Martingale - the sad truth | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Martingale, Anti-martingale, and Compounding 40 replies

Martingale, Reverse Martingale, Modified Martingale, Maths 1 reply

Martingale Calculation EA (also reverse martingale) 3 replies

How to change this Martingale to Reverse Martingale ? 3 replies

Martingale vs. Non Martingale (Simplified RoR vs Profit) 1 reply

Martingale - the sad truth

  • #1
  • First Post: Edited Oct 21, 2012 10:26am Oct 20, 2012 9:53am | Edited Oct 21, 2012 10:26am
Martingale or d'Alembert systems seem to be in high favor with some traders, so let's just estimate after how many trades a martingale system will blow your account. The system increases - usually doubles - your stake after every loss. The theory is that you can not lose, as after so many losses eventually you'll win and get all your money back.

Assume a system with a 50% win rate. You invest 1% of your equity per trade. After every loss you double your investment: 1%, 2%, 4%, 8%, 16%, 32%... that's a total 63% after 5 losses, so the sixth loss will give you the margin call.

With a 50% win rate, the probability of 6 consecutive losses is

0.5^6 = 0.015625

The estimated number of trades until the margin call likeliness exceeds 50% is

0.5/0.015625 = 32

(precisely it's log(0.5)/log(1-0.015625), but that needs not bother us here). With one trade per day, your account will last about 2 months. A higher win rate, such as 75% or 90%, won't help because then the average trade loss would be accordingly higher and the required loss streak shorter.

By the way, if someone wants to sell you his system or EA, you can often see from the equity curve if it's a martingale system. The equity curves of martingale systems have telltale sawtooth dents.
  • #2
  • Oct 20, 2012 11:22am Oct 20, 2012 11:22am
What about if you do what nanningbob and some others recommend and start with 0.1% of your account?
  • #3
  • Oct 20, 2012 11:33am Oct 20, 2012 11:33am
Quoting Hugh Briss
What about if you do what nanningbob and some others recommend and start with 0.1% of your account?
The number of trades prior to a margin call becomes 10 instead of 6, so the numbers become .00097656 and 512, so such a system with a precise 50% win/loss ratio would last around 2 years before the account would be blown.
  • #4
  • Oct 20, 2012 12:36pm Oct 20, 2012 12:36pm
Quoting Andy B
The number of trades prior to a margin call becomes 10 instead of 6, so the numbers become .00097656 and 512, so such a system with a precise 50% win/loss ratio would last around 2 years before the account would be blown.
Fair enough.
  • #5
  • Oct 20, 2012 1:03pm Oct 20, 2012 1:03pm
An even sadder truth is that no amount of proof will dissuade some of these people from martingale.
  • #6
  • Oct 20, 2012 1:16pm Oct 20, 2012 1:16pm
How'd this thread get commercialized? Did I miss something?

We need more anti-martingale threads!
FXGears.com
  • #7
  • Oct 21, 2012 10:28am Oct 21, 2012 10:28am
Quoting Jack_Larkin
How'd this thread get commercialized? Did I miss something?
Commercial members apparently can not post in other forums - at least I can't, so I have to post here.
  • #8
  • Oct 21, 2012 10:35am Oct 21, 2012 10:35am
Quoting jcl365
Commercial members apparently can not post in other forums - at least I can't, so I have to post here.
Oh, I thought it was posted somewhere else first then moved here... and I didn't see anything commercial about it.

Mm..
FXGears.com
  • #9
  • Oct 22, 2012 8:31am Oct 22, 2012 8:31am
Do you consider a system that:

 

  1. Only goes long.
  2. Adds to losing positions but not doubling or tripling them.
  3. Add space between orders as long as the draw-down continues.

to be a martingale ?



I'm talking about my system. If you want to get the details:
100$ Trading



Kind regards.

Simplicity is the Ultimate Sophistication.
  • #10
  • Oct 22, 2012 10:21am Oct 22, 2012 10:21am
It's not a literal martingale, but if I understood it right, it works similar. The martingale relies on loss streaks not to exceed a certain limit, and your system relies on prices not to exceed a certain movement range. As both assumptions are only valid for a limited time, both systems will collect a lot small wins and then blow the account. At least that's how I understand it.
  • #11
  • Oct 22, 2012 12:11pm Oct 22, 2012 12:11pm
Yeah, it does suffer from some of the same issues as martingale. All you've done is increase the distance price can move against you in exchange for requiring additional distance price has to move in your favor to become profitable.

The big flaw comes when you start holding large baskets with many layers of trades, as each additional trade moves the break even point less in your favor (diminishing change.)

There's no problem with averaging into a trade, so long as you pre-planed to do so and the total risk of all entries doesn't exceed what you were prepared to lose in the first place. In other words, don't just keep adding blindly, have a reason to go long/short and have a price level where it's clear that you're wrong and must close your position. Doing a little math ahead of time will tell you how many trades and how much size each trade should be to make that stop level within your risk tolerance.

Of course, I'm not suggesting you trade like this, just detailing one method should your trade plan (and edge) benefit from it.
FXGears.com
  • #12
  • Oct 23, 2012 1:28am Oct 23, 2012 1:28am
Thanks jcl and Jack for your contributions.

I would just add that, Yes, I knew that I had some of the same issues as martingale, but nobody and no system is perfect right ?

Also, Maybe you forgot that playing AUD/JPY which pays some swap, is also part of the strategy. So, each day, each week, each losing position loses a little less. Time is my ally, not my enemy... at least today. This is important as it allows me to replace the hope, when I have losing positions, by just patience.

And yes, Jack, I don't just keep adding blindly. It would be stupid to do so when being at historical highs or just after the RBA has lowered its interest rates. The trader judgment and some TA is important here.

As you suggested, I have also done the math and make my trades according to my plan which allows me to go for a loss of 50% in the pair from the first buy.


Kind regards.
Simplicity is the Ultimate Sophistication.
  • #13
  • Last Post: Edited Oct 24, 2012 2:20am Oct 23, 2012 11:51am | Edited Oct 24, 2012 2:20am
Just for anyone further interested in Martingale and its derivatives, here's a small Zorro script for Martingale trading:

Inserted Code
function run()
{
  Spread = 0;
  Slippage = 0;
  Stop = ATR(100);
  Profit = ATR(100);
  Lots = pow(2,LossStreakTotal); 

  if(NumOpenTotal == 0) {
    if(random() < 0)
      enterShort();
    else 
      enterLong();
  }
}

And here's the equity curve with AUD/USD - looks almost perfect, doesn't it?

http://zorro-trader.com/Log/Martingale_AUDUSD.png

Mind the tiny downwards peaks...
0 traders viewing now