Is there a profitable EA on mql or are they all scams? - page 3
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
I think there are a few good robots here and serious developers. Look for robots with a long tracking history.
An EA is just automation. Automation in and of itself does not guarantee profits or loss.
It's like a person or machine making a doughnut, the machine is not inherently going to make better or worse doughnuts just because it automated the task, but rather it depends on the instructions and ingredients.
An EA is just a set of instructions, then issue comes in where probably 90+% of EA's are programmed by unprofitable traders. You can't automate a task you fail at doing and then expect it to magically work.
There's also the issue of some EA's and systems do work, but only under specific parameters. Take something like mean reversion, there's a huge difference between trading on a 1-minute chart while looking back 60 bars, vs a daily chart that looks back 120 bars. One covers one hour, the other covers half a year. Both can be automated, by the same exact EA, but one method might work and the other might not.
TLDR: Are there profitable EA's on MQL Marketplace? Sure. Will most make you a multi-millionaire out of the box? Absolutely not. At best you might find an EA that can trade certain pairs and certain time frames, but an all-around magic money machine isn't something that exists. You can have a look at websites like [edit]Sets to get an idea of what's realistic, you'll quickly see that even a well optimized EA is only going to work for a few pairs, at best, and maybe one- or two-time frames. There's none that are just going to make you endless money with every asset, it has to be programmed to do the exact task at hand. Ex: the doughnut machine above can't be used to make burgers; and the EA built for EURUSD isn't going to trade Microsoft stock.
I agree with what James wrote. An EA is like a toaster, most likely the toaster is doing what it's supposed to do and it's not a scam. But you are the one who turns the dial and puts the bread in...
If you haven't done your due diligence in backtesting, and if you haven't tried real ticks backtesting, then you just basically scammed yourself. If the EA is following instructions correctly, it can't be a scam. More paraphrasing...you can't blame a calculator for giving incorrect results, because you are the one who put the formula into the calculator...
I agree with what James wrote. An EA is like a toaster, most likely the toaster is doing what it's supposed to do and it's not a scam. But you are the one who turns the dial and puts the bread in...
If you haven't done your due diligence in backtesting, and if you haven't tried real ticks backtesting, then you just basically scammed yourself. If the EA is following instructions correctly, it can't be a scam. More paraphrasing...you can't blame a calculator for giving incorrect results, because you are the one who put the formula into the calculator...
We can say for sure that if a system performance can be affected by differences of few ticks, proving it's robustness is almost impossible and beautiful backtests are nothing else than result of over fitting.
Also, 99% of people uses dukascopy ticks, but almost no one trades on that broker. This make all backtests made with these ticks not replicable in real time.
Working with prices of closed bars and not using Ask/Bid in the strategy (like trailing stop, just for naming one) means to have more robust systems by their nature.
You will also save huge amount of time because for testing you can totally trust 1 minute HLOC method and data provided by the broker you will use for real money trading.
Not all systems are ticks sensitive.
Well yes this is true, but for any EA that is built to trade on the trend for timeframes lower than M5, it will inevitably be tick sensitive (the trend is always less stable on the lower timeframes), and because you can only buy at ask price, and sell at a bid price. Even if you program it to buy or sell at a close price, it will still open the trade on the available tick price. OHLC is basically just a representation for the candles. The order book only processes bids and asks
Well yes this is true, but for any EA that is built to trade on the trend for timeframes lower than M5, it will inevitably be tick sensitive (the trend is always less stable on the lower timeframes), and because you can only buy at ask price, and sell at a bid price. Even if you program it to buy or sell at a close price, it will still open the trade on the available tick price. OHLC is basically just a representation for the candles. The order book only processes bids and asks
I agree with what James wrote. An EA is like a toaster, most likely the toaster is doing what it's supposed to do and it's not a scam. But you are the one who turns the dial and puts the bread in...
If you haven't done your due diligence in backtesting, and if you haven't tried real ticks backtesting, then you just basically scammed yourself. If the EA is following instructions correctly, it can't be a scam. More paraphrasing...you can't blame a calculator for giving incorrect results, because you are the one who put the formula into the calculator...
actually , with an overfitted model even testing on real ticks can be misleading (misleading concerning the outcome and also towards the author , not misleading in the sense of being a scam).
An overfitted model that the author of the product does not necesserily know it is overfitted
I believe we are seeing a lot of that over the years . Overfitted models whose authors are blindsided by the overfit
I mean its even possible to overfit with paper trading and testing it is not exclusive to neural nets or the genetic algorithm inside the tester.