sas >> White's test of heteroskedasticity

I have been reading White's 1980 article. I have had trouble discerning
exactly how the test for heteroskedasticity is conducted. How is White
consistent estimator of the covariance matrix of beta compared to the
ordinary least squares estimator of the covariance matrix of beta? I
attempted to hand calculate the chi-square using White estimate of cov
(beta) and the ordinary least squares estimator estimate of cov(beta)
(which was obtained using COVB)on the printout. The hand calculation of
chi-square and df did not match the printout.

```Hi all,

Under the "PROC REG" statement (model options), we could specify the ACOV
to get the asymptotic covariance matrix of estimates assuming
heteroscedasticity. However, the standard errors on the main printout of
parameters are not adjusted. So, may I know is there any efficient way (or
how to) to get the corrected standard errors/t-values printout using SAS
steps to extract the diagonal elements of the covariance matrix for these?

Thank you very much!
Thomas
```

```Hi everyone,

Can anyone tell me how to do white correction for calculating the standard
errors in a OLS regression if heteroscedasticity is assumed to be present?
I've been looking to the PROC REG and it seems that is only possible to get
the test for heteroscedasticity...

Thanks for any help!

Nuno
```

```I need to run the robust Huber-White t-test on my regression parameter
estimates.  I am using pooled (time-series, cross-sectional data) with
potential cross-sectional correlation.  Is there a SAS procedure that will
provide me this t-statistic?

Krishna

Krishna R. Kumar, Ph. D.
Professor of Accountancy
The George Washington University
Washington DC 20052
Ph (202) 994-5976
Fax (202) 994-5164
```

```Hi All

I need to run a fixed effects model on a panel data set which has
heteroskedasticity and autocorrelation. I use the following code to run
the fixed effects model {Code in curly parenthesis are my comments}

proc glm data=c.datae;
class cusip {variable identifying the firm} year {variable identifying
the year};
model con=cusip year size tenure/solution;
run;

Is there any way of correcting for heteroskedasticity and
autocorrelation in SAS. I normally use the Newey-West correction with
proc model using the kernel correction (Shown below)

proc model data=Merge2;
parms Int CoeFF_LnMVE CoeFF_ROA CoeFF_RET;
Comp= Int + CoeFF_LnMVE*LnMVE + CoeFF_ROA*ROA + CoeFF_RET*RET;
fit Comp / gmm kernel = (bart,1,0.383439 {Bandwith calculation});
instruments lnMVE ROA RET ;
run;

I would be most grateful for some clarification.

Thanks
Guy Fernando
PhD Candidate (Accounting)
Syracuse University

```