Systematic US Equities Investing

Discussion in 'Journals' started by mahras2, Jan 19, 2006.

  1. bubbrubb

    bubbrubb

    i have plently of fundamental data.

    i was curious where your friend got his data, as some of the databases suffer from survivorship and look ahead bias. . .
     
    #71     Feb 24, 2006
  2. mahras2

    mahras2

    I have no clue where my friend may have gotten the data from.

    As for the model, all positions were closed off at 2.30PM EST. The results for the month of February:

    Rank 1-10: +3.1%
    Rank 1-20: -2.12%
    S&P 500: +0.33%
    NASDAQ: -0.86%
    DOW: +1.49%

    So Rank 1-10 easily beating the benchmarks, while Rank 1-20 did miserably after realistic fills/slippage/commissions.

    Lets see what the second month holds. Will post the list over the weekend.
     
    #72     Feb 24, 2006
  3. mc107

    mc107

    Very nice results on rank 1-10. Do you often see such difference in monthly return between rank 1-10 and rank 1-20?
     
    #73     Feb 26, 2006
  4. mahras2

    mahras2

    Thanks.

    No typically you do not see such a large difference between the results. But as a rule of thumb Rank 1-10 will outperform Ranks 1-20 (49% annualized vs 41%).

    Rank 1-10 for the month of March:

    PBR
    URI
    TLAB
    AH
    CMCO
    AOS
    AP
    OATS
    IM
    FTO
     
    #74     Feb 26, 2006
  5. sorry if question is redundent but I assume you are putting equal $ amts vs. equal shares? Thanks donna...BTW very interesting thx
     
    #75     Feb 26, 2006
  6. mahras2

    mahras2

    Welcome. Yes equal dollar amounts for every pick.
     
    #76     Feb 26, 2006
  7. cnms2

    cnms2

    I don't recall if you already mentioned it, but I guess you're not using any stops.
     
    #77     Feb 26, 2006
  8. mahras2

    mahras2

    No at this moment I do not set any stops. However, cat stops approximately 25% away from price doesnt seem to have any harmful affect on the performance of the system.
     
    #78     Feb 26, 2006
  9. cnms2

    cnms2

    Maybe you should backtest using an allocation based on the historical volatility of each equity on your list.
     
    #79     Feb 26, 2006
  10. You might want to avoid earning announcement, as in my experience, they usually lead to disappointment.
     
    #80     Feb 26, 2006