Is martin so bad? Or do you have to know how to cook it? - page 11
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
That is, if there is a maximum series of eight losing trades and a maximum series of one profitable trade, you can get profit using this method if the initial deposit is large enough? It is only necessary to calculate the risks and take into account the moment when the series of losing trades may be larger.
not really. By D'Alamber's method (or a variant) we bet 1 unit (assuming that it loses completely), and when we lose, we add another unit and so on until we win, and then again.
The total for 8 consecutive losses we get the costs:
1-й -> -1
2-й -> -2 - 1 = -3
3-й -> -3 - 3 = -6
4-й -> -4 - 6 = -10
5-й -> -5 - 10 = -15
6-й -> -6 - 15 = -21
7-й -> -7 - 21 = -28
8-й -> -8 - 28 = -36
then we bet 9 (hence the minimum deposit requirement - more than 9+36 = 45 times the initial bet) and win our bet (9) and another same amount. In total our loss from this series is -27 initial bets (as you can see in the picture by troughs) - not a plus at all.
Here the profit/loss ratio is important. If it is 1:1, the profitability of the method will end at a series of 4 successive losses (the 5th bet will return 5 units, the loss has already been -10 by that time).
If it is 2:1, then the profitability ends with a series of 6 consecutive losses (the 7th bet will return 14 units, while the loss by that time is -21).
By "profitability ends" we mean that this series of losses cannot end in "+" according to this method.
So, if Profit\loss = 1:1, then for profitability, you should have most losing series not exceeding 3 consecutive losses (if 4 - already a loss) - that is exactly the code given by 220Volt - he made one mistake in the series length in the previous post.
If profit\loss = 2:1, then profitability will be maintained for most series of 5 losses.
And so on; the higher the Profit\loss is, the longer "most loss series" can be allowed.
And the fact that a system with negative MO above the spread and with at least one positive trade - with MM and sufficient funds using "history", one can make it profitable - no doubt (take martingale, or cunning MM - minimum lot at a loss (we use "history") and loss +1 at gain - theoretically such coincidence is possible in real life, though the probability is very low).
But, as Urain stressed, if results of some trades depend on others, such a system can almost always be improved. R. Vince says that ideally it is possible to apply different MM methods only to systems with Z-count close to 0 (though not even with Z-count much larger). If any dependence is detected, it can be eliminated by means of "exploitation" and only then the new system can be studied for MM.
Well, here I am at 5......
As correctly noted by tol64,
the first problem with martingale is the long series of losing trades. If, for example, a losing series of 10 consecutive losing trades, the volume would have to be increased by 1024 times, which is unacceptable for any system.
Therefore, measures must be taken to limit the length of losing trades.
Approach 1:
We know that the length of losing trades depends on the number of deals evenly distributed throughout the history and is proportional to its logarithm with base 2. Increasing TP and SL, we decrease the number of trades.
For example, for TP=50 and SL=50 (4x sign) we will have about 4 deals per day, for 10 years - about 8000 deals. On this history we will theoretically have one series of 13 losing trades, 2 series of 12 losing trades,
4 series of 11 losing trades, 8 series of 10 losing trades etc.
If we increase TP and SL up to 300 points, the number of trades will be reduced by (300/50)^2=36 times, it will be 8000/36=222 trades somewhere. In this case we can theoretically expect 1 losing series of 8 losing trades,
2 series of 7 losing trades, 4 series of 6 losing trades, etc.
Approach 2:
Splitting series into sub-series. Suppose there is a series of trades "-P-P-U-P-U-P-U-P-U-P-U-P-U-P-U-P", where P is a profit and U is a loss. We have a series of 10 losing trades.
It can be well broken down into sub-series of 5 trades: -PPUPUPUPUPUPUPUPUPUPUPUPUPUPUPUPUPUPUPUPP. In this case the lots will not be changed by the principle 1-2-4-8-16-32-64-128-etc. but according to the principle
11111-22222-44444-88888- etc.
There are other approaches, to solve this problem...
Another problem of martingale is that the whole series of losing trades is attempted to be brought to profit by just one profitable trade, which leads to geometrical growth of trading volumes.
However, it does not necessarily have to be done in one trade, it can be done in two, three, etc. - the next series.
For example, a series of losing trades 1U-1U-1U-1U can be taken in a profit series of 2P-2U-2P-2P-2P.
For instance, a well-known martingale based on Fibonacci numbers from the following series: 1-1-2-3-5-8-13-21-, etc. The trick of this type of martin lies in the fact that if one trade turned out to be in profit, this profit is used to settle 2 previous trades.
For example, the deal opened in profit with 13 lots, we can destroy 2 previous losses - 8 and 5, not counting the spread.
Martin based on Fibonacci numbers is more soft, unlike the usual Martin.
There are other approaches to solving this problem...
From my first acquaintance with Forex (this event is 10 years old) I got it from some "authority fathers" saying: "Martingale is the trader's absolute evil" as an axiom. I have never used this method till now.
But recently I`ve been reviewing forum messages and saw the question "Where can I get a good Martingale?" and in the comments to the question I got a comment that there are no good and bad martingale.
I decided to get acquainted with the question and try it out for myself, so to speak.
I will go ahead and firstly tell you my conclusions, and then illustrate my findings. So, I concluded that one can and should use martingale, but on one obligatory condition: "The Expert Advisor must not be initially losing, otherwise it will only accelerate losing". That is, without martingale, the advisor should still show profit on most trades, and then martin can slightly compensate for losses from unsuccessful entries (with a certain degree of probability).
Well, this is my personal opinion, but I wonder what percentage of traders are really friendly with martin? I also wonder how many of the future participants of the Championship 2012 will use this method in their EAs? I wanted to post a survey on this topic, but I don't know how to do it...
In the files for illustration:
Report EasyTrend-minlot-
This is the tester report on a simple trend Expert Advisor where every new bar has a signal to buy or sell (depending on how the trend is defined) and a minimum lot is opened with the specified TP and SL.
Report EasyTrend+Martin-minlot-
Report of the Strategy Tester for the Expert Advisor that differs from the previous one in that the martingale is used (usual lot multiplication scheme) .
Report EasyTrend+Martin+MM-
Tester report by the Expert Advisor with martingale and simple money management (test for compatibility).
According to the results of the first 2 tests we can see that Martingale has increased our profit by 24% (is it a little or a lot?, com), the balance curve looks better and some indicators have improved.
The result of the third test shows that the martingale and money management system (according to the current balance size) do not conflict with each other.
So maybe it was in vain, that I treated Martin badly for so many years?
1. Well, here I am at 5......
2. As correctly noted by tol64,
the first problem of martingale is long series of losing trades. If, for example, a losing series of 10 consecutive losing trades, the volume would have to increase 1024 times, which is unacceptable for any system.
Therefore, you need to take measures to limit the length of losing trades series.
There are other approaches to this problem...
1. Congratulations again! :-)
2. ... e.g., increase lots not by multiplying by 2, but by the same Fibo numbers, by other schemes, e.g., ratios more aggressive with downside: 5-4-3-2-2-2..., calculation of dynamically changing channel width by the APR indicator...
And if entries are not close to random (at least with minimal TA), then the sense of the TS with MM on Martini is lost, IMHO - because of "rare" entries first of all. It turns out that in the case of the TS on a fixed lot with MM just above zero, it is more successful to trade with MM on the basis of the same percentage of DEP or whatever, but not with martin with its minimum entry rate, which is still at the "solid TA or whatever other kind of analysis we have to wait long time (for market entry signal), as opposed to random entry on a small TF-m.
In general, what, IMHO, is the task of TS on the martin, using min. What is the task of TS, for example, to enter the market on the basis of an indicator, make profit with minimal lot or 0,1% of deposit on a Martin, then make its size in the channel (ie, I mean the over-rotation Martin), say, the same numbers of Fibo, and then increased volumes, when prices leave the channel in a certain direction - take profit with some kind of trawl... But again, price turbulence is high on small TFs, so it won't spread long... If the market will exit on the trawl from the profit level, for example, trawl on fractals is not bad.
All, IMHO. If you are interested, I can look at my posts and post links from the foursome here on these issues.
The thing is that the last 8 months I work only with him, why? Because I do not trust delayed orders. 2 The Expert Advisor (ShokBar v1/1) allows you to change profit margins within any reasonable limits, depending on the deposit and the person's employment in the market.
//---
I conducted a number of experiments and still found a grain of truth in using this method. I found that this method can be used to make a profit on random inputs and when ~90% of the optimization results are simply lost.
//---
But of course it is not recommended. At the beginning it is better to work on the strategy when ~90% of optimization results are in the profitable zone.
//---
That is when this method will be a very good addition to the trading strategy. That said, the martingale can only be part of the money management system. It can be combined with Fixed Proportion or something else (you have to experiment). And also hedging, diversification, etc. are not cancelled out. All this should be used in trading.
Notused ,DmitriyN ,R0MAN thanks for the variants.
//---
I conducted a number of experiments and still found a grain of truth in using this method. And absolutely on random inputs and when ~90% of the optimization results are simply plummeting, this method can pull in profits.
//---
But of course it is not recommended. At the beginning it is better to work on the strategy when ~90% of optimization results are in the profitable zone.
//---
That is when this method will be a very good addition to the trading strategy. That said, the martingale can only be part of the money management system. It can be combined with Fixed Ratio or something else (you have to experiment). And also hedging, diversification, etc. are not cancelled out. All this should be used in trading.
It would be better to write the phrase: "But of course it is not recommended to do so" in red and poisonous colours, and several times bigger font, and duplicated at the beginning and at the end of the topic ... Because I have already imagined, how many of them, who have seen only one idea (the one that easily finds resonance in soul), rushed to make martins on the basis of "signals of a coin".
Well then your highlighting is very appropriate. ))
I will also add that even if 90% of optimization results are in the profitable zone, no one is secure from a rapid loss, or at least a significant part of the deposit. )))
Notused ,DmitriyN ,R0MAN thanks for the options.
//---
I conducted a number of experiments and still found a grain of truth in using this method. I have found a grain of truth in using this method, even though it is absolutely random and when ~90% of the optimization results are simply plummeting.
//---
But of course it is not recommended. At the beginning it is better to work on the strategy when ~90% of optimization results are in the profitable zone.
//---
That's when this method will be a very good addition to the trading strategy. That said, the martingale can only be part of the money management system. It can be combined with Fixed Proportion or something else (you have to experiment). And also hedging, diversification, etc. are not cancelled out. All this should be used in trading.
I sincerely congratulate you in approaching one of the GRAAL variants!
By the way, random entries (or by candle colour) show not bad results... Preparing my variant of martin for real.
I sincerely congratulate you on approaching one of the GRAAL variants!
By the way, random entries (or by candle colour) show not bad results... Preparing my variant of martin for real.
Thanks, but I am still at the very beginning of my journey and I think I need to work hard, because I have already come up with lots of ideas and variants before I even start. Now I will have to test them all before drawing any conclusions. ))