I’ve asked myself this question about 1,5 years ago. So - as I don’t believe in backtests - I put around 105 different EAs on demo accounts and made my own statistics.
These are my results:
1.) there are indeed tradingbots that do profitable work.
2.) nevertheless the majority of the 105 are not working, at least not constantly
3.) There are technical solutions in order to stop the trading of specific bots in certain market conditions (e.g. news stops or volatility stops) - so setup and forget is truely a thing.
4.) big advantage is that your emotions are definitely kept out of trading.
5.) big danger is to install too many bots at the same account - so the emotion „greed“ is a thing. Even with automated trading.
Conclusion: you have to make very precise statistics about your EAs, most of all about possible drawdowns and trading length in order not to overtrade your account.
If you really stick to your numbers, it’s absolutely realistic to setup (multiple) accounts that won’t make 100% per month, but something between 10 and 50%. But it’s really hard work that you have to do before on valid statistics.
Successful automated trading is realistic, but it‘s long term and hard work.