Upgrades and downgrades are inevitable with A.M. Best’s new rating approach
A.M. Best is currently in a ‘Request for Comment’ (“RFC”) period for proposed changes to its rating methodology. It is also just finalising an update to one crucial part of that, namely the capital model it uses to evaluate non-US insurers (known as the “Universal BCAR”).
As a regulated ratings agency the RFC process follows a defined script. The agency has released its proposed changes and is receiving – and will subsequently publish – comments from market participants along with its response to them (the initial RFC period ends on June 30th). A second round is expected once Best makes the changes it chooses following the RFC feedback: this is likely to run through Q3 and may extend into Q4. Meanwhile it continues to update its ratings based on its existing methodology. Once the required process of comment and review is completed Best will ‘flick the switch’, and all subsequent ratings reviews must then be executed using the new methodology. The expected date for that is early 2017.
S&P followed a similar path in 2013, with a similar goal, namely to create a more transparent ratings process.
A.M. Best stresses that the goal is not to move their ratings up or down in the aggregate; they are happy with the overall current distribution of rating outcomes. They also note that some individual rating changes are pretty much inevitable. In 2013 S&P said the same, and the outcome was that 8% of ratings changed as a direct consequence of the introduction of the new methodology, with roughly as many winners as losers.
Rating agency regulation however adds some spice to the process. It requires that when an agency moves to a new methodology it must notify the markets of those ratings it believes may be impacted.
There is only one way that can be done in practice: all ratings reviewed in the run-up to the launch date, while based on the existing methodology, must also be being privately reviewed by the agency using the new methodology. And sometimes that must mean two different conclusions being reached.
Simply put, rating committees will periodically be privately concluding “based on how we currently do things this company is an ‘A’, but based on how we will probably do things come early 2017, it would be an ‘A-‘”. Since rating reviews are typically annual, we must assume this process is already taking place.
At Litmus we would expect that, given there is both an overall methodology change (as per S&P) and a change to the capital model, then at least an S&P level of rating changes is likely and in all probability more (especially outside the US).
More than 10% of ratings being impacted outside the US would not surprise us at all. Nor would half of those being downgrades.
Any A.M. Best rated company therefore not explicitly focussed on how it will look (and should consider presenting itself) under Best’s new methodology is risking (to borrow a cliché) “preparing to fail by failing to prepare” in terms of managing its rating.
Whilst we await the new Universal BCAR to get a handle on that part of the new analysis (due for release to rated firms in the second half of 2016), it may not be the prime cause of those upgrades and downgrades that emerge.
Indeed, an explicit part of Best’s goal in making the methodology change is to clarify quite how much of the ratings analysis is not explicitly driven by the BCAR outcome, or even the wider review of capital adequacy overall. (A common misconception among rated firms and, sometimes, their advisors, is to presume that the BCAR outcome in essence defines the rating).
The graphic below summarises Best’s proposed new ratings methodology. For those familiar with Best’s Financial Strength Ratings (FSRs) scale it should be noted that they use the (S&P like) Issuer Credit Ratings (ICRs) ratings scale in this process (in practice Best produce FSRs for operating insurers via a mapping of the ICR scale to their unique FSR scale).
Note: A ‘notch’ refers to the difference between ratings levels, e.g. there is a ’one notch’ difference between ‘A’ and ‘A-‘.
The analysis starts with the BCAR outcome, which is then subject to various sources of modification (for example due to risks associated with the insurer’s domicile, strengths/weaknesses in its holding company’s financial profile, the quality/nature of the reinsurance protection it buys etc.)
This leads to an initial ICR outcome reflecting purely ‘Balance Sheet Strength’ (and any country risk impact on that).
Critically this outcome can be no higher than ‘a+’. That outcome then gets modified up or down by the subsequent areas of the analysis: operating performance, business profile, ERM and a final one notch rating adjustment (intended to be rare) if the rating committee feels that somehow the sum outcome of the previous analysis was not quite right.
So, no matter how strong the BCAR result, it can only ever lead to a Balance Sheet Strength ‘base-line’ outcome of ‘a+’ at most. Thereafter, if all non-capital factors were maximally negative, the final rating could be ‘b’.
In practice the credit profile an insurer would need for that journey from its initial BCAR result to take place is almost inconceivable. But Best’s proposed new criteria certainly makes it very clear that the BCAR outcome is indeed not the rating. For those Best rated firms for whom having a rating starting with an ‘a’ is important, there is not a lot of downside protection if the capital component can only get you to ‘a+’ at most. And, as noted above, privately Best must already be running reviews based on this.
To us as rating advisors that implies: get focussed on how your operating performance, business profile and ERM support your rating. And start doing that now if you want to have the rating you believe you deserve.
+44 (0)20 3651 5044
First published in Insurance Day on Friday 3rd June 2016