OK, so I see what synthetic bars have to do with backtesting but I still don't know what it has to do with automation. Richard
Ok, well I am not Anekdoten - and frankly I have an idea what Anekdoten is eluding to. But I don't want to ask any question's cause I don't want to possibly taint any edge that I myself may have by doing something similar. The thing is we don't know what we are both "exactly" are flirting with in public discussions. So I will just take a chance that I'd rather safeguard what little edge I have rather than talk about it in order to perhaps gleam some in-site how or what someone else is doing to the end benefit of improving my own. Did ya get all that? mb generally talking we either filter data or don't use real data to trade with the latter of which is classified. define:real data - as using other data to trade another data. or morphing the real data as so it no longer resembles the real data. or another idea which is nether of the above but you would have to create your own data server to collect and store that data which is currently not available historically. so you could make a system out of it. and if all this doesn't prove i'm a loony then you all must be crazy.
In automation edges are extremely rare and protected for that reason alone I never discuss the little I know in that department. The second your little edge becomes common knowledge it begins a cancerous process until it dies. Anek
you know there are few great advances in mankind. very few advanced ideas and or thinkers. but once a man sees what another can do - then also himself can duplicate. i used to have the domain blackboxsystems.com before i sold it to microsoft and my email address was pandora@blackboxsystems.com - yea pandora for good reason. don't let that lid ever open - ever!
holy @&% I cannot believe this. My system has an almost identical result. If you tell me how to generate that png file, I will post mine. unbelievable!!! See my posts at http://www.elitetrader.com/vb/showthread.php?threadid=113260
Read your thread. Posted the code for a similar system to yours about a year ago in AFL format. Very tradeable, you might find it interesting. http://elitetrader.com/vb/showthread.php?s=&threadid=87281 To generate PNG, I run a report and capture with a program called SnagIt. If you wanna compare, consider using $30K and not compound, also IB commissions and a Tick slippage. --jeff
Hi all, interesting thread, I'd just like to share how amazed I am after a glance to these PNG files. Your system doesn't buy up-breakouts and sells down-breakouts, but -seems to me - follows trends very well for a proudly defined "counter trend" system... In other words, you enter on temporary weakness, but - too many times to be explained with randomness - in a matter of days your entry seems really in line with the overall trend. In other words, it does enter counter-trend, but seems able to switch to "trend mode", when a sizable trend is in sight. Without this, the pain of countertrend, can be very high: if you give space to your position (e.g. no stops, or very larger ones), you'll have to stand painful DD's. If you use not-so-large stops... well.... profitability will decline (not always a bad price to the sense of "security" gained). Just to make this clear, I join a little simple "real" countertrend system: just look at the sell signals on March and August, that's what expected in a "plain and simple" countertrend system: no pain.... no gain. PS this little system makes about 1700 S&P points on 1 lot S&P futures from 1982 on (NO comm & slipp), but - for reason too long to explain here, it is not a workable system, because - IMHO- it is exposed to a "conceptual flaw".
thanks for the comments and the picture - i have updated the pictures below to include the 30 yr. bonds now. i only did this exercise to show that there is no turtle system that can even come close to keeping up with modern day systems. the turtles do outclass everything in the marketing department though LOL! http://www.yafeu.com/pic/ also your system shows a great deal of promise, i know what happens it makes a flurry of winners and then hands it back with a large looser. this is the nature of counter trend trading. it's how you prepare for that big looser that can make the difference for you. you have to understand quartiles and how it can be applied to your trading. so lets say 75% of the time what is the maximum open gain that a trade will have prior to it winning or loosing? using this you can defend from the big losers and flip them around to actually trend trade, but only as a last resort. mb
doing great UNTIL 10/17/08 and then i took a 2500 per contract hit on the minis. have recovered about half that as of today. i have churned about 17k in gross profits to net about 3500 in over about a month. so a whole lot of trading for little gain. but didn't loose my a$$ at least. but took some heat.