Mixing TA with FA ? or not ?

Discussion in 'Strategy Building' started by jacksmith, Mar 16, 2009.

  1. - Yes it is OK to mix TA and FA. As long as you're cognizant of which indicators are lagging and which indicators are more predictive.

    - As an example of when fundamental info is not included in the price, Zacks believes that raised EPS estimates for next calendar year aren't factored into price for 1-4 weeks or more. Again it pays to realize that some indicators are lagging and some are more predictive.

    - Some indicators I typically take into account are:

    sales growth 1yr
    pct below 52wk high
    on balance volume
    peg ratio next yr
    price chg 52 wks
    rating chg last 4wks
    forward pe
    price cashflow
    chg in next yr est
    proj long term eps growth
    avg surprise last 4qtrs
    moneyflow index
    williams pct r

    - A detailed listing of some indicators to consider, including statistical analysis, is in the paper:

    A Non-Random Walk Down Canary Wharf
    Emanuele Canegrati
    August 2008
    http://mpra.ub.uni-muenchen.de/9871/1/MPRA_paper_9871.pdf
     
    #11     Mar 23, 2009
  2. which are some predictive indicators ?

    Thanks.
     
    #12     Mar 23, 2009
  3. Don't waste your time with FA. Markets have increased in tempo too much while companies still report quarterly.
    Entire teams at institutions with superiour information do a pretty poor job at this kind of analysis.
    Your a fool if you think you can do anything alone with inferior information.

    All FA will do is screw up how your looking at the chart because of bias.
     
    #13     Mar 23, 2009
  4. I hate to disappoint but, other than Zacks' idea mentioned above, that kind of information isn't free. It took me 4+ years of backtesting and forward walking different systems to figure out. And still I'm probably only about halfway there to developing a tradeable system. I will say however that I currently use MiniTab for multivariate statistical analysis. If you or someone you know knows MatLab, R, or SPSS, one of those could probably be used as well.
     
    #14     Mar 23, 2009