Did Martingale work for someone? | Page 4 | Forex Factory

Options

Go to
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

Attachments: Did Martingale work for someone?
Exit Attachments

Did Martingale work for someone?

  • #61
  • Dec 15, 2014 3:56pm Dec 15, 2014 3:56pm
You have a point there Hanover...what i wrote there was from my perspective and how I would utilize it if I were to use it...

I believe there were quite a few institutional traders who used the Martingale system in the past..most of them, didn't do well at the end
  • #62
  • Dec 16, 2014 2:50am Dec 16, 2014 2:50am
Compounding is the new martingale...
  • #63
  • Dec 16, 2014 4:56am Dec 16, 2014 4:56am
Quoting Olarion1975
Compounding is the new martingale...
New? A secret?

Savvy traders have been using fixed frac sizing to compound their gains since time immemorial. And without the need to use any kind of martingale.
  • #64
  • Dec 16, 2014 5:53am Dec 16, 2014 5:53am
I believe there are lessons we can learn from the failures of the martingale system

1) averaging down is kind of cool, but we don't need to scale in as aggressively as martingale.
2) we can't just keep on hanging on to losing trades, at some point we should cut our losses
3) inversely, we shouldn't limit our profits either.

I think sticking to the above rules you could tune something that survives the last 10 years of price data. Take on half as much risk as what that shows you and you should have a pretty good chance of surviving the next 10 years of price data.
  • #65
  • Dec 16, 2014 6:00am Dec 16, 2014 6:00am
at some point all systems need a human being to step in and say 'no more!' all systems!

Traders want to find a system that has no uncertainty in it - an impossible task and the major element in losing traders.

The only difference with systems are the level of risk - the more you increase the risk the greater the potential reward.

Martingale would work if you had the ability to kill your losses at the point the market decides this particular run will not come back - of course you can not quantify that with analysis - you can never ever quantify what will happen next. Phrases like 'high probability' just put a smoke screen over the complete uncertainty of it all.

The ironic part of it all is martingale is popular because people don't want to lose but taking a loss at some point is the only thing that will make it work
tradewith60
  • #66
  • Edited 1:47pm Dec 16, 2014 1:33pm | Edited 1:47pm
Quoting 60minuteman
The only difference with systems are the level of risk - the more you increase the risk the greater the potential reward.
That’s true, but with martingale you’re increasing risk, and exponentially so, merely to recover to breakeven.

Quoting moomooforex
2) we can't just keep on hanging on to losing trades, at some point we should cut our losses
Quoting 60minuteman
Martingale would work if you had the ability to kill your losses at the point the market decides this particular run will not come back
Agreed, but the trader who has sized at 1+1+1+1+1...... will always have a smaller loss to recover from, than the trader who has sized at 1+2+4+8+16...... (or 1+2+3+5+8...... or whatever progressive staking is used). The trader who used martingale has an unnecessarily large loss to recover from.

Having said that, the irony with martingale is that, after losses have become significant, taking “just that one more step” may offer the only realistic chance of recovery; but it also plunges the trader exponentially deeper into drawdown, if the next trade is a loss. It's a no-win situation.

Quoting moomooforex
I think sticking to the above rules you could tune something that survives the last 10 years of price data. Take on half as much risk as what that shows you and you should have a pretty good chance of surviving the next 10 years of price data.
Martingale defies this. The more rare the event, the larger the data sample required for the testing, in order to gain a statistically meaningful result. In the case of martingale, if the probability of getting enough consecutive losses to cause ruin is infinitesimally small, then the data sample must be infinitely large, otherwise the statistical error becomes prohibitive.

Please excuse the morbid nature of this analogy, but I believe it is apt: Is a person who has previously driven a car for 10 years without an accident more likely to die in a car crash, than one who has been driving for 20 years without an accident? That’s akin to the problem of backtesting a martingale. Even if no ‘black swan’ has occurred in 20 years’ worth of data, there’s no way of calculating statistically whether it’s more likely to happen tomorrow, or whether you’re more likely to escape ruin for several years. Historical testing becomes meaningless.

Quoting 60minuteman
Traders want to find a system that has no uncertainty in it - an impossible task and the major element in losing traders. ...... The ironic part of it all is martingale is popular because people don't want to lose
Agreed. Same-pair hedging, recovery trading, adding to losses, martingale — these are all born out of the human desire to avoid exiting at a loss.


Alas, I can not find a single positive aspect to martingale. It is simply the worst possible kind of MM.
  • #67
  • Dec 16, 2014 7:59pm Dec 16, 2014 7:59pm
Quoting hanover
Please excuse the morbid nature of this analogy, but I believe it is apt: Is a person who has previously driven a car for 10 years without an accident more likely to die in a car crash, than one who has been driving for 20 years without an accident?
I understand the point you are trying to make, but the reality is we aren't given a choice of only two strategies. We have a choice of infinite strategies and need some way to decide which one to choose, or inversely, which ones to eliminate.

Quoting hanover
Even if no ‘black swan’ has occurred in 20 years’ worth of data, there’s no way of calculating statistically whether it’s more likely to happen tomorrow, or whether you’re more likely to escape ruin for several years. Historical testing becomes meaningless.
Doesn't your point apply to all strategies and not just martingale?

I strongly disagree that historical testing is meaningless. I think it is very meaningful in determining which strategies you should not run. If I had to choose between two strategies:
- one that worked for 10 years that mathematically has a very small but certain chance it will explode; or
- one that works very well for only 5 out of the last 10 years
... tell me why I should choose the latter?

All strategies are flawed from the beginning as we are immediately at a loss due to commissions and spreads. In the best case scenario we would have no commissions or spreads and we would still only have a 50/50 chance of winning. The name of the game is to find the elusive edge. When found, your odds magically become better than 50/50 and although this is mathematically impossible people still throw millions of dollars into it. (Hanover, with respect, you are the expert here so please correct me if I'm wrong).

I think if we're willing to ignore the maths when we believe we have found an edge, then you can do the same when it comes to martingale as long as you are prepared (mentally and financially!) for what may come.
  • #68
  • Edited Dec 17, 2014 12:04am Dec 16, 2014 11:37pm | Edited Dec 17, 2014 12:04am
Quoting moomooforex
I strongly disagree that historical testing is meaningless.
I don't believe that all historical testing is meaningless. However, testing for the likelihood of a 'black swan' event would require a vast amount of historical data, due to the extreme rarity of the event.

I'm sure you're aware of the way martingale operates: it guarantees recovery until one unexpectedly long losing sequence suddenly causes ruin (here is an example: note how the account balance rises consistently until the system suddenly implodes). Given that such a losing sequence (for example, 10 consecutive losses) is a very rare event — in the example, it took more than 5,000 trades before it happened — using historical data to try to determine how likely it will occur sooner, rather than later, is close to futile, due to the high statistical error.

Given that price movement is close to random (50/50), if a trader trades for long enough then, at least theoretically, he will eventually encounter every conceivable price pattern, including the one that causes the 'death sequence'. It might happen tomorrow, or it might happen at a point beyond the end of the trader's lifetime, in which case his martingale is (fortuitously) a resounding success. But my point is that you have no way of knowing, and no way of testing, when it's more, or less, likely to occur. It's like playing Russian roulette — you don't know which chamber has the bullet, but if you pull the trigger often enough, you'll eventually encounter it.

In my opinion, that is very poor risk management.


"If I had to choose between two strategies:
- one that worked for 10 years that mathematically has a very small but certain chance it will explode; or
- one that works very well for only 5 out of the last 10 years
... tell me why I should choose the latter?"

An interesting dilemma. I think Van Tharp once put forward a similar question, something like: would you rather (1) lose a guaranteed $5,000, or (2) have a 1% chance of losing $200,000, and a 99% chance of losing nothing. Mathematically, the second option is better (on average, you will lose $2,000). However, people insure their homes, and other valuable assets. That suggests to me that people would rather have a guaranteed small loss (paying the insurance premiums) than have a tiny risk of a losing everything.

However, in your scenario, you don't say how much capital I have in the account, and that would likely determine my choice. If the account size was trivial, e.g. $500 that I could easily afford to lose, then I'd run with your first strategy, in the hope that it would pay off big without exploding.

However, if the account size was large, e.g. 50% of my net wealth, and I was forced to choose one of the two strategies, then I'd run with the second strategy, and accept a small loss, if necessary. But if I was allowed to choose neither strategy, then that's what I'd do.


"I think if we're willing to ignore the maths when we believe we have found an edge, then you can do the same when it comes to martingale as long as you are prepared (mentally and financially!) for what may come."

I respectfully disagree. If you've found an edge, why not trade the way the pros do, i.e. protect your capital with conservative MM, to give the edge the maximum possible opportunity to prevail over the long term? If the edge is robust, losses will be recouped as a matter of course. Martingale, with its 'death sequence', is a completely unnecessary risk. To state the obvious, that's why none of the pros use it.
  • #69
  • Dec 17, 2014 12:08am Dec 17, 2014 12:08am
Quoting hanover
If you've found an edge, why not trade the way the pros do
That question answers itself, plus what you imply here...

Quote
To state the obvious, that's why none of the pros use it

Lol, the Factory never changes. Still fighting the (good?) fight.
Congratulations ScalaFX for coming out of the closet!
  • #70
  • Dec 17, 2014 1:28am Dec 17, 2014 1:28am
Quoting Chicky
Hope someone may still be alive to answer this question.
I am repeating my answer after all these years.

Will renew one more time in 2018.

Did we yet find a person still alive martingaling forex?
The Thief of Wall Street
  • #71
  • Dec 17, 2014 4:53am Dec 17, 2014 4:53am
Quoting Chicky
{quote} I am repeating my answer after all these years. Will renew one more time in 2018. Did we yet find a person still alive martingaling forex?
I'm currently using the Martingale lot sizing in my trades successfully. And I use original martingale (BUY, SELL, BUY, SELL...) strategy. The basic idea of my strategy is to open a position if and only if the candle breaks out a level (closes on the other side of a target level). So, depending on the distance between the previous position's open price and that of the current position, I calculate the proper lot size (not necessary a multiple of two). If the number of trades becomes 3, I watch the net profit to become zero (breakeven) or more to close the trades. If the number of trades becomes 4 I watch the net profit and as soon as it hits my (virtual) SL target, I close all the trades. Another point is that, I always take the breakout of Asian session high and low levels. I have written an EA based on this strategy. So far it has worked very well for me.
Trading is the hardest way to make easy money...
  • #72
  • Edited Dec 18, 2014 2:41am Dec 17, 2014 1:34pm | Edited Dec 18, 2014 2:41am
Quoting nubcake
Lol, the Factory never changes. Still fighting the (good?) fight.
Yes, still here. Whenever I feel bored or stressed, it motivates me to write, even if it's on the same old topics. Don't know why.

Good to see that you're still around. Compliments of the season to you, my friend.

Quoting Chicky
{quote} I am repeating my answer after all these years. Will renew one more time in 2018. Did we yet find a person still alive martingaling forex?
Chicky,

Forexhard uses (a capped) martingale with his SSBO system, and as far as I know, he's still alive, and one of the most successful traders on FF.

While a capped martingale can work, and deliver consistent returns while it does so, it's still too risky for my personal liking. The textbook approach (build an entry/exit method around a directional or behavioral bias in price movement; and size at a consistent 1%–2% risk per position) still makes the best sense to me.

Compliments of the season to you, also.

David
  • #73
  • Dec 18, 2014 2:06am Dec 18, 2014 2:06am
Quoting hanover
.... The textbook approach (build an entry/exit method around a directional or behavioral bias in price movement; and size at a consistent 1%–2% risk per position) still makes the best sense to me.....
That's too boring though, there's no apparent challenge in it. Fortunately for those who need losers to feed the market, most traders here are far too smart to trade like that
si hoc legere scis nimium eruditionis habes
  • #74
  • Dec 20, 2014 11:08pm Dec 20, 2014 11:08pm
Quoting hanover
{quote} Chicky, Forexhard uses (a capped) martingale with his SSBO system, and as far as I know, he's still alive, and one of the most successful traders on FF. While a capped martingale can work, and deliver consistent returns while it does so, it's still too risky for my personal liking. The textbook approach (build an entry/exit method around a directional or behavioral bias in price movement; and size at a consistent 1%–2% risk per position) still makes the best sense to me. Compliments...
Thanks Hanover, I don't know how you dig out what is relevant from these countless threads.
Will read the detail in this extracted thread later. I believed in martingale until 2007, it always worked for me, always, on Usd/JPY, even in those times when there were no charts connected to trading platforms/screen whatever. In 2008, I lost my account trying to martingale that 700 pip down eur/usd move. It again comes back to "its good if it does not happen in the lifetime of a trader", but the fear is always there and this fear never lets me gain big. I also tried some capped martingales, but not very seriously. Maybe I go back and check it one more time.

Thanks and Happy Holidays.
The Thief of Wall Street
  • #75
  • Edited Dec 21, 2014 3:24am Dec 20, 2014 11:39pm | Edited Dec 21, 2014 3:24am
Most Martingale traders have low probabilities in their result to start with. in this case, it's guaranteed to fail when you encounter a series of losing streak to burn through your pocket.

If when probability is high enough, Martingale would work well.
Below was my result in Martingale + Grid in the past although several investment bank portfolio managers didn't like me using the Martingale approach by its characteristic and you can't really define a fixed risk % to your account when you are in bad luck situation of having losing streaks one after one.
Attached Image (click to enlarge)
Click to Enlarge Name: Capture.PNG Size: 321 KB

There are plenty article online explains what probability you would need to be consistently profit in Martingale. Look it up.

If you are on the path of wanting to get recruited by any FX trust banks, Martingale is NOT preferred.
If u luv happiness, start w/ discomfort, make friends w/ pain and failure.
100 Hits 0 Miss All Time Return: 35.8%
  • #76
  • Dec 21, 2014 1:27am Dec 21, 2014 1:27am
How did you have a best trade of 1159.8 pips when you are using a 50 pip grid? (Note average win/loss/expectancy).
  • #77
  • Dec 21, 2014 3:10am Dec 21, 2014 3:10am
Quoting moomooforex
How did you have a best trade of 1159.8 pips
oh, since you noticed that... I was actually in a broker (IBFX) contest (who gets the most pips in one position in a week), and I let one of my GBPNZD position ran loose (with profit) and won the contest for the week.

Quoting moomooforex
when you are using a 50 pip grid? (Note average win/loss/expectancy).
Naw... exit strategy has always been letting the winner run as much as it does, and split exists.
If u luv happiness, start w/ discomfort, make friends w/ pain and failure.
100 Hits 0 Miss All Time Return: 35.8%
  • #78
  • Mar 1, 2015 4:07pm Mar 1, 2015 4:07pm
Quoting FXSayWhat
Most Martingale traders have low probabilities in their result to start with. in this case, it's guaranteed to fail when you encounter a series of losing streak to burn through your pocket. If when probability is high enough, Martingale would work well. Below was my result in Martingale + Grid in the past although several investment bank portfolio managers didn't like me using the Martingale approach by its characteristic and you can't really define a fixed risk % to your account when you are in bad luck situation of having losing streaks one after...
Hello, are you still trading martingale style ?
  • #79
  • Mar 1, 2015 7:05pm Mar 1, 2015 7:05pm
Quoting metta87
{quote} Hello, are you still trading martingale style ?
hi, no i have already replied it's not something for the long terms.
http://www.forexfactory.com/showthre...28#post7956928

Thanks.
If u luv happiness, start w/ discomfort, make friends w/ pain and failure.
100 Hits 0 Miss All Time Return: 35.8%
  • #80
  • Mar 2, 2015 5:05pm Mar 2, 2015 5:05pm
martingale works very well until the market starts to go way off in the other direction, then you just end up getting margincalled,
the result is allways the same..
Bulls are stupid Animals!especially when Im short!
0 traders viewing now