We're continuing to test the reliance of mean reversion versus momentum strategies on TLT...the 20 year Treasuries ETF. For the past several months the 3 day mean reversion pullback approach has been highly reliable versus a similar 3 day momentum tactic. But now a new dynamic has emerged and I scratched my head over these metrics trying to decipher what they meant. I suspect the parity of the 2 models is a function of the narrow range periodic cycling of the current market as reflected in the Bollinger Band studies profiled yesterday.
Note the behavior of each model along the 2 year RSQ equity line.
Also note the 30 equity curve of each model supporting our mean reversion preference for TLT.