Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • CAPM-testing using fund characteristics

    Dear all,

    I want to test if there is a difference in returns between fund characteristics (think number of holdings). I observe these just once for each fund and the whole time period:

    .................. .................. Fund Name.... Portfolioretuns ......Benchmarkreturn...... Number of Holdings
    01.01.2005 .................. Fund 1 .................. 0,02 .................. 0,01 .................. 5
    02.01.2005 .................. Fund 1 .................. 0,03 .................. 0,04 .................. 5
    01.01.2005 .................. Fund 2 .................. 0,04 .................. 0,01 .................. 8
    02.01.2005 .................. Fund 2 .................. 0,07 .................. 0,04 .................. 8

    For the CAPM approach I am using:

    Code:
    reg Portfolioretuns Benchmarkreturn, robust
    How should I test now for the influence of the characteristic? I thought about time fixed effects...

    Thanks!

    Matthias
    Last edited by Matthias Johnsen; 14 Jun 2018, 15:18.

  • #2
    You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex. Also, try to simplify your posting to the critical issues. Also, assume we don't understand your area so present your question in general statistical terms.

    I don't see how fixed effects would "test if there is a difference between returns..." Fixed effects make sense to control for unmeasured stable fund characteristics but if the dv doesn't change within panels, you can't use fixed effects. They also drop iv's that don't change within panels. I would suspect you want to see if a characteristic influences the fund's alpha. You get alpha by running your regression by fund (and maybe time), and saving alpha in the observations for the fund. But that is just my guess - I'm not in finance and the field undoubtedly addresses this question in a more complex way. Actually, the first thing you should do is go to the literature and see how others have tested similar propositions.

    Comment

    Working...
    X