Performance Of Leveraged Asset Funds

Abstract

Leverage plays an important role in the practice of today's portfolio management industry. Leverage can be part of the funds' investment policy, long term strategy, or can be employed to meet temporary liquidity demand. Although the non-normality created by leverage is much the same than those created by options, its effects on the fund's performance are still not widely discussed. The present paper investigates the ranking characteristics of conventional and non-conventional performance measures, when evaluating leveraged portfolios. Because most asset funds contain some level of leverage, we focus on 1) whether the overall used performance measures as well as particular alternative measures account correctly for the (additional) non-normality, created by the capital structure decision, and 2) whether using leverage leads to superior performance. Within an experimental analysis, based on simulated and real data return distributions, a number of alternative methods are employed to analyze the ranking characteristics of different performance measures on asset funds with and without leverage. Using expected utility as benchmark, we compare conventional composite measures like the Sharpe Ratio, the Sortino Ratio, etc., as well as performance measures based on the use of replicating portfolios as in Berenyi (2001): the Excess Return, Modified Sharpe Ratio and the Efficiency Ratio. Rank correlations between the rankings with expected utility (with varying risk parameter) and those of performance measures are calculated. Our results show that, with some reservations, the non-traditional performance measures can handle the non-normality resulting from leverage better. Following this, also the consistency of those performance measures is investigated. In addition, we investigated whether there exist a difference in results between the full-portfolio-maximizing and equity-value-maximizing strategies.