In its complexity and its vulnerability to market volatility, the CPDO might be viewed as the poster child for the excesses of financial engineering in the credit market. This paper examines the CPDO as a case study in model risk in the rating of complex structured products. We demonstrate that the models used by S&P and Moody's would have assigned very low probability to the spread levels realized in the investment grade corporate credit default swap market in late 2007, even though these spread levels were comparable to those of 2002. The spread levels realized in the first quarter of 2008 would have been assigned negligibly small probabilities. Had the models put non-negligible likelihood on attaining these high spread levels, the CPDO notes could never have achieved investment grade status. We conclude with larger lessons for the rating of complex products and for modeling credit risk in general.