Picking Stocks for Portfolio

Discussion in 'Automated Trading' started by BlackPhoenix, Dec 20, 2023.

  1. How do you pick stocks you will trade each day, and how often do you update your portfolio? Do you have some kind of automated process for it or do you do this manually? Also, how many stocks you trade daily?
     
  2. ph1l

    ph1l

    Weekly for ETFs (this could also apply to individual stocks), I run scripts to find a set of symbols such that each symbol's correlation to every other symbol in the set is below a threshold as described in this post.
     
    murray t turtle and BlackPhoenix like this.
  3. @ph1l Ah yes, correlation check is something I need to also implement for my portfolio generation. Recently watched this video by Darwinex about checking correlation across assets:

    Do you run backtesting to pick the best performing ETFs from your ETF universe?
     
  4. Karin42

    Karin42

    It seems to me that the best option in casing of picking stocks for portfolio is to do it manually. Maybe some automation systems exist but i am personally not aware of them.
    The classical approach to selecting stocks include the evaluation of a set of parameters a company has. The most important one, in my opinion, is P/E. Some other things include: company's stock price history, industry trends, news checking and verification of annual reports.
    It's advised to revise the portfolio once a year or maybe once per quarter.
    So, this is everything that concerns investments with long term prospects.
     
  5. %%
    Amen;
    ETFs in my notebook, about 20-21 like SPY,qqq + such............................................
     
  6. ph1l

    ph1l

    My strategies try to use generated rules (backtested and out-of-sample tested) on ETFs such as those posted here, here, and here.
     
  7. Interesting, I'll need to look into genetic programming in more detail :) I saw this video where he mined candlestick patterns with a genetic algorithm. Not that candlestick patterns are necessarily a good predictor overall, but more as an inspiration how genetic algorithms can be used:

     
    Last edited: Dec 22, 2023
    ph1l likes this.
  8. ph1l

    ph1l

    I like the video -- the narrator gets to the points quickly and shows out-of-sample results.

    If you are interested in genetic programming "Linear Genetic Programming," by Markus F. Brameier and Wolfgang Banzhaf is a great book.


    Another good reference is the attached "Dynamics and Performance of a Linear Genetic Programming System" by Frank D Francone which is the basis for the commercial product Trading System Lab http://www.tradingsystemlab.com
     
    BlackPhoenix likes this.