Application of Portfolio Theory in Decision Tree Analysis

¹ University of Pennsylvania, New Bolton Center, Center of Animal Health and Productivity, Section of Epidemiology and Health Economics, Kennett Square 19348
² Cornell University, Department of Animal Science, Ithaca, NY
³ North Carolina State University, School of Veterinary Medicine, Raleigh 27606

A general application of portfolio analysis for herd decision tree analysis is described. In the herd environment, this methodology offers a means of employing population-based decision strategies that can help the producer control economic variation in expected return from a given set of decision options. An economic decision tree model regarding the use of prostaglandin in dairy cows with undetected estrus was used to determine the expected return of the decisions to use prostaglandin and breed on a timed basis, use prostaglandin and then breed on sign of estrus, or breed on signs of estrus. The risk attributes of these decision alternatives were calculated from the decision tree, and portfolio theory was used to find the efficient decision combinations portfolios with the highest return for a given variance). The resulting combinations of decisions could be used to control return variation.

Key Words: portfolio theory • decision tree analysis • prostaglandin

Submitted on July 9, 1990
Accepted on December 13, 1990