I would say that to create such a "super bot" would almost certainly require the kind of market knowledge that would make you competent at trading manually anyway.
Also, anyone who makes are really successful bot is very unlikely to give it away/sell it.
Therefore I don't think there is any getting around learning the basics.
The only difference with neural nets in particular is the skill comes in the design of the net and the selection of training data you give it. Why it works is of less consequence.
I would add that a bot is also a TOOL and like any tool results will vary depending on how people use/set it up.
Many people do automate trading as we see evidence of them everyday on the horse markets.
Now factor in the time,effort and money it took that person to:
a.)Learn to program to a sufficient level(try a few years)
b.)Learn to spot profitable market patterns and triggers, then convert them into machine identifiable form
c.)Test and finance the bot throughout it calibration period.
d.)Adjust its behaviour over the longer term as markets change.
Its clearly no small undertaking to do all this. Hats off to anyone who has and been successful.
What's the pay-off for this additional effort? Well besides the profit, they have served their time in front of the monitor every day in creating and commissioning the thing. It will require much less day to day interaction than manually trading. You could even make a smart phone app to login and check its performance/adjust parameters during the day when your out and about possibly in employment or just enjoying yourself instead of being glued to at the adverts.
Tash
