Frosty, plot a volatility measure alongside your equity curve and see the results. The volatility indicator will trail market data and your periods of weak performance if there's a connection between weak performance and volatility. The past few days have been more volatile than a good bit of this year.
Frost, You already know this but just to remind you it is pointless to compare P$L each day for each of your systems. You'll drive yourself crazy! The fact that System #2 lost less $$ today is meaningless. Keep the bigger longer-term picture in mind. Compare weekly performance of each system maybe - or monthly even better. All the best!
frostengine, changing systems or contract in the middle of the action is equivalent to quitting and starting a new system. From experience developping ATS (just over 3 years, not much but still I think I can give you some feedback) I can tell you that second guessing your system will end up costing you a lot. You should have PREDEFINED criteria for: - when to change the trading contract, stock,... - when to move from system1 to system2 to system3 to ... - when to stop completely... The whole game should be PRE-planned, PRE-defined, and preferably coded before you do anything. I have a fully Automated Adaptive Trading System that have a life on its own. It is running 24/7 and no manual intervention is required ever. Everynight it even logs into my brokers accounts, pulls the daily logs and verify that the trades in its database match the ones in the brokers, and should alert me if there is any discrepency). Theoretically, I can leave for vacation for months and when I come back it should be still trading on its own (I never do it thou, because I worry what if both servers crash? what if... The only parameter I change at anytime T are: Staring now at time T+1 second (irrespective of what happened before T), what Profit Factor, % of winning trades, maximum $ loss, and other proprietary statistical measures... before it shut itself off. DO NOT try to make decisions after a big loss, a big gain... Have everything well defined before you even turn it on and live with it. Good luck! StocksSniper
frost, I follow your thread with great interest. Hopefully you just chose the very wrong time to go live, like you said, so that it is not the system itself that is not performing well. Hope you get out of this weak period soon so we will get to see some nice upside too!
Stocksniper. May I suggest get a blackberry and write your own Java app that lets you monitor your system remotely. Then vacation all you want
Yes, or just http://www.logmein.com Remote connection to your pc, it works really well, I can recommend it. Although it wont help avoid hardware issues, you can monitor your system and control your computer from any PC with internet connection. No need to install any software on the PC you use on your vacation, so you can do it from internet cafés or just get a small laptop if you have internet connection on your private island!
STOP What you're doing is called bleeding in trading venacular, and you have to stop to it and attempt to find a system which is profitable. As it is, you've entered the red with a series of steady losses after an outlier day which produced outstanding results. A lot of good advice given on this thread by professional traders, you should take it. Regards, JJ
who wants to waste time at internet cafes or sitting on computers? (not to mention the implecations of logging onto your secure ATS servers from public PCs) carrying a blackberry around is the way Im going to go. monitoring only. if I need something done. ill call someone to do it. JJ whats this 'line in the sand' chart all about. iv seen these charts posted by day traders over the past few months of reading ET. It seems like they get all giddy with a line tool and like to connect dots and pretend like its meaningful. Maybe the 'line in the sand' is a S/R level. looking at that graph, I would not be comfortable assuming it is not just going to break through that line like it has through countless lines over the past year that could be drawn with the same reasoning. and whats the deal with posting chart without any explaination why they are there.
I don't think the vanilla fortune cookie advice with regards to systems applies when you are in the implementation phase. The objective is to get a profitable system up and running at minimal cost. There are inevitable costs associated with implementing a new system on a new platform. The first thing you need to figure out is if the order entry is happening with normal allotted slippage, and is functioning as it should without trade errors. The second thing to evaluate once the order entry mechanism works flawlessly is whether or not the system itself is performing as it should be performing. For instance, comparing forward testing results to live testing results to see if real market assumptions are accurate in your forward testing or if the introduction of live market data / fills is somehow causing the system to perform not as expected. And then we get to whether the system logic itself is sound and has a profitable advantage, is robust etc. Curve fit systems (and systems with other fatal system flaws) often crank out losses right away when newly implemented and tend to produce losses that are way out of line with historical averages and maximums.