How to spot
FAKE EAs
Introduction
Picking an EA to buy from the mql5 market is no easy matter, especially considering there are several 'tricks' savvy developers can employ to make backtests look amazing or generally deceive unsuspecting buyers. I've outlined a series of questions and answers you can employ to help navigate the sea of EAs on the marketplace.
1. Do they ask you to backtest using a particular modelling quality (particularly 'Every Tick')?
If the author requests you only backtest using the less precise options of 'Every Tick' or '1min OHLC', rather than 'Every Tick Based on Real Ticks', they are almost definitely hiding something. Any developer who requests you do not backtest using the 'Basedon Real Ticks' option has created a fast scalping-style EA which tricks the backtester that quick, safe profits are made from their EA. I too have accidentally created EAs like this - thought I was on to a winner, then realised it only works with these imprecise modelling qualities. The culprit of successful tests like these are very small take profits. Most commonly seen in scalpers or bots where the take profit is less than 3pips. In real-money live trading conditions these bots do not work.
2. Do they ask you only to backtest during a specific time period?
If the author states 'test from between XY date and YZ date only, they are likely hiding something. The only exception is the middle of 2017, when tighter forex regulation changes were implemented. Many bots don't work before mid 2017, and this may be because of the insider trading scandals associated with forex prior to this. Some bots have worked great from 2010, 2005 and earlier, but this is rare. It's reassuring if this is found, but the only other significant date is mid 2017. If the author states 'backtest from '2020.05.04' to current, they are hiding all trades before then. This is a very blatant lie. If the bot has NOT worked from at least mid 2017 / Jan 2018, it isn't going to be worth anything in the future in all likelihood.
3. Does their signal start AFTER the product was released?
Usually the signal for a product is started BEFORE the product's release to show its efficacy. Signals can be started and restarted indefinitely. If the author launched the product three months ago, but the only signal they link to on their sales page is 2 weeks old, for example, it's likely the bot took a large loss after launch, and the author has deleted the trade from the backtest and restarted the signal, to present it as 'working flawlessly', while you aren't able to see the evidence that the bot already had a significant failure and the author has erased it from history.
4. Do they have a signal? If so, which broker are they using?
Why would an author, confident of their bot, not have a live signal running? I can't think of one good reason! I can only think of one obvious one; their bot probably doesn't work!
Also if they do have a signal, which broker are they using? If you click on the name of the broker under the signal, and NO OTHER SIGNAL exists on mql5 using that broker, I'd be very suspicious! Are they using a commonly / widely used and accessible retail broker? If not, why not?
5. Do backtests look 'too good to be true'?
Let's gain some perspective. We're all secretly hoping someone random developer has found the 'holy grail' of EAs. Too many of us are secretly hoping we can 'get rich quick'. As time and time again tells us, this is not possible!
It is possible to get a relatively clean backtest from an EA, but sometimes they appear 'too good to be true'.
The tell-tale sign of this is not a very high win-rate, but the 'symmetry of all trades'. If all trades appear IDENTICAL in the backtest, it's more than likely that history reading is taking place, as in, the EA does not work.
Long winning streaks are acceptable, but if there is no 'mess' at all to the backtest's drawdown, if every trade is BEAUTIFULLY CLEAN, without any 'messiness' to the green equity line, it's very likely the bot is just a history reader.
Some bots are coded more efficiently than others, but I would also be hesitant if the backtest runs more than 2X / 3X faster than most other EAs. This could either represent efficient coding, or it could represent that there's no real logic happening within the code / that it's just history reading.
In my experience, I've not found or seen a legitimate strategy that can comfortably make more than 300% in a year (on its own - EAs that combines multiple strategies are different / can perform better), and even when it does, it is usually operating on a very, very risky setting.
Remember - hedge funds, professionals, managing trillions of dollars for the biggest financial institutions in the world, seek to return 20%pa, with drawdowns up to 25%. And they charge huge fees for this! Any bot that can outperform hedge funds by a factor of 2 or 3 is a really good EA (e.g. 60%pa with 25%DD) .
We can always hope to find a bot that can make 1000% in a year with 10%DD, but just think - if someone really had developed a bot that outperforms everything else that's ever been seen, the internet would break. This has, to my knowledge, NEVER happened before! Please remember this!
In my humble experience, a good benchmark is a bot that can make 100%pa (non-compounded) with 30%dd. This is a good bot.
I would be highly suspicious of any bot that claims to comfortably make more than this (e.g. 500%pa with 15%DD with one single strategy).
Conclusion
Identifying and avoiding manipulated Expert Advisors on the mql5.com market is essential to protect your investments. By thinking about the questions in this article, conducting thorough research, scrutinizing developer profiles, and reviewing user feedback, traders can significantly reduce the risk of falling victim to fake EAs. Remember, forex trading is already challenging enough; don't let other people further complicate your trading journey. Staying vigilant and making informed decisions is the key to success in this dynamic landscape. Happy trading!
My EAs
Check out my new EA, The Ultimate Bot! It works across all brokers, uses incredible risk-management and is highly profitable!
*NEW* Click here for the Ultimate Bot MT5!*
Click here for Prosperity MT5
Click here for Prosperity MT4
Click here for EA Drawdown Limiter MT5
Click here for EA Drawdown Limiter MT4
Thank you for taking the time to read this carefully. I wish you well on your journey!