Mathematical Finance Seminar

September 21, 2000 , 5:30 PM to 7:00 PM

Giovanni Barone-Adesi, Universita della Svizzera Italiana

On the Informational Content of Changing Risk for Dynamic Asset Allocation

An investor with quadratic utility invests amounts changing with his perceptions of risk and expected return in a market with changing risk. Optimal investment policies are derived under several hypotheses for expected returns. These policies are combined in a Bayesian framework to yield an optimal policy. Historically this policy performs better than the 'buy and hold' policy in our mean-variance tests, except in the case of the FTSE index. Simulation evidence shows that performance gains are modest in mean-variance, but they are substantial in the mean-VaR space. The positive skewness of dynamic policy returns accounts for that, leading often to stochastic dominance of buy and hold policies.