DTI Employs A Rules-Based Methodology
To determine whether a sector should be long or short, the DTI compares a sector’s close at month-end to a seven-month weighted moving average for the sector. If the sector closes above its average at month-end, it will be held long for the forthcoming month. If the sector should close at month-end below its seven-month weighted moving average, it will be short in the upcoming month (except for energy, which would be flat). Sectors are rebalanced back to their base weights (as highlighted above) on a monthly basis. The example above assumes the energy sector is long; if the energy sector is flat, the weightings would be different than those shown in Figure 1. The underlying components of multicomponent sectors are only rebalanced annually. Using a weighted moving average method places a greater weight and importance on prices that have occurred more recently.
One of the big dilemmas that creators of trend-following strategies face is determining the appropriate length of time to establish a trend. Short-term trend-following strategies aim to generate favorable returns at the risk of eroding profits by excessively trading and repeatedly investing in false trends or losing strategies. Longer-term trend-following strategies seek favorable returns by identifying and correctly trading long-term trends in the futures markets at the risk of overlooking profits from short-term volatility.
In short, the determination of the length of the moving average line involved trade-offs, but historical simulations of the strategy’s success were not dependent on just this one specific seven-month average. Other moving-average periods could also be used to form the basis of the buy/sell decision without compromising the nature of the strategy.
Characteristics Of DTI
The DTI has been calculated live in real time since 2004. Over that period, it has been negatively correlated to both U.S. equities and U.S. bonds: -0.21 and -0.17, respectively (through Dec. 31, 2011).11 The negative correlation is a key element in the diversification that managed futures strategies can bring.
There are few asset classes that can provide meaningful negative correlation. Commodities was an asset class that many assumed would provide noncorrelation, and for a large amount of time, that was true. During the financial crisis, however, one can see that the correlation of a commodity index such as the S&P GSCI Index spiked higher and started to approach 0.8 from a negative correlation in 2005.11
The DTI rolling three-year correlation also shows mostly negative correlations on a three-year basis. However, in the middle of 2011, the rolling three-year correlation was impacted by the spike in overall commodity correlations, as commodities do represent 50 percent of the DTI.11 (See Figure 2.)