Learning and Portfolio Decisions for CRRA Investors

We maximize the expected utility from terminal wealth for a Constant Relative Risk Aversion (CRRA) investor when the market price of risk is an unobservable random variable and explore the effects of learning by comparing the optimal portfolio under partial observation with the corresponding myopic policy. In particular, we show that, for a market price of risk constant in sign, the ratio between the portfolio under partial observation and its myopic counterpart increases with respect to risk tolerance. As a consequence, the absolute value of the partial observation case is larger (smaller) than the myopic one if the investor is more (less) risk tolerant than the logarithmic investor. Moreover, our explicit computations enable to study in detail the so called hedging demand induced by parameter uncertainty.