Hi Aaron, Could you tell us what software you use for constructing and backtesting your systems? Runningbear
We primarily use Tradestation 2000i. Before 2000i came out we had to use C++ because Tradestation 4 can't handle long periods of intraday data. I'm sad that Tradestation has gone the way of a brokerage firm rather than stand-alone system testing software. I expect we will eventually migrate to a different platform and are keeping our eyes on E-signal's backtesting capability and on Wealth Lab. We also use Excel for quick backtests where the data will fit on a single worksheet. We get our historical data from Tick Data.
ddefina.. a question. How many contracts did you backtest the system with? This $100 Avg/Trd is this per contract or for the total contracts? Is there Commission included? Just wondring.. thnx, alain
I've been using Wealth-Lab for a couple of months now and it's pretty good. I like the fact tha tyou can test over an entire portfolio rather than just a single stocks like TS. And they have futures support as long as you have the data for it. It's also improving everyday. 99
Hi Aaron, A little off topic, but I thought this fit well in a "Systematic Performance Stats" thread... Congratulations on your performance! I just finished reading my current issue of Futures, and I saw you at the top of the list. Well done, and continued success!
ddefina: welcome to the beautiful world of optimization! Impressive results, too good to be true, IMO. What's the slippage and commissions? I agree with dottom, who said that systems should be tested with severe slippage. If they survive this test, they're good.
Thanks, DaveN! If you or anyone else would like to keep up with our performance, private message or email me your email address. I'll send you a monthly email of our return and the S&P500 return for the month. Your email address won't be used for anything else.
ddefina your system has a max drawdown -0,64%, and total net return +95%... very, very impressive ratio! I just can't believe my eyes.
Aaron, What time frame does your NQ trading system trade? ie 5 min bars, 15 min bars? With only 44 trades over a month period, it must be a 5 min or so? Your average profit/trade of about 2.1 pt/trade for a 5 min frame would drop if equated to a 1 min frame. But more trades on a 1 min frame would equate to the same profit. Just trying to get a feel for my mechanical automated system which is doing 1.5 pt/trade on a 1 minute bar chart, but does 2000 trades ave. per year. Thanks,