Last week I wondered whether I should be worried about being with the consensus, as revealed in the Commitment of Traders report.

I have checked this question in quick-and-dirty way by running three strategies in fifteen futures markets over the past five years:

  1. A simple breakout strategy. This makes a return of 181 risk units.
  2. The same strategy with the added feature that it will not trade if the net non-commercial position is beyond 50% of its most-extended level (long or short) since the start of the period (i.e. five years ago). This returns 62 risk units.
  3. The opposite of 2: only trade in the direction in which net non-commercial interest is beyond 50% of its most-extended level. This returns 107 risk units.
Conclusions:
  • Adding COT data detracts from returns (unsurprisingly — if the system works, having a switch that turns it off for long periods of time is likely to reduce returns).
  • It is better to be with a strong consensus than not. 
I say that this method is quick-and-dirty because it does not consider risk-adjusted returns — it only gives a final profit figure. However, given the magnitude of the differences between the return numbers, I have some confidence in the conclusions.
Advertisements