One of the most controversial aspects of rating agency activity seems to be dying a slow but inevitable death.
On Tuesday S&P announced the withdrawal of almost all its entire North American and EMEA ‘pi’ (public information based) insurance ratings. This amounted to 38 ratings withdrawals in EMEA and 131 in America.
The agency has been steadily reducing these ratings over recent years. Typically its rationale has been that insufficient information was publicly available to support the analysis but, in this case, it states a lack of market interest. The agency receives no direct payment for ‘pi’ ratings and so needs to believe that producing them either enhances the value of its insurance products to subscribers and/or its ratings franchise.
While, no doubt, interest in some of the ‘pi’ rated carriers was marginal, the premise that all of these ratings were of no market interest surprises us. At least in part we see this as a policy decision to cease unsolicited rating production except where the agency feels a truly fundamental need to do so.
All three of the other main insurance rating agencies have also either substantially reduced or fully ended ‘unsolicited’ rating coverage (not all of which carried a ‘pi’ type subscript depending on the agency) within the insurance sector in recent years. Moody’s did so at the back end of the last decade for example.
Fitch remains the most active in publishing unsolicited ratings in the sector. Of the 43 groups that we follow in our Litmus Ratings Review none have a major carrier with an ‘unsolicited’ rating other than from Fitch (who publish these on carriers for 12 of the 43 groups). Yet in the last year even Fitch have withdrawn these on 3 of the 43 (Everest Re, Fairfax and Platinum).
In the reinsurance, specialty and large commercial lines markets the great majority of carriers have one or more ‘solicited’ ratings (although the largest groups often have some smaller carriers not assigned an interactive rating, some of which previously received an S&P ‘pi’). In these cases the market impact of a withdrawal of an ‘unsolicited’ rating is typically negligible. However, where there is no rating on a significant carrier and buyers lack the ability to assess an insurer for themselves, the onus tends to fall back on brokers to make explicit recommendations about security. Not a position they typically relish.
Indeed this highlights a fundamental but often misunderstood truth about ratings use in the re/insurance industry. It is brokers as much as anybody that have driven the growth in the desire for carriers to be rated, for the simple reason that the prospect of having to use unrated (or uncollateralised) markets on any scale is loaded with issues for them.
This is something of an irony given brokers are often to the forefront in complaining (informally) that the agencies have too much power!
A firm called ISI (subsequently acquired by S&P) began publishing insurance ‘pi’ ratings on non-US re/insurers in the mid-80’s.These became a focus of market attention later in the decade following the turmoil in the London Market and elsewhere that resulted from a host of catastrophe losses coinciding with increased recognition of historical asbestosis-based exposures. At that point brokers started to embrace ratings as a means of carrier selection and since then carriers in many sectors have sought to engage with the rating agencies in order to tell their stories about their financial health.
From a position where the vast majority of non-US re/insurer ratings were unsolicited and based solely on public information, we now have nearly all reflecting a paid-for engagement with the rating agencies and exchange of otherwise confidential information. There are many different opinions as to whether this is a positive or a negative…
(For more on the history of ratings see the Litmus Ratings Guide here).