Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

# Re: st: Comparing two timeseries regressions using esttab

 From Robson Glasscock To statalist@hsphsun2.harvard.edu Subject Re: st: Comparing two timeseries regressions using esttab Date Mon, 18 Apr 2011 12:49:29 -0400

```Hi Victor,
If I'm understanding you correctly, you are trying to conduct a
hypothesis test to see if there is a difference between two parameters
obtained from separate regressions. Let's call the first parameter B1
and the second parameter B2. Under the null hypothesis of parameter
equality B1= B2, so  B1-B2= 0. To find the statistical significance of
the difference, you need estimates of the variance of each parameter
and the covariance between them.

The test statistic will be (B1-B2)/ [Var(B1) + Var(B2) + 2Cov(B1,B2)] ^.5

The covariance between the parameters should be 0 if they are from
separate regressions, so you can drop the third term in the
denominator.

best,
Robson Glasscock

On Mon, Apr 18, 2011 at 11:03 AM, Victor Kihlgren <21693@student.hhs.se> wrote:
> Thank you for the tip. Is significance level of differences possible to obtain using only t-statistics? I am trying to replicate an essay comparing two different set of funds exposure to risk-classes.
>
> many thanks
>
> victor
>
> ________________________________________
> Från: owner-statalist@hsphsun2.harvard.edu [owner-statalist@hsphsun2.harvard.edu] f&#246;r Charles Koss [hqtiger@gmail.com]
> Skickat: den 18 april 2011 16:31
> Till: statalist@hsphsun2.harvard.edu
> Ämne: Re: st: Comparing two timeseries regressions using esttab
>
> use Excel to get the difference. Why do you want to obtain the
> significance level of each difference?
>
> Charles
>
> On Mon, Apr 18, 2011 at 2:19 AM, Victor Kihlgren <21693@student.hhs.se> wrote:
>>
>> Hello,
>>
>> I would like to compare two timeseries regressions using esttab. Ive managed to tabulate the results but i would like a column showing the difference between each beta and the significance level of each difference, how is this possible?
>>
>> Thanks
>>
>> Victor
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/statalist/faq
>> *   http://www.ats.ucla.edu/stat/stata/
>>
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>

--
Robson Glasscock, CPA
Accounting PhD Student
Virginia Commonwealth University