The contribution of asset allocation policy to portfolio performance

Swiss Society for Financial Market Research (pp. 219233)

WOLFGANG DROBETZ AND FRIEDERIKE KHLER

THE CONTRIBUTIONOF ASSET ALLOCATION POLICY TO PORTFOLIO PERFORMANCE

Wolfgang Drobetz is an assistant professor of financeat the University of Basel. Friederike Khler is a masters studentat the University of St. Gallen (Switzerland).

Corresponding author: Wolfgang Drobetz, Department of Finance, University of Basel, Holbeinstrasse 12, CH - 4051 Basel, Switzerland, phone: [phone], mail: [email protected].

We thank Heinz Zimmermann for valuable comments.

1. Introduction

In their seminal studies, BRINSON, HOOD, and BEEBOWER (1986) and BRINSON, SINGER, and BEEBOWER (1991) document the overwhelming contribution of asset allocation to the return performance of a sample of pension funds. They disentangle total plan returns into three components: (i) asset allocation policy, (ii) market timing, and (iii) security selection (stock picking). Asset allocation is usually defined as involving the establishment of normal or passive asset class weights. In contrast, market timing is the process of managing asset class weights relative to the normal weights over short periods of time. Security selection refers to the decision of how an asset class portfolio should be invested in each of the securities making up an asset class.[1] The aim of deviating actively from the passive weighting scheme is to enhance the managed portfolios risk-return tradeoff. Using time-series regres-

sions of fund returns on benchmark returns, BRINSON, SINGER, and BEEBOWER (1991) show that asset allocation explains, on average, 91.5% per cent of the variation in quarterly total fund returns. In other words, they conclude that total fund returns are largely unrelated to the level of active management.

In two recent studies, SURZ, STEVENS, and WIMER (1999) and IBBOTSON and KAPLAN (2000) argue that this result has led to a number of misinterpretations among investment professionals. The BRINSON et al. studies have been applied to questions that they never intended to answer. In fact, their approach measures the importance of asset allocation to explain the variability of returns over time. However, an equally important aspect might be to assess the importance of asset allocation to explain the variation of performance among funds. As strongly emphasized by IBBOTSON and KAPLAN, the BRINSON et al. studies did not address this question. SURZ, STEVENS, and WIMER (1999) argue that neither a time-series nor...