Recent rating downgrades and the risk of this continuing have focussed the attention of many brokers on an old problem; how, if they use ‘A-‘ as a minimum level of automatic acceptability, should they react to a downgrade of a former ‘A range’ carrier to the ‘BBB range’?
One response is that they should form their own conclusions on security anyway; but for most brokers that is a big ask, especially in terms of analysing a larger or more complex group.
While broker approaches to market security vary, for many relying on a ‘rating floor’ (typically ‘A-‘) OR requesting specific client approval for carriers rated below this (or not rated at all) is the norm.
But what if the carrier was previously ‘A range’ but drops below this? In most cases requesting client approval is really only tenable from sophisticated buyers and, if there is higher rated alternative capacity available at an affordable cost, it is easy to imagine how the reaction can be instinctively risk adverse (‘move my business’). That is to say that if the broker, as the ‘expert in the chain’, is unwilling to continue to recommend the security, a natural consequence is for the policyholder to be concerned.
A broker could, of course, substitute the role of the rating with the fact that a carrier is regulated and licenced to trade. However, neither the PRA in the UK or any regulator throughout the EU runs a ‘zero failure’ regime; they set prudential rules that accept the premise that regulated insurers may fail and that the regulatory process is designed to limit this risk but not completely preclude it.
In that context a broker’s duty of care to policyholders may make them feel that they need more than simply the fact of a carrier being regulated for them to continue to propose it (indeed that is why ratings are used by brokers in the first place).
The uncertainty around how to deal with downgrades in part derives from a limited understanding of what ratings actually are. They are opinions about the future; in other words forecasts.
Expert forecasts are important contributors to economic and business decisions but they should never be treated as ‘facts’. So, the binary use of ratings (above a certain level, fine; below that level, a problem) is conceptually flawed (as the confusion caused by ‘A-‘ ratings on negative or developing creditwatch highlights).
Moreover, like most forecasts, ratings are really expressing a view of probabilities. Indeed the rating agencies publish data on exactly this. For example S&P’s data tells us that the historically observed probability of an ‘A-‘ defaulting over a one-year period is 0.07%, whereas for a ‘BBB+’ the historical probability is 0.14% (see below *).
Whether that extra degree of implied credit risk is reasonable for any given policyholder is something the broker should consider (and perhaps discuss with the policyholder). But, it seems to us, providing such advice is no different from considering the risk to the policyholder of different types of cover, exclusions, limits or other terms and conditions, as are related ratings issues such as where a given carrier is rated below the level of other members of the group OR where one rating agency has a significantly higher or lower rating on a carrier than another OR whether different views of desired rating should exist for different lines of business.
So, while it may be impractical for most brokers to have the internal resources to do their own full security analyses, it seems entirely reasonable to expect brokers to have a good understanding of what ratings mean in practice and, hence, how they can constructively advise their clients as to what they imply.
*Source: Standard & Poor’s. Litmus Comment; the application of generic bond default data to the risk of rated insurers not being able to pay claims is challenging in that the point at which an insurer is in default in terms of claims payment is hard to define. Moreover ‘financial strength’ ratings address the ability to pay valid insurance claims, not willingness to pay. Nonetheless the data is generally seen as being a good proxy for insurer default risk.
Stuart Shipperlee, Analytical MD, Litmus Analysis
Stuart – Great article – however I think that sovereign ratings also need to be taking into account. Until c. 2001 the rating agencies considered the sovereign rating as the ceiling that no other corporate or subnational borrower was likely to pierce. While this is no longer the case the conditions to pierce the country ceiling are drastic and therefore not often attained so the country sovereign rating still has to be a consideration when reviewing markets.