Discussion in 'Trading' started by psytrade, Oct 23, 2005.
Indicate system timeframe, annual return, drawdown, market traded, #trades, and leverage/margin
This post misses an important point about system testing.
You can over-optimize any system to within an inch of its life and tweak out the last 0.001% of profitability. And then go & try to trade it - it falls flat on its face.
Also, to increase profitability, just increase the length of time you are optimizing over.
I counter that it is less important to find a system that yields 500% returns with 20%maxDD, than a system that yields 40% with a 10% maxDD and can consistently be expected to work for the next 6 months.
That's what I am after.
Did you mean systems that TESTED well or TRADED well?
-please note if your system rules were optimized. simple yes / no is suffucient.
-also noted if this is on the bench or a traded system
-Have you found a system with only a 10% drawdown earning 40% returns?
Try this one:
This is a very broad question -- try narrowing it further.
The idea is to get an idea of the average number of traders with systems. I already use a backtesting environment and have a system.
What number of traders stick with existing systems such as aberattion, etc? Is the average Elitetrader user a developer of systems or user of systems?
I'm using two different systems I developed myself. One is a dip buying system that I did completely from scratch the other is a trend following system based on something I found on the wealthlab website.
The trend following system is customized for both the Russell 2000 and S&P 400 midcap e-mini futures. Either version of the system works well on the other future, but I do have the parameters changed to get the maximum historical profit.
The dip buying system works across a fairly broad array of markets without changes, although I'm now considering whether or not to optimize for some currencies that have had problems in the past.
I don't see a problem with some optimization, but if the system didn't work before the optimization then optimizing it to give a good profit is a bad idea. At least in my opinion to be a candidate for optimization a system should start with a profit factor of at least 1.5.
Separate names with a comma.