Thanks, seen those date problems myself. Busy at present so no time to correct/upload yet just yet, will do so later.
For the Ocroft sheets they're both hard coded (entered), one looked at the first (standard) AIM and just tallied sequential trades and entered the same total $ trade size of all of the accumulated sequential trades in the month where the buying (selling) stopped (so yes Ocroft applied to both buys and sells). The other entered used the same spreadsheet, no reference to the first one and applied trades as/when the same direction trades stopped (i.e. maintained its own PC, #Shares etc. Of the two the former was the better, but that might very well be down to that particular sequence of share price motions being more favourable the first method over the second method. Looking at a plotted 'cash %' chart I see that the former rose to 90%+ and then dropped down to around -30% in one single dip i.e. was fortunate in its 'timing'
Having a model that works well in a dive/rebound is useful in that that model can be diversified with other models, potentially against the same stock, such that you have a portfolio of models better suited to a range of events (share price sequences). i.e. perhaps something like that first Ocroft model 50% weighted and buy and hold 50%
Clive.
