Replies to post #35932 on A.I.M. Users Bulletin Board (AIMUSERS)

[Conrad]

Interesting Issues Karw:

You need enough data to get a stable sigma, preferably data over all economic 'climates'.

I think this would be a reasonable basis for a "design" for a case in which a Data Set exhibits a repetition to the future. One can then use the metaphor "What is good for the goose is good for the gander" J

This then would approximately apply also to an optimisation. . .if there is a near repetition of the price history the optimised parameters would also be nearly optimum for the future.

If I an interested in a short history. . .which I am now. . . then I can at least notice if the current trend more or less maintains itself. . .If it does then the sigma should be ~stable.

I don’t see how the different days would impact this? In Vortex you could start with the average price as SP,
so you are then date independent.


Suppose I started SPY on 29-05-12 at a price of $ 133. Thus the Start PC/sh = 133
If I start on 04-06-12 PC/sh = 127,44
Sigma = for this set is 5,1

HZ 1 = 5,1/(133-5,1) = 0,0399. . . .4 %
HZ 2 = 5,1/(133+5,1) = 0,037. . . . 3,7 % Total = 7,7

For the next calculation I have removed the 133 price and added a new price on at the end 146,83(previous last Data Point);145,87 <----New Data so that the Data Set has the same number of points.
Sigma has increased to 5,28

HZ 1 = 5,28/(127,44-5,28) = 0,0432. . . .4,3 %
HZ 2 = 5,28/(127,44+5,28) = 0,0398. . . .4 % Total = 8,3

You may well be right that Average Price would be a good compromise.

With the method above I move the Optimisation Data Set forward in time with the new Portfolio. . .Hoping that with each optimisation I have a better chance that I keep the optimised parameters at reasonable values.

I just optimised the set for the new Data Set and I got these Hold Zone values:

Buy HZ = 2,1 % Previously I got 4% with 133 as starting price.
Sell HZ= 2,2 % Previously I got 13% with 133 as starting price.
I do not understand fully why the Sell HZ is reduced that much from 13 % to 2,2 %. Possibly the 2,2 % result is a lower local optimum point. I should try the optimisation in the neighbourhood of Sell HZ = 13. . .Just hold on. . .I do it right now. . . .tra la la. . . tra la la. . . .Done . . .it only took 20 minutes:-). At 13% it is no longer the optimum

At
S HZ 6%
B HZ 0%

There is here an Optmised point with a
ROI = 106% yr
B & H ROI =48 %
4 Trades

These results are not quite consistent.
Maybe I should rethink this and just pick the high one . . .hahahahaha.

For these latest HZ sets 2,1 % and 2,2 % the simple ROI for the short run = 39% :-)
Of SPY continues this way it looks good:-)

On the formula with the sigma and Starting Price PC/sh I am still scratching my head.

Time to get some dinner now! It is 9:35 PM
Regards,