MM needed for a yearly target
I'll assume your math is correct. However, your trading model is a quasi-martingale system. As such, it means that if you "fall behind" your formula increases the size of the trade to "catch up" to your fixed goal, or conversely if you are fortunate and "ahead of schedule" the formula will then decrease the size of the trade so it softly lands on the 300% target.
I will assume for the moment you do not have a working EA, but are considering ways and means for its money management aspect.
I think a really import formula for helping you design an EA algorithm is this: Preliminary Profit Factor = win%/loss% x reward/risk
Lets say your win% is 65% and your Reward-Risk ratio is 2:1 the Preliminary Profit Factor (PPF) is: 65/35 x 2/1 = 3.7
Anytime the PPF is over 3.0 you have a very stable EA ------- if you do not abuse the size of your trades.
You can have a stable EA with a winning percent below 50% provided the Reward-Risk ratio is sufficiently high, for example:
PPF = 40/60 x 125/25 = 3.3 ; where TP=125 and SL=25
EAs with a winning percentage below 40% are hard to make stable
Personally I do not like martingale systems because they put you at unnecessary risk. You are far better off living with the stochastics that your EA alogrithm delivers instead of applying some martingale MM principle where you try to catch up if you fall behind. Essentially your EA is what it is ---- don't try to put lipstick on the pig to trick yourself. There are other ways to fix an under performing EA.
The remainder of my comments apply to non-scalping EAs. Scalping EAs rely on thin margins just beyond the Spread and my following comments will not work on thin margins. However. if you develop an EA that has some working room (ie large TP and SL), consider the following.
1. It's very hard to write an EA that does not produce some kind of bias one way or the other, provided the trade rule(s) you apply are "trend related". What do I mean by "bias"? An EA that has no bias produces 50% wins and 50% losses. Therefore utilize some indicators that purposefully put some bias in the trade selection algorithm and swing the bias to 60/40 or better. The key aspect that goes hand in hand with that stochastic bias is the associated Reward-Risk ratio (refer back to the PPF formula).
2. You can create "reasonably" stable EAs when the PPF is above 1.8 preferrably over 2.0, but you really have to dumb down the size of your trades to avoid large drawdowns.
3. Play around with many TP/SL configurations in the backtester (even weird configurations like TP=40 vs SL=400) and see what kind of bias your EA shows you. (Note: scalping EAs typically use weird TP/SL configurations that is why it is worth a look --- more on that later)
A moment ago I said "there are other ways to fix an under performing EA" You'll love this story.
The first EA I produced was an absolute dud; I could not get it to produce a Profit Factor over 0.72 in the backtester. But I kept playing around with various TP/SL configurations ---- even weird ones like the TP=40 and SL=400 configuration mentioned a couple sentences ago, and the Profit Factor was abysmal with the winning percentage of only 16%. The results were so bad, I thought what if I do a contrarian switch, after all, a 16% winning percentage coupled with a 10:1 reward-risk ratio produces a PPF of 1.9.
Take note: a contrarian switch does NOT mean you simply switch the TP and SL around. Instead, it means you change the OrderSend() function to issue BUY orders when the algorithm says sell, and issue SELL orders when the algorithm says buy. By doing so, you associate the 16% winning percentage with the TP=400.
Did it work? Yes, the contrarian switch worked; but I did not achieve the PPF that I calc'd before making the switch. The Profit Factor was 1.42.
At least now instead of having an EA that is an absolute dud, I have one that is a very low performer. I have continued working with various TP/SL configurations after making the contrarian switch. I have been able to achieve Profit Factors ranging from 0.90 to 2.72 depending on what year of market data it tests and what the TP/SL settings are. I have also discovered many TP/SL configurations that produce winning percentages in the 40's so I think I can make this EA reasonably stable if the size of the trades are really dumbed down to say the account size divided by 75 per trade.
I actually have more to say but this post is already quite long so will stop here.
Cheers
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Hello
I have had an idea of setting a yearly percentage increase on my account. one where the loses are spread accross the year and turned into a TP
example
account open $300
Percentage increase 300%
Target $900
Tick Value GBPJPY 1.022
Yearly pip target 881
monthly 73
weekly 17
Daily 4
If the first trade was a loss of 75 then you would take $75 / 1.022 = 73 and add the 73 to the 881
giving a 954 target and 5 pips per day target. and if you manage to keep your losses to a minimum then you should in theory maintain your percentage increase.
Let me know what you think.