By Barney Wanstall, Director, Insurance, PwC
As a result, and in practice, the question is much less “ is the standard formula appropriate?” and much more “is it inappropriate?” given few firms, especially with years of ICA experience, would choose to model themselves in the way the standard formula requires.
It is also worth remembering that this requirement is not a PRA regulatory add on but a core Directive requirement for all firms to assess where your business (materially) deviates from the Standard Formula assumptions.
In advance of this, the PRA has helpfully shared some guidance on where it has so far seen some indicators of inappropriateness. For Life firms, this includes those who have significant equity portfolios, very “credit risky” assets, “non standard” longevity exposure (e.g. deferred/ enhanced/ impaired annuities). For non-life firms, this includes those with material periodic payment order exposure, large amounts of non-European catastrophe risk or complex reinsurance arrangements (such as stop losses or shared programmes) which the standard formula is not set up to model. For all insurers, pension risk and significant outsourcing / complex operations may also be indicators.
Many firms have been busily working on this over the summer months using the ICA as a start point for the assessment. Whilst this makes sense conceptually, it is far harder to do in practice:
-
Firstly, the ICA is of course not calibrated to the Solvency II based 1 year time horizon, but a ‘run-off to ultimate’ time horizon. This difference requires some careful thought, particularly for non-life firms, before a proper comparison can begin.
-
Very often the ICA has not been modelled using the same risk categories as the standard formula – significant amounts of re cutting / re grouping of risks is therefore required to enable an effective comparison.
-
Typically the ICA has a larger amount of diversification, since it is prepared at a more granular level.
-
The ICA may also contain some prudence in some areas – firms may then find they are “over” in some areas and “under” in others – it is difficult to ignore this even if overall you think the comparison is sensible.
-
In the London market, as warned by the PRA, Cat risk does indeed often look very low in the standard formula compared to legacy ICA’s and catastrophe model outputs. The difference however appears far starker with non-European natural catastrophe than man-made cat events.
The assessment of appropriateness has, in most instances, needed to go much further than just using the ICA as a tool to perform this assessment. Many firms have found the EIOPA document, which details the basis for the underlying assumptions behind the standard formula, a useful tool to appraise the applicability of the calibrations to their firm. It is also useful to consider how well generally your business is likely to fit into the industry average – e.g. how well your business classes map into Solvency II classes of business and align to their definitions, whether your investment portfolio contains atypical investments for that investment class and whether your operations differ from those of a typical insurer (e.g. above average levels of outsourcing for instance). Doing this well enables you to more quickly hone in on the “different” aspects of your risk profile so they can be assessed specifically.
Doing this assessment properly will enable firms to confidently articulate their “level” of standard formula (in)appropriateness – if then necessary, this enables a firm to proactively propose a voluntary capital buffer above the official standard formula SCR. This confidence must come not just from the strength of the analysis, but from management’s intimate and demonstrable understanding of how the standard formula fits. Whilst this might convert to a formal capital add-on or request for a partial internal model, it is far more preferable to what may very well be a harsher regulator driven conclusion where the firm’s analysis is flawed. It is an exercise well worth the time therefore.
|
