|
PureBytes Links
Trading Reference Links
|
Davidd,
On Sat, 18 Oct 1997 13:13:29 -0400, you wrote:
>By "frequency effect" I assume you're refering to the exaggerated effect
>of re-investing when you buy on margin. As far as I can tell, margin is
>handled by the system tester and is not unique to the MS' MPS. With
>commission set to zero and margin set to 50, both the MPS test and the
>formula I posted get the same results. I've only experimented with this
>on a few charts, but when I look at the report details, both tests only
>seem to re-invest from trade to trade. Neither test appears to take into
>account the possibility of selling and re-investing half-way through a
>long move.
Imo, this (trading) "frequency effect" is independent from any margin.
So it also works with margin=100. It is a general effect that can be
very important in a true Maximum Profit System and also in practical
trading situations. (I have seen a paper on this topic from last
year's european conf on "intelligent computing". The solution for a
MPS presented there really was VERY complex. So I am interested to
know, what MS has implemented.)
>Maybe someone from Equis can jump in here and give a definitive answer
>as to how the Maximum Profit System test works.
That really would be very helpful, if Equis could give some detailed
description of the MSP implemented in MetaStock.
>20/20 refers to a standard eye test in which someone has 20/20 vision
>(clear or "perfect" eyesight.) I don't know the exact specifications,
>but it is something on the order of - from 20 feet you can read 20 point
>type. In other words, "clear, almost perfect hindsight."
Thank you for these hints! (I had interpreted "20/20 hindsight" as
something like the knowledge of prices for the next upcoming 20 days.
If "20/20 hindsight" is an unrestricted view of the upcoming prices,
MPS things are very much more complicated.)
mfg rudolf stricker
|
| Disclaimer: The views of this user are strictly his own.
|