Who got their money burnt by EAs? and why? What caused it?
Hybrid EA, Sniper EA, Super sniper EA, 90% winning EA, everyday there's a new EA with a name that would make any serious trader run away. Who picks these dumb names? This gives bad rap to MT5 EAs ecosystem and serious algo traders don't want to associate.
Anyway, whoever has tried EAs or currently trades with one, what are your results? Describe result simply as total net monthly profit vs actual max drawdown, potential worse case drawdown that could happen and equity exit point over a time horizon.
And if you blew up trading with an EA, what exactly did the EA do to cause that?
My observation: From what I notice, most don't understand algotrading ( including EA developers ). For example:
1: Talked to a dev and he said he cannot reveal his proprietary formula, asked about performance, turns out he doesn't trade. Both red flags. No matter how proprietary you go, it can only create marginal improvement. Think about it. If you switch out RSI for some secret oscillator, how much can you really improve as at the end of the day, you either buy or sell in a game of chances. So indicators are not some secret sauce, at the core of it and important but probably less than 20% of the picture.
2: Just like you send a kid to university, he passes exams, gets internship and then gets hired at a job. Same way, you train a bot on data ( university ), then you forward test ( OOS data - the exam ), then you run live demo ( the internship ) and then live trade ( gets hired )
3: Now, as a business, you hire 1 employee ( a bot ), do you expect this business to be successful or a business with 100 employees? You cannot succeed with an EA without diversification in terms of strategies, as well as assets. Creating portfolios is a must, that is how institutions do it. Understand correlations and volatility, not double down on Gold and yolo
4: A bot can fail, in fact it will fail at some point in time given a long enough horizon so all the optimisations and backtesting, forward walk is akin to weather forecasting. The best strategy can have a 20trade losing streak and an average strategy can have a 20trade winning streak and we don't know which will do what, and at what point? This is where calculated betting comes in, if you dont manage position sizing and exit points, you will eventually be blown by variance ( same goes for manual trading )
5: Most traders trying EAs are chasing a 100% return straight away. Just like manual trading, algotrading requires practice and patience. Was it Buffett who said that markets are a mechanism of wealth transfer from the impatient to the patient? If you run before you learn to walk, you will fall. Why? If you run small numbers for a year, for example, say you make 3% / month, now month 1, drawdown was 1%, m2 still 1%, m3 2%, m4 1%, m5 1.5%, m6 3%, m7 2.5%, m8 8%, m9 6%, m10 5%, m11 4.5%, m12 4%. Now, you ended the year with net profit since dd is 4% and you made 3% every month, you understood the variance of the drawdown, peak was 8%, you know your breakeven point and with all this, you have confidence in your approach. Now with this data and experience, you can scale numbers with higher odds of success because you know your numbers inside out so with this awareness you know that if you scale up these numbers and shoot for 20 or 30% / month, what risks are you exposed to and you can find opportunities to do this within windows. If you did this prematurely, you would've blown your account long before this point, if you did this on a prop account, you would've failed the challenge.
Bottomline: Correct position sizing can only come from live trading performance once you know your numbers over a time period. If you oversize prematurely, no strategy will work, manual or auto. Sizing correctly for a long enough period is crucial to eventually scale position sizes and accounts, provided edge is there, and it is executed with consistency.