Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd Logo
Upload

Welcome to Scribd!

  • ch2 Capm

    Uploaded by

    vgh
    9 views4 pages
    The document summarizes applying the Capital Asset Pricing Model (CAPM) to different datasets. It replicates tables showing regressions of asset returns on market risk premiums. It finds low but statistically significant beta and high alpha coefficients when applying the model to returns from the Madoff Investment Securities hedge fund, which seem improbable.

    Original Description:

    Econometria -

    Original Title

    ch2_capm

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document
    The document summarizes applying the Capital Asset Pricing Model (CAPM) to different datasets. It replicates tables showing regressions of asset returns on market risk premiums. It finds low but statistically significant beta and high alpha coefficients when applying the model to returns from the Madoff Investment Securities hedge fund, which seem improbable.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    Download as pdf or txt
    9 views4 pages

    ch2 Capm

    Uploaded by

    vgh
    The document summarizes applying the Capital Asset Pricing Model (CAPM) to different datasets. It replicates tables showing regressions of asset returns on market risk premiums. It finds low but statistically significant beta and high alpha coefficients when applying the model to returns from the Madoff Investment Securities hedge fund, which seem improbable.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    Download as pdf or txt
    of 4

    ch2_capm

    September 26, 2021

    1 Chapter 2
    1.1 2.7 Illustration: The Capital Asset Pricing Model
    THE CAPM model predicts that

    E(rjt − rf ) = βj E(rmt − rf )

    where rjt is the risky return on asset j in period t, rmt is the risky return on the market portfolio
    and rf is the riskless return. The parameter βj is a measure of systematic risk. Under certain
    assumptions (see Section 2.7) is is possible to consistently estimate βj by running the regression

    rjt − rf = βj (rmt − rf ) + ϵjt

    Next, we apply the CAPM model to data on three industries (food, consumer durables and con-
    struction) using data for 660 months from January 1960 to December 2014.
    First we need to read the CAPM dataset. We need to change to the folder where the dataset is
    located:

    [3]: cd "C:\Users\Administrator\Google Drive\My Classes\Verbeek\datasets\ch2"

    C:\Users\Administrator\Google Drive\My Classes\Verbeek\datasets\ch2


    and now read the file. The option clear erases anything from memory.

    [4]: use capm5, clear

    We can do some descriptive statistics:

    [5]: summarize

    Variable | Obs Mean Std. Dev. Min Max


    -------------+---------------------------------------------------------
    dateid01 | 660 10028.84 5803.509 0 20058
    cnstrrf | 660 .566803 5.934099 -29.79 25.29
    durblrf | 660 .4207879 5.652569 -25.89 29.23
    foodrf | 660 .6980152 4.336142 -18.85 19.85
    hml | 660 .3709697 2.810302 -12.61 13.88
    -------------+---------------------------------------------------------
    jan | 660 .0833333 .276595 0 1

    1
    month | 660 198706.5 1588.659 196001 201412
    rf | 660 .3947424 .2544793 0 1.35
    rmrf | 660 .5057121 4.44066 -23.24 16.1
    smb | 660 .2091667 3.043928 -16.4 22.02
    We are now ready to replicate Table 2.3. First we need to run the regressions. Notice the option
    nocons after regress. Also quietly is used to suppress the output to the sceen. It is an example
    of a prefix command.

    [6]: quietly regress foodrf rmrf, nocons


    est store t231
    quietly regress dur rmrf, nocons
    est store t232
    quietly regress cnst rmrf, nocons
    est store t233

    And we are ready to produce the table.

    [7]: est table t231 t232 t233, b(%4.3f) se(%4.3f) stats(r2 rmse)

    --------------------------------------------
    Variable | t231 t232 t233
    -------------+------------------------------
    rmrf | 0.755 1.066 1.174
    | 0.025 0.027 0.025
    -------------+------------------------------
    r2 | 0.590 0.706 0.774
    rmse | 2.812 3.072 2.831
    --------------------------------------------
    legend: b/se
    Table 2.4 is similar but the regressions have a constant.

    [8]: qui reg foodrf rmrf


    est store t241
    qui reg dur rmrf
    est store t242
    qui reg cnst rmrf
    est store t243
    est table t241 t242 t243, b(%4.3f) keep(_cons rmrf) se(%4.3f) stats(r2 rmse)

    2
    --------------------------------------------
    Variable | t241 t242 t243
    -------------+------------------------------
    _cons | 0.320 -0.120 -0.027
    | 0.110 0.120 0.111
    rmrf | 0.747 1.069 1.174
    | 0.025 0.027 0.025
    -------------+------------------------------
    r2 | 0.585 0.705 0.772
    rmse | 2.796 3.072 2.833
    --------------------------------------------
    legend: b/se
    Finally, we estimate Table 2.5 where a January dummy is included (variable jan).

    [9]: qui reg foodrf rmrf jan


    est store t251
    qui reg dur rmrf jan
    est store t252
    qui reg cnst rmrf jan
    est store t253
    est table t251 t252 t253, b(%4.3f) keep(_cons jan rmrf) se(%4.3f) stats(r2 rmse)

    --------------------------------------------
    Variable | t251 t252 t253
    -------------+------------------------------
    _cons | 0.400 -0.126 -0.077
    | 0.114 0.126 0.116
    jan | -0.971 0.081 0.605
    | 0.393 0.433 0.399
    rmrf | 0.749 1.069 1.173
    | 0.024 0.027 0.025
    -------------+------------------------------
    r2 | 0.589 0.705 0.773
    rmse | 2.786 3.074 2.831
    --------------------------------------------
    legend: b/se
    The CAPM is often used to evaluate the performance of professional money managers. A positive
    intercept (“alpha”) reflects superior skill (or information) of the investment manager. We now
    use the CAPM model to analyze the returns of the Madoff Investment Securities hedge fund. For

    3
    this we use data on returns on Fairfield Sentry Ltd, which was one of the feeder funds of Madoff
    Investment Securities. The data cover period December 1990 until October 2008 (T=215).

    [10]: use madoff, clear


    label var fsl "return (in %) on Fairfield Sentry"
    label var fslrf "excess returns"
    label var rf "risk free rate"
    label var rmrf "excess return on the market portfolio"
    label var hml "excess return book-to-market factor (Fama and French)"
    label var smb "excess return size factor (Fama and French)"

    We replicate the results in Table 2.6

    [11]: regress fslrf rmrf

    Source | SS df MS Number of obs = 215


    -------------+---------------------------------- F(1, 213) = 14.54
    Model | 6.44404169 1 6.44404169 Prob > F = 0.0002
    Residual | 94.4286913 213 .443327189 R-squared = 0.0639
    -------------+---------------------------------- Adj R-squared = 0.0595
    Total | 100.872733 214 .471367911 Root MSE = .66583

    ------------------------------------------------------------------------------
    fslrf | Coef. Std. Err. t P>|t| [95% Conf. Interval]
    -------------+----------------------------------------------------------------
    rmrf | .0408859 .010724 3.81 0.000 .0197471 .0620246
    _cons | .5049538 .0456993 11.05 0.000 .414873 .5950347
    ------------------------------------------------------------------------------
    The beta coefficient is very low at 0.04 while the alpha has a value of 0.5% per month. These
    numbers seem “too good to be true”.

    [ ]:

    You might also like