Beliefs, Doubts and Learning: Valuing Economic Risk

This paper explores two perspectives on the rational expectations hypothesis. One perspective is that of economic agents in such a model, who form inferences about the future using probabilities implied by the model. The other is that of an econometrician who makes inferences about the probability model that economic agents are presumed to use. Typically it is assumed that economic agents know more than the econometrician, and econometric ambiguity is often withheld from the economic agents. To understand better both of these perspectives and the relation between them, I appeal to statistical decision theory to characterize when learning or discriminating among competing probability models is challenging. I also use choice theory under uncertainty to explore the ramifications of model uncertainty and learning in environments in which historical data may be insufficient to yield precise probability statements. I use both tools to reassess the macroeconomic underpinnings of asset pricing models. I illustrate how statistical ambiguity can alter the risk-return tradeoff familiar from asset pricing; and I show that when real time learning is included risk premia are larger when macroeconomic growth is lower than average.