Two very important research areas in macroeconomics are consumption theory and asset pricing models. These areas are linked by the optimal consumption profile for the the individual, which can be used for the determination of asset prices. However, when tractable specifications for the preferences of the so-called representative agent have been used the implications of this theoretic linkage for consumption and asset prices have been rejected empirically. Specifically, the assumed utility function does not even begin to replicate the behavior of consumers who have both a very strong desire to smooth out their consumption profile and a moderate degree of risk aversion -- a plausible description according to most of the available empirical evidence. The supposition of the research undertaken here is that the standard structure of preferences used in this estimation is not flexible enough so that the data are rejecting the specification but not the "theory". What is proposed, therefore, is to employ a new more general class of preferences based on the work of Kreps and Porteus, which will allow for the separation of attitudes towards risk from the intertemporal substitution effects. This separation of risk aversion from intertemporal substitution will shed new light on unexplained empirical regularities and advance our understanding of existing theories of consumption and asset pricing. The implications for consumption theory of Kreps-Porteus (KP) preferences will first be examined. The Hall (1978) random walk hypothesis will be reformulated and tested to confirm that KP preferences help account for the sensitivity of consumption to current income. The permanent income hypothesis when both labor income and interest rates are random will be studied by using a KP model of infinitely-lived risk neutral agents to determine the informational content of changes in income. The effects on consumption volatility of risk persistence will be characterized and calibrated with empirically estimated risk aversion and intertemporal substitution parameters. The determinants of asset prices under KP preferences will then be analyzed. The fact that KP preferences theoretically predict the better empirical performance of the market portfolio based on the capital asset-pricing model (CAPM) will be assessed empirically. It will be shown that separating risk aversion from intertemporal substitution helps explain part of the equity premium "puzzle". Finally, the effects of risk aversion and intertemporal substitution on the volatility of stock prices when dividend risk is time-varying will be characterized.
