Where Now For Long-Term Guarantees And Solvency II?

On 14 June 2013, EIOPA published its report setting out its technical findings on the Long-term Guarantees Assessment (LTGA). This report was eagerly awaited as it is one of the key steps in a succession of events leading to the implementation of Solvency II. In this Insights, we consider the implications of the assessment for insurance companies and the wider path to Solvency II.

Summary of EIOPA's report

Solvency II is designed to be risk-sensitive and, as a result, solvency positions will vary under different economic conditions, which was observed in the LTGA results. While EIOPA recommends some changes, their overall conclusion is that most of the measures proposed under the LTGA are generally appropriate.

The report has had a mixed reception from the industry. On one hand, it is a step forward in helping lawmakers to understand the implications of the Solvency II proposals and provides valuable insight on the impacts and sensitivities of the various components of the LTG package. However, it does not fully address the main concerns of the industry, which relate to increased balance sheet volatility under Solvency II, and how this might cause pro-cyclical effects which reduce insurers ability to provide long-term products and be long-term investors. This is particularly acute in certain Eurozone markets given that EIOPA has largely resisted calls to allow more national variation to reflect the significantly different government bond yields in different Eurozone countries. Given the proximity of the current Solvency II implementation date, it is far from being as conclusive as the industry might have hoped. In particular, some fairly fundamental changes to the package were proposed but there now appears to be limited ability to conduct any comprehensive consultation on them.

For the industry as a whole, post-SCR solvency ratios were lower in all 13 scenarios compared to the equivalent Solvency I ratios. As expected however, this varied significantly by country.

From the report, it is clear that without applying the LTG package, an unacceptable proportion of the industry would be materially underwater – the weighted average SCR coverage ratio at December 2011 would only be 77%. Under the 'base' scenario, this increases to 143%. However, 29% of life (including composite) undertakings would still not meet their SCRs.

Had Solvency II been in-force on 31 December 2011, a surprisingly large number of companies would have failed to cover their SCR and hence would have been subject to supervisory intervention. This implies that a significant proportion of insurers are under-capitalised under what EIOPA argues is an economic basis. However, given the experience of insurance companies through the various phases of the recent economic crisis, this appears to be a fairly harsh assessment of the industry's capitalisation levels and risk management strategies.

Modified proposals

Based on its assessment of the companies' results and internal analysis, EIOPA has proposed some changes to the LTG package. Key among these are the exclusion of the Extended Matching Adjustment and replacing the Counter-cyclical Premium (CCP) with a new measure, the Volatility Balancer.

EIOPA proposes that the Extended Matching Adjustment is excluded for a number of reasons, the main one being that it could encourage insurance companies to back relatively liquid liabilities with a high proportion of illiquid assets, thereby exposing themselves to the risk of unexpected voluntary leavers. Overall, the measure was deemed to be detrimental to policyholder protection and difficult to supervise.

The loss of the Extended Matching Adjustment will cause concern for a significant proportion of companies, such as those in France where just under 90% of technical provisions would have benefited from it, as well as the majority of the business in Italy and Belgium. It will also be a disappointment to some non-life companies with long-term liabilities (see below).

This is partially mitigated by some modifications to the Classic Matching Adjustment. EIOPA proposes that the principle of substance over form should be applied, so that legal classifications alone should not automatically disqualify some types of business from applying the Classic Matching Adjustment. So for example, cross-border business, reinsurance and non-life business may be eligible for the Classic Matching Adjustment, provided that they meet the criteria. However, the criteria are stringent and will only cover a relatively small subset of the business that was previously eligible for the Extended Matching Adjustment. It also suggests that business with immaterial amounts of mortality risk should also qualify for the Matching Adjustment.

EIOPA acknowledged the industry's concerns with the issues within the CCP, which are:

• The need to hold capital against the risk that it is subsequently withdrawn resulting in a limited or even negative impact in some countries.
• The lack of a smooth trigger process, which could increase rather than decrease short-term volatility thereby causing adverse financial stability implications.

They therefore recommended that the CCP is abandoned and replaced by a new, simpler, Volatility Balancer. Similar to the CCP, this is based on a spread derived from a currency-specific reference investment portfolio but would be permanently available. However, it would only take credit for 20% of the spread within the reference portfolio. EIOPA justifies this as needed to take account of the 'risk associated with implementation' without explaining what risks they are referring to. If there are extreme variations between national spreads within a common currency zone, that is, the Eurozone, EIOPA proposes that a small adjustment be made in the Volatility Balancer to reflect this. However, this and the use of only 20% of the spread mean that the end result is likely to be a similar balance sheet impact to the CCP, which EIOPA themselves say was insufficient. The Volatility Balancer is therefore unlikely to address the concerns of the industry, particularly in Eurozone countries experiencing high government bond yields to swap spreads.

EIOPA appears to favour a longer convergence of the risk-free rate to the ultimate forward rate, suggesting that a 40-year period may be appropriate for Euro-denominated liabilities. This is significantly longer than the 10 years used in most of the LTGA scenarios. This, combined with the definition of the last liquid point, is an important assumption for insurance companies in some major jurisdictions, such as Germany, where business has been historically written with relatively high levels of guarantees.

Whether a faster or slower convergence rate improves the balance sheet surplus or not, depends on whether the forward rate at the last liquid point is higher or lower than the ultimate forward rate of 4.2%. At the current time, interest rates in many major European countries are significantly lower, which would make the modification unfavourable for many insurance companies.

In addition to these modifications, there remain a number of areas which continue to be described as incomplete and still requiring additional analysis, such as the volatility of default probabilities, the application of ring-fencing for the Matching Adjustment and the floor for fundamental spreads.

Key implications of the LTGA

Investment management

Investment strategy continues to be a big issue for companies, particularly for those portfolios which apply the Classic Matching Adjustment. Although we welcome the relaxation of the constraints on liabilities and the decision against a closed list of applicable assets, there is little movement in the stringent criteria in order to apply the adjustment. As a result, the main concerns still remain, in particular:

• The cost and effort of managing a ring-fenced portfolio of assets to back the liabilities.
• The loss of diversification benefit due to the rules applicable to ring-fenced portfolios.
• The low materiality limit for the cash-flow matching requirements.
• The potential ineligibility of material portfolios of assets currently used to back the liabilities, including callable bonds, floating rate notes, residential mortgages and equity release mortgages.
• The cost and effort of ascertaining the eligibility of each asset and liability against the criteria.
• The potential 'cliff' effects of assets falling in and out of the eligibility limits.

The asset restrictions will be of concern to some companies applying the Matching Adjustment, as in many cases, these form a significant proportion of the assets backing the liabilities. Callable bonds, in particular, currently make up a sizable portion of investment for life insurance business with long-term guarantees. Under the current definitions, these will now fall foul of the third party control limitation as the cash flows are not completely predictable.

The proposed principles based approach for determining asset admissability is potentially helpful for equity release assets as their bond-like, predictable and high quality characteristics1 may be recognised. However, these will need to be assessed on an individual asset basis. Where these are written directly by insurance companies, the terms of the contracts may be under the control of the company, which may wish to consider taking legal advice or changing the legal wording of their contracts in order to comply.

More generally, the report acknowledges that there is likely to be significant effort required to analyse each asset and policy against the Solvency II requirements. Companies may now want to start considering Matching Adjustment eligibility now in their purchases to avoid large-scale portfolio movements later.

Business which might have qualified for the Extended Matching Adjustment but which will not qualify for the Classic Matching Adjustment face a significant issue. The report anonymises the pre-LTG package solvency ratio analysis but companies that fall within this category are likely to need some kind of reprieve. In the short term, they are likely to depend on some form of transitional arrangements. In the longer term, depending on the specific jurisdiction, they may be forced to de-risk their asset portfolios towards government and other supranational bonds at the expense of corporate bonds and equities, and consequentially have to accept lower long-term expected returns.

The Matching Adjustment is an effective tool for mitigating short-term volatility and will therefore be attractive to insurers. However, it is very complex and the criteria for its use are onerous. In particular, if the asset eligibility requirements remain as currently proposed, this is likely to result in a significant change in the attractiveness of assets, from a Matching Adjustment perspective, in the period leading up to Solvency II implementation. Insurance companies are major investors in most jurisdictions and although the impact of the market movement cannot be predicted, it is unclear that the consequential adjustment to the balance of supply and demand is desirable.

Balance sheet management

In a number of cases where the LTGA rules prescribe modifications to market-derived data, such as the extrapolation rules and the transitional arrangements, EIOPA acknowledges that these call into question how market-consistent the measures are. Article 75 of the Solvency II Directive requires that assets and liabilities should be valued at the amounts for which "they could be transferred, or settled, between knowledgeable willing parties in an arm's length transaction". It could be argued that in some cases, this requirement could not be met.

There are then various practical implications of this. In particular, it creates management issues for companies. The approach results in potential differences between the Pillar I balance sheet and their economic balance sheet. It has been suggested that companies will need to calculate both, the former for regulatory reporting and the latter for ORSAs, resulting in extra cost of calculating, reconciling and managing the different balance sheets. As a result, companies will need to make conscious decisions as to which they wish to optimise under certain conditions. In practice, this manifests itself through difficult decisions that companies will need to make with regard to investment strategies, hedging decisions and other risk management strategies. Although not an impossible task, it is unclear that the extra overheads add value to companies or their policyholders.

Contract boundaries

Although not formally tested under the LTGA exercise, companies did need to interpret contract boundaries for their business in preparing the results. The guidance provided within the question and answer process implied that companies should take a strict legal interpretation of the rules, resulting in onerous and potentially significant negative impacts for companies' solvency positions.

We recommend that companies investigate their legal policy documents with a critical eye and consider whether contract boundary requirements are likely to have significant adverse consequences. If so, a number of options are possible, such as monetising the value falling outside the contract boundary definitions and modifying the terms of the contracts. Both of these options are likely to require long lead times to implement.

Products

Conventional wisdom dictates that insurance companies would seek to take advantage of the most beneficial rules. As the Volatility Balancer is likely to be less beneficial than the Matching Adjustment, it is possible that more companies in more jurisdictions will seek to redesign their products so that they are eligible, where this is possible. However, as the Matching Adjustment comes with stringent criteria that limit the features of the products, an unintended consequence might be that policyholders will have less choice in the type of products offered and that the new generation of insurance products will not necessarily meet all of their needs.

Non-life business

There are certain types of non-life obligations that could benefit from the LTG package, for example, annuities stemming from non-life insurance business (for example, workers' compensation) and health insurance similar to life insurance business. The LTGA results appear to reinforce the generally held belief that non-life business is not materially affected by the LTG proposals. However, we believe that the situation may not have been adequately investigated within the exercise.

For example, in the UK motor industry, a number of non-life companies now have material exposure to annuity-type liabilities in the form of periodical payment orders (PPOs). These are long-term annuity payments made to victims of motor accidents with severe injuries. As a share of premiums, these are small but comprise a material and growing proportion of the reserves. The average term is 40 years, which is longer than for retirement annuities. In addition, PPOs are linked to the Annual Survey of Hourly Earnings (ASHE), a measure of earnings inflation.

 It had been hoped that PPOs could benefit from some form of the Extended Matching Adjustment, but EIOPA is now recommending that this be removed from the LTG package. PPOs are unlikely to qualify for the Classic Matching Adjustment due to the stringent matching criteria attached to it. Due to the very long-term of these liabilities, PPOs will be disproportionately affected if liquidity cannot be allowed for. Transitional measures and extension of recovery period

Although the intention is sound, transitional measures have proven more difficult and controversial than initially thought. EIOPA states that "... though there are a number of drawbacks to the LTG transitional measures, there do not seem to be real alternatives to implementing these measures for existing business in cases where the change of regulatory framework results in significant negative impact on solvency positions." Indeed, the transitional measure tested under the LTGA resulted in an overall reduction in surplus and SCR/MCR ratios for some countries.

A key flaw in the proposed approach is that adjustments are made to liability values without also considering asset valuation differences between Solvency I and II, that is, it only considers half of the balance sheet. Also, the dynamic component is derived using a weighted average liability discount rate of the then Solvency II discount rate and the Solvency I discount rate, fixed at the outset of Solvency II. This could cause significant artificial mismatch volatility in countries such as the UK, Ireland, Denmark and Holland who have a more market-based implementation of Solvency I. Assets would move in line with economic conditions but the liabilities would only partially do so because of the weighting to a past discount rate. Significant further work is needed in this area.

Where does this leave the implementation of Solvency II?

The implications of the LTGA go beyond the current impact assessment, as they have longer-term implications for companies' financial strength under Solvency II, financial reporting, risk management, investment and hedging strategies, product design and pricing, and also internal models and validation.

The balance sheet volatility attached to long-term guarantees under market-consistent reporting has been intensely debated since the onset of the financial crisis. Over the course of the political debate, the terms of the mitigation measures were moved into Level 1 legislation which, unfortunately, meant that the lack of agreement was holding up Solvency II implementation more generally. The LTGA exercise was intended to provide a basis for the legislators to agree on the exact terms prior to the Omnibus II vote.

Having now seen EIOPA's LTGA report, what are the main scenarios for the implementation of Solvency II?

Although it is an important step forward, the LTGA still leaves many unanswered questions. It should be noted that besides the long-term guarantees issues, there remain a number of other unresolved hot potatoes such as the look-through basis, equivalence, catastrophe risk, treatment of pension schemes and contract boundaries.

As a result, full implementation of Solvency II in the near future is unlikely. January 2015 has been touted as the earliest possible date provided everything progresses smoothly from now, but the historical pace of progress suggests that is fairly unlikely. 2016 is cited as a more likely date but there is a healthy contingent which believes that it will be implemented much later than that.

There is an added dimension of uncertainty in that the 2014 EU elections will now be held before the implementation of Solvency II. The outcome of this, of course, is unclear; particularly if the economic crisis continues unabated. If there is a major overhaul in the membership of the European Parliament, Council and Commission, it is possible that the drive to completion may ebb in the immediate aftermath.

There is some evidence that EIOPA and national regulators, like the insurance industry overall, have lost patience with the political process and have publicly expressed concerns about the costs and the lack of benefits to policyholders.

Although nothing can be ruled out completely, a total collapse of the legislation is unlikely. This would not only be politically unacceptable given the amount of money that has been spent by the industry on meeting the new requirements, but would call to question the credibility of the European political process.

Most parties agree that the patchwork of old regimes does not meet the standards for an increasingly complex insurance industry and there is much in the proposed regime that is fairly uncontroversial. For example, a large proportion of the Solvency II Directive articles merely consolidate existing Directives such as the Life Assurance Directive, Insurance Groups Directive and Reinsurance Directive. And many of the new proposals have been well received, such as those in regard of the Own Risk and Solvency Assessments (ORSA) and co-operation between national supervisors.

Viewed in its entirety, the implementation of Solvency II is really being held up by a small, but critical, number of issues. In particular, those related to the valuation of long-term guarantees. Although these are difficult and fiercely debated, there is little reason for them to continually postpone the parts of the Solvency II regime that are generally recognised to be good practice, for example, risk management and governance.

There therefore appears to be some renewed political will to move the Solvency II project on in the near future. Ahead of this, various national supervisory authorities have already effectively instigated phased implementations of Solvency II and this will be seen even more when EIOPA concludes on its consultations on guidance preparing for Solvency II later this year.

What happens next?

The EIOPA LTGA report is one in a chain of events leading to Solvency II implementation. Following this, the Commission is to provide its report to the Trilogue by 12 July 2013. This second report is intended to be a more strategic report based on the Commission's view of EIOPA's technical findings.

The Trilogue will then base their decisions on the reports provided to them. However, historical precedent has shown that they do not always act in line with EIOPA's advice. As such, there may still be some surprises in store for insurance companies.

At the current time, the Parliament vote on the Omnibus II Directive is scheduled for 22 October 2013. This means that there is effectively very little time for the Trilogue to agree on the details of the LTG package, particularly the proposed new measures such as the transitional arrangements, the Volatility Balancer, modifications to the current proposals and the many areas which EIOPA still says requires consideration. In doing so, the Trilogue will need to deliberate on the appropriate balance between protecting policyholder interests and retaining a strong and vibrant insurance industry in the long term.

This is particularly difficult while European countries have still not fully emerged from the economic crisis and the long-term outlook is still unknown. Specifically, while interest rates remain low in some parts of Europe and high in others, it is unclear that any LTG package that is implemented will prove to be robust in the long term.

What does it mean for companies?

Since the Solvency II delays, many companies have wound down their Solvency II programs and transferred at least parts of it to their business-asusual teams. If Solvency II were to be reinvigorated, some parts of the process will need to be ramped up again. Setting this in motion, of course, takes time and given the implications for the industry, needs to be done in an organised and considered manner.

It was also clear, particularly from the participation rates of some of the LTGA scenarios, that there were still significant difficulties in complying with the various requirements. The exercise was performed on a 'best endeavours' basis and in practice, many companies have employed manual processes and judgement to provide the numbers that were required. However, adaptation of many insurance company models and processes require a long lead time to implement. To do so in a robust and reliable manner that would meet the standards of financial reporting is a significant task for most companies. It is therefore imperative that the requirements are finalised as soon as possible to give companies the best chance to do so.

Conclusion

It is critical that Solvency II is appropriately implemented. The on-going debate over the LTG package is causing serious delays, which are both damaging for the industry's reputation and costly. In the past few years, the industry has put an enormous amount of money into building models, setting up governance and data systems, and reorganising themselves in pursuit of a pan-European solvency regime which would create a harmonised, level playing field for insurance companies across the region. It was hoped that we would be left with a modern regime to replace the patchwork of out-dated Solvency I rules and a risk-oriented framework that would encourage companies to understand and manage the risks within their businesses.

The intentions were sound but in practice, Solvency II has proved to be an enormous distraction from the day-to-day task of managing the businesses. Resources are limited and it could be argued that companies have not had the chance to implement some of the good risk management practices or strategic corporate aims that they may have wanted to. It would also not be inaccurate to say that some companies may have held off taking some actions that they may have otherwise have wanted to while Solvency II rules remained unclear, in order not to be adversely impacted by the rules when they were confirmed.

In combination, these have not been beneficial for the insurance industry and if they were to left to continue, could prove to be fairly damaging. In order to avoid further negative impacts on the industry, it is important that the outstanding issues are clarified and a clear path to Solvency II is bedded down with certainty.

EIOPA's report can be found on this webpage:

https://eiopa.europa.eu/consultations/qis/ insurance/long-term-guarantees-assessment/ index.html

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.