Department of Statistical Science
Duke University
presents:
Jose Quintana
Bluford Putnam
Bankers Trust
"Global Asset Allocation: Stretching Returns by Shrinking Forecasts"
Abstract: The standard tool for constructing portfolios of financial assets is Markowitz mean-variance optimization and its driving force is the stream of predictive means and variances. The key to avoiding the typical extreme allocations that make money managers apprehensive, is to use a consistent integrated model for generating the predictive means and variances, as opposed to employing cut and paste forecasting procedures. Consistent relative multiple-factor currency exchange rate models that are invariant to currency base perspective have been developed recently (Putnam and Quintana, 1994). These models use a framework of analysis based on relative comparison between fundamental factors of the two countries that underlie each exchange rate. This formulation facilitates the use of shrinkage techniques to further enhance their performance. In this paper we consider natural extensions of these models to include global financial assets (stock and bond indices) in addition to currencies. The extension is based on the observation that the price of, say, a stock index is the exchange rate between a unit of the index and the local currency in which it is denominated. An assessment in terms of portfolio performance is presented for models with various degrees of refinement to verify the premise that the more sophisticated the model the better the investment achievement.
Key Words: Bayesian Forecasting, Bayesian Shrinkage, Global Asset Allocation, Multiple-Factor Models, Portfolio Management
Friday, April 19, 1996
11:45 am - 12:45 pm
130 Sociology/Psychology Building Any questions concerning the seminar may be addressed to Cheryl McGhee @[919] 684-8029, e-mail cheryl@stat.duke.edu, or finger seminar@stat.duke.edu