Notes on the Role of Transaction Costs in Portfolio Analysis
- Friday, November 14 from 3:30 to 4:30 in Room 552; Refreshments 3:00 in Room 502
In traditional portfolio theory, following Markowitz, it is assumed that an investor can trade immediately and costlessly to obtain a portfolio that optimally expresses her views on future security returns (i.e. her forecasts) while controlling risk exposures. This assumption results in a particular approach to the sort of analysis routinely done on trading strategies: statistical evaluation of forecasts, attribution of portfolio returns to individual forecasts, examination of risk exposures, and sizing of positions. Over the last fifteen years, however, it has become incresingly evident that transaction costs must be considered in studying any trading strategy in which the forecasts are not essentially constant over very long horizons. The transaction costs introduce a dynamical component to the analysis, as the investor must balance her eagerness to adapt to the changing forecasts against the costs of trading rapidly. I will discuss how the traditional approaches to risk analysis, forecast evaluation and attribution, and capacity need to be modified in the light of these considerations.
Jerome Benveniste was a member of the Quantitative Trading Group at Highbridge Capital Management, LLC for twelve years, the last six as Managing Director and Portfolio Manager. He was involved in nearly every aspect of Highbridge's quantitative business, including forecast generation, risk modeling, transaction cost modeling, and optimization. Before joining Highbridge, he was a mathematician working in the areas of differential geometry, Lie theory, and ergodic theory and was on the faculty of Case Western Reserve and Stanford Universities. Jerome holds a Ph.D. from the University of Chicago and an A. B. from Harvard University, both in mathematics