More on APS and out_of_sample testing

Discussion in 'Strategy Building' started by goodgoing, May 9, 2010.

  1. OK, got it. Let me ask,

    1) If I get the demo http://priceactionlab.com/Demo/demo.html -- what data are needed to make it work? Are data supplied?

    2) What's the connection between the software and the Harris book ? Are the book's examples developed with priceactionlab?

    Thanks.
     
    #21     Aug 5, 2010
  2. #22     Aug 5, 2010
  3. I did a demo of this product couple of years ago, this was before the p-indicator. Also bought Michaels book profitability and systematic trading. I thought it was an interesting tool.

    What i find somewhat disconcerting is:

    1) the old company/website trading patterns totally vanished

    2) the price seems to oscillate quite a bit, it used to $4500, then just last week $4000 but with special discount $3000. Yesterday it was priced at $2500 per year.

    What do you guys think?
     
    #23     Aug 18, 2010
  4. ronblack

    ronblack

    Tradingpatterns was in business I think for about 9 - 10 years. I bought APS about 3 - 4 years ago; I don't even remember when exactly, time flies. It was recommended to me by a friend who bought some pattern code from the company back in 2001 or 2000 in a trade show. I think this is a fair lifespan for a small company. Then you have to change name, trademarks, do some renovations and stuff.

    I recently upgraded to the new product Price Action Lab. It is more or less the same program but with some minor but useful new features. Price is OK I think. I wouldn't worry about the oscillations. Do some bargaining. This is a buyer's market. I do not recommend this product unless you have some experience with statistics and you are willing to go one step beyond what it does. The p-indicator is very interesting. I am following the journal for SPY on the new website and so far their calls (presumably Michael Harris') are spot on.
     
    #24     Aug 18, 2010
  5. If my understanding of what you do is correct, I have comments/questions:

    1. There are theoretical formulas that give the results you are looking for, and you do not need testing to find the edge, but testing can be useful to quantify the edge.

    2. What happens to the trades that never reach neither the stop nor the target, if target=stop. My guess is that they cancel out on average, and they are equal in number and in the average win and loss.
     
    #25     Aug 29, 2010
  6. So what's your result? I found patterns for Eur 15 min and back testing result is good. I did forward testing for several days but the result is disappointing.
     
    #26     Aug 31, 2010