A Wall Street Journal Survey of how the credit ratings firms have performed in prediciting looming defaults. The Journal's Matt Wirz looked at 35 years of data. He found the ratings firms did not do an effective job with predicting defaults in the 12 month period before an actual default. Of the 15 government defaults since 1975 tracked by S&P, S&P's sovereign ratings division rated 12 of the countries single B or higher in the 12 months preceding the default. S&P says a single-B rating on sovereign debt signifies that the government has only a 2% average default rate in the next 12 months. For Moody's Investors Service the figures show that of the 13 governments rated by Moody's, 11 were rated B or higher one year before an actual default. By contrast the investment grade ratings of the credit ratings firms have worked better- as no government defaulted within 15 years of having a tripe-A, double-A, or a single-A rating. Ratings firms say that the ratings indicate a relative default risk for countries and not an actual default probability, a rank ordering for different countries and their relative risk. Research chiefs at investment management firms point out that once a crisis develops the ratings firms are not much help. They also say the ratings firms use static indicators like current account balances and other critical indicators for countries in emerging markets such as political sentiment and bank deposit flows get less attention. Historically bond yields have priced in higher risk premiums into government bonds before a default and investors look at the bond yields in assessing risk conditions, and not at the ratings which change only slowly. Brazil and Argentina both had a double B-minus rating in Jan. 2001. A year later Argentina had defaulted....