Johnathan: I fixed the testing multiple stocks. You don't need the Pro thing for that. With applying the template I can make the results visual (thanks to your help)

I don't like retro fitting (by optimizing) a system to a particular stock. Stock graphics are no fractals that repeat themselves over and over again. So I find tailoring an indicator to particular stock or a particular time frame a waste of time and a matter of fooling myself: the past will never be repeated the same way in the future.

Riding the trend with moving averages is producing the same results with optimization as best, but you can't optimize the future. There is tot much lag effect and there are to many losses in sideways moving markets. My system works with fixed parameters. The same shortcoming is resting on the VST system: nice that you can back test 6x6xYxZxX different situations, but you have to make your buy decision on one!

How do you work: fitting every stock with a separate set of parameters and trade with that?

@Toblakay: you can upload though Image Schack or service like that and paste the HTML code after clicking The HTML icon is the post window. You can also let Metastock export a screen as a webpage in HTML

your remark: "............. we would all have had the perfect automated system a long time ago." means: we are all a bunch of alchemists looking for the stone that turns led into gold?

Remaining are perceptions and emotions?

Anton,

>>I don't like retro fitting (by optimizing) a system to a particular stock.

Normally, I'd have to agree. However, given the following claims:

* Automatically and instantly finds optimal Long and Short strategies for any chart, based on 90 VST strategy combinations.
* Displays optimal Market Trend Filter type, as well as optimal entry and exit conditions.
* Unique optimization robustness factor (based on three parameters) adjusts optimal results for any curve-fitting.

The VSTopt module adds intelligent signal optimization for 90 VSTpro band combinations in each security.
For more robust optimal results, VSTopt optimizes signals based on the Profit/Risk ratio rather than the traditional profit-only method which ignores risk. VSTopt's unique method also incorporates sa[censored]uards to geatly reduce the curve-fitting problem common to other optimization processes.

Being familiar with Jose's work over the years, so if he says it's worth it -- and has gone to the trouble of coding it -- that weighs very heavily, in my book.

Some others in the private VSTpro forum have tested it with good result. I may try it in the future, depending on my tests -- after I finish current projects.

Anton,

>>If you need 90 combination to try before optimizing the retro fit, how much garantuee do we have that it also will "future-o" fit?

Um, I suppose the same guarantee that a high-Beta stock will continue to follow the trend of a market average, and a low-Beta will not.


>> Because I can only make money with future quote data and trends and not with past trends.

That's not quite accurate -- if a stock tends to high volatility, historically, then a trading method based on volatility -- Bollinger Bands, Envelopes, VST volatility bands, etc. -- should capture gains.


>>You might try the VST with tailored parameters in the systemtester...You should be able to test the hypothesis above in one evening.

Not the way I'd do it<G>. To avoid hindsight bias, it would be a bit of work.

Since the VSTopt is customized to each security, first you have to pick a test "universe" -- say, the DJ30. Then decide whether to use MTF. Next, choose duration and past start date (say, MTF turning points). Delete data AFTER that date, to get proper settings for each security at the start date. Re-download data and record trades as they happen, using customized indicator settings for each chart, until the next MTF downturn. At the next upturn, repeat the process.

Given the dynamic nature of the VSTopt module, I see no way to do the above in the System Tester, TradeSim or VSTpro. Maybe not quite as statistically rigorous as Aronson, but should provide reasonable confidence.

By the way: Aronson says with I said: just test and compare enough stocks and different periodes with your system and compare it with the standard to beat: buy and hold or "riding the waves"
Dare to make a comparison as I did in the spreadsheet: testing with 50 stocks for 25 or more periods should provide consistently better results or it;s worthless. I am curious to the Monte Carlo routine but don't know where tot look for it.

This review I found covering must remarks in this discussion thread (even the title of this tread "Alchemy" and the relation with gold at the end of this review fits)


My experience with "technical" trading books has been that they are
either pseudo science of the quasi-astrological variety (eg, wave
theory or fibonacci hocus-pocus) or entertaining and vague
reminiscences of some successful trader(s). The "serious" literature,
instead, is written in statistical/mathematical formalisms which you
may be able to understand given an adequate background, but even then
serious effort and time is required and the sheer distance from the
formalisms to the pits is fraught.

Professor Aronson has done an admirable and unique job of dispensing
with the hocus-pocus and applying the requisite rigor to the many
difficulties associated with analyzing the results of historical
backtesting and datamining. His bibliography seemingly covers all
serious (and some not so) references of interest to the algorithmic
trader. Months after having consumed the book, I still refer to the
bibliography to glean interesting sources.

As an earlier reviewer noted, the early portion of the book in which
he debunks TA and establishes basic statistical literacy for later
chapters may or may not be of interest to many readers, but his
chapters five and six are genuine contributions to the field and
should be read, studied and implemented by anyone looking to derive
tradable strategies based on historical testing.

Anyone who has spent time attempting to optimize parameters or
otherwise evince trading rules from historical data will have quickly
learned that if you look at enough cases, you will always find
something that looks good but (alas!) really isn't. Fool's gold,
indeed! This is the first trading book which directly addresses this
issue in a rigorous yet accessible fashion.

This is not the book for someone who is looking for the magical key to
a kingdom of riches (I recommend Voltaire's "Candide"). But if you
are an algorithmic trader looking to get a statistical grounding and
some concrete methods for assessing your backtested results, there's
simply no better book available.

I am not smart enough to grab the point, or if I think if I have grabed it, I would have drawn it differently. But that's maybe because I am distorted used to viewing X Ray's, You always image an object from two (perpendicular planed) sides, to get a a good idea about the whole (3D) picture. :-)

(You managed to get the picture here [:D])

Oops, wasn't ment a an insult.

In Dutch we say "meten is weten" translating "measuring is knowing" That what this discussion was about.

>>You managed to get the picture here
Now, THAT pun is so bad, my funny-bone hurts...

On your last remark: I fully agree.

Thoblakai: thank for your input but I am afraid you would have missed a nice ride. :-)

What should I have done at point two?

(my point is: you can wait everyday but never make really reliable decisions unless you do it at a wide scale and know the probabilities of each choice. At 2 you know that you are in an upgoing trend but you still don't know what tomorrow will bring: continuation or a turnaround) But I have a continuation image that shows what really happened after 2 if you are interested.