A great book on testing and optimisation - page 2
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
It would be nice if at least 5% of the forum members would not just download this book, but also read it... This could significantly improve the quality of testing/optimisation compared to what we see now. Otherwise, I'm sick and tired of these endless cardboard grails and the bewilderment of their authors taking offence at the results in real life... .
Thanks for the link! I'll definitely read it!
Mark Conway, Aaron Behle: Professional Stock Trading: System Design and Automation. http://www.amazon.com/Professional-Stock-Trading-System-Automation/dp/0971853649
Looking for a book in electronic form, can anyone suggest one?
Mark Conway, Aaron Behle: Professional Stock Trading: System Design and Automation. http://www.amazon.com/Professional-Stock-Trading-System-Automation/dp/0971853649
Conway and Bele's book is very - perhaps even overdone! - is full of technical details describing several specific systems. The whole code is EasyLanguage, although it is explained in quite some detail. The instruments are stocks, and their TA is still different from that of currency pairs. I did not find any recommendations about testing and optimization comparable to what is described in Pardo's book. The book could be interesting for a professional who expects 5-10% a month on the stock market with a starting capital of about $100K.
The link doesn't work, how can I download the book?
http://bitzi.com/lookup/CFMSHFSFKLFF22H7QB4GYYAMTFEG6SKL
The whole procedure of installation and downloading took 20-30 minutes.
chv, what's wrong with the direct link I gave? It's working right now. It saves as usual: you right click and choose the save directory. But it's really not the book I was talking about in the beginning...
2Mathemat:
Please help me understand the panacea of the forward test. Here I am reading:
Page 26 (in Deja Vu 22) last paragraph: "It can definitely be said that a weak or tweaked strategy will not show a profit in forward analysis."- Where does this conclusion come from? How are we suddenly deprived of the ability to hit the tweaking parameters for the whole population? Is there anything other than chance?
Page 133 (in Deja Vu 75) 3rd paragraph: "The main purpose of the forward test is to determine whether the performance of the trading model being optimised is real or whether it is the result of a tweak." - That's how a forward test does it? See, the author on the same page 26 practically stipulates: "Of course, a forward test is not real trading, but it is a much more realistic simulation." Imitation of what? Trading? Doubtful, because the purpose of the forward test is specified above, and a less realistic trading is not the problem that the forward test solves. What about the simulation of finding parameters with predictive power? - In my opinion, yes. This is what I firmly believe so far - the forward test is a profanation on the subject of predictive power of parameters. Please push me from my position with a more convincing argument than, since we have found parameters that have performed well on a series of forward tests (of the past), then by analogy, they should perform well in the future. Or is that the whole argument?
Suppose, according to the methodology, we have found the cherished right parameters with "predictive value" etc. However, at the same time, an unsophisticated optimiser did the same thing without the forward test, using the same data as we did, but in the aggregate. He, of course, got the wrong, tweaked parameters. So I have a question for you, what assures us that our parameters are not the same as the unsophisticated optimizer's?
Mathemat, share your thoughts on what inspires confidence in your forward test? If so, of course.
P.S. P.3 on page 34 (in deja vu 26) about reasons for systems failure is hilarious "3. Exceptionally unfavourable and untested market conditions have arisen". Doesn't that happen every time a trader starts trading? ;)
Vita, you make a good point. I myself have big doubts about the sufficiency of the forward test to conclude about the sustainability of a strategy. It is only a necessary test, which essentially says: "If the strategy is sustainable, the forward test will be good". The equivalent statement is: "If the forward test is bad, then the strategy is bad".
The law of meanness: this is just a negative statement, which allows you to screen out unsustainable strategies - but does not allow you to guarantee the quality of sustainable ones. It's still not bad either, as 90-95% of blatantly bad strategies are screened out...
I have been thinking about it for a long time but haven't managed to write it as an article. Frankly speaking, I'm tired of writing it, because new problems keep popping up. But I can tell you that my own approach is very different from Pardo's.
Here one must clearly understand that the result of such a thorough check (according to Pardo) is statistical. Even if it guarantees the quality of the strategy by 95%, there is still a 5% chance that a thoroughly tested strategy that has passed all the tests will fail. Maybe there is a need for diversification by strategy, to be at least certain of something!
And in general, Vita, sufficient conditions ("If a certain group of tests is good, then the strategy is good") is an unattainable goal, you understand. What positive guarantees can there be? They do not and will not exist... But you can improve as much as you like, adding more and more negative criteria (necessary) to ensure that bad strategies are cut off...
Nevertheless, I think this book is a very good introduction to the extreme complexity of the problem. The author doesn't actually solve it, he only shows the state of existing approaches to it, but it's much better than single optimization, the results of which grail creators post here.