For an asset class that represents just 1.4% of insurer’s asset holdings1, equity release mortgages (ERMs) have consumed a remarkable amount of firm and supervisory time. A decade or so ago, the regulatory challenge of this asset class lay on the conduct side. More recently, however, and not without some irony, the main mitigant of these conduct risks, the no negative equity guarantee2 (NNEG), has switched the focus primarily onto the inherent prudential risks of equity release, namely its illiquidity and, owing to the NNEG, the long term exposure it brings to the fortunes of the housing market without further recourse to the borrower.
The PRA has recently published supervisory statement SS3/17 on illiquid unrated assets and ERMs in the context of the Solvency II Matching Adjustment (MA). SS3/17 (summarised in the Annex to this blog) follows a discussion and consultation process stretching over some 21 months - a reflection no doubt of the technical debates, complexities and sensitivities attaching to the underlying prudential issues. It shows, in our view, an ongoing and careful balance between prudential caution and guarded support for what the PRA broadly recognises to be an appropriate asset class for insurers with annuity books, and which has potential wider economic and societal benefits.
Overall, SS3/17 tightens the valuation and hence capital treatment of equity release and places a sharper emphasis on the responsibility of the board to ensure that the level of MA benefit being claimed is justified. However, the PRA has also left considerable room for pragmatic supervisory judgement firm-by-firm.
This blog discusses three key messages on the PRA’s overall position on this asset class and the approach firms looking to invest in it can expect from the regulator. In overview, these are:
- The PRA is concerned about the insurance sector’s increasing exposure to UK property, with ERMs being a growing part of this trend.
- The PRA is further concerned to ensure that firms are not claiming undue MA on illiquid assets through internal rating and valuation processes that do not take full account of all relevant risks. Consequently, the PRA is increasingly likely to use verification by external independent specialists, whether through s166 or other mechanisms.
- The PRA will also expect boards and senior managers to be able to justify ratings and valuations comprehensively, and to demonstrate that they have gone through robust governance challenge and review.
Why is equity release of concern to the PRA?
Measured by the value of lending, the UK ERM market has almost trebled in size in the last five years3. Around 25% of annuities are currently backed by “direct investments” such as ERMs and infrastructure funding, and this is set to increase to 40% by 2020 as insurers continue to search for yield4. David Rule starkly illustrated the PRA’s concerns on insurers’ growing exposures to UK property in a speech to the ABI in July, estimating that a house price shock of 30% with 0%-0.5% growth thereafter (a stress roughly aligned to the Financial Policy Committee’s banking system stress test) could lead to industry-wide losses of £2 billion to £3 billion. This is a significant figure, given that the total surplus own funds for the 74%5 of the UK life sector participating in the 2016 EIOPA stress test was £25 billion6. In practice, of course, this £2 billion to £3 billion of losses would be concentrated more narrowly among firms with the greatest exposure to the housing market and ERMs.
ERMs are typically written at LTVs in the region of 25-30%7. Nevertheless, as potentially a decades-long product with interest perpetually rolling up, LTV cover can deteriorate relatively quickly. A realistic assessment of the risks posed by the NNEG is therefore a key regulatory priority, to ensure that:
a. restructured notes are not valued and/or rated too highly, so attracting excessive MA benefit; and
b. there is adequate capital cover for the risks of the NNEG in a severe housing slump.
Technically, these issues can prove controversial depending on one’s view as to:
a. the housing market’s current cyclical state and prospects; and
b. whether there is an enduring risk premium in residential property as an asset class that it is prudent for insurers to capitalise up front via the MA, in the form of a substantial reduction in discounted liabilities.
Three key trends in the PRA’s approach
I. A cautious outlook for UK property
Future house price growth is arguably the most critical assumption for valuing ERMs. SS3/17 sets out four principles (discussed further in the Annex to this blog) that the PRA will employ to verify that firms assign an appropriate (i.e. sufficiently high) value to the NNEG risks retained by the insurer in the valuation of the ERM notes8. These principles are necessarily complex and we expect they will allow for a range of interpretation and supervisory discretion in practice.
In our view, the third principle (that future possession of a property cannot be more valuable than current possession) is likely to attract the most future debate. Very importantly, this principle implies that assumed future house price growth cannot exceed the discount rate applied in the valuation. In simple terms, insurers who do not already take account of this principle in their house price growth assumptions may see a higher valuation of retained NNEG risks and hence a lower MA benefit and a corresponding higher capital charge.
II. Use of independent assurance and benchmarking
SS3/17 sets out a detailed qualitative and quantitative framework that the PRA will use to review firms’ valuations and credit assessments of illiquid unrated assets and ERMs. This includes the use of quantitative thresholds to guide the intensity of supervisory review (similar to the PRA’s “quantitative indicator” approach for insurers’ internal models9), and deconstruction of the economic value of ERMs and ERM securitisations.
In practice, deriving suitable thresholds and gaining assurance over valuations are likely to be highly complex and judgemental exercises (for example, the particular complexity of the NNEG prevents straightforward comparison to reference instruments), and are likely to require extensive development of analysis tools and perspectives by the PRA (for example, considering rating agency-type assessments). The PRA is clear that it is prepared to commission external reviews where insurers cannot provide sufficient assurance using their own resources.
III. High expectations for management understanding, skills and oversight
SS3/17 notes the responsibilities resting with Senior Insurance Management functions and its expectations for senior management to understand the risks to firms investing in illiquid assets and ERMs10.
In our view, this topic is likely to attract far more of the PRA’s attention than the limited coverage in SS3/17 might suggest. David Rule in his ABI speech flagged further PRA work and on-site review on governance and risk management over direct investments, and made clear that insurers will require “different skills” (for example, “the capability to renegotiate and restructure the debt”). We would expect firms investing in ERMs and other direct investments to see an increased level of scrutiny and questioning from the PRA, with the bar set very high for management’s understanding of the valuation of such investments.
And what of the FCA?
Despite the product’s difficult conduct history, the FCA has struck a generally positive note on ERMs in recent years. This largely reflects a compelling social case for the product as a means of unlocking housing wealth to mitigate the problem of limited pension savings. As recently as 2015, Chris Woolard was highlighting the benefits of the products in these terms11. However, possibly reflecting in part the PRA’s developing prudential concerns, more recent FCA statements have become notably more nuanced. For example, in a speech to the Gleneagles Pensions Conference in September 2016, Andrew Bailey sounded a note of caution about the complexity of ERM contracts resulting from the NNEG and its prudential and conduct consequences, while recognising potential benefits of ERMs for individuals and retirement savings.
Overall, we expect the FCA to be supportive of ERM products in principle. However, we expect that conduct and vulnerability concerns will receive increasing attention as the ERM market continues to grow. In particular, as ERM books age, we expect conduct issues around contractual obligations on borrowers to maintain and insure properties to receive increasing scrutiny.
1David Rule speech at the ABI, July 2017.
2Unlike for a conventional mortgage, the NNEG limits repayment of the loan to the value of the underlying property.
3From £789m seen in 2011 to reach over £2.15bn in 2016 (Equity Release Market Report Spring 2017, Equity Release Council, page 4)
4David Rule speech at the ABI, July 2017.
52016 EIOPA stress test report, page 14
62016 EIOPA stress test report, page 22, converted at £1:€1.08
7Equity Release Market Report Spring 2017, Equity Release Council, pages 10 and 11
8Recognising that an ERM contract cannot provide the fixed cash flows and contractual repayment date required by the MA rules, the PRA allows firms to securitise ERMs internally to create MA-eligible senior tranches. While this is a pragmatic way of allowing a prudent level of liquidity premium and avoided a severe “capital crunch” in the treatment of these assets (the UK’s former ICAS capital regime allowed insurers more or less to attach the same amount of liquidity premium to an ERM contract as they would a higher quality corporate bond), this approach has nonetheless brought fresh problems in its wake: insurers incur artificial restructuring costs, and the restructured assets and cash flows are made more complex and harder to value, and therefore more difficult to supervise.