Backtesting manipulation, not easy for the buyers to identify before buying
Please search 🔍 before you post.
Forum on trading, automated trading systems and testing trading strategies
Fernando Carreiro, 2022.06.07 14:40
Be it by fake agents, brokers, signals, or EAs, there is one weapon to beat them all — knowledge! Your knowledge is the key to prevent you from being scammed.
Invest and take time to improve your knowledge, to gain experience and improve your skills. Learn to do things for yourself and not to rely on others so much. You don’t have to be a master or even be very good at it, but the more you know and understand, the easier it will be for you to detect when something is not what it seems to be.
Learn how brokers work and how they are licensed and regulated. Learn about the different markets, trading conditions, types of accounts, leverage, and when you don’t understand something, ask first about it and get answers from different sources. Don’t just blindly accept things.
Learn to trade manually, even if you intend to use signals or EAs. You don’t have to be very good at it, but you need to understand all the different concepts of how orders are placed, how they are triggered, how to manage them, how to calculate and manage your risk, and all those pesky little details required to trade manually. Only then can you better evaluate the signals you wish to follow or the EAs you wish to use.
If you are going to hire someone to code your EA for you, then take some time to understand how coding works and some basic knowledge about it. Again, you don’t have to be good at it, but the more you understand, the easier it will be for you to communicate with the developer and provide valid, realistic requirement specifications and to understand how much work is involved as well as its value and how much to budget.
Also, learn to use the Strategy Tester properly and how to interpret the results. The better you understand it, and the more you use it, the easier it will be for you to evaluate the results from EAs you are considering.
So, in conclusion, don’t be a greedy idiot quick to jump onto anything “shiny”. Instead, be wise. Take your time. Invest in your knowledge and skills. Choose carefully.
Please consider which section is most appropriate — https://www.mql5.com/en/forum/172166/page6#comment_49114893
you as buyer can't know in advance whether the EA is good or not.
Backtesting will only show what the seller wants to share with you.
But if you try the EA about two weeks before purchasing in a demo account,
then you will have all the information and you can purchase or decline.
the sellers are always satisfied because as a buyer you no longer have to get
your moneyback
Hello everyone,
this is a survey because many sellers are willing to cheat buyers by manipulating backtesting
and we as buyers lose a lot of money when buying EA.
what can we do to prevent fraudsters?
- I think that if the desired EA is testing for two weeks in a demo account before buying,
there will be no manipulation in that case.
IMHO, only 2 weeks is a bit presumptuous. About 80% of the EA's posted in the Market are one form or another of a Martingale strategy--originally designed for a roulette wheel in a casino. In fact, I recently coded an anti-Martingale Strategy at the request of another trader, and it blew up his demo account even though it was stacking trades in profit instead of loss. Any Martingale style strategy could blow an account in 2 years or even 5 years. It's basically a ticking time bomb.
Be cautious of any EA that had a sister Signal in the Market that is now removed from the Market. Although the optimal trading results (fixed in time) will remain posted on the EA page in this case, the Signal likely went bust in real-time--causing the Seller to completely erase it from the Market. So, what are we buying if we buy that EA?
Another thing to consider is that few, if any, EA's are set-and-forget. A strategy designed with the USDJPY pair in mind is probably not going to work well on the GBPJPY pair for example. This is without even considering other markets, broker-dealers, timeframes, etc. Then we have to optimize the EA settings periodically. Having said that, there is such a thing known as "overfitting" with machine learning/AI EA's which is not profitable.
Therefore, look for an EA that:
- Trades only one position at a time,
- Has a trade history of 10 years or more,
- Has never had a sister Signal removed,
- Has been traded in your market,
- Has been traded on your symbol,
- Has been traded with your broker-dealer, and
- You have profitably tested on 10 years or more of data (including periodic optimizing).
IMHO, only 2 weeks is a bit presumptuous. About 80% of the EA's posted in the Market are one form or another of a Martingale strategy--originally designed for a roulette wheel in a casino. In fact, I recently coded an anti-Martingale Strategy at the request of another trader, and it blew up his demo account even though it was stacking trades in profit instead of loss. Any Martingale style strategy could blow an account in 2 years or even 5 years. It's basically a ticking time bomb.
Be cautious of any EA that had a sister Signal in the Market that is now removed from the Market. Although the optimal trading results (fixed in time) will remain posted on the EA page in this case, the Signal likely went bust in real-time--causing the Seller to completely erase it from the Market. So, what are we buying if we buy that EA?
Another thing to consider is that few, if any, EA's are set-and-forget. A strategy designed with the USDJPY pair in mind is probably not going to work well on the GBPJPY pair for example. This is without even considering other markets, broker-dealers, timeframes, etc. Then we have to optimize the EA settings periodically. Having said that, there is such a thing known as "overfitting" with machine learning/AI EA's which is not profitable.
Therefore, look for an EA that:
- Trades only one position at a time,
- Has a trade history of 10 years or more,
- Has never had a sister Signal removed,
- Has been traded in your market,
- Has been traded on your symbol,
- Has been traded with your broker-dealer, and
- You have profitably tested on 10 years or more of data (including periodic optimizing).
I think that those are very good targets when looking for an ea, however, the backtest, rarely shows a drawdown period. Number 8 should be "Proof of an equity stop feature used in any backetest. If equity dips to XX%, then, the tests is ended". Most eas that backtest well, will NOT when you add an equity stop to the code.
First the scams were tick volatility scalpers that exploited the default price data of MT4
then there was EAs that stored optimal trading times within the executable so the backtests were flawless
Nowadays we have martingales that are overfitted to show very good results in backtest, and pretty good in forward test (because of the martingale aspect, also they use very high lots on small accounts to match the incredible returns, if one account blows, they start another, or they could even have 5-10 different EAs on demo accounts so that they always have at least one "working" that they can link to)
it's all scams
rule of thumb is :
if it's too good to be true then it's probably is
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Hello everyone,
this is a survey because many sellers are willing to cheat buyers by manipulating backtesting
and we as buyers lose a lot of money when buying EA.
what can we do to prevent fraudsters?
- I think that if the desired EA is testing for two weeks in a demo account before buying,
there will be no manipulation in that case.