Even though I'm not that familiar with Stata, i doubt there is a routine to estimate VaR, which is applicable for everyone and without strong assumptions. Even CreditMetrics or CreditRisk+ require you to heavily calibrate to the portfolio (with most likely different results).