Martingale Madness

So, I have been playing with the dreaded martingale for several years now. Tons of blown accounts (which is what the martingale is famous for).

Since I look at Forex through the eyes of a gambler and not an investor. This dashboard was designed around that philosophy. Similar to playing multiple seats at the blackjack table. Trading multiple pairs at the same time instead of hands of cards.

Recently, I came up with the idea to limit the number of iterations that the martingale could use. I settled on 8. This way, I could control loss to a maximum as set by the base lot size. .01, .02, .04, .08, .16, .32, .64 and finally 1.28. If all of these orders lose, then we would lose a maximum of about $1,200.00.

That being said, I started playing with different entries and exits that would (hopefully) never exceed that fatal 8th loss.



Here is what I have come up with...

3 Year back test. Same settings on all pairs so I do not fall into the curve fitting trap.

1 EA to manage them all. Starting with simple 60 pip SL and 60 pip TP. 2x Multiplier on losing trades. 21 Pairs tested, but only those that use 8 iterations or less to recover get to move forward to the next round on a demo account. As you can see, there are MANY settings I have played with. I will be doing some tweaking in the future to see if profitability can improve.



Beyond each pairs doing its own thing, I wanted to be able to watch overall account health and take profit when it happens.

CLOSE PROFIT button will close all trades in profit and reset their martingale calculation on their individual pairs to base lot size.
CLOSE ALL will close all trades and reset their martingale calculation on their individual pairs to base lot size.

Back Test info to follow with a Trade Explorer to start on demo account after the first of the year.

Update 12/30/20 - Backtest completed on 21 pairs. I am choosing the pairs in green that each went through 3 years of trading without needing 8 trade iterations to recover.



Since we all know that historical results are no guarantee of future results, proper withdrawal schedule should be implemented to keep the money flowing.

My personal goals are to pay off debt and build a retirement plan that I can control.

Happy New Year all!

Update 5/2/2021: Revamped EA. New Backtest start at post #14. Trade Explorer added with start date of 4/24/2021 when I started this EA on live account.

EURCAD

EURAUD

EURUSD

GBPJPY

GBPUSD


EURNZD


GBPCAD


GBPNZD


USDJPY


NZDCAD

NZDUSD

AUDUSD


AUDCAD


EURGBP


AUDJPY

Phase 1 complete. 21 Pairs tested on latest version with Martingale.

Number of consecutive losses / Win Percentage


I am choosing the pairs in green that each went through 3 years of trading without needing 8 trade iterations to recover. That demo account will be started on Monday and Trade Explorer posted here on FF.

Have a safe and happy new year!

So, I took all of the strategy tester data and placed it on a spreadsheet. I then highlighted any trades that took 7-8 trades to recover. I wanted to see if trading 15 pairs on a single EA on same account would over leverage the account if more than one pair was in drawdown at the same time. The good news is, in 3 years, that did not appear to happen but once. Although lining up the dates on the spreadsheet is challenging. So the randomness on the trades SHOULD limit the number of large trades at the same time. 7 Trade draw down only happened 16 times in 3 years. And ALL recovered within those trade sequences.

I have attached my data if you want to view it.

It looks like...
6/26/2018 USDJPY
8/9/2018 EURNZD
10/1/2018 NZDCAD
1/2/2019 EURUSD
4/5/2019 AUDCAD
4/22/2019 AUDUSD
5/16/2019 EURCAD
11/5/2019 EURCAD
11/18/2019 GBPJPY
12/17/2019 AUDCAD
12/17/2019 EURAUD
1/17/2020 GBPJPY
1/30/2020 EURNZD
2/12/2020 NZDCAD
4/15/2020 GBPJPY
7/2/2020 AUDCAD

Based on this information, there should be a minimum of 2x Equity and Drawdown in the account to ensure 8 trades can be covered for at least 2 pairs that go into draw down at the same time.



These tests show us the possible risk. Now, we have to wait and see what actual trading does to see if the reward is worth the investment.

If you took the gains from each pair over the 3 years and added them all together, there would have been a profit of $7,824.39. I would suppose, what the broker's leverage is would dictate the account size and then you could calculate growth. Assuming of course that the next 3 years were similar to the last 3 years. But it is all a guessing game at that point.

A 1000 or 500 to 1 Leverage account would have the best return as the account triples in 3 years without any compounding. Whereas a 50-1 leverage account would earn about 50% in the same time frame. Not a bad return. The opposite would be true with risk. IF there were a losing sequence, the loss would be about $1200.

On a $3000, 500-1 leverage account would equate to a loss of 40%, whereas a 50-1 leverage account with $15,000, the same loss would equate to 8%. Hopefully it would not happen, but it could and should be prepared for by moving early profits to a recovery fund or not taking out profits from the account until there is at least $1200 in profit to cover a possible loss.

Well, after a week, I am not impressed with the Martingale. My Martingale is on the right, when my personal trading is on the left. I think Ill just stick with the one on the left. But, I'll let it run and see what happens.

Going back to my Gambler's Dream Algorithm. Targeted trades based on Price Action and Price History with Martingale recovery.

Created a new Hedging model. Will adjust TE to 2/25/21 tonight. Closed all active trades in the process

Who does not take risks, he does not drink champagne

EA mechanics seem to work properly. Adjusted Risk to aggressive mode .01 per $1000 of account balance and changed TE date to start today.

In the event of a sequence fail, should to about 18-20% loss. The question will be, does the algorithm generate enough to cover that and more. Or, in my planned usage, to generate a steady second daily income that can be pulled out to enjoy.

Since the back tests of this thread pertain to an initial version before adding the hedging model, I am starting a new thread to avoid confusion.

https://www.forexfactory.com/thread/post/13427087

I have been tweaking my algorithm and think I have something I would actually go live with. Back Test results to follow. Same settings for each pair.