If you combine several stocks into one portfolio then you essentially create an index, right? This index will behave quite different than anyone of the stocks and has also lower volatility than any one of the stocks. What do you gain?
I presume, of course, that with stock selection you pick stock that you feel comfortable with for AIMing anyway. I would AIM & optimise each stock separately. This is what I do in the Vortex-AIM. I make up various portfolio's and each stock is treated separately. Thus also the optimisation would be done separately(manually or automatically).
On top of that one might look at the portfolio and use some sort of stock analysis method to optimise the relative weights that each stock should be assigned. I can imagine that for this various types of optimisation objectives could be defined. I think Mark Hing's idea of combining AIM with the MPT is an example of this.
In this case tweaking the portfolio parameters would then be quite apart form tweaking the AIM parameters.
That's what I see as an objective. The question now is: What types of different Portfolio Optimisation objectives can we define to optimise the portfolio, other than the MPT as a start?
I would appreciate some handles on portfolio optimization ideas. For example I can think of weight optimization on risk evaluation so that the high-risk stocks are reduced with respect to a percentage of the total.