Message-ID: <68b1e2610906100814u39169e0an92a8d149dfa1cbd1@mail.gmail.com>
Date: 2009-06-10T15:14:07Z
From: Liviu Andronic
Subject: Vector autoregression with Newey-West standard errors
In-Reply-To: <4A2FBD20.6050901@gmail.com>
Dear Matthieu,
On 6/10/09, Matthieu Stigler <matthieu.stigler at gmail.com> wrote:
> by:
>> attr(equation[[colnames(yend)[i]]]$terms, "intercept") <- 0
>
> instead of:
>> attr(equation[[colnames(yend)[i]]]$terms, "intercept") <- NULL
>
Replacing "1" (in the published package) with "0" indeed solves my issue. Now,
> coeftest(temp.varest.lm, df = Inf, vcov = vcovHC)
z test of coefficients:
Estimate Std. Error z value Pr(>|z|)
yldave3.l1 0.9294 0.0685 13.57 <2e-16 ***
spread5.l1 0.0273 0.1245 0.22 0.83
gdpc1LogRetLag1.l1 0.0632 0.0433 1.46 0.14
const 0.2347 0.3786 0.62 0.54
---
Signif. codes: 0 ?***? 0.001 ?**? 0.01 ?*? 0.05 ?.? 0.1 ? ? 1
Thanks a lot for you help. Best,
Liviu