The Noise Is The Signal

[exp]

Tom, you have an interesting system. I am wondering if you have done any backtesting of your system to determine its expected performance during bull and bear markets and also its maximum drawdowns as a percentage of the peak account values. Here are my thoughts on your system.

General points:

(1) It's a trend following system using relatively infrequent switches from long to cash/short based on the Intermediate Composite Model NYSE/NASDAQ Index (ICM Index) reading. It is designed to capture longer term uptrends and avoid longer term downtrends.

(2) The system uses a reasonable level of leverage (1.5 x Profunds sector funds) and a moderate degree of diversification (typically, various correlated tech sectors with greatest momentum)

(3) The system is fully mechanical since daily switches are completely determined by the ICM Index-based long/cash/short signals and by the Sector selections using the 100 day Relative Strength Composite values (RSC values).

Specific points:

(1) The first key aspect is the quality of the long/cash/short signals. Some systems are anticipatory ("early" entries) and some systems are reactive ("late" entries). It will be interesting to see into which category your system falls based on a reasonable track record.

(2) The second key aspect is the quality of the position sizing algorithm. Your system appears to employ automatic gradual position increases in 25% increments ("pyramiding") which should be a sound approach. However, this also depends on the quality of the long/cash/short signals and the resulting average gain from the time the ICM Index moves above 0 until the time the ICM Index moves below 0.

(3) The Sector selection using RSC values appears somewhat less important since highest momentum sectors will be usually various tech sectors which are both highly correlated to each other and to NDX. Yet, it should provide some additional gains due to relative outperformance of highest momentum sectors.

(4) The system assumes that the 100% aggregate long position size will be maintained when the ICM Index is above .20. There is some question whether this is the optimal strategy given Kelly's formula for fixed-fraction betting and given the related probabilities of gambler's ruin outcome. Moreover, the 1.5x leverage translates the 100% aggregate position into a 150% aggregate position in terms of the total account size.

George