Style indices

As stock traders are fond of noting, it is a market of stocks and not a stock market. (This author has arrived at a similar conclusion for long-only commodity indices.) A fair objection at this point is both the Barclay and Stark indices aggregate many currency trading styles into one index and include only those CTAs selfdescribed as currency traders.

However, most CTAs diversify across markets and few eschew the currencies. This raises the interesting question whether investors should expect — or indeed, demand — more from CTAs who claim a specialty. After all, isn’t it reasonable for a cardiologist to be better at open-heart surgery than a general practitioner?

All is not lost. In the world of modern finance, you can find an index and a performance benchmark for everything. Sometimes these indices are real, have long trading histories, and underlie actual trading instruments such as index futures and options or exchange-traded funds. The presence of these indices and trading instruments alters the behavior of participants in the market; consider how a stock being added to the S&P 500 will jump in price as the indexers are forced to buy it. (This phenomenon is the subject of an upcoming article in the July 2007 issue of Active Trader magazine.)

Sometimes, however, indices are hypothetical “backcasts” designed to describe something after the fact and perhaps get participants to think about trading styles and approaches in a different manner. Such is the case with a set of four “currency trading style” indices maintained by ABNAmro. The respective styles were described by James Binny of ABNAmro in a Winter 2005 article in The Journal of Alternative Investments titled “Currency Management Style Through the Ages.” Let’s use Binny’s definitions, without comment:

Value forecasting: This uses fundamental factors external to the currency market, such as relative inflation rates, bond yields, or productivity growth to define an equilibrium or “fair” value of a currency as defined by the Purchasing Power Parity model. The trading style is simple: Sell the overvalued currencies and buy the undervalued currencies.

Trend-following: The strategy employed here is a simple moving average crossover; when the five-day moving average exceeds the 40- day moving average, buy the currency and sell it when opposite. Simple filters, not described in the article, exit extreme over-bought or over-sold positions and block entry in sideways markets.

Carry: This strategy borrows (sells) low-yielding currencies and lends (buys) high-yielding currencies. The positions taken are proportional to the size of the interest-rate differential relative to the average interest rate. The various convergence trades popular in European currencies prior to the advent of the euro (such as buying the Italian lira and selling the Deutsche mark) are excluded from the backcast as they are no longer available.

Volatility: This strategy compares the implied volatility in each currency to its three year average. If at the end of a month volatility exceeds this average, an at-the-money straddle is sold. The positions are reevaluated monthly.

These styles were reconstructed from the end of 1986 and therefore match the Barclay Currency Traders Index’ lifespan. The returns are “pure alpha,” or excess return to the risk-free rate. It depicts the indexed returns for the valuation, trend-following, and volatility styles on its left axis and the carry style on its right axis.

Average annual returns are posted within the legend. (For comparison, the average annual return on an index of three-month T-bills over this period was 4.88 percent.)

Over a 21-year period, the excess return for three out of the four trading styles underperformed the average annual return of the most commonly accepted definition of risk free returns. The fourth style, the carry style now under scrutiny for those who worry about the yen carry or Swiss franc carry trades, outperformed Treasury bills, but only at the cost of a 68.6-percent retracement of gain between August 1992 and February 1993.

Most veterans in this business either have produced simulated trading records or have seen others’ simulated trading records, or both. Let’s just say a performance gap between real trading and simulated trading exists; you are free to decide in whose favor.
All of these trading records are before fees. If professional traders cannot outpace Treasury bills over a 21-year period, the U.S. government should start charging a 1-percent management fee and a 20-percent incentive fee for holding their paper, for example. It is the ultimate principal-protected investment, after all, and advocates on both sides of the political aisle can claim credit for reducing the federal deficit therewith.