Two recent public events, a speech to the ABI by PRA Executive Insurance Director David Rule, and an evidence session to the Treasury Select Committee (TSC) enquiry into Solvency II led by PRA Deputy Governor Sam Woods, with messages reinforced in a subsequent speech by Sam Woods on 20 March on the PRA insurance objective, provide important insights into the PRA’s latest view of Solvency II. But there were also a few straws in the wind as to how the PRA, if given a free hand post-Brexit, might want to adjust the UK prudential insurance regime. This would not involve any wholesale departure from Solvency II, which the PRA thinks is essentially working well, but might involve a limited move towards a less prescriptive, principles-based regime slightly more in the mould of its ICAS predecessor. Notable features of any adjusted regime might include:
- Some streamlining of data collection and transitional recalculation;
- More flexibility on matching adjustment eligibility, particularly on cash flow fixity;
- A continuing role for the risk margin (albeit using a much less interest rate sensitive calibration); and
- Continued application of the quantitative indicator (“QI”) actuarial framework, involving more granular sectoral coverage, to underpin future model approval and change judgements.
The lines taken in the speech and in the TSC evidence were highly consistent. Importantly, the PRA team giving evidence included one of the PRA’s insurance non-executive directors, David Belsham, one of the UK’s most respected actuaries. The evidence can therefore be taken as broadly representative of the PRA’s executive and non-executive “insurance” view.
Overall, ten key messages emerge as to the PRA’s latest view on Solvency II:
1) While certain design features need to be reformed, most notably the risk margin, PRA considers Solvency II overall a “sensible” and “good” regime.
Solvency II incorporates much of the UK’s former ICAS economic capital regime, which was always a key UK objective in the original Solvency II negotiations. Whilst disclaiming any role as “cheerleader” for the Directive, Woods expressly repudiated any notion that (net of transitionals) Solvency II had materially increased capital requirements or that, viewed “in the round”, Solvency II was “heavy handed” or “seriously flawed”. Nor does the PRA accept that Solvency II is responsible for the decline in annuity rates, which it ascribes almost entirely to the fall in risk free rates.
2) That said, PRA would ideally prefer a less prescriptive, less “complex” and less “rigid” regime as characterised the ICAS predecessor. This is particularly the case on matching adjustment (MA) and specifically in the treatment of equity release, where Solvency II has introduced a much more restrictive regime.
Notably, in discussing the MA, Belsham acknowledged that he “would prefer principles based regulation in this sort of area”, albeit recognising that Solvency II had introduced greater “rigour”. Woods struck a similar note: if he had “complete control”, he would want to do something “slightly different” and be “a tiny bit more flexible” on MA requirements such as fixed cash flows and unexpected defaults. As to equity release, Solvency II had added “more complexity than I would like to have and that place where matching adjustment and securitisation rules meet is not an entirely happy place…”
3) The PRA is supportive of insurers increasing their investment in infrastructure and does not accept that Solvency II capital requirements are hindering this.
Rather, the PRA considers that the lack of availability of suitable, attractively priced assets is the key block. Belsham’s view was that levels of insurer investment in infrastructure under Solvency II would be “very similar” to the previous ICAS regime.
4) The risk margin is flawed in calibration terms, but not as a concept and would therefore likely still have an important role to play in any future UK regime.
This is a subtle but key distinction in the PRA’s line on the risk margin. The PRA has consistently voiced its concerns about the risk margin’s excessive interest rate sensitivity and hence likely pro-cyclicality and it did so again, liberally, in both the speeches and TSC evidence. But that should not be taken as indicating that the PRA would want to dispense with the tool altogether. That is because UK prudential insurance regulators have on occasion found it challenging to find “white knight” purchasers of even moderately distressed life funds given the valuation risks and uncertainties than can arise. Equitable Life has been cited by the PRA frequently as the most public example of this.
Accordingly, PRA views the concept, if not the present calibration, of the risk margin as a necessary policyholder protection tool in any resolution situation, particularly in the absence of formal insurance resolution powers. That perspective clearly underlay Woods’ comment to the TSC that: “The best estimate will not capture what gets done if someone needs to take over those liabilities. That will require a little bit extra for the uncertainty”. As to the role of the risk margin in future solvency regimes, Belsham noted that it features in IFRS 17 and the ICS and is now “widely seen as part of a sensible solvency framework”.
5) Pro tem, because of concerns about setting wider precedents, and the impact on the UK’s overall negotiating position in Europe, the PRA has concluded that the risk margin problem can only be fixed at a European level (i.e. via the Commission, with technical advice from EIOPA) following the same process, and to the same timeline, that PRA had set in train prior to the Brexit referendum.
Woods acknowledged that the PRA board had a “long and very difficult” discussion as to whether the UK should introduce an immediate local fix, principally through accepting the industry’s argument that the availability of management actions, essentially reinsurance, mean that longevity risk is not un-hedgeable and should therefore not attract rm. He explained that the PRA board had been swayed by the arguments that this would “allow the genie out of the bottle… in all sorts of other parts of the framework” and “undercut” the UK’s wider negotiating position in Europe.
6) Meantime, PRA has sensibly reiterated its important message in the run up to Solvency II that the transitionals necessitated by the risk margin should be regarded as high quality capital.
Here Rule’s speech disclosed that the aggregate sector level of risk margin has now risen to 50% of Solvency Capital Requirement compared to 35% at the time EIOPA conducted its impact study. (Since this is a sector wide figure the level will presumably be much greater for firms with substantial longevity exposure, illustrating strikingly the present scale of the risk margin burden.) Rule added that since Solvency II was not intended to raise capital levels on the stock of existing business, accordingly “the PRA has complete confidence in TMTPS (transitionals) as a means of achieving a glide path to the new Solvency II requirements”. This stance gives welcome continued clarity to the market as the era of Solvency II “Pillar 3” disclosures begins.
7) The PRA considers that too much longevity risk is being reinsured in order to avoid the risk margin. Given the policyholder protection concerns this creates, the PRA will be reviewing firm by firm practical controls over reinsurance risk, such as collateral management.
8) Model approval will continue to require the PRA to have its own view on calibration; consequently, the “quantitative indicator” (“QI”) framework remains essential.
In the run up to Solvency II, there was some expectation that the Directive’s emphasis on model methodology rather than output would limit or preclude the PRA’s longstanding use of quantitative benchmarks. These indicators have been criticised on the grounds that they replace firms’ judgements with that of the regulators, are not generally used by other EU regulators, and force all firms onto an inappropriately high level of capital cover. However, the PRA robustly defended their continued use. Specifically, Belsham noted that he had independently explored the QIs with the PRA actuaries and had concluded that they were in a “reasonable range”.
9) In a similar vein, “model drift” and model complexity remain major concerns for the PRA justifying continuous strong supervisory oversight of models.
This stance is likely to be most keenly felt where firms apply to make model changes. Rule in particular highlighted the “reality” that “firms have strong incentives to reduce capital requirements in order to report higher coverage ratios, increase return on regulatory capital and ultimately pay higher dividends”.
10) Finally, while PRA has no appetite to dismantle the reporting regime, it may be open to some streamlining, particularly of quarterly reporting requirements, which Belsham described as “rather onerous”.
PRA considers that Solvency II data reporting has brought important gains for insurance regulation, particularly in developing peer outlier analysis and monitoring insurer exposure to banks. Woods commented that “we need a decent amount of data to do the job. If we go all very generalist and do not have the data, it will honestly be a disaster”. Nevertheless, in subsequent correspondence with the TSC he committed to conducting a review to identify ways of reducing the reporting burden
In sum, the PRA has given a clear and strong public affirmation of its fundamental support for Solvency II’s design architecture and experience to date. PRA has seemingly wanted to give a signal that, post-Brexit, it will not be looking to introduce anything other than marginal reform of the insurance capital regime. And in that regard, PRA will no doubt be mindful of the constraint that equivalence considerations may place on any substantial departure from the Solvency II requirements in the current absence of globally accepted insurance capital standards.