Recent changes among ratings criteria at Standard & Poor's represent "significant" repercussions for collateralized debt obligations (CDOs) and residential mortgage-backed securities (RMBS), according to the ratings agency. The changes will make high ratings on securities in sectors troubled by poor credit performance "more difficult" to receive, S&P said. The changes aim to enhance the comparability of ratings on these securities with ratings on credits in other sectors. "More than any other kind of institutional change, changes to criteria directly affect our credit analysis and our ratings that result from that analysis," S&P said. "Indeed, criteria is the exact spot where the rubber meets the road for a rating agency. By reading our criteria, investors can gain a deep understanding of the nature and levels of risk expressed in our rating opinions." The ratings agency recently adopted stress scenarios for use as a tool to calibrate criteria, meaning assigned ratings ought to be able to withstand higher levels of economic stress without defaulting. The recent weak performance of CDOs and RMBS prompted S&P to revise its criteria in order to improve rating performance and comparability. The new US RMBS criteria establish a 7.5% credit enhancement level for a security backed by an "archetypical" prime mortgage pool in the triple-A rating category, a "substantially higher" level than previously established. The ratings agency said some RMBS risk features like low borrower credit scores or slim home equity not accounted for in the "archetypical" scenario could trigger adjustment mechanisms in the criteria to allow for higher credit enhancement levels. S&P indicated the implementation of the new RMBS criteria resulted in few downgrades, since many outstanding RMBS deals already faced downgrades over poor performance. The ratings agency also updated the corporate CDO criteria to add both qualitative and quantitative tests to a default simulation model already in place. The model addresses the loans or bonds backing a CDO from a mathematical framework, calculating probably behaviors and statistics. The tests added to the model addresses the "model risk" inherent in a probability-based model, S&P said. "In addition, we recalibrated the simulation model to achieve stresses based on Depression-era experience," S&P added. "The calibration method that we used makes it easier and more transparent for investors to understand our ratings and to relate them to their investment objectives." Write to Diana Golobay.