Is Martingale really dangerous?

I have experienced two problems with this strategy (1) how to cut losses - (2) when to add to a bad position - for the first - the problem is that the nature of the idea is to never exerience a loss on a trade - if one trades long enough eventually one has to make this choice - how when everything you have is tied up in being right about the original trade set-up do you make this choice. My biggest losses came about from having to stop because I couldn't handle the pain of the market moving agianst me for days.

On the second some of my best pay days have come by using this method as part of my trading plan - It is worth looking at when entering the market and looking where your stop loss is going and think about doing a limit entry instead of a stop loss. when you do that then you look at one more level of price action against you which then helps you see the larger market. The problem is doing this is can give you the yips - unable to pull the trigger - and throw you off your game.

Bottom line - doing this can work it requires one finding the depths of their soul. That area of self we seldom explore - our core needs and fears and learning to live with them fully expressed.

Hi,

what you state in your first paragraph has nothing to do with Martingale.

Adding to an already open position is something completely different.

Martingale means you always bet on a specific outcome during a sequence of try's. Until that outcome arrives, you double the bet size each time.

Its not about being in a trade that has already gone way agains you and then averaging on the way down to get a better break even point.

regards
daytrading

Exactly. The sequence of try's are my different trading setups...

what about margin? risk/ratio? is this not in question?


Best Regards,
Marcelo Marques

I think we assume that people do the math before attempting to use or discard these strategies.

regards
daytrading

You need a REALY good indicator to do this, because, if you enter in wrong place , you can lost all the capital...
Some people make statist about some indicator that how many errors can ocurred, so, study the lots value to dont stop because of margim


Best Regards,
Marcelo Marques

Marcel,

...not sure what you're getting at. I don't use any indicators at all - simple math, thats all.

I know precisely when to enter, exit and how much per trade - no need for indicators the way I trade.

regards
daytrading

Yeah, indicators are for noobs! I do not mean to offend you. But there will never exist an indicator that will turn 10k into 1mil!

I trade with discretion and use various support and resistance levels to enter a trade only in the direction of the trend!

I'm just updating my money management a bit and that is why I posted this....

Read Ryan Jones's "Money Management".

He does not advocate the Martingale Method or its derivatives. As mentioned above, a string of losses can quickly wipe your account out. Assuming you're trading for the long run and not using it as a get-rich-quick scheme, you should risk no more than 1% of your account (or up to 10% if smaller than say £10k). As this account size falls through losses, so will your actual amount risked.

Averaging down is also dangerous for the newbie. There is likely to be a reason why a price is falling (assuming you are going long). That reason is likely to be "because you are wrong". Accept that you are wrong for this trade and live to fight another day.

MM...MM...MM is the key!!! Simply looking at the daily charts indicates that about 50% of the time if today was an up day, then tomorrow will also be an up day, or vice a versa. If one simply follows the daily trend, a study of the charts will show that you stand at least a 50% chance of being correct. With that said, with proper MM you can be very successful!

With proper MM, one can have a system that is successful only 40% of the time (or even as little as only 25% - 30% of the time) and still reap in about 400 pips per month. For example, with the gbpjpy (sl-100; tp-200), you could be wrong 6 out of 10 times on your daily trade and still earn $4k per month on a $10k account.

Now, who wouldn't be satisfied with a 40% ROI? Martingale adds an unnecessary degree of difficulty to good MM practice. I have used martingale with mixed results. Personnally, I would double up only once or twice at the most. However, once you get into martingale it get's increasingly more difficult to get out with each progressive trade... until you hit the wall and your account is blown. Again, IMHO... MM is the key... maybe on a sporadic basis I would be willing to use martingale for 1 or 2 progressions.

Thanks,

Walt

Today i was watching on the news about a trader that lost 7 billion using the martingale. He would double up every time he lost thinking he would come back and he never did.

His managers and the the trader were all fired once the bank found out.

So this martingale stuff you gotta watch out for it can get you out of a hole or it can dig u in even deeper.

There you go. It might work until you hit that one time when you're wrong and it just keeps going.

Or, just like the last few days, the markets have been a little crazy IMHO. This morning I took two trades that were not bad trades, but I lost them both. It just happens. Had I reversed, I would have come out ahead for the day, but that doesn't fit my plan. I'll wait.

Tomorrow is a new day, I trade my plan as always.

The only way somebody learns the martingale lesson is when they blow all their profits (and/or account) to hell. Here is a article that's worth a read http://articles.mql4.com/472.

Even if a trader has a high win rate like 60%, the martingale would still eventually wipe out the account. The reason is that for any finite account equity, there is SOME maximum number of times you can double the trade amount before you bankrupt the account. And that string of losses is not as uncommon as people think.

Suppose you have $1000 and you start with a $1 trade. How many times can you lose, doubling the amount of the next trade each time, before running out of funds? Let's see...

1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, uh ohhhh!

10 losses in a row and you're done.

But then you think, "But how likely is it that I'll get 10 losses in a row? I have a 60% win rate after all."

Well, let's see. The probability of 10 losses is 0.40 to the tenth power. That's a pretty small number: 0.0000105. That's about 1 time in every 9,500 trades. Or about 9-1/2 1KT traders. : So for every 10 1KT traders who use martingales, chances are, one has gone broke doing it.

Another way to look at is if you're a scalper and do 100 trades a week, you'll go broke about once every 95 weeks, or about once every couple of years.

This isn't exact of course, but it gives a good idea of the realities involved.

Doesn't take much bad luck at all to bankrupt an account with old Martingale....

Sidus, I made a lot of experimenting with every kind of Martingala derivatives I could imagine. So far I've found only one strategy that seems promising. I'm going to describe it here.

First, you need a trend following system in order to use this strategy. You need to employ a trailing stop loss for every position. As every trend following system you will have a lot of small losses, a lot of small wins and few very big wins.
This generates an equity curve that I don't like. Yes the big spikes up when you win big are good, but indeed most of the time the equity curve has a descent (even if very small) that I don't like. So I tried this MM strategy that seems to solve this problem.
What you do is this: you decide a fix percent of your capital to risk by default on every trade (let's suppose 1%). Then when you start a new trade you look at your last X closed trades (I've found 5 is a good value for X). If the sum of the P&L of these last X closed trades is bigger than zero then on your next trade you will risk 1%. If it's less than zero you will use 2%. No other sizes are never allowed.
I've found that this strategy produces a way more smooth equity curve than simply using always the same fixed percent risk.
The only reason I'm still not using this strategy live is because I've still not understand why I never get bigger drawdowns with this strategy compared to the fixed percent strategy. In fact even when I have 12 consecutive losses this strategy ends up with smaller draw down than the one with fixed percent.
I think, but I'm still investigating it, that the answer lies in the fact that the real reason trend following systems have those descents in their equity curve between one big winner and the next is not because of the great number of consecutive losses that could incur, but, indeed, because of the great number of small losses mixed with small wins together which usually amounts to a number way bigger than the sole 7 (or 9 or 12...) consecutive losses that you could get.
What I'm saying here is: 10 consecutive small losses do not hurt that much your equity, but 10 consecutive small losses + 1 small win + 3 other small losses + 1 small win + 2 small losses and so on... you see, they adds up in a way that generates that kind of descents that I don't like. By doubling (only one time of course) your risk when your last X trades (I use 5 for X) sum up to a loss you give your equity the opportunity to better capitalize those small wins here and there that you get between the small losses. This generates an equity curve that is a lot more smooth than the classical approach. And then, of course, you'll have the big winner, which could also happen that you will ride with the double sized position if you are lucky.
Let me know what you think.

I use a method that I guess could be considered an inverse Martingale, but only on trades where I firmly believe in the long-term direction (yearly trends, not weekly/monthly). I am holding a trade using this method right now that survived last week's destruction of emerging markets.

Here's the method:

Open new trades in the same direction of a losing trade in front of known S/R levels and increase lot size as you get further away.

example:

EUR/USD is at 1.48. Known support: 1.47, 1.455, 1.44, 1.425, 1.41

Trade 1: LONG 1 lot at 1.48, T/P @ 1.50
Trade 2: LONG 0.1 lots at 1.4725, T/P @ 1.50
Trade 3: LONG 0.25 lots at 1.4575, T/P @ 1.50
Trade 4: LONG 0.5 lots at 1.4425, T/P @ 1.50
Trade 5: LONG 1 lot at 1.4275, S/L @ 1.4075, set T/P @ 1.48 for all open trades

The intended goal is to constantly improve the breakeven point without putting too much strain on the short-term volatility as in the typical Martingale. Here's how the average price would change:

Trade 1: 1.48
Trade 2: 1.4793
Trade 3: 1.4752
Trade 4: 1.4664
Trade 5: 1.4527

I'll rarely trade averaging down though and believe it can only be done with margin and leverage under tight control.


1.479318

hi youalreadykn
Lets see the possibilities and ca be make an EA to make some back tests in a volatilite market.
Thanks and good luck.



Best Regards,
Marcelo Marques

Yes, martingale really is very dangerous. but it's the only way I've found so far that guarantees a profit everyday.

Trading is a losing game. high probability methods with >50% win/loss ratio and bigger average winner than loser do not exist. at least I haven't found one so far.

instead, I've found that with the right MM, any method could work, especially the trend following ones.

I have been daytrading for the last few months. I've noticed that on lower time frames, market trends almost everyday. by trend I mean a satisfactory move in one direction that gives enough pips for profit taking. so if one could come up with a not too sensitive way to follow price action and combine it with a proper MM, it's really not impossible to make a profit everyday.

martingale is probably the easiest applicable MM. I have been using it live for the last 3 weeks, and it has made me a profit every day. I have backtested it up to 8 Oct 2007, it also produced a profit everyday except the three Christmas days (24,25,26).

however it still is a very dangerous money management method. because it completely relys on market making a sizable move everyday. but nobody can guarantee that. Oct 2006 is a very good example, that many small range days could have killed an account many times over, had martingale been employed.

so what's the solution. can we really take money from the market without taking too much risk? I don't know. but I do know that we need big winners and small losers to profit in the long run. this brings us back to the old cliche - cut losses short, let profits run. I don't know about others, but I've found this very hard to apply in real life.

the Pascal's Triangle introduced by twoblink offered some hope in solving this problem. it's the 1st systematic and consistent way I've found so far to "cut losses short, and let profits run" . I'm still playing around with it. but the results are already convincing me that pyramiding is a lot safer than martingale, and the profitabilities are quite comparable.

Hi all,

..as I mentioned in one of my previous posts during this thread, the Martingale strategy is not just black and white.

When I trade intraday, I have a very high win rate 80%+ but with very small targets and very small (mental) stops. So, in the event that I have a loser, I don't strictly apply Martingale regardless of the circumstances.

What I mean is that some of the approx. 2 to 20 trades I make during a day have a slightly higer probability of succeeding. When those setups arrive AND I had a loser on the previous trade, I apply Martingale.

The reason why it works is simple math: even if the Martingale trade fails (first stage = double the original size), it means that instead of the original lost trade, I have technically now 3 lost trades. (1 lost with normal size and 1 lost with double size = 3 lost normal size)

With a ratio of approx. 8 out of 10 trades winning, or better 800 out of 1000, I can apply Martingale to a certain EXTEND.

Using the strategy does not force you to over-do it. Its a question of risk management. If the first stage Martingale (double the original position size on the next trade) fails, I don't proceed to the next stage (double the double), I go back to the original size.

Also, to use Martingale, you need to have a lot of constant parameters in a trading method - it does not work if targets and stops are different for every new trade.

regards
daytrading