# Re: st: reverse equation estimation

 From Luciano Lavecchia To statalist@hsphsun2.harvard.edu Subject Re: st: reverse equation estimation Date Tue, 26 May 2009 03:00:47 -0700 (PDT)

```I've come up with an idea which I will verify tomorrow morning with my professor:

STATA is not able to identify in the reverse equation each single coefficient but I can compare the builted ratios from forward equation with the "naturally" estimated coefficient by STATA. According to Hahn and Hausman, under strong instruments, there should be no difference. indeed under weak instruments I should find different results..

what do you think?

--- On Tue, 5/26/09, Luciano Lavecchia <luciano.lavecchia@yahoo.com> wrote:

> From: Luciano Lavecchia <luciano.lavecchia@yahoo.com>
> Subject: st: reverse equation estimation
> To: statalist@hsphsun2.harvard.edu
> Date: Tuesday, May 26, 2009, 8:37 AM
> Hi,
> I'm try to test how weak instruments can affect Taylor
> Rules in the euro zone. basically I'm running a TSE GMM with
> Newey-West kernel with 4 lags of bandwitdh, regressing
> euribor on constant, output gap, inflation and a lag of
> euribor i.e
>
> euribor=
> (1-c(4))*(c(1)+c(2)*inf+c(3)*outputgap))+c(4)*euribor(-1)
>
> Now,  I would like, following Hahn & Hausman (2002) to
> obtain reverse equation estimates (as this would give
> insights on the
>  presence of weak instruments) i.e. estimate something
> like
>
> outputgap=-(1/c(3))*(c(1)+c(2)*inf-(1/(1-c(4))*(euribor -
> c(4)*euribor(-1)))
>
> as you can see there is no "free" regressor which I can use
> to estimate with nlcom the others... what can I do? It's
> funny because E-views (which, by the way, doesn't support
> neither TSE nor CUE with GMM ....) has this useful equation
> editor, while STATA is not so "friendly". I know the command
> "constraint" but is not allowed under ivregress and ivreg2.
>
>
> Thanks!
>
>
> luciano lavecchia
> P.S: I'm aware of Stock & yogo if somebody would like
> to suggest me of their test!
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> *
> *   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/
```