By Christopher Faille
Passive and active investments are often contrasted as if the distinction is self-evident. It isn’t. Even for an unambitious long-only equity indexed fund, trades have to be executed in order to maintain the desired balance, and these trades can be executed either well or poorly, in ways that help or hurt the investor. As Oriental philosophers say: there is work within passivity and rest within activity.
For certain asset classes, notably currencies, the active/passive distinction disappears entirely, because there is no accepted benchmark. That is one of the insights of a recent paper by Michael Melvin and Duncan Shand, both of the Global Market Strategies Group at BlackRock.
Their paper, “Active Currency Investing and Performance Benchmarks,” appeared in The Journal of Portfolio Management earlier this year. It begins with the observation that there is no “market portfolio” in the FX world. In other fields, notably equities, the Capital Asset Pricing Model (CAPM) suggests a market portfolio: a weighted sum of every stock in the S&P 500, for example. No close analogy exists in FX, simply because every trade involves buying one currency while selling another, so an investor “is naturally long and short across the chosen currencies” and must constantly rebalance to suit any given strategy as market conditions change.
Alpha Can Be Tricky to Measure
In lieu of any long-only or passive benchmark, how does one find alpha? Analysts have tried to create benchmarks based on the three most common strategies in the FX world: the carry trade, momentum, i.e. trend following, or value investing, where value is usually defined by purchasing power parity (PPP).
Efforts to benchmark each of these strategies have produced disparate results. Consider trend following strategies: Melvin and Shand graph three distinct indexes devoted to that strategy, one created by the Centre for International Banking Economics and Finance (known as the AFX Currency Index), the second created by Credit Suisse (CS) and the third by Deutsche Bank (BD). A look at the chart shows that they aren’t tracking the same homogenous fact.
DB ranks the G10 currencies, and then calculates a benchmark based on a long position in the top three currencies and a short in the bottom three.
AFX establishes a long position when the current spot rate exceeds a moving average value for a given number of days, with three different measuring periods, and establishes a short when the spot rate is below the moving average value under any one of the three rules. Then it averages the results of the three sets of positions established by the three moving averages to create its benchmark.
CS introduces an element of carry into its definition of the trend-following strategy, because the returns of the carry trade within a particular currency pair constitute part of the definition of a trend.
But Alpha Does Exist
Some scholars have suggested, in Melvin and Shand’s paraphrase, that “only returns beyond those explained by the generic factors of carry, momentum, and value should be considered as true alpha.” But since there is no “standard accepted definition” of any one of these factors, it is not obvious how they can be combined into a benchmark either.
Another way of trying to measure manager performance, while minimizing definitional disputes, is to examine drawdown performance at times of crisis. Melvin and Shand “calculate the periods of major loss associated with each factor and then check if managers have skill in moderating the loss….”
Carry, momentum, and PPP were each associated with considerable fund drawdowns in the period from Nov. 15, 2005 to May 18, 2009. The largest drawdowns were associated with the generic carry strategy, and the smallest with the momentum strategy. Only two managers had drawdowns during this period worse than generic Carry, although fifteen managers had drawdowns less than those of generic Momentum.
Melvin and Shand conclude that there is alpha in the currency field, despite the difficulties in measurement. “[M]any managers have ability in avoiding worst-case drawdowns that are associated with a mechanical implementation of the generic factors … and this is a skill that is rewarded in the market.”