I'm guessing it's time to get out of Frosty's thread and create a new one like: "What it takes to succeed as a systematic ATS trader" The basis of the discussion seems to go back and forth between, general consensus and Frosty's level... I started talking from what Frosty should be doing around his competency level, but he doesn't offer much. With questions and comments, the thread is starting to get a mix of ones who are professionally inclinded and not... Added: I actually went ahead and made one: http://www.elitetrader.com/vb/showthread.php?s=&threadid=99383 I'll keep myself in the new thread, I seem to be a big distraction in here. Have fun and best wishes Frosty.
Since the bot is currently offline and new strategies being tested/developed. I figure we can possibly open this up to some of those issues... Specifically exits... As stated before through this thread arbitrary stops and profit targets I have always found to severly hamper a strategy. So with that said I have generally only used stop and reverse type exits. Also in my previous strategy I had some exits based off of ADX values to get out when I thought it was becomming choppy. What are some other potential exit strategies?
For a trend following bot, maybe it could exit when a fast and slow moving average cross or when they go flat. It seems that is an indication that a trend has at least pulled back and price may be heading for choppiness or heading for a reversal. If the bot exited under those conditions, then it could wait until there was a clear indication that a new trend had started.
my deal is.. exits are there to capture profits from all those nice entries my system produces. hence..their importance cannot be understated. i'm always looking for, or developing a new exit trailing stops hard profit & stop loss typically, i surround my entries with all of em not saying this is correct or proper. i havn't the experience to verify it. just the way i have been building mine. bb
I set targets by average range, a simplifiction of average true range wherein I just measure bar ranges, [high -low], and average those figures for anywhere from 7 to 14 bars back, whatever backtests the best. I update the target as Average range changes while I'm in a trade. I set stops on trend reversal because "the trend is your friend until it bends". Regarding testing on the IB sim account, I played with it. I could send 2 market orders a second apart and get fills 20 pts apart on the YM, that is way off from any kind of reality. I'm porting everything to NinjaTrader starting tomorrow. NT has the good simulator, the C# compiled code to run quick, less limitations than Tradestation/Multicharts in ways important to what I'm doing, mulitiple data sources, several ways to autotrade... I put it off because I did not want to wade into C# but it's my best choice at this point.
I ran JSystemTrader's 'TrendFollowing' strategy today. Get JSystemTrader. True: I got chopped-up, but I don't believe there is any way that JSystemTrader's trend follower would give you the kind of losses you experienced with your system.
Swing trading; close above/below daily pivot, cross above/below zero line CCI. intraday cross above/below zero lineCCI, cross of LRC.
Good clues in 'New Market Wizards' as regards time-frames. A 'Guru' who posts on this site got most of his 'Ideas' from reading the various Schwager books imo.
Hey Jonny - why make the post so cryptic? Would be good to know who this "guru" poster is. Or even what you are refering to in NMW. If these are trade secrets, fair enough but then why bother posting at all?