Using proprietary position sizing and stop methods I show a 333% gain since Oct-2001 with a 5% MaxDD net commissions/slippage. 2 years of this data is out of sample. This does look very nice, however there are some problems: 1) Backtesting is not reliable. Needs to have a Monte Carlo simulation run. 2) It only works that well on the Russell and not on any other contract. 3) Now everybody knows about it so you cannot expect it to work for much longer. As for now the idea is valid.
Waiting for a pullback ruins the system ( I tested it ). You are assured a fill on losers and will only sometimes get a fill on winners. A version of the system tests very well on ZN (10 year note). Now all I have to do is make sure my ZN data is accurate. fan27
Actually yes I do believe it. Over the years I have seen it numerous times. The entire idea behind trading is finding an edge and exploiting it. When everybody finds your ideas you no longer have an edge. Maybe you have had different experiences but I have seen it happen more than once.
samson is right. you take a simple system that works, then bastardize it with optimal-f, optimal-that and optimal-other until nothing works. i wonder why wyckoff, livermore and the rest were so succcessful - probably because they never had computers to balls it all up for them!
While there is some truth to simplicity I think it worthwhile to point out that IB's and hedge funds are not paying quant programmers $500K+ a year for their own health.
How are those hedge funds doing this year? How has 99% of the bellweather systems performed in the last 2 or 3 years? Michael B.