Factor scores calculated through*principal component analysis

[Kseniya Levykina]

Dear Statalist community,

I would like to get your help in figuring out how to calculate resource slack variable based on factor scores calculated through principal component analysis.
For this, I am following a paper from Fang et al. 2008, that says "we use factor scores, calculate through principal components analysis from two financial ratios—(1) retained earnings to total assets and (2) working capital to total assets—to assess resource slack".

For now I have gone just for average between 2 ratios, but suspect that "principal component analysis" should be done somehow differently in Stata. Would appreciate a hint here how to do that.

gen rslack1 = re/at
gen rslack2 = wcap/at
gen rslack = (rslack1+rslack2)/2

Thank you!

[Kseniya Levykina]

Anybody who could help here?...

[Dung Le]

Hi Kseniya,

This is to just give you an example of how to do PCA using household durable assets. Let's make this simple and ignore your context. In this example we have a set of household durable assets, say car, motorbike and bike and all of them are dummies (1=yes and 0=no)

factor car motorbike bike [household weight], pcf factor(1)
    predict wealth_score // this is a wealth score predicted by the first component of PCA