Trading is more a hobby for me. I spend spare time developing strategies and testing ideas. However I'm not much interested in trading them myself (ironically, my other business is a better place to invest my own money). The question is: what is the best way to optimally monetize on profitable models I periodically find? All purely statistical/mechanical, no snake oil or subjective reading whatsoever. Would appreciate advice.
You never have a profitable model unless it has a real money, real time and independently audited track record. Then you get to say you have a profitable model until then it's hypothetically profitable.
Segment your trading account; fund the segment with sufficient funds to work your idea; ask someone to trade it for you, according to the rules you establish; pay them a fixed fee; sleep well.
Good point. However, if something is purely mechanical and doesn't have hidden slippage/commission costs, it's as close to real as possible. You advice to trade something on a real account and acquire a track record before it has any value, correct?
Problem with most stats/quant models is you don't really know what's behind an edge you found. Without that knowledge it may suddenly stop working and you don't know why, which can lead to serious drawdowns. The real holy grail would be something I understand on a deep level from the very basic reason why price does X. Have yet to happen.
Unless your data has problem(s) or you fell for one of many biases inherent in backtests. It's a good first step, but I would start forward testing asap.
How would you monetize a forward tested strategy provided you have a better vehicle for your own capital?
Are you saying you would rather sell/pawn your ideas/strategies than trade them yourself? I don't know what your vehicle is but a strategy worth real money should scale far beyond what any other vehicles can offer (except illegal ones of course). The only way I would buy any strategy (track record or not) is in some kind of installment plan based on real money performance. I would want the strategy pay itself off.