Page 8 of 11

### Re: Fama-MacBeth regression

Posted: Mon Nov 17, 2014 12:43 pm
You will have to modify the code to do this yourself.

### Re: Fama-MacBeth regression

Posted: Tue Nov 18, 2014 12:51 pm
difficult for me to modify the program i'm not an expert like you
it's very important for me to use this Add-in in three methods OLS, GLS and GMM.

### Re: Fama-MacBeth regression

Posted: Sat Feb 14, 2015 2:22 am
how can show the graph of linear regression for the first and second step in the spool object.

Kind Regards. MED

### Re: Fama-MacBeth regression

Posted: Fri Feb 20, 2015 3:24 pm
What exactly do you want to show?

### Re: Fama-MacBeth regression

Posted: Sun Feb 22, 2015 9:40 am
1-show regression line with scartter: pr11 c rmkt, pr12 c rmkt,...etc (the first step); it takes a long time to do this.
2-show regression line with scartter: avrgrate c beta01 (the second step).
3-how can store the avrgrate in the series and calculate it by the command (i can do this manually)?
i hope that you understand me now.
Kind Regards. MED

### Re: Fama-MacBeth regression

Posted: Mon Feb 23, 2015 12:57 pm
Are you asking how to do this with commands?

For #1 and #2, use

Code: Select all

`group g1 rmkt pr11g1.scat linefit`

And so on. This code fragment can go in a loop.

For #3, look in the code to see how avgrets is calculated.

### Fama-MacBeth regression

Posted: Fri May 29, 2015 4:26 pm
Hi guys!! I am new here at forum.
I was told to run fama-macbeth and take the fitted/predicted values of the regression(about debt and leverage factors) :D = pr +s+ gr+ tng+ nbts+ dr+ liq and then, use them as proxy for the target debt of each firm. How i can take the predicted/fitted values of the aforementioned regression by using fama-macbeth ?
Thank you in advance!

### Re: Fama-MacBeth regression

Posted: Sat Aug 29, 2015 11:13 am
Hi!

I'm hoping someone can help. I'm trying to use the Fama-MacBeth add-in to test the relationship between abnormal returns (alpha) and expense ratios of mutual funds (see for example Carhart 1997). So I have the raw return data for the mutual funds, alongside the return data on the Fama French factors (exrm, smb, hml) and Carhart's momentum factor (mom). Hypothetically, if I were to also obtain the expense ratio data, how would I use the Fama-MacBeth add-in to analyse this relationship?

### Re: Fama-MacBeth regression

Posted: Sat Aug 29, 2015 11:46 am
kwasiyeboah42 wrote:Hi!

I'm hoping someone can help. I'm trying to use the Fama-MacBeth add-in to test the relationship between abnormal returns (alpha) and expense ratios of mutual funds (see for example Carhart 1997). So I have the raw return data for the mutual funds, alongside the return data on the Fama French factors (exrm, smb, hml) and Carhart's momentum factor (mom). Hypothetically, if I were to also obtain the expense ratio data, how would I use the Fama-MacBeth add-in to analyse this relationship?

The example in the add-in only assesses the sensitivity of all portfolios to one column vector (rmkt). But I want to assess each portfolio's sensitivity to it's own expense ratio? I'm not sure if the Fama-MacBeth can do this or if I should be looking at the loop/roll programming stuff?

### Re: Fama-MacBeth regression

Posted: Tue Sep 08, 2015 1:59 pm
It's not clear to me what you want to do. The first step of the Fama-MacBeth regression will regress your portfolio returns on whatever factors you choose. The resulting betas (see Equation 1 in the doc) are reported in the "bhat" object.

### Fama-MacBeth regression

Posted: Sun Sep 27, 2015 3:39 pm
Hi,
i'm trying to run the Fama-MacBeth Regression. I have one dependent variable and 10 independent. However, I receive the message 'near singular matrix in rowplace(g, @transpose(@inverse(@transpose(design) * design) * @transpose(design) * retvec), j)'. Then it creates an object called retsm01. What is it? How can I solve it?

Thanks

### Re: Fama-MacBeth regression

Posted: Mon Sep 28, 2015 11:13 am
Can you post your workfile?

### Re: Fama-MacBeth regression

Posted: Tue Sep 29, 2015 2:07 pm
EViews Rebecca wrote:Can you post your workfile?

### Re: Fama-MacBeth regression

Posted: Wed Sep 30, 2015 6:03 am
Belca wrote:
EViews Rebecca wrote:Can you post your workfile?

The dependent value is cash_assets and factors are the ones starting with F*

### Re: Fama-MacBeth regression

Posted: Wed Sep 30, 2015 4:05 pm
I noticed that your workfile only has one dependent variable. Fama-MacBeth is intended for a portfolio of assets. Using only one asset will lead to problems in the cross-sectional regression step. See Equation 2 in the doc.