Have two basic questions on the Core Reading/Acted notes for Chapter 7 -thanks in advance!: Pages 12 and 17 lists limitations of the S&P approach to review a firm's ERM. What do these two limitations mean?: (i) "There is no explicit mention of agency risk" and (ii) "There is reference to "complicated and powerful simulation models"; this a highly subjective and maybe problematic". What is the purpose of the table in section 1.6: S&P ERM capability assessments (page 13)?
Hi Bill, 1. For (i), it is simply that, there is no assessment of agency risk mentioned in the S&P note when considering the ERM capability, when agency risk would exist. Agency risk is further discussed in Chapter 11, including agency risk involving credit rating agencies (ie that they are being paid by the companies that they are being asked to rate). For (ii), this is explained at the bottom of page 9 of the S&P note and relates to the increased model risk as a result of using these models to estimate predictive risk measures. 2. This table describes S&P's definition of 'Excellent' , 'Strong' , 'Adequate' and 'Weak' ERM for various areas. These can also be found on page 4 of the S&P note. Please let me know if you have any followups from this. Alvin.
For (ii), this is explained at the bottom of page 9 of the S&P note and relates to the increased model risk as a result of using these models to estimate predictive risk measures.[/QUOTE] Thanks for all helpful responses. But why does the Core Reading describe S&P's approach of mentioning model risk (and the challenges of running and auditing extensive predictive models) as 'problematic'? -is it simply that actuaries, insurers, banks etc couldn't survive without complex models
I must admit, when I read it I just took it as referring to problems with developing and running complex models such as time, cost, expertise, explaining outputs, reviewing/auditing etc . I imagine there might also be problems with using subjective assumptions / methodologies eg comparing between different models and increased model risk!
