What do the proposed Solvency II matching adjustment reforms mean for private assets?


In April 2022, as part of the UK’s post-Brexit review of Solvency II, HM Treasury (HMT) issued a consultation on its proposed reforms. This was accompanied by a technical discussion paper issued by the Prudential Regulatory Authority (PRA). Through these reforms, HMT says, it hopes to "spur" the UK insurance industry and create better outcomes for policyholders as well as the wider UK economy.

Specifically, HMT and the PRA are keen that the review supports capital deployment in long-term productive assets, including those which will support the country’s journey towards net zero. Upon further examination of the technical detail, however, the matching adjustment reforms (see below) may just be counterproductive.

The review

The two key areas of Solvency II being reviewed are risk margin and matching adjustment. Broadly speaking, the expected impact of the reforms are:

  1. A reduction in the risk margin, strengthening insurers’ solvency positions. The risk margin is a provision which insurance companies are required to hold under Solvency II, and is the amount the insurer would be required to pay to transfer the business to another firm.
  2. A reduction in the matching adjustment, worsening the solvency positions of insurers who apply a matching adjustment. The matching adjustment, as we’ll see below, is an increase to an insurer’s liability discount rate under certain conditions. This can reduce the present value of the insurer’s liabilities, benefiting an insurer’s solvency position. A reduction in the matching adjustment would therefore reduce the size of this benefit.

HMT stresses that these reforms should be viewed as part of the same, net beneficial package, with an expected 60-70% reduction in risk margin and expected overall capital release of 10-15% for long-term insurers.

Part of the case for matching adjustment reform, the PRA says, is the fact that insurers are increasingly investing in newer asset classes to back long-term liabilities. These include assets such as real estate lending, infrastructure debt and infrastructure equity.

These are key components of the real economy, and have the potential to deliver on climate change and wider ESG objectives. Although it is encouraging to see the regulator engage with these asset classes, the mechanisms being proposed are unlikely to incentivise investment in them.

Fundamental (spread) issues

In a nutshell, the rationale for the matching adjustment is that by matching asset and liability cash flows well, the risk of insurers not realising the spread on their assets is reduced. Insurers can therefore recognise some of this spread in their discount rates.

Under current regulations, the spread on an asset is split into two parts – the fundamental spread and the matching adjustment. The fundamental spread is the part of the spread which reflects risks that insurers are assumed to retain, and therefore cannot contribute to the insurer’s discount rate. The remainder of the spread is the matching adjustment.

The current Solvency II regulation requires the fundamental spread to reflect the risks of default and downgrade, and it is floored (for non-sovereign bonds) at 35% of the “long-term average spread” of comparable assets.

Current matching adjustment levels are heavily driven by this floor. HMT believes that this is indicative of two issues:

  1. the core matching adjustment calculation methodology is insufficient; and
  2. the fundamental spread is slow to react to changes in market conditions, including signals of credit deterioration.

In addition, HMT and the PRA highlight the further issue that:

  1. the current regulation does not adequately account for different risk characteristics across different asset classes.

Credit risk premium

So far, the most concrete proposal to address these issues is the introduction of a new component in the fundamental spread calculation: the credit risk premium.

The PRA believes that the fundamental spread should incorporate two elements:

  • Expected loss due to default, and
  • The credit risk premium, or compensation that a willing, arm’s length third party would demand for taking on the risk.

In other words, in addition to accounting for the expected loss, firms should also account for the uncertainty around the expected loss.

The PRA believes that the credit risk premium needs to be calibrated to at least 35% of credit spreads (cycle average). Although this resembles the current “35% of long-term average spread ” floor, the PRA argues that the current floor is backward-looking and slow to react to market changes. In this way, the credit risk premium is intended to address issues 1 and 2 highlighted above.

One potential formulation of the credit risk premium is the “index-spread approach”: 

credit risk premium =

   X * (average spread of reference index over n years)

+ Z * (difference between spread of asset and that of a reference index)

where X, Z and n are parameters to be calibrated, and the reference index is to be confirmed.

Here, the first term links the credit risk premium to market spreads, where the n parameter will need to be calibrated to provide sufficient sensitivity without causing excessive volatility, and leading to pro-cyclical behaviour. The PRA has suggested n = 5 as a candidate. There are not yet any indications of the calibrations of X or Z.

A closer look at illiquid assets

The second term of the credit risk premium is intended to address issue 3 – the fact that some asset classes appear to be more “matching adjustment-efficient” than others.

This fact in itself should not be surprising. After all, the matching adjustment exists to allow insurers to take credit for parts of the spread that are unrelated to the risk of default or downgrade, notably the illiquidity premium. As many private asset classes are less liquid than corporate bonds, a higher illiquidity premium and hence matching adjustment is expected.

The issue with the current fundamental spread calculation, according to the PRA, is that it is calibrated using historical corporate and government bond data, and does not accurately capture the different risk characteristics of other asset classes. Evidently, a BBB-rated infrastructure debt will not have the same risk of default or downgrade as a BBB-rated corporate bond debt.

The second term of the “index-spread approach” is an adjustment to the fundamental spread to close some of the gap between the spread on the individual asset and the reference index. It is intended to remove some of the advantages that currently come with certain illiquid assets.

In effect, assets with higher spreads relative to other assets with equivalent credit ratings – notably private, productive investments – will be penalised through a higher fundamental spread. The size of this penalty will depend on the calibration of Z. The higher the Z, the more the incentive to invest in these assets will be reduced.

The PRA has not explicitly stated whether there will be a single reference index for all assets, or asset class-specific indices. It is also still considering the choice of the reference index (or indices). Many have already pointed out this link of the fundamental spread to a market index is likely to increase the volatility of matching adjustment balance sheets. For insurers with significant illiquid asset holdings in their matching adjustment portfolios, there will be an additional valuation mismatch if the PRA were to require a a corporate bond index to be used for all asset classes.

For illiquid assets, for example infrastructure debt, it is not clear that corporate bond index spread movements necessarily indicate a change in the credit risk characteristics of the individual infrastructure debt. These movements would cause a change in the asset’s fundamental spread (and hence matching adjustment) without – depending on the insurer’s asset valuation methodology – necessarily changing the valuation of the asset. Without consistency between the fundamental spread and valuation approaches, such a credit risk premium could introduce additional balance sheet volatility.

In this vein, the PRA has also hinted that asset valuation is another area it is looking into: “The current PRA view is that private lending […] presents additional challenges for firms in terms of both asset valuation and the identification and monitoring of risks”.[1]

While the new methodology can in theory either increase or reduce the matching adjustment at any one point, over an economic cycle we would expect the proposed methodology to lead to a reduction in matching adjustment. This is because, as stated earlier, the current methodology floors the fundamental spread at 35% of the long-term average spread (average spread over 30 years), and typically bites. An updated methodology would include a credit risk premium of 35% of average spreads (over a different period), plus the expected loss from defaults. The final proposed credit risk premium may also be calibrated to above 35%.

According to the PRA’s analysis, based on year-end 2020 submissions, a methodology based on a credit risk premium of 35% of spreads would have led to a reduction in matching adjustment for all credit quality steps.

Closing remarks

HMT is keen to frame the reforms as having an overall positive impact for insurers’ balance sheets. The flagship 60-70% reduction in risk margin is one such contributor, and is expected to reduce balance sheet interest rate sensitivity and pro-cyclical behaviour.

Further positive matching adjustment changes are also proposed, such as the removal of the “BBB cliff”. Currently, the matching adjustments on sub-investment grade assets are not allowed to exceed those on comparable (by duration and class) investment grade assets; this can discourage investments in BBB or lower-rated assets, and lead to pro-cyclical behaviour. The change would remove some of the barriers to investing in such assets.

HMT has also signalled a broadening of asset eligibility, in particular targeting assets with prepayment risk such as callable bonds, real estate lending and infrastructure debt, but stops short of providing further details.

Other key proposals include a simplification of the matching adjustment approval process, especially in relation to new asset classes, reducing the reporting and administrative burdens on firms, and taking a more proportionate approach to matching adjustment breaches. HMT is further considering extending liability eligibility to other products, for example income protection, with-profits annuities and deferred annuities.

It is clear that HMT is trying to communicate a desire to foster an environment which supports productive investments and investment in the country’s transition to net zero. Even the changes to the fundamental spread calculation are indicative of a willingness from the regulator to better engage with newer and more illiquid asset classes in the future.

However, the proposed reform to the fundamental spread has attracted considerable criticism. The UK’s Institute and Faculty of Actuaries, for example, has said in its official consultation response that it expects the fundamental spread reforms to bring additional, pro-cyclical volatility to insurers’ balance sheets, increase annuity prices, and make asset-based offshore reinsurance more attractive.

From an illiquid asset perspective, we believe that the proposals will make private assets less attractive as matching adjustment assets, and is likely to be counterproductive to HM Treasury’s stated objective of driving capital to “long term productive investments, including infrastructure”. [2]

References

[1] Annex to PRA DP2/22, https://www.bankofengland.co.uk/-/media/boe/files/prudential-regulation/publication/2022/solvency-ii-review-matching-adjustment-and-reforms-to-the-fundamental-spread

[2]https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1071899/20220328_Review_of_Solvency_II_Consultation.pdf


Contributes to
Unstructured Learning Time

CPD Accredited