Search found 3321 matches

by startz
Sun Jul 15, 2018 12:31 pm
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

Normality is of almost no importance for estimating an impulse response function. Sometimes people transform variables, for example taking logs, to get more normal residuals. But then of course you have to translate the transformed impulse response function back to the variable you actually care abo...
by startz
Sun Jul 15, 2018 8:09 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

My question is why are you concerned about non-normality?
by startz
Sun Jul 15, 2018 6:13 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

What do you think the problem is?
by startz
Sun Jul 15, 2018 6:12 am
Forum: Data Manipulation
Topic: Correlation
Replies: 2
Views: 29

Re: Correlation

Look at the function

Code: Select all

@movcor(x,y,n)
by startz
Sat Jul 14, 2018 10:45 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

No, the VAR is fine with highly correlated variables.
by startz
Sat Jul 14, 2018 10:15 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

Perhaps run with one lag and run with two and see if it makes any difference. The real issue is that only having 20 observations limits what you can do.
by startz
Sat Jul 14, 2018 9:24 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

This probably suggests adding one more lag to the VAR. But you have very few observations, so it's not entirely obvious whether you should.
by startz
Sat Jul 14, 2018 8:37 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

The article is just wrong.

A VAR is just a collection of least squares regressions. Normality and heteroskedasticity matter for coefficient tests, but usually no one is interested in doing those for a VAR. Autocorrelation does matter-it's usually handled by adding lags to the VAR.
by startz
Sat Jul 14, 2018 8:16 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

I would just proceed without the test.

And why do you want to test for heteroskedasticity anyway?
by startz
Sat Jul 14, 2018 7:58 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

Even if there is heteroskedasticity, the estimated equations should be unbiased. What would you differently if you found there is heteroskedasticity?
by startz
Sat Jul 14, 2018 7:15 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

I believe the White test has more right hand side variables than you have observations, in which case the test can't be done. It's not a very informative error message.

By the way, the White test is a large sample test and using it 20 observations probably isn't valid.
by startz
Sat Jul 14, 2018 6:50 am
Forum: Estimation
Topic: Near Singular Matrix Error
Replies: 33
Views: 29007

Re: Near Singular Matrix Error

You'll probably have to give more details of exactly what you did. And possibly post your data.
by startz
Wed Jul 11, 2018 9:57 am
Forum: Programming
Topic: Scalar of average of last 12 months
Replies: 2
Views: 45

Re: Scalar of average of last 12 months

See if @movav helps.
by startz
Sun Jul 08, 2018 8:12 am
Forum: Econometric Discussions
Topic: Non-linear Fixed Effects
Replies: 13
Views: 195

Re: Non-linear Fixed Effects

Look in the built-in help system under Wald test/coefficient restriction
by startz
Sun Jul 08, 2018 7:28 am
Forum: Econometric Discussions
Topic: Non-linear Fixed Effects
Replies: 13
Views: 195

Re: Non-linear Fixed Effects

No, this is just a test of coefficients. You can do it either with fixed effects or without fixed effects.

Go to advanced search