Does the martingale system really work?

hello guys , i just want to know is that martingale works?
is that system could be use for long time?
-could anyone who already use this system give comment?
-which pair is the best for trading this system?

thanks for your attention.

Short term--- maybe. It is possible to do well with martingale over a short number of trades.

Long term--- account destruction. Eventually, you will hit a losing streak that will wipe you out. It may happen in the first week, or it may take a few months. Rest assured, it will happen. It is a mathematical certainty.

Regards.

Your luck might work for some time, but you'll certainly run out of it.

It's very important to fully undestand how powerful and dangerous martingale is.

Let's take 10 000 account - if you start with just 0.1 lot (1:100 leverage), how many consecutive losses that account can take if you use just 20 pip SL?
Maybe surprise but only 8.
1) 20
2) 40
3) 80
4) 160
5) 320
6) 640
7) 1280
8) 2560
That's it, your account is gone.

Wikipedia will help you more with mathematical side if you are still interested:
http://en.wikipedia.org/wiki/Marting...etting_system)

In a highly leveraged market like FOREX....a Martingale strategy has the potential to wipe you out. As many have mentioned, you might be able to get away with it 70% of the time. But that 30% will have you "cryin' uncle" in the worse way and your broker will be thanking you for your generosity.

Martingale strategies increase lot size after previous losses.

ONLY if you have a trading system & market conditions which make the higher lot size transactions have a higher expected win (normalizing for lot size!) does it make sense.

Otherwise, what they do is to give the illusion of a larger expected trade win than the underlying truth.

Statistically there is the true "unknown" expected trade win which you will experience after infinite number of trades.

You never have this, you have only the actual results, a finite set. The question is how well can you determine your edge (exists? and how big?) from finite data. Martingale strategies mask this by falsely appearing to have a greater empirical Sharpe ratio over a finite data period, interspersed with rare huge losses. The convergence to the true expected win is much longer.

Personally I think one ought to see at least 20 fully independent losing trades.

Professional hedge funds typically use anti-martingale strategies---they increase risk when they are near all-time NAV highs and decrease in drawdowns.

thank you for the reply. for example i deposit US $10000 and begin with 0.01 lot (10 cent/pips). I think it's save enought to trade this strategy. what do you think?
Is is possible use this strategy for eur/gbp or eur/chf pair?
Anyone using martingale strategy could give explanation too. thanks.

What about a martingale that only adds five times? Wouldn't that limit the risk?

So you double 5 times. If the fifth trade loses, you are down 1+2+4+8+16 = 31 units. So here are your choices:

a) Cut your losses, and revert back to a position size of 1 unit. Now you have to make a sequence of 31 wins to compensate for 5 prior losses; OR

b) Double again, and risk being down 31+32 = 63 units, but at least it will only take 1 win to compensate for the 5 losses. But if you keep doubling with each loss, it will only take one sufficiently long losing streak to blow your account.

Increasing position size to recover losses is gambling. No position sizing method can give your trading an edge. Only a solid knowledge of how markets operate, and fashioning a suitable method of entries and exits around that knowledge, can ever achieve that.

Having said that, there's no law against using forex as a vehicle for gambling. But you need to decide carefully what kind of risks you're willing to take. It's a personal decision.

FWIW, I posted about Martingale in more detail here.

Re which pairs, Martingale is a betting (sizing) strategy, not a trading strategy, so it makes no difference. You might use different entry and exit strategies to exploit different price behavior in different pairs, but that has nothing to do with sizing.

The smaller your starting size relative to your total capital, the more doubles your account will survive. But you're crippling your returns exponentially. As a VERY ROUGH example:

Trader 'A': account = $10,000. Starting lot size = 0.01. One trade per day, profit taken at 10 pips. Annual return = $1 per day x 250 trading days per year = a 2.5% return on a $10,000 account. Starting risk per trade = 10 pips = $1. Hence the account can survive 1+2+4+8+16+32+64+128+256+512+1024+2048=$4095, i.e. 12 consecutive losses. The probability of 12 consecutive losses, within any 250 trade sequence, with a 40% win rate**, is 41% (see the attached table).

Trader 'B': account = $10,000. Starting lot size = 0.1. One trade per day, profit taken at 10 pips. Annual return = $10 per day x 250 trading days per year = a 25% return on a $10,000 account.Starting risk per trade = 10 pips = $10. Hence the account can survive 10+20+40+80+160+320+640+1280+2560=$5110, i.e. 9 consecutive losses.
The probability of 9 consecutive losses, within any 250 trade sequence, with a 40% win rate, is 91%

[** Let's assume that the win rates of both traders is 40% (for scalping 10 pip moves, for example, it will be somewhat less than 50%, due to the spread).]

Do you think that either of these equations represents acceptable equation?
-- 2.5% annual return, with a 41% probability of irretrievable drawdown in the first year
-- 25% annual return, with a 91% probability of irretrievable drawdown in the first year

Put another way, is it worth crippling returns by 10 times, when the risk of ruin is only a little more than doubled?

Of course, these figures are a rough guide only, and they assume 10 pip scalping, no real entry/exit strategy, 1 trade per day, no compounding of wins/losses, etc. But hopefully they illustrate the kind of thinking you need to undertake before you put money on the line.

Footnote: values in the table are rounded to the nearest whole percent. Hence the yellow areas mean > 99.5% and < 0.5% respectively. The calculation assumes that there is no causal link between trades, i.e. that each trade is an 'independent' event.

I used Martingaling before and came to
conclusion that is is able to burn any account.

Hanover,
How about this scenario:

Trends are stronger at the beginning (which is one thing that distinguishes trading from casinos), and eventually reverses, so seems to me a position sizing strategy based on that would provide an edge. So, let's assume a trend signal is generated on a chart, and compare 2 strategies for same length of a trend:

1) Decreasing lot size as trend continues:
a) .4, TP @ 50p
b) .3, TP @ 50p
c) .2, TP @ 50p
d) .1, TP @ 50p, and TP every 50p until trend reverses

If one of the above loses, 2 options will produce viable results:

a) continue decreasing lot size until reversal signal.
b) close trades and start over again with .4 when new reversal signal is generated.

2) .1 lot size only during the entire trend.

So, for #1:
1) will have a much higher profit.
2) any losses will have high probability to be more than made up for in next trend.
3) if use #1, option a, and .4 trade loses, will still have profit at end if remaining trades during that trend are profitable.
4) any losses will not have a "death" trade effect as with a martingale strategy.

The 50p TP I used, starting lot size of .4 (and subsequent lower lot sizes) are arbritrary and are used for comparison purposes only.

There are a couple martingale EA's out there that you can put on a demo account and let it run to see how it works..

Here is one that is not a full blown martingale but as the same effect on the account if it does not work. Power SM

What you're describing is a type of 'pyramiding' rather than Martingale. The answer is that success will depend on how well you're able to pick trends, and (as with all methods) how accurately you adjust your position (or in this case, add positions) relative to price reversals.

A pyramid is nothing more than a group of individual component trades. Each of these trades needs to be on-balance profitable in its own right, otherwise it is undermining overall bottom line. Simply adding positions on the basis that the existing trade is in profit won't work in itself. If a pyramiding method is on-balance profitable, that is because of the way in which it's exploiting the trending nature of the market, not because of the MM. The proof lies in the fact that if price was a completely random walk, EVERY method would have an expectancy of zero. Hence the success of any method ultimately depends on how accurately it exploits inefficiencies in price behavior.

I included some specific examples of the rationale behind 'pyramiding' in my post here.

Secret 1) Capital must be enough to accept maxDD, maxDD is the max price go before meeting Good Entry(GE).
Secret 2)Entry signals must be an Edge,: meaning no more than 5 Wrong Entry(WE) before price meet GE.

as long as you can find both secret, you can safelyuse martingale EA in Forex but of course no guarantee.

Secret 3) You must bank in certain target gain till 100% gain. Then you will be consider having an Edge martingale EA.

I have found my Edge martingale sometimes ago.

Secret 4) Keep your secret safe.

Good Luck!

martingle fails, averaging works.

martingle works on a model that double drawdowns
averaging works on a model that builds positions overtime to attain a desired limit.

two very similar concept but with very different risk modelling

(1) What happens if you get 6 or more consecutive Wrong Entries before you obtain the 100% gain needed to bank your target?

(2) If you bank your target, doesn't that mean that there is less money in the account, to survive a later sequence of Wrong Entries?

(3) If the entry signals provide an edge, why not simply trade consistent position sizes, and eliminate the risk I alluded to in (1)?

"you can safely use martingale EA in Forex but of course no guarantee"

(4) If there is no guarantee, then how can you say that 'you can safely use martingale'?

thanks for all of your comment. anyway we should make martingale , by using other indicator too. we saw the trend first, not only martingale every time the price against you. so in my opinion , it's better trade manually not using an EA for martingale.

thanks for your opinion. btw it's almost the same between martingale and averaging ,right? so do you mean you open with same lot with the previous trading( you don't double the lot).

martingale never works unless you have an unlimited account

as I said before I deposit for $10000 and start with 0.01 lot (10cent/pips) so it's 100000 pips is your margin. I think it is strong margin available. do you think it's almost unlimited account ?