So many responses and everyone misses the point. It doesn't matter if you can identify the causation as long as you can find a correlation that has enough signficance to trade. Obviously a sample size of 6 is much different than 60 or 600. I'll trade a black box any day, as long as I have enough data to fully evaluate significance, robustness, risk vs. volatility, etc. Of the thousands of posts I've read on ET, from tired old cliche that no one actually tests; to blind guru faith thrown at ppl like LBR whose great strategies sell books but have never tested to show an edge (and many a quant have tried variations of her ideas); to all this psyco-babble talk about psychology. This game is very simple. 1. Find an edge. 2. Trade it with % of bankroll to satisfy your risk criteria (e.g. usually 1-2% risk per trade) Compound that edge and live the good life in a few years.
Bwwwwwwwwwaaaaaaaaaaahhhhhhhhhhhh the kind of thread I feel like to make a long answer so I will prevent from doing it maybe this week end rather