1. is it based on a reasoned, tested hypothesis?
Or
2. is it based on observation and testing of patterns, without actual understanding those patterns?
In other words: could the strategy be based on a pattern that occurs but there is no specific reason or mathematical basis for it?
Interesting question -- but I think what I re-worded is more of what you meant. Yes?
As a one-time tick-scalper, then almost 100% index credit spread writer, and now 70%+ a rules-based trend exploiter (call that, "algo trader" if you wish).... I go back and forth between wishing/needing to form an explicit hypothesis, and just not caring. It *might* be based on the time horizon. It could be that,
• in writing options, I need a hypothesis which suggests that the market will stay between a given high and a given low, for a period two to three weeks out. (I don't care what happens *within* that high and low, just that the market stay bounded, and outside of my short positions further down and higher above.)
• in exploiting a trend, I am only working a pattern, without {much} regard for hows or whys. I enter (long or short) on a shout, and exit on a whisper. (I enter on a multiple of strong indications; I exit on the first hint of trouble. In older terms, I let winners run, I cut losses early.)
Yeah yeah yeah. I like that. Option writing is time related -- you have a necessary market thesis that requires a given {range of} behavior. Trend-following only requires entry+exit rules. Otherwise, it's a hunt for new patterns, whether you can explain them or not. (If they are not robust, your exit rules will take you out.)