Why martingale EAs look profitable… until they aren’t (especially on gold) - page 4
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
Grid and martingale systems are easy and fast to code. They do not require a real market model and simply assume that markets range most of the time. The logic is basic: when price goes against you, add another trade. Because this behavior works often enought in renging conditions, the results look good quickly. It is the lazy solution: simple logic, fast developement, easy to sell.
Pyramiding is the oposite. It is hard to design and harder to test. It requires trend detecton, regime filtering, volatilty scaling, and strict risk control. It fails freuqently before it works, and the equity curve looks unattractive for long periods. That makes it difficult to market.
In short, martingale is popular because it is easy to build. Pyramiding is rare because it is hard to do right.It's a type of martingale, isn't it?
To be technically correct, no. Martingale was invented by a gambler playing a roulette wheel in a casino and it excluded decreasing the bet size following wins.
[O]nly you take the maximum risk immediately instead of increasing it as the price moves against you.
No, not "only," but that is merely one of the elements of pyramiding.
Wait, what if the price immediately goes against you? Accept a huge loss?
As a caveat, you need substantial capital, high leverage, and/or a nano (cent) account to trade in this way.
"Huge" is subjective and relative to this: ☝
You can see from the lot size sequence in Post #29 that the initial lot size is the largest. In an Egyptian Pyramid, the largest layer is the first layer laid. As with any strategy, you have to determine the best settings (including grid size), entry conditions, and exit conditions. Also as with any strategy, you will experience consecutive stopouts and must test and adjust to find the max consecutive stopouts that you're willing to accept. I should note that an initial fixed stop and a dynamic/trailing stop are assumed. If your entry conditions don't catch price runs in a way that is statistically significant during testing, then you need to rethink them. Remember, pyramiding is not a hail mary disaster recovery attempt--it's a cumulative reward for catching price runs (proportional to the quality of each run).
Again, the overarching concept is that it makes more sense to pyramid profitable price runs than it does to excavate losing price runs.
One more thing...
The simply stated purpose of a Martingale strategy is to refuse to accept any loss whatsoever. We all know how that goes. ⏱💣
One more thing...
The simply stated purpose of a Martingale strategy is to refuse to accept any loss whatsoever. We all know how that goes. ⏱💣
"Huge" is subjective and relative to this: ☝
No, if you use such a strategy (or intend to), then you probably know the size of the initial stop loss as a % of the balance.
[edit]
Yes, if the previous order is already in profit, you get an equity cushion for the next one. But if after the initial entry the price immediately goes against you, then you are in trouble (because, judging by the decreasing volumes, you most likely have a high risk in % of the balance).
[edit2]
The Martingale "lives" until the first prolonged trend against you without pullbacks. Your "anti-martingale" lives until the first series of several unsuccessful initial entries in a row. This is why I don't consider the "anti-martingale" to be less risky than the martingale or vice versa.
You're confusing a scaling-in feature with a Martingale strategy. In contrast to a Martingale strategy which is "blind", scaling-in can be incorporated into a strategy with proven entry conditions--if the original conditions remain valid, the underlying concept is adding to the position at a "discounted" price. Without those proven conditions, there is no statistical edge. Again, a gambler defined the Martingale strategy where it is not selectively applied.
On a related note... if you have to have a bipolar opposite of pyramiding, that opposite would more closely be scaling-in. They are both merely features/components.
The Martingale "lives" until the first prolonged trend against you without pullbacks. Your "anti-martingale" lives until the first series of several unsuccessful initial entries in a row. This is why I don't consider the "anti-martingale" to be less risky than the martingale or vice versa.
No, if you use such a strategy (or intend to), then you probably know the size of the initial stop loss as a % of the balance.
Also as with any strategy, you will experience consecutive stopouts and must test and adjust to find the max consecutive stopouts that you're willing to accept.
No, you don't evaluate based on "initial" anything. Again, you must test and adjust either via backtesting or demo testing. Only then will you have meaningful statistics on which to potentially trade.
Yes, if the previous order is already in profit, you get an equity cushion for the next one. But if after the initial entry the price immediately goes against you, then you are in trouble (because, judging by the decreasing volumes, you most likely have a high risk in % of the balance).
Again, nothing is based on "initial" anything. You must test and adjust and follow the statistics collected. Note the warning that I posted:
As a caveat, you need substantial capital, high leverage, and/or a nano (cent) account to trade in this way.
Obviously, pyramiding is not for small accounts because a small account won't be able to handle 10 standard lots let alone consecutive losses. Furthermore, adjusting the starting lot size down is at odds with the 33% reduction factor.
Your "anti-martingale" lives until the first series of several unsuccessful initial entries in a row. This is why I don't consider the "anti-martingale" to be less risky than the martingale or vice versa.
No. A Martingale strategy is a "blind" system unto itself. Assuming that you're actually posting about your experience with pyramiding, you have to test, adjust, and collect statistics.
Respectfully, we may have to agree to disagree as neither of us are allowed to post proof without code.
By the way, by martingale I meant a grid of orders. If you open a new position (with an increased volume) after closing the previous one at a loss, this is more likely a case of a series of losses in a row, rather than a protracted trend (since nothing prevents you from opening the next position in the opposite direction).
Got it. Right. That's the blind Martingale strategy.
Edit: I do wonder whether everyone is reading the thread before posting. 🤔
Obviously, pyramiding is not for small accounts because a small account won't be able to handle 10 standard lots let alone consecutive losses. Furthermore, adjusting the starting lot size down is at odds with the 33% reduction factor.
I don't see the point in discussing flat order volumes. Because they don't take into account your balance (and its changes over time, by the way). I believe that when you plan money management, you should operate with % of balance, not flat volumes. This way, you have strict and transparent rules that are always clearly interpreted.
Respectfully, we may have to agree to disagree as neither of us are allowed to post proof without code.
If you don't agree, then you should disagree. We're just having a conversation, after all. If everyone had the same opinion, there would be nothing to discuss.