Quote from Newmoney24:
who can manually do a quality strategy manually though besides manually being available 24/7?
at the least, some basic back testing should be in line (either independently or in line w/ the team)
Running a team here.
It is not "trading it", it is validating it. Programmer time is valuable and - programming you do not SEE The markets. Watching it for some time - even in replay, even only part of the day - gives the strategy designer ideas how to trade it better.
Backtests and optimizations do not show a lot of things. You focus - time wise - on results and implementation, not thinking how the strategy behaves. Manually running the strat, manually validating the entries, means you focus on the strat and the market details, possibly seeing new ideas.
It is not "have an idea, spend a month programming, money", it is "have an idea, validate it, program, valdiate, spend serious time looking whether everything makes sense or it was just a bad idea".
Not saying backtests have no value - damn, I have close to 200 cores permanently in action doing them now, and right as I speak there are 26 jobs queued.... but: They come quite at the end of the work progress. First you check whether the concept at least theoretically makes any sense. Manually.