Currency Overlay in Performance Evaluation

The aim of performance evaluation is to make judgments about the success of portfolio managers in asset allocation and security selection for investment portfolios. This article presents a method of performance measurement that uses the interdependence of market management and currency management in terms of allocation processes but allows their separation in terms of selection processes. As a consequence, performance evaluation is more reasonable than in previous approaches and produces the same results whether the investment process starts with market management followed by currency management or vice versa. Currency overlay is an investment management hedging strategy designed to separate, as far as possible, the currency management component from the management of the nondomestic assets in an international portfolio. In conventional performance evaluation for global portfolios that use local-currency and exchange rate returns, the interrelationship of market management and currency management is ignored, which leads to inaccurate results when performance attribution is carried out. I describe a method of performance evaluation that uses the interdependency of these management stages in terms of the allocation processes (to markets and to currencies) but allows their separation in terms of the selection processes (security selection and currency selection). As a consequence, performance evaluation of a portfolio is the same whether the investment process starts with market management followed by currency management or vice versa.The aim of performance evaluation is to make judgments about the success of managers in allocating investment funds to assets and in selecting investment securities and strategies. In the past 20–30 years, many approaches have been developed for directly comparing actively managed portfolios with passively managed, benchmark portfolios. During the late 1980s, the main goal was to find and define attribution variables (allocation and selection variables). Beginning in the 1990s, the theoretical focus shifted to global investment portfolios and adding currency management to the decision variables. By focusing on local-currency and exchange rate returns, however, and ignoring interest rate differentials, these conventional performance evaluation systems continued the separation of market and currency management.This article improves upon earlier methodology in two ways. First, within performance measurement, I provide a mathematically correct method of evaluating return variables. Second, I derive appropriate formulas for the attribution variables for the return premiums developed in that method. In this way, a new portfolio evaluation system is provided that consists of an established method of portfolio decomposition, a new method of performance measurement, and also a new method of attribution analysis. All the definitions, mathematics, and interpretations are clarified with numerical examples.The first part of the article deals with performance measurement. The aim is to find mathematically correct formulas for return and weight variables that are compatible with practical demands. Thus, all returns are initially calculated in simple, instead of continuously compounded, terms. Overall performance is defined by multiplying market and currency returns rather than adding them. Furthermore, for arbitrage reasons, forward currency returns are expressed as the difference between two Eurodeposit returns divided by a discount factor. This kind of performance measurement allows a breakdown of overall performance that suggests definitions of return variables as return “premiums”—that is, excess returns. By focusing on market and currency premiums, this article provides portfolio managers a more reasonable interpretation of performance attribution variables in global investment portfolios. Moreover, it reveals that these two management tasks are interdependent in terms of allocation but separable in terms of selection processes.The second part of the article considers attribution analysis. Here, the aim is to identify the individual contribution of each market and currency decision to the overall excess return of the portfolio (excess return over a benchmark portfolio). For this purpose, a passive benchmark portfolio is created and the success of active management in allocating and selecting is measured against it. Once return and currency premiums have been calculated, the derivation of attribution formulas is straightforward. Each attribution measure is composed of the return difference of the created active and passive portfolios. The returns differ in the weight or return only of the variable for the allocation or selection decision to be evaluated. Even though the derivation of attribution variables is thereby still influenced by the order of market and currency management within the investment process, interpretation of the variables leads to the same performance evaluation judgments regardless of when the management phases were carried out.