A.I.M. Users Bulletin Board

[Conrad]

Hi Adam, you appear to still have some spare time on you holidays, so you schedule is not fully loaded up! Take all the time you need to dig into this issue!

I try to give you my take on this testing and you remarks:

Any of the spreadsheet methods is going to give very different results depending on the security selected to test on. Your security goes from 25 down to 13 and then rebounds to 50. That is a very fortuitous choice for our methods, which buy on the way down and is even great as for buy and hold depending on the time scale. Unfortunately it's not representative of most of the stocks/funds we deal with. And it does not address the problems that AIM has with some stocks.

I am not sure what you mean. Do you mean any two AIM spread sheets for different securities will react very different? Or two spread sheets with different methods will react differently on the same security? Either way I would expect that to be the case.

This testing started out with a couple of charts that Clive presented on a characteristic behavior for an AIM Account. The chart, as I understood it, presented an AIM-feature without respect to a specific fund (it shows how the “gears” of the “machine” work and how it determines the output”). See for example the chart I refer to:

http://investorshub.advfn.com/boards/read_msg.asp?message_id=28451186

I admit that I did not fully understand that AIM feature on Professor Clive’s Charts but I suspect that many AIMers did not understand it either . Clive's analysis goes to the heart of the AIM technique and addresses a theoretical aspect that escapes me as yet but the subject rekindled my interest on HOW AIM is fundamentally different from Vortex. . .I wondered how Clive’s Chart would look if Vortex was analyzed in the same way.

I know quite well how Vortex works but I could not really put my finger on it in relation to AIM from a fundamental consideration, and that triggered my curiosity still more: How would Vortex react different from AIM to an arbitrary price data set if I tune Vortex so that at the First Trades would be the same . . .(as a response to a +/- price change). . . with identical initial conditions being identical. That would at least be an interesting thing to discover. I suppose I compared it to a car race in which from the start signal both cars were tuned so that there would run neck-on-neck. . .same acceleration and then the question is how the would perform thereafter on the same track with all sorts of curves with different curvatures. Both cars would have the same power and respond with the identical acceleration on the straight starting stretch. How do the two cars take the curves????


The selected QQQ Price Data Set was provided by you, but the volatility was so low that I simply magnified the ups and downs to create great volatility. . .AIMers always like to talk about high volatility in that it is good for AIMing, so I thought I give the test some volatility to get AIMers talk about it

It worked!

As for the Comparative Testing a stock does not need to be any specific stock. . .certainly an arbitrary price behavior is as good as any, and the more volatility there is the better it is for a test such as I proposed. If you consider the other extreme. . .taking a stock that does nothing then AIM and Vortex would sit there waiting for Tom Veale to be 10 years old again. . .I did not want to wait that long. My conclusion was that a volatile Data Set would be perfect. If you like we can run this test with 10 different price sets of your choice.

So my take is the results don't really tell us much on how well these methods do with most real-life stocks/funds.

I disagree. First of all AIM has been operation since 1977 in its Lichello form and with a number of derivative. . .Judging by the functioning of this Forum AIM has done well generally over the years (much to the credit of smart users that know not to give AIM the freedom to make ALL the decisions. . .(for example when to bail out when the going gets rough or when to start buying when the herd starts to stampede). So if you asked me AIM and Vortex work fine in the right hands for typical stocks. I myself have done also quite well with highly volatile funds between 1986 and 1999.

That the testing we have done does not address" the problems that AIM has with some stocks" might be true but the Comparative Testing was not done to address "problems". . . I do not even know which problems you mean. What are they?

I used to see only one problem with AIM and I called it The Lichello Flaw: producing a Residual Buy Advice without having had a price change. I found THAT an extremely irritating feature and I kid you not, that "problem" was my primary reason to create an AIM Version that would give and Advice=0 after a trade was executed.

Anyway, I have said that to my satisfaction I have discovered. . .from a theoretical perspective. . .how AIM and Vortex are very different and that looking forward there is no easy way to predict the best parameter setting for Vortex unless I understands the "mechanics" of my system much better. This test has made me understand AIM as well as Vortex better, but I still do not get the meaning of Clive’s Charts

It would give me great satisfaction if AIMers have come to understand AIM better as a result of this Comparative Testing.