Hi Karel, on the Vortex.
You understand the Vortex AIM probably better than anyone! I remember your discussion on a scalar vs a translation for the respective mathematical name-calling. If I read you right you see the differences probably in a deeper way than I could explain them right now. I appreciate that but the analogies such as memory, hurdles and turnstiles introduce dimensions that I can not place immediately in what happens in Vortex, but that OK.
I tend to think that the power of Vortex lies in the flexibility rather than in the fact that there is a little fundamental difference between the calculation of the Buy Order.
Also if I set the Sell Aggression Factor=0 the Vortex acts very similar to the Lichello AIM(No PC Update for Sells for one thing).
Now, the fact that for AIM BTB there remains the Residual Buy is an interesting difference which I interpret mostly as a necessity for a logical Buy/Sell System. Its operational effect is mathematically interesting but I have not fully been able to interpret its operational impact other than it causes a delay factor if the stock starts rising. Maybe you call it memory but that escapes me as yet. I achieve the necessary delay simply with an asymmetric Minimum Buy/Sell factor of an asymmetric split Resistances. In this way we achieve the same effect that a Stock Buy will be retained until it is profitable!
What is most important for me is that the performance of Vortex on the 6 month SYMC stock run for which the optimised AIM BTB achieves a 1,13% Loss while the optimised Vortex achieves 49% Gain. This Gain is due to the optimisation of only 4 parameters(since I did not use a SAFE), but only the minimum buy/sell(A limitation of the spread sheet!
The other point is that simply by choosing the parameters one can make the portfolio stock holding grow or shrink at will, while maximising the total value. I believe that this flexibility completely overshadows the small difference in calculation procedure of the Buy/Sell Orders.
In my book I address the option of decoupling the Buy Factor and the Aggression Factor so that the Vortex becomes even more flexible for automatic optimisation with at least 8 parameters. Such a system would be unmanageable to optimise manually, but also it would remove the requirement I have for PC=Y after every action. That would be OK. So, you see that I come around on that Residual Buy Issue: For the sake of a better optimisation result I would sacrifice the PC=Y demand, and certainly choosing a completely different Buy/Sell Function as I have now would then also be acceptable.
The reason for this is that the word Vortex has no significance in the Vortex AIM Structure. The fact that this AIM variant sucks up the dough lies at the root of choosing the name!
Now, money being the root of something is a different topic. Maybe that is your turf as well?
Conrad
Market Data 
Markets 
AI
