In the tests, the minimum decision time is 1 bar. Since a bar in these tests is about 30 seconds, it follows that the shortest decision time is every 30 seconds. But as said: one can decrease the bar time upto 1-sec, but then things take very long. From a mathematical point of view it should not make much difference. OTOH: the shorter the timeframe, the more is possible. Each bot trades independently of the others, but they report after each bar their result. The master (ie. system) coordinates the risks of the bots. The bots reach their targets in about 5 days on average. This is highly dependent on the input params. Of course the shorter this time is, the better it is.
In GBM harsh market conditions occur as well, and I spent much time to find a solution for this problem. Finally I solved it. But the difference is: it was single tickers who fall big, not all tickers at the same time, ie the whole market. I would not offer this system if I only had the play money myself. So I need to find capital. I thought this offer could be a good solution to the problem.
I don't know what other context to take your statement. Unfortunately you are convinced that your system is not flawed. Good luck finding someone to lease it. More likely you will waste a lot of your time on something obviously wrong.
What's the maximum days for a trade? 8 days ? 10 days? Perhaps a distribution chart would be useful. When a system can make decisions every 30-second, that means many of 30-second trade is possible. I would think the maximum days for a trade must be very large, in order to produce an average 5 days per trade.
Sorry, I can't give you a satisfactory answer. I just know it is very well possible. According to your logic there should be not a single entity that makes some profits, ie. an equilibrum in the markets, but this is not the case as you know. Maybe take this example: you have say 50 tickers in your basket, each deviates up and down say 2% daily. Now imagine you can capture say 1% every x days or so from most of them...
I'll try to make such a distribution table/chart, but will take some time, maybe tomorrow. True, but an instrument doesn't necessarily move much in that time. Not necessarily. If it really takes longer, then it closes this pos and tries another ticker (BUT: it carries over the losses to the new pos... As said: it tries to detect and avoid flat markets, ie. if a ticker does not make any movements over a period.