AIM Testing Results
I have made a short list of the AIM Testing Results I have seen
so far. It is impossible to list all the data here...this list
below took me more than an hour to make it presentable.
I suggest that later on I send the Excel Matrix to you by e-mail
if you want to get it.
At this time the results of Egro Sum are the highest. For many
results a lot of data is not listed as I did not see it on the
Board as the data came in.
My personal preliminary conclusions are:
1) With the many variables its difficult to find the highest
peak in the multi-dimensional coordinate system(N-Space). There
are several peaks in such space and without going through all
the realistic settings one can not be sure the best combination
of parameter settings is achieved.
2) The Yield isvery, very sensitive to all parameter
settings. An infinitesimal change on for example Buy/Sell
Multiplier can make the Yield jump down (or up) very
drastically. This is caused by the fact that the Buy/Sell
Resistance(or SAFE) are not continuously variable, and as the
Buy Advice crosses the threshold-value it can throw the End
Result out of whack, if the other parameters are not optimum
for the new calculation results...It throws the end-result into
another place in the N-Space. This makes manual optimisation
very tedious.
3) Because of (3) it is not certain that the lower results are inferior to the higher results. One would need to test the full
range of parameter settings to find out any fundamental difference.
4) Very interesting is the fact that there are two extremes with
respect to cash-equity distribution. Some AIMs have
maximised the cash while others have maximised the equity!
This could be interesting information with respect to what the
investor wants to achieve: Equity Growth or Equity
Liquidation? In this Test I believe that the distribution
that we see is accidental, as none of us, I think set
out to maximize either the cash or the equity.
This cash/equity distribution is also important in relation to
how the share price is developing. If the stock price starts to
drop again after week 85 then a cash-risch AIM will be a good
thing for you….If you believe the stock is sound! If the price
starts rising again then you will be better off with an equity-
rich AIM. So, my conclusions is that with the stock price
running at an average even keel of 100 the best
cash/equity ratio to AIM for is the 50/50 that we started out
with. Being an idealist(on top of being a pragmatist) I would
also venture to say that if you can manage to control the cash/equity ratio to follow the Idiot Wave then you might
have the best parameters settings that you can hope for.
5) At this time I would conclude that aside from the fact that
for any price distribution we can find an optimum set of parameters, it is very difficult to find the best set parameters
for a particular stock pattern. This would make it even more
difficult to find a good set of parameters for a stock
with an unknown pattern and unknown volatility(as was more or
less already expected).
6) In order to obtain useful information from this testing the
data needs to be studied more carefully, taking into
consideration all the parameter settings. But then we would
need to optimise the testing so that we are sure that we have
the highest peak in the N-Space. Such information would
disclose specific difference between the AIM variants that are
presented in this test.
Short List AIM Test Results
Data Cash 50/50
Capital 20.000
Interest on cash 0,05 Compounded Weekly
Trading Cost
Fixed Fee 20
Commission 0,006
Interest on cash 0,050
Initial Trade Cost 80
Equity to start 10.000
Netto Cash to start 9.920
Name Value PC(W=85) ROI Equity Cash Trade Interest
Cost
AIM Guy 22951 9,03 8151 22951
Vortex 1 28040 8392 24,59 20743 7297 802 961
Vortex 2 28696 10988 26,60 18615 10081 567 831
Vortex 6 25350 17278 16,37 3712 21638 346 346
Lost Cowboy 30143 31,03 Synchrovest Experiment
AXP 1 24847 14,83 At 30% Cash
AXP 2 At 50% cash?
Pragmatico 25835 17,85 5154 20681
ICT 32232 37,42 17456 14776
Ergo Sum 32610 50,68 32170 4400
Labestul SA 24309 15793 13,18 9309 15000
Labestul TA 24367 17136 13,36 7767 16600
Labestul HY 24150 15907 12,69 9150 15000
DonC1 25464 16,71 7406 18058
DonC2 24964 15,18 6519 18445
DonC3 30836 33,15 Synchrovest Experiment
If those of you that have submitted your testrun data want to
add the parameter definitions and values in the Result Matrix
the I invite you to enter then on this Board or e-mail them to
me: eng@vortex,demon.nl
In order for me to process this data properly give me the name
pf the parameters and what it does...(23,5,6,34,78,...) means
nothing to me.
Regards,
Conrad
Conrad
AI
