete News and Asset Prices'', co-authored with Jessica A. Wachter, provides a quantitative model to address the macro-announcement premium. Empirical studies demonstrate striking patterns in stock returns related to scheduled macroeconomic announcements. A large proportion of the total equity premium is realized on days with macroeconomic announcements. The relation between market betas and expected returns is far stronger on announcement days as compared with non-announcement days. Finally, these results hold for fixed-income investments as well as for stocks. We present a model in which agents learn the probability of an adverse economic state on announcement days. We show that the model quantitatively accounts for the empirical findings. Evidence from options data provides support for the model's mechanism.