25 January 2021
Carlos Montalvo Rebuelta analyses EIOPA's advice to the European Commission on the Solvency II 2020 review, highlighting the points of contention between regulators and policy makers
When I saw last week that my comment piece on the Solvency II 2020 review was amongst the most read article on Insurance Asset Risk in 2020, I thought that such interest deserved a follow up. And what a better way to start 2021 than sharing with you candid views around the so called 2020 Review of Solvency II, and EIOPA opinion on it.
In particular, allow me to share preliminary thoughts around EIOPA's, but also around the process and the timing of the review - as they matter. Let me as well underline the obvious, even at the risk of being redundant: today´s reality, from negative rates to technology, from Covid19 to the relevance of sustainability, differs from that of 2016, let alone of 2009; and the review must ensure that Solvency II remains up for the task, also in the current environment.
Starting with the conclusion, it may well be one of "No need to panic, not now". It is fair to say that changes proposed bring more prudence in most areas addressed. But it is also true that in other areas, such as the approach to the risk margin, the regulators acknowledged that improving design deficiencies should prevail over considerations on solvency impact... and that the willingness to improve the framework, including in areas where it means reducing prudence is a step in the right direction.
A clear understanding of the process is key to interpret where we are and where are we heading towards: EIOPA is not a decision maker, but a technical expert body. On this basis, EIOPA´s advice will be one of several inputs to the Commission´s regulatory proposal. Furthermore, the Commission´s proposal will not only be based on EIOPA input but also on its own consultation process, giving policy makers the freedom to deviate from EIOPA's advice.
In other words, as relevant as it can be (and EIOPA has done a massive exercise to assess the impact of what it proposes), the regulator's advice is still part of the input to the proposal, and not the proposal itself. Plus, this proposal will then be discussed, amended and finally adopted by the co-legislators (Council and European Parliament).
I mentioned that there is no need to panic. Indeed, what EIOPA is proposing may not be seen as balanced by some (even by many?), yet let me be provocative by posing a very politically incorrect question: Do you expect the input from the three EU Institutions (Commission, Council, Parliament) to bring more or less prudence to EIOPA's proposal? If, as Lord Byron said, " The best of the prophets for the future is the past", then we got a clear answer!
EIOPA chair Gabriel Bernardino told InsuranceERM that the overall changes produce a "balanced outcome". Allow me to help interpreting this message, as referring to outcome may be misunderstood by many who are not familiar with the process. Perhaps referring to a proposal, rather than an outcome, would have been more accurate?
In terms of content, the advice focuses mostly on the Call for Advice by the Commission to EIOPA. Please note that I have used the word mostly, as EIOPA covered all the areas within the Call for Advice in their response but also added additional elements they deemed relevant to either enhance the framework, its convergence and consistency or its supervision.
Interestingly enough, while many think macro-prudential considerations, recovery and resolution or insurance guarantee schemes (IGS) were outside the scope of the Commission original request, they were in fact explicitly included.
Long term guarantees (LTG)
The LTG measures were not part of the original Level 1 Directive, as adopted back in 2009, yet can you imagine Solvency II without it?
Solvency II without them would have been a failure, creating problems to many insurers in terms of regulatory compliance and even insolvencies. But it is obvious that keeping the LTG measures within Solvency II is a must; improving them, too. And we see how the proposed changes on extrapolation, volatility adjustment (VA) and matching adjustment (MA) will make the measures both more effective and more usable.
In particular, it keeps surprising me to be honest, how little attention the MA has received in the debate. It should be the most efficient way to allow insurers to invest long-term (earning in its entirety a most needed illiquidity premium) and to offer guaranteed products to policyholders that don´t want to bear all investment risk, two key objectives for the Commission.
Yet it continues to be overlooked, perhaps because it is still seen as a tailored solution to UK and Spanish products, rather than as a mechanism, built on sound ALM, to earn 100% of the spread, and one that could be the basis for designing new products. Why don´t we reset and start considering this as a new opportunity to meet policyholders demand for guarantees, even at a time of negative rates?
Interestingly enough (as well as surprisingly, as innovation usually tends to come from insurers rather than from the regulatory community), from my conversations with both regulators and industry, it is the supervisory community which is more open and positive to design products compatible with the MA (underlined in EIOPA's suggestion to remove the current limitation to benefit from diversification in MA portfolios).
I just wonder why they are not more vocal about it, as there is a risk that the MA becomes a lost opportunity for EU insurers.
Ignoring a problem is always a bad thing... but providing a response that exacerbates such problem is certainly not a good one, Is it? The approach taken towards extrapolation is an excellent example for this.My personal opinion is that EIOPA has found the right balance in this -complicated and very sensitive- case. There is a problem with the current methodology that leads to under-provisioning in the current negative rate environment, as well as to relaxing risk management around it.
There can be an equally big problem if we changed the approach to the Last Liquid Point and replace the current 20 year for 30 or even 50. When looking at the numbers provided under the impact assessment undertaken by EIOPA, such change would create more problems than it would solve. However, the new proposed methodology (see box out) serves as a reminder that risk management and a clear understanding of the real value of guarantees matter for the business, regardless of what a regulatory number says, but it also limits the impact of the change to a level that can be assumed by the players.
EIOPA's proposed amendment for the extrapolation methodology:
EIOPA suggested replacing the LLP with a First Smoothing Point, but keeping the 20 years, and the 40 years convergence period, extrapolating up to the Ultimate Forward Rate through the use of a weighted average of forward rates and the use of a convergence parameter; then completing the approach with a sensitivity analysis and a mechanism to limit the effects of the change of methodology in periods of very low rates.
Concerning the VA, a new design is proposed by EIOPA, one that differentiates between a so-called permanent VA and a macro VA, combined with a revised general application factor (from 65 to 85%) and the addition of two additional application factors to deal with over-shooting and illiquidity of the liabilities.
These changes reinforce the positive effects for its users in terms of available own funds, as it will reduce the value of technical provisions, but expect a lot of attention and debate moving forward, as it doesn't meet the expectations of industry around it (nor does it bring any comfort to a relevant number of supervisors). However, don´t expect such debate to cover the enhanced expectations from supervisors, be it for VA or MA, around liquidity planning... this is here to stay. Furthermore, it is also entering in new areas such as recognition and allowance of long-term equities with a 22% capital charge rather than the 39-49% standard charges.
Talking about the treatment of equity risk and the concept of long-term equities, my take is clear: regulators may still dislike the proposed approach (EIOPA wording is pristine around that, referring to a 2010 advice it still backs, and expectations around well-diversified portfolios only reinforce such idea), but they acknowledge that this is not the right battle to pick, taking into consideration the strong political support to the idea by key Member States, be it when it comes to the capital charge or to conditions for considering such investments as long term.
The proposed approach, namely Show me that you can hold these assets for long and not only that you want to do so, makes sense... as it also does, in an environment of negative rates, to create the grounds for insurers to further invest long-term in these assets as a way to enhance returns (if, and only if, they are well positioned, from a risk management, solvency and liquidity standpoint, to do so).
EIOPA proposals on Technical Provisions focus on enhancing convergent treatment by insurers, in areas such as boundaries or expenses - This is consistent with the 2021 action plans of a number of supervisors, EU or UK, sufficiency of technical provisions and accuracy around their calculation stands as a key theme.
Most important is the acknowledgement that the Risk Margin has to be amended, in particular regarding its sensitivity to changes in interest ratesa decrease in rates increases significantly the amount of the Risk Margin, specially for long term products, penalizing them), where a floor is settled, and the dependence of risks over time when projecting future SCRs. Both corrections do make sense and are steps in the right direction.
Let´s please avoid, if I may suggest so, that the debate focuses on the other elephant in the room, i.e. the Cost of Capital factor, currently at 6% and on which EIOPA calculations show it should be higher, which is why it proposes to keep it.
EIOPA insists on the need to change Interest rate risk calibration, and the wording it uses throughout its opinion (e.g. "strongly advises to change", "the module severely underestimates the risk") can be seen as an indication of the fact that this matters to them... a lot.
Aware of the impact recognizing negative rates would have here, EIOPA is also willing to introduce a phased-in approach for 5 years, to make the change more acceptable. My view is it's difficult to challenge the idea that negative rates have to be accounted for, at least intellectually. I expect this proposed change to end up in the final text, and I am also convinced that the final outcome of the review process will not lead to an increase of the overall solvency requirements (more to the opposite, if we consider changes to technical provisions or LTG measures).
The pending question is to identify in which other areas further relaxation may come... and spread risk treatment immediately comes to my mind, considering that the design of it, with shocks to spreads with a probability of occurrence of 1 year, may lead to an overestimation of the capital requirements and discourage long-term investing (which happens to be one of the political goals of the Capital Markets Union).
Let me be politically incorrect, for a second time in just one article, so two too many: at the end the underlying question is whether political goals should entail setting regulatory incentives that are not risk sensitive in a risk-based regime. For example, EIOPA claims that there is no evidence of disincentives to invest in corporate bonds, but a different way to look at it is whether they should be an incentive.
Without any relevant news around Own Funds (beyond the recognition of the existing differences between Insurance and Banking frameworks), the next two areas I wanted to reflect on have in common the acknowledgement that things can and should be improved here; I refer to proportionality and to disclosures.
On reporting and disclosure, I still have to meet a policyholder that has read a Solvency and Financial Condition Report (SFCR), currently on average a document between100 and 200 pages. So, the suggestion by EIOPA to have an SCFR for investors and another one for policyholders can be a good step towards providing the right information rather than more and more information.
But of course, there is the risk that this only leads to more burden to insurers... and such outcome has to be avoided. The Why, and the What make full sense; now let´s get the How right. And such change should be compatible with more work around other reductions of information that is not needed, or not used... and such type of information currently exists.
Let me share an idea that may help here: What about identifying areas where the request of the information would be "on demand" only? In the short term it would not reduce the cost to insurers, as they would keep collecting such data; but it would create a proper incentive for regulators to reduce the level of reporting, and to be open to doing so in more areas - as they could still ask for it - and it should pave the ground towards a permanent reduction in the medium term (kind of a phasing out model).
But the aforementioned does not address what in my view remains one of the biggest failures around Solvency II, namely that we are focusing too much on a number, the SCR ratio, and too little on how you get to that number, a situation that has led to comparing apples with pears (as an SCR of 200% with transitionals, surplus funds, high level of expected future profits, etc. has nothing to do with another one of 200% without these features).
The best example to illustrate this point is the debate on the use of transitionals after day 1, in particular those for technical provisions: what the market should have rewarded as a signal of strength and solvency (i.e. not asking for them) has ended up penalizing the comparison of those entities that decided not to use them.
Perhaps if the regulators had been more active in terms of acknowledging these situations - I was a regulator at the inception of Solvency II and I also failed on this - we all would be in a better situation now, spending time in understanding how you reach a given number rather than being complacent with the number and ignoring the underlying assumptions that build such number. An idea I launch, even if too late, to my former colleagues: factor in positively, for those entities that do not use these measures, this approach when it comes to dividend distributions. Perhaps a way to get back to what we had in mind when designing the framework?
Regarding Proportionality, the intention to enhance it should be supported, as the extent to which the principle of proportionality has been extended and used is not a success story, let´s be honest. But please allow me to raise a word of caution: I have heard that there is a possibility currently being considered by the EU institutions to make proportionality application "automatic": in other words, that if you meet the criteria to be considered a low risk profile undertaking, no supervisory discretion should apply.
The reason why I caution on such approach is simple: EIOPA proposal as it stands would allow for applying simplifications (e.g. regarding calculation of the Best Estimate or BEL via deterministic methodologies) also to undertakings not meeting the aforementioned criteria but for which it would make full sense to extend the simplified calculation.
If the decision is to remove discretion, rest assured that this will operate both ways, and discretion to apply simplified methods to undertakings not considered as low risk profile will vanish. As per the reporting comments, let´s ensure that proper focus is also put on simplifying quarterly submissions, removing templates...
The proposed changes regarding Group supervision aim at responding to a factual situation as of today: there is a lot of uncertainty regarding the scope of group supervision, regulators are giving different responses depending on the country or on the transaction, and this is not good for the business. Extending the scope all the way up to the ultimate holding company, capturing horizontal groups or ensuring clarity around transferability and fungibility may imply a change in the Directive to clarify it, but de facto it does not change what some regulators are already requesting, either directly or indirectly. And regulatory certainty is always good for business... and maybe the starting point for waiving requirements that do not add to effective supervision, but only to administrative burden.
Content wise, introducing macro-prudential policy considerations in Solvency II, will be one of the biggest changes to the framework, and please note that I use the word "will", as I don´t see any reason that will lead to not introducing this dimension in the framework, also considering that Europe is a key member of the IAIS and the global trends are heading into this direction. Having said so, and trying to forecast how this will be introduced, I don´t expect that EIOPA request to set up a capital surcharge for systemic risk, or the setting up of countercyclical buffers stand a chance to end up in the final text.
A different outcome, however, can be expected regarding Liquidity risk. In my view, they are both useful tools, and should enhance risk management of insurers. Ideally, liquidity risk management plans should also support decision making around dividend distribution... including for insurers to reinforce their case towards the supervisor and (positively) affect the "exceptionality" assessment that, as per EIOPA advice, supervisors should take into consideration before limiting such distributions. It should also be a welcomed approach that liquidity risk does not become a Pilar 1 risk for which a capital charge applies.
On pre-emptive recovery and resolution, I should repeat myself... full support to the idea of recovery as a key risk management tool to avoid gone concern situations; strong doubts around the extension of resolution plans and requirements, that remain the responsibility of the supervisor, to all insurers.
Putting it in a descriptive way, working everyday on making your marriage a happy one should be more effective towards preserving it than preparing and agreeing on a sound divorce package should this happen. I see this logic applying to resolution and recovery as well. Having said so, I am convinced that recovery is here to stay, and that the approach taken will not be fundamentally different to the one we have already seen in Banking.
Regarding Insurance Guarantee Schemes, where EIOPA proposes that all EU countries have an IGS in force, I have mixed feelings, as I see and share the Why, but have serious doubts regarding both the What and the How (ex-ante vs ex-post funding, scope, Home-Host model... just too many open questions where the responses provided don´t give me assurance that we are getting it right).
Having said so, I remain a strong believer regarding the fact that the design of Solvency II should allow for failure of insurers.
To conclude, there is a key element I wanted to raise, as it has to do with the timing of this review. Where all changes would be channeled via Level1 (the Directive), and working on the assumption that the Commission will present its proposal around my birthday (month of July), standard Brussels process would mean an expected application date around 2025.
The big Question mark is: Will the Commission propose that Level 2 (Regulations) changes are aligned time-wise to the Level 1, or will we see a "fast track" Level 2 proposal, addressing inter alia Sustainability or long-term investing considerations?
My personal view is clear: there are areas (linked to the EU strategic agenda) where we cannot afford to wait; therefore, where this is legally possible (i.e. there is room to change Level 2 without amending Level 1), we all should push for it.
Good advice may render ineffective if implemented too late, and in this particular case, there is an option to accelerate certain changes via Level 2, so why not embracing it?
Carlos Montalvo Rebuelta is a partner at PwC. Previously he was executive director at Eiopa.