Suppose that I have a long-only strategy with an edge. Any thoughts of how I can know when my edge is fading?
IF the market is going down, and it is a long-only strategy, then the P & L would be going down. However, how to know if this is going to continue or the P L would go up again when the market recovers?
I would say if you hit Monte-carlo worst case DD something is not right with the system ... I expect my system to win at least 42% If over 2000 trades the worst case DD was not worse than 600 USD I know once I hit this I have a problem. The system that uses 2R for profit-target does not have 42.69% Win rate anymore Maybe too primitive approach but works for me
Statistics. Compare your current stats with past/properly-backtested stats during similar situations (bull-run to bull-run/bear to bear/high volatility to high volatility/etc.). Average equity curve pull-backs. Frequency of pull-backs. Max pull-backs. % profitable rate of return sharpe ratio sortino ratio correlation to sp500 etc. etc. Can you tell, in advance, whether the fade is permanent? or temporary? Doubtful.
Slope of monthly equity curve. Going back so many years, you can look for "mean" drawdowns, I use 20-30 years to back test so I can get better idea of slopes and loss pullbacks. Too many IMHO do not test far back enough of what one can expect. Am guessing that so long Indexes are rising, your method will continue, but did you test it 2007-2008 years? You might be able to see better how it is did in strong downturn.
If the pnl of n trades is below -2 sigma (based on stats I have), then I start looking in more depth what has changed or if anything went wrong. Process control.... For example:
How to know if my edge is going away? -When you are not making profit anymore and just continuing to make losses? Seriously, is this a question?
Don't worry, you'll know. Signed, Guy who has religiously followed his trading system for that last 8 months and hasn't made a dime over that period.