Can a method of averaging be optimized in a simple system of trading the same contract on 2 accts simultaneously but in opposite directions? On the winning side (long or short), Average UP 1, 2, 3 ⦠contracts (variation of Inversed Martingale Model) adding 1 contract at each level and a cycle having at most 2 levels for scalping (i.e. having 3 contracts at most on the winning side anytime) Levels are based on N/2 ticks increment where N could be 10. i.e. +1@ N/2, +1@ N, take profit -3@ 3N/2, repeat +1@ 3N/2 + 1 tick (level 0) i.e. Cycle of 1,2,3 contracts accumulated -> taking profit at end of the cycle then revert back to accumulating 1,2,3 contracts if the trend continues. [re-entering 1 contract 1 tick (level 0) higher after taking profit at 3N/2] On the losing side(short or long), Routine: Just average +1+1+1+⦠at N, 2N, 4N, 7N, 11Nâ¦levels If the trend reverse and the winning side starts to retrace, For 1 contract accumulated, start the losing side routine when it retrace by N. For 2 contracts accumulated, take profit of 1 contract immediately and start the losing side routine when it retrace by N. For 3 contracts accumulated, take profit of 2 contracts immediately and start the losing side routine when it retrace by N. When it retrace beyond the average of the original losing side, that direction will become the winning side such that you have to clear all the original accumulated contracts first before following the winning side cycle again. The clearing of winning contracts would be done 1@N/2, 2@N, remaining contracts@3N/2 from the average point which applies if it reverse back again. When the spread between the 2 averages > 5N ticks, look at total profits accumulated â total unrealized loss = Net. Close both sides and restart. Close both sides also when it exceeds 8 contracts on the losing side if your total $50k<capital <$100k even though your net is a profit for less exposure. The goal is at the end of the day/week, I want the Total Nets to be positive Let me know if this system can be further improved. PS: brokerage fee is like $1.5 per side for a mini-Dow contract YM using IB (and 1 tick is $5) so don't bother about it. esp if your N=>10 ticks Also, it would be easier if this system can be automated i.e. BOT Trading /Programmed.
Can anyone recommend more literature on betting strategies or averaging methods implemented in trading systems? Maybe Shannon and Keller were right, it could be possible to beat the market...
Can't beat Kaufman for detailed info on just about ANY method - including theory of runs/gambling stuff. http://www.amazon.com/New-Trading-S...ef=sr_1_2/002-8513457-5264002?ie=UTF8&s=books
That's an excellent book but did it have a section on averaging method and gambling theory implemented in a trading system?
I just have the 3rd edition of Trading Systems and Methods, but it has some info on runs and martingales-antimartingales starting on p 565.
Thanks, have you tried implementing any of those methods? What are you trading? Want to share some of your ideas? Can PM me if you are shy
Hey Wayne, What do you think of this? http://www.elitetrader.com/vb/showthread.php?threadid=78779 Looks pretty viable to me. Regards, JJ
I agree with the conclusions of austin and 40yo. If you have a positive expectation system, and you strictly control trade size/risk, the "average down" or whatever you want to call it, is just another trade. And if you are a believer in reversion to mean, it could be a trade with a higher probability than the previous, losing, trade. Your mileage may vary...