Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk> |
To | <statalist@hsphsun2.harvard.edu> |
Subject | st: RE: Re: Calculating Hansen-Hodrick standard errors using Stata |
Date | Mon, 7 Mar 2011 17:34:16 -0000 |
Khabara, > -----Original Message----- > From: owner-statalist@hsphsun2.harvard.edu > [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Khabara > Sent: 07 March 2011 17:07 > To: statalist@hsphsun2.harvard.edu > Subject: st: Re: Calculating Hansen-Hodrick standard errors > using Stata > > Dear Mark and Eric, > > Thank you for your reply. > > The problem is the following: I've got monthly data on > zero-coupon bond prices with maturities ranging from 1 to 5 > years, from which I recover yields, forwards, yearly returns > & yearly returns in excess of a 1-year yield. > > First, I regress average (across maturities) excess returns > on all the forwards and a constant (using -ivreg2- and > specifying "bw(12) kernel(tru) robust" in options). > > Second, I regress excess returns on each bond on the fitted > values of the dependent variable from the first regression > (no constant - otherwise, the same command). I am a bit confused here. Do you mean you are doing IV/2SLS by hand, where the 2nd regression is the 2nd stage of 2SLS? If you do 2SLS by hand, the SEs in the 2nd stage are wrong. Perhaps this is the problem? What happens if you estimate the second equation setting the average excess returns at the dependent variable and the regressors in the first equation as your instruments? > In the first regression I recover EXACTLY the same HH-12 lags > st. errors as the author of the original study. > > However, in the second-stage regression, the standard errors > reported by the author are VASTLY different from those > calculated by me (mine are way larger and increase linearly > with the maturity of the bond - they do not in the original study). > > Also, when I try to recover SEs robust to autocorrelation > only (omit "robust"), I obtain SEs different from the ones > obtained by the author, even though the difference is not large. Might the difference be a degrees of freedom adjustment? Perhaps try the -small- option of -ivreg2-. If that doesn't do it, can you contact me off-list with the details and we can investigate further? --Mark > Any ideas on where the discrepancies might come from? > > Thank you very much in advance! > > -- > View this message in context: > http://statalist.1588530.n2.nabble.com/Calculating-Hansen-Hodr ick-standard-errors-using-Stata-tp6094922p6098074.html > Sent from the Statalist mailing list archive at Nabble.com. > * > * 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/ > -- Heriot-Watt University is a Scottish charity registered under charity number SC000278. * * 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/