garyfritz wrote:Iamshakey wrote:3. If the trade goes wrong, and price moves short, the EA opens another long trade at the next S&D box, at an increased lot size. This pattern continues.
4. The Bot now has X number of losing trades open, and the final winner will cancel out the other losers, therefore accepting the smaller losses and keeping the bigger profit.
Seems logical to me that this should work. Sooner or later, price WILL break out in your favor.
Maybe, maybe not. And your account may not survive long enough to see it. This is just a type of Martingale strategy. It can make the equity curve look great, popping you right out of losses -- but it can also blow up. One strong fundamental-driven move could kill you.
At a minimum I would want to limit this to opening no more than a specified number of trades. If those continue going into the red, you should give up and close the positions. And it will hurt.
I agree with you completely with problems with martingale, and also agree with your solutions of limit trade level and shirt protection.
When left unbridled it will leave you broke, and sitting in a dark corner with a half-empty bottle of scotch every time. The only success i have ever had martingale is by interfering manually. BUT, this is better than most. Closing or turning the EA off at selected times, might save it.
I have been trading this over the last week manually, Opening and closing the trades by the rules of the EA, but managing the trades manually as well. I have been doing this with several pairs, usually having 2-5 pairs going at once. Sometimes, I manually close the initial profit trades at 20-30 pips if I think it will return to BE or move to loss. Sometimes I close combinations of 2-3 pairs when the math shows a modest and overall profit. Sometimes, close the entire basket at once when profit is around 1% of account balance.
This is on 15 min time frame with Supp-Dem set with forced TF of 30 min. I have opened the traded pretty much as mechanically as possible according to the EA, but have skipped over some signal which "just didn't look ritht to me.
Also traded all Sessions (I have an odd schedule that allows me to trade each session a couple times a week). Have even had good result early Asia.
Here are this weeks results. (I can't post actual statement since this is a little live account, and I can't figure out how to remove account # etc).
Closed Trade P/L: 13.77
Details:
Gross Profit: 18.18
Gross Loss: 4.41
Total Net Profit: 13.77
Profit Factor: 4.12
Expected Payoff: 0.22
Absolute Drawdown: 0.00
Maximal Drawdown: 1.46 (0.87%)
Relative Drawdown: 0.87% (1.46)
Total Trades: 64
Short Positions (won %): 34 (70.59%) Long Positions (won %): 30 (83.33%)
Profit Trades (% of total): 49 (76.56%) Loss trades (% of total): 15 (23.44%)
Largest
profit trade: 1.94 loss trade: -0.97
Average
profit trade: 0.37 loss trade: -0.29
Maximum
consecutive wins ($): 11 (5.38) consecutive losses ($): 3 (-1.14)
Maximal
consecutive profit (count): 5.38 (11) consecutive loss (count): -1.14 (3)
Average
consecutive wins: 4 consecutive losses: 1
Trades were usually started with 0.01 lot. Usually escalated lot size by double, but occaisionally by only on additional base lot.
Traded mostly major crosses, with the AU, AJ, and UCan thrown in.
I think using multiple pairs at a time help the winners offset the losers. The most the martingale escalated (that I remember) was, 0.01 0.02 0.04 0.08
I used Steves MTMP set on basket mode with 5% acct balance to close entire basket in case of disaster (it will probably need 10% if too many orders get larger).
Also noticed that baskets seem to move together well.
I realize that such manual judgement is almost impossible to code. If it give Ribelo any food for thought, and would like more details, let me know. I will play around with this for awhile, and may trade this more seriously once I find out what the WCS (worst case scenario) is.
Shake
