That's all for him to figure out.Automation of the process of backtesting rule-based trades upon price fluctuations, for such purposes effectively random data. The key is to data-fit to transient clustering of events in data sets small enough to not completely eliminate randomness. I could see how somebody that does not understand this might have the expectation of some predictive outcome.:eek:
Automating this process is indeed a more efficient way to arrive at a disastrous conclusion. What could possibly go wrong?
