Thank you for inviting me to speak today at the 2021 Insurance Asset Management Conference.
In the past year my colleagues Anna Sweeney and Gareth Truran have spokenfootnote  about the role insurers play in supporting the real economy; specifically as (1) providers of protection, (2) as guarantors of retirement income, and (3) as long-term investors.
It is the third of these that I want to talk about today. The regulated financial sector plays an important part in long-term investment, including supporting the government’s macroeconomic objectives, and the pursuit of net zero economy. Many commitments to such investments have been made. The Association of British Insurers’ Climate Change Road Mapfootnote  in particular signals what insurers can do to finance the transition to a net zero economy, and calls on regulators and government to play their part.
Public discussion of potential Solvency II (SII) reforms has focussed on financial incentives for investmentfootnote . Today, I would like to discuss the investment process, which has so far been less in the spotlight, but where I believe there is plenty of scope for improvement. Changes, in part through the SII review, could remove barriers to delivering those commitments to sustainable investment.
Of course, insurers can only invest if they are able to write policies and receive premiums, which provide the funds for investment. Annuity and with-profits business in particular can channel funds into long-term productive assets that support the real economy and its sustainable growth.
Since these investments back policyholder liabilities it is essential that investment risk is well managed, to ensure firms are able to honour their policyholder commitments, and maintain their own safety and soundness. And so as I set out how we hope to support sustainable investment, I will also discuss how the Prudential Regulation Authority (PRA) will continue to look to firms to demonstrate that they have the necessary capabilities to manage the consequent investment risks.
Annuities as a source of funds for investment
As a society we need long-term, productive investments. Annuity portfolios with their long-term liabilities are a natural home for such assets, and the bulk purchase annuity market in particular is channelling large quantities of funding into the insurance sector. Of course, much of this is not new money: the funds are transferred from defined benefit pension schemes, albeit usually with some additional injection of funds from the sponsoring employer to extinguish its liabilities and covenant. The liabilities may be unchanged, but there is a difference between the investment incentives of a fund backed by an employer covenant, and those of one supported instead by the capital base of an insurer. What we see in practice is a transfer from pension funds at least in part investing in equities, to annuity writers investing in long-term illiquid assets with fixed cashflows.
Suitability of assets to back annuities
Since the introduction of SII, firms have expanded the range of assets they include within their investment portfolios, particularly to back annuity business, including a wide range of internally valued and rated assets, which make up a large and growing proportion of assets in firms Matching Adjustment (MA) portfolios. In many ways, the SII regime accommodates such assets, including through allowance of internal ratings and securitisations.
The PRA has responded positively to this development – within the constraints of the SII regime – but doing so has often involved us acting as a gatekeeper for investments. This has meant not just extensive publication of expectations, but also lengthy supervisory review and approval processes. We are working to improve and streamline such processes for both insurers and ourselves. We have had good responses to the information requests we have madefootnote , and of course, frequent feedback from the firms we supervise. If we are to change our gatekeeping approach, changes will also be needed in some firms to strengthen their investment risk management capabilities, including governance – particularly where there is less experience of the assets whose performance provides security to annuitants.
Many firms have told us they would like to invest more in certain infrastructure assets, including green infrastructure, to back annuity business as part of their commitment to sustainable investment. There is nothing inherent in the MA that precludes this, and indeed firms do already include infrastructure within their MA portfolios. However, we understand that at present sometimes otherwise suitable assets may not meet existing MA eligibility criteria.
We are open and ready to listen to suggestions on the specific barriers you face when considering the MA eligibility of these assets. We need to understand novel risks and what may need to be done to allow these assets to attract the MA benefit. I will talk today about reforms we are considering that may tackle some of these barriers and risks, including where this involves lengthy construction phases or prepayment risk. However, we are open to further views on what it would take to overcome any further barriers that remain, for these and other potential new assets.
While annuity writers are a natural home for assets with long term characteristics, they may not seem the most likely investors in innovative asset types. However, where risks are appropriately managed and mitigated as part of a well-diversified portfolio of matching assets, insurers can help capture the huge positive consequences of such investment in the fight against climate change.
For some assets, there may be a disconnect between the actual level of risk posed on a forward-looking basis, and their measured risk under the current regime. In some cases, actual risk will be higher than risk measured on the basis of historic performance – investments in carbon-intensive or fossil fuel extraction industries, for example. As the Governor has said before, “investments that look safe on a backward look may be existentially risky given climate risks. And investments that might have looked speculative in the past could look much safer in the context of a transition to net zero”.footnote  As regulators, we must be alert to this, and challenge where risk is not fully recognised.
However, where regulation makes measured risk higher than actual risk, this may represent unnecessary barriers to investment, and it is right that we explore ways to remove these, particularly where this will help to ‘unlock’ investment in green assets. By reforming and simplifying the regime for approving internal models and model changes, we hope to allow firms greater latitude to use forward-looking risk assessments when data is lacking or does not tell the whole story.
Such an approach should remind us that the question of whether a given asset is suitable for a particular firm is not only about the asset’s features, but also about the individual investor’s ability to manage the risks of the asset throughout its life. This will be a function in part of their skills and expertise in managing the asset, and their capacity to mitigate risks that may crystallise from their exposure to that asset, including work out capability. This is particularly important where asset features may be new, or complex, or where exposures are material.
One tool available to us here is the Prudent Person Principle, which I have talked about previously.footnote  We expect firms to be able to demonstrate that they have the capabilities and skills to manage the assets they invest in, including through appropriate consideration of the risks arising from uncertainty in valuations and/or potentially excessive exposures. The assessment of individual firms’ risk management capabilities is as important a part of our processes as the more abstract assessment of MA eligibility.
We are considering a number of reforms, which may expand the range of assets firms can include in their MA portfolios.
A particular issue with cashflow matching has been the need to restructure some types of assets which themselves would not meet the MA eligibility requirements but might otherwise be suitable to back annuities. We worked hard to find a way to ensure that such assets can be accommodated within firms’ MA portfolios in a safe way.
We recognise the burden that internal securitisation has brought to insurers, particularly through the costs of setting up and maintaining ‘special purpose vehicles’. But we also recognise the benefits of these structures – particularly the identification of a senior tranche of cashflows that can be credit rated and hence risk adjusted for their inherent uncertainty: this is essential to the way that the MA works.
Securitisations are typically used for larger exposure pools, and this size increases both the potential MA benefit and the prudential risks. We consider that internal securitisations will still have a useful role to play as one way for firms to make some ineligible assets eligible, and firms should continue to engage with their supervision team at the earliest opportunity where looking to take this approach.
MA eligibility remains a key area of focus for us. As I said in Aprilfootnote , fixity of asset cash flows, and matching, result in a higher level of confidence that they will be available as the corresponding liabilities fall due without the need for reinvestment, asset sales or financing in future unknown market conditions. Insurers remain exposed to changes in credit conditions, and the level of uncertainty over credit losses, but it is this elimination of the necessity of transacting in order to meet liabilities that allows a proportion of spread to safely be capitalised up front as MA.
If an asset does not meet strict fixity conditions, can we nonetheless find a way of assigning a prudent amount of MA to it? Asset prepayment risk, and deferral risk, is a particular area we are looking at. For example, the existing SII approach to callable bonds may result in an unduly prudent outcome. Our ability now to reform SII permits us to reconsider the approach, without jeopardising the prudence and simplicity of the fixity requirement.
It would not be in policyholders’ interests for their benefits to be backed by an asset pool where sudden deferrals of redemptions could imperil cashflow matching. However, potentially there is scope to allow such risks with appropriate limits, much in the same way as the MA portfolio may carry residual interest rate, inflation and foreign-exchange rate risk.
We are currently asking ourselves whether a risk of such an approach is that any call premium within the credit spread could be inappropriately captured as additional MA, and if so, whether an appropriate response would be to mitigate this through a more risk-sensitive MA. This serves to illustrate the interconnectedness and complexity of drawing up an overall reform package.
As well as larger potential reforms, we are taking the opportunity to consider other changes to the current implementation of the MA. For example, clarifying our expectations for liquidity management for certain construction phase assets may allow firms to invest in productive assets of appropriate term rather than holding cash or gilts. We are also considering whether a softening of the restrictions on cash flow recognition for construction phase assets can be accommodated within the prudential framework.
We are reconsidering the liability risks that the MA portfolio can include. These are currently set out in the SII regulations and most notably exclude morbidity risk. There are issues to consider with including such policies in MA portfolios including the treatment of future premiums, but we recognise that close cashflow matching may be possible.
Similarly, the current MA regulations do not allow splitting of individual liability contracts, and this may be unduly prudent where some benefits are contractually guaranteed, with annuity-like characteristics. Again, where close cashflow matching may be achieved, we will continue to consider what is possible.
PRA approval process
MA conditions refer to both eligibility of assets and liabilities as well as firm processes. So it is important that firms continue to discuss with us in a timely manner how they plan to manage applications where introducing new features to their MA portfolio. Assets and liabilities with similar features on paper can present different levels of risk to different firms depending on their internal processes. This needs to be captured by the PRA’s review.
Good risk management need not entail lengthy or cumbersome processes. We know that applications can be time consuming, particularly where new asset features or asset types are proposed, and this can become resource intensive both for applicants and ourselves. Indeed, sometimes it may be better for the PRA to say yes earlier, but with safeguards, rather than a lengthy application assessment. Or – equally – to say ‘no’ earlier in the process if there is a large gap between an asset’s characteristics and what is needed to make the MA work, or between a firm’s risk management capability in the relevant areas, and a reasonable minimum standard.
Changes to streamline approval processes, where appropriate, are being considered. This could involve a ‘two-lane’ process – wherein complex or precedent setting assets are subject to thorough review before approval, and a more streamlined and fast-track review is applied to less complex assets. Under this approach, certain aspects of the review work, including capital modelling and internal ratings, could be deferred until after initial approval. This might be in combination with temporary safeguards to capture the additional uncertainty in the calculation of the MA benefit pending the further review work. We will also look to build our understanding of firms’ investment risk management capabilities through the BAU supervisory review cycle.
Currently MA regulations on breaches of approval require revocation of MA if compliance is not restored within two months of identifying a breach. This seems an unreasonably harsh sanction. We are considering whether there is an appropriate alternative that softens this cliff edge, while retaining the emphasis on effective portfolio management.
Where new assets are being brought into firms’ MA portfolios this can necessitate changes in their internal models, which can introduce another source of application process. On internal model approvals, we are looking at ways to improve and streamline the framework, including adding additional flexibility in modelling approaches where appropriate. Safeguards may again be effective to facilitate the reformed approach, while ensuring that overall modelling standards are not compromised and any uncertainty or risks are adequately captured. This could achieve a more dynamic model framework that responds more quickly to changes in firms’ risk exposures and business strategies.
Other potential reforms recognise that some of the burden of managing the MA portfolio could be lessened without a material reduction in policyholder security. These operational burdens, for example on the payment of surrender values that exceed the best-estimate liability, are being carefully considered and where we can reduce those burdens in a pragmatic way we shall endeavour to do so.
A particular case of operational burden of interest to this audience is the limitation on trading of assets in the MA portfolio. SII requires maintaining the assigned portfolio of assets, and the PRA’s view has always been that trading for good risk management is appropriate. We believe an alternative approach to ‘buy and hold’ that gives greater scope for trading of assets could be justified, subject to firms holding at all times the assets generating the MA benefit they are taking credit for. This may reduce an operational burden on asset managers. However, as with asset eligibility, it is likely that some safeguards will be needed to ensure portfolios are not being aggressively turned over to the detriment of policyholders.
Facilitating investment into the insurance sector
My remarks so far have focussed on insurers as investors. We are also looking at a number of ways in which we can facilitate competition within the insurance sector, including through easing entry. We are pursuing reforms aimed at facilitating new investment into the insurance sector, which should flow through into supporting the real economy.
We have heard the message clearly that our one-size fits all processes can, at times, prevent market players from grasping the chance to invest in an underwriting opportunity in a timely manner, in particular in the London Market, where opportunities can arise to inject substantial capital to take advantage of relatively short lived opportunities. And international competition for those opportunities is strong.
Improving the responsiveness of the regulatory environment, whilst maintaining proper standards cannot be achieved solely through PRA actions. But there are some steps we can take. We are in dialogue with market participants, including through an industry roundtable this year to discuss the particular circumstances of the wholesale insurance market. As a result, we are considering how we can make some improvements, within the constraints of the current legal framework, to how we engage with applicants who may want to bring traditional capital to the market, or innovative capital through the Insurance Linked Securities regime.
On the retail side, we previously announcedfootnote  that as part of the SII review we have been working on developing a mobilisation regime for insurers, based on our tried and tested regime for start-ups in the banking sector, and adapted to fit the insurance market. Under an insurance mobilisation regime we would look to apply a time-limited and proportionate set of requirements to new entrants to allow them build their capabilities and secure investment capital, before being subject to full SII requirements.
The PRA strongly supports insurers playing an important role in the real economy. We will assess insurers’ ambitions here in the context of the risk they pose to our objectives. Any investment incentives, including that provided by the MA, must work to encourage insurers to invest well. Crucially this includes ensuring that they are able to honour their commitments to policyholders, whose financial relationship with the insurer may far outlive that of the investment manager. Where we have seen good investment risk governance, this has involved boards taking a long-term view of the risks to meeting these commitments.
The SII review gives us the chance to step back and consider how we might pursue reforms which smooth investment processes – both in MA portfolios and in entering the market – while maintaining adequate risk management practices that are in step with appetites for strong investment returns. These, we hope, should position insurers well to perform their key functions in the real economy, including investing in the transition to net zero.
In the coming months we will analyse the results of our Quantitative Impact Study and Qualitative Questionnaire, and will discuss firm’s responses with them to ensure we have understood the feedback they have provided. We continue to work closely with HM Treasury to draw up reform proposals which will be brought to consultation next year.
I would like to thank Cameron Westwood, Alan Sheppard, Anu Ralhan, Saoirse Carberry, Terrence Venables, Francesca D’Urzo and Dean Minot for their assistance in preparing these remarks.