Duma,
thanks for clearing it up.
If I read it correctly, I'm impressed. It would also be easily verified.
But ... from a practical point of view, accounting for the mechanics of executing trades, it is very difficult to always get the opening price. For starters, I'll assume one has a broker where limit orders can be placed where the limit price is set at the opening price and not a specific preselected number (because you don't know the opening price ahead of time). A strict limit order to sell in the morning can easily go unfilled for the whole day because you will be asking for opening price or better, which would be equal-or-higher in the case of a sell order. The other type of limit order turns into a market order when triggered, so the order would surely get filled but very often at lower prices, maybe much lower.
At the least, the information you gave helps validate what I found and that it makes a great deal of sense that my system was modified to allow jumping the gun and buying on an opening partial gap prior to price crossing the trigger level.
But I'm very thorough about my backtests. So I also tested what happens if one jumps the gun and buys because of the PG, but later in the day finds that price did not close above the trigger. On average, over time that strategy looses money. So you should only do that, within the other limitations of my system, if you are sure that the Close will be above the Trigger level. Most of the time there is very little doubt -- price usually gaps for a reason that is not difficult to figure out. But for backtesting purposes, my software can't perform that reasoning, so that's why I incorporated the closeness test using ATR (which is the poor man's version of VIX, except it has the advantage that it can be easily calculated on any stock without performing mathematical gymnastics).
