UK Risk Settlement. Market pricing

Transcription

1 Aon Hewitt Retirement & Investment UK Risk Settlement Market pricing The annuity market has continued to offer pricing at historically favourable levels of pricing over the autumn, reflecting successful asset sourcing from insurers across a range of asset classes, and competitive pressure. The chart below shows recent pensioner pricing. Solvency levels for many schemes have been improving significantly, driven not just by improved annuity pricing, but also positive investment returns over the past few years and record amounts of deficit recovery contributions from sponsors. This has led to a significant increase in demand from schemes since August Accordingly, 2018 is now expected to be a milestone year for the annuity market. The insurers are showing interest in annuities at all November 2017 sizes and are increasing their capacity to provide quotations with continued investment in teams and IT. The strong capacity for longevity reinsurance, and attractive yields from more illiquid assets, will also continue to support the market in We would expect most schemes to have scope to benefit from annuity purchase at recent prices. However, we continue to see a wide variation in how annuities are captured in asset flight plans and integrated risk management, meaning many schemes are not well set-up to be first in line for next year s opportunities. This does not mean the current price opportunity will be lost for schemes not in the market but it is important to plan the auction carefully. How to read this chart This shows the return from a bulk annuity for pensioners, relative to the yield on a comparable gilt portfolio Annuities shown as 'Cheap' if giving a better return than gilts This comparison ignores the material value from annuities giving a better hedge including longevity cover Expected pricing for a typical scheme is shown by the blue line Best prices typically fall in the darker shading, some auctions fall in the lighter shading. Pricing outside the shading typically represents an unusual liability profile Chart sourced from Aon's Risk Analyzer

2 Buy-out without full funding Pension schemes do not have to be fully funded to secure an annuity policy. Whilst it is common for schemes to secure an annuity for part of the current pensioner population, it is also possible to design annuities that address a greater part of a scheme s risks while an annuity to cover all liabilities remains unaffordable. Under one example structure, an annuity is purchased to cover all risks except for the earlier years benefit payments. The responsibility to fund the payroll over a short fixed period is retained by the scheme. With many sponsors making material contributions in respect of past benefits, it may be possible to meet the shorter term payments from a mixture of short-term investments and the use of these future contributions. The bulk annuity purchase then removes all of the longer-term risk from the scheme, including almost all longevity risk. In this way, the sponsor can fix its costs at an acceptable level, with very little scope left for cost variation, while the trustees and member benefit from the substantially increased security which follows from the bulk annuity purchase. This contract structure may be particularly attractive where the sponsor covenant could weaken over a longer time period, as it sets out a pre-agreed contribution schedule in the short term. In any case, insurers are able to accommodate extra flexibility in the contract so that the amortisation period can be extended if the financial circumstances of the scheme should make this necessary. The downside of extending this period is the corresponding delay in a full buyout and scheme wind up, so the costs of keeping the scheme running continue to be incurred. For many sponsors, a buy-out will mean a loss in the corporate accounts which may be easier to absorb at some times than others. Securing a bulk annuity which covers the vast bulk of the scheme liabilities, but where the timescales for a buy-out are less prescribed, and can be agreed to fit in with a sponsor s objective, may well be a desirable solution. Recent pricing reflects strong insurer capacity to absorb current levels of demand from schemes. As the annuity market becomes more active, in tandem with the increasing maturity of DB arrangements, the competitive pressure may decrease in time and prices increase as a consequence. A well-funded, well prepared scheme with a clear buy-out objective may then have little reason to delay approaching the market to establish what may be available buy-out may well be closer than you think. Risk Settlement Bulletin 2

3 Setting your annuities strategy The number of schemes becoming repeat buyers of annuities, covering different tranches of members over time, has increased in recent years. In 2016, these phased transactions made up more than two thirds of the 10.7bn of annuity business placed. However, a phased approach to securing scheme liabilities with insurers is not suitable for all schemes, and we are often asked what strategy is the most effective to meet a scheme s objectives. Should the scheme secure liabilities gradually over time ( phased approach ) or secure all liabilities in one transaction at the point of buy-out and wind-up? The answer is very scheme specific, but there are a couple of golden rules that will apply in the majority of cases. Scheme size In theory, transactions are possible at all scheme sizes. In practice, for smaller schemes (e.g. total assets less than 20m) it is likely to be more cost effective to complete one transaction rather than have a phased approach if the scheme circumstances allow. Investment flexibility - Entering into an annuity policy is an irrevocable decision, and the scheme will only get assets back in very limited circumstances. It is therefore critical that any assets used to pay an insurance premium for a phased transaction are not going to be needed by the scheme in the future, mainly for growth, liquidity or collateral. Treating deferred members consistently For members that have not retired, one of the impacts of buy-out is the transfer of the responsibility for setting member factors, from the scheme to the insurer. These factors are used to set transfer values and the terms for taking cash on retirement, for instance. Schemes normally secure all deferred members with one insurer, typically in one transaction, to ensure consistent factors across the membership. This is not an issue for pensioners, so different annuities for different groups of pensioners are often secured over time. Reinsurance Most annuity providers will need to access longevity reinsurance terms, as a critical aspect of reaching a competitive price. So far, longevity reinsurance has been difficult to access for an annuity transaction where the majority of the liabilities (70%+) relates to members that have not yet retired. A scheme securing a series of annuities should consider the likely earliest timing of its final annuity purchase in order to ensure that an appropriate mix of both in payment and deferred members are included. Otherwise there is significant risk of a toxic final tranche of liabilities which may be uninsurable. Use technology to capitalise on opportunities How can we help you decide on your asset and de-risking strategy? When considering long-term planning of investment strategy, our clients are increasingly using our Viewpoints framework to agree objectives: Anonymous gathering of views on key issues from key stakeholders, using a series of carefully targeted questions; Feedback sessions to discuss the implications of the range of views; Interactively illustrating the impact of different objectives and asset strategies suggested by the discussions, using our Risk Analyzer software; Agreeing the resulting objectives, so that that preferred strategy can be fleshed out in detail and adopted. Obtaining clarity on your objectives is key, as only then will you know what success looks like. An example objective could be Reach full buy-out funding in 10 years, whilst minimising the Company contributions required over the 10 year period, subject to being at least 50% invested in annuities. In this case: 10 years is likely to relate to a key point in the scheme s membership evolution or potential covenant support, and The desire to be at least 50% invested in annuities represents the value that the Trustees and Company place on making Risk Settlement Bulletin 3

4 progress towards their end goal and the matching characteristics of annuities. It won t always be possible to define your objectives neatly in advance, and often discussing the output helps to focus the mind on what is or is not important in the scheme s strategy. Even if your scheme is some way from being fully funded and able to buy-out and wind-up, it is worth agreeing your strategy for securing liabilities with an insurer now, and periodically checking that strategy remains appropriate over time along with monitoring progress against key metrics for a transaction. Hidden benefits of buy-out? Scheme managers who have considered their end goal have often reached the conclusion that they will either target buy-out, transferring the pension risk to insurer, or target a low-risk "self-sufficient" status that meets the fiduciary obligation of paying members with minimal ongoing reliance on a sponsor's covenant. Schemes take a variety of different routes to get to these end states. The choice to favour self-sufficiency over buy-out often comes down to the perception of cost. Buy-out is viewed by as a more expensive target, with expectations being that there must be some regulatory arbitrage in running on the scheme under a pensions regime rather than paying for the greater security demanded of insurers. However, buy-out and annuity strategies have a number of advantages over running the scheme on that emerge over time, some of which are set out below. Long-term asset risk - Self-sufficient asset strategies assume investing until all benefits are paid indefinitely perhaps for the next 50 years maintaining the associated asset default risk over this period. Long-term covenant risk - The strategy may effectively assume that the sponsor remains around for this 50 year duration, and be willing or able to bear the responsibility for residual risk over this period. Potentially reducing efficiency - Over time shrinking schemes get less efficient to run with expenses becoming increasingly disproportionate. Asset optimisation - Many schemes will adopt a relatively narrow range of asset strategies or funds to limit the governance requirements on the scheme. They may be less able to access some of the best-yielding assets that match liabilities outside of the traditional bond markets. In contrast the insurers can often source assets within their business, tailoring the income from them to suit pension payment commitments, such as payments from commercial mortgages or infrastructure projects. The larger size of asset pools held by insurers also helps them to maximise risk diversification, and to apply more resource to risk monitoring and asset sourcing. Longevity - Schemes that do not address longevity risk under annuities and longevity swaps still carry a material risk that is not expected to be rewarded. Future change - Pension schemes may experience future legislation that tightens current funding and investment freedoms, reducing the difference between the pension and insurance regimes. Sponsors also continue to carry the risk of legacy data issues coming to light eventually. Concentration risk For most schemes, larger value members will influence a scheme s risk profile considerably more than in an annuity fund (the funds are typically 15bn- 50bn in value). For these reasons, we expect most schemes to gradually accept that a self-sufficiency target is ultimately an interim position, with the sponsor seeking buy-out once the scheme reaches a position of much greater maturity, and possibly with a series of buy-ins along the way. Risk Settlement Bulletin 4

5 Longevity swap activity picks up Having taken some time to reflect the latest mortality trends, reinsurers in the market are now better reflecting the heavier recent deaths in their pricing. This means that pension schemes looking to hedge longevity risk can do so with reassurance they will be paying a fair price for the insurance cover based on up to date information. This improvement in pricing means that greater activity is expected over the coming months and indeed into next year. In particular, there is pentup demand from a number of pension schemes who paused longevity broking processes when it became clear that reinsurance pricing was lagging behind the most up to date information during the second half of 2016 and the first half of In addition, there continues to be strong demand from UK insurers to hedge the longevity risk relating to both new bulk annuity business and existing back-book exposures (driven by the implications of the Solvency II capital regime). All of this means that 2018 could be one of the busiest years to date for the longevity hedging market. It is also interesting to note that there have been a number of sub- 500M transactions over the past few years as the market has become increasingly accessible to varying sizes of pension scheme. However, with significant demand from multi- bn pension schemes and the insurers themselves, smaller pension scheme transactions will need to be carefully positioned in the market to ensure provider engagement. Aon continues to advise schemes of all sizes, starting with providing a realistic assessment of the cost of hedging longevity risk and unbiased advice on the most appropriate method for doing this. In a busy market, it is vital that all stakeholders are enabled to make informed decisions, so that when the time is right to approach providers for pricing, you can demonstrate a well thought through business case with clear targets for success. If you would like an assessment of longevity risk in your scheme and details of the hedging options, please get in touch with one of our specialists. Mortality projections model CMI_2017 a first look The Continuous Mortality Investigation (CMI) published their latest mortality projections model (CMI_2016) in March This model incorporated mortality data for England & Wales up to the end of Because it reflected the higher than expected mortality rates that have been seen in the population in recent years, its adoption typically led to reductions in life expectancy compared to other recent mortality projections models. For example, using the same long-term rate for both models, our preferred calibration of the CMI model typically led to reductions in liabilities of around 2.5% compared to CMI_2014 and around 1% compared to CMI_2015, although some individual schemes will have seen reductions either higher or lower than these. We believe that the adoption of the next model (expected to be named CMI_2017 and to be published early in 2018) is likely to lead to a further reduction in liabilities compared to CMI_2016 our analysis suggests that the impact is likely to be a reduction of around 0.5% to 0.75%, based on information available to date. More detail on our approach is set out below. On a weekly basis, the Office for National Statistics publishes statistics on the numbers of deaths registered in England & Wales. Using this data, it is possible to come up with an estimate of the next version of the CMI s mortality projections model. Mortality experience over the first three quarters of 2017 has again been heavier than projected by the CMI_2016 model, which is why we estimate CMI_2017 is likely to produce lower liabilities compared to using CMI_2016. Our analysis also suggests that the largest reductions are likely to be seen at the oldest ages. It is important to note that this is based on assumptions about population-level experience for the end of We also assume that the CMI will not change the structure of the model between CMI_2016 and CMI_2017. To produce our estimate, we have used actual recorded deaths up to the end of September 2017, with a neutral estimate of the number of deaths Risk Settlement Bulletin 5

6 that might be expected in the final quarter of the year. In practice, the actual deaths experienced in England & Wales at the end of 2017 may be higher or lower than the neutral estimate, and these actual figures will feed into the new model. One reason why mortality rates might be expected to be higher in the late part of 2017 / early part of 2018 than a neutral estimate is if the flu season is more severe than average, leading to increased pressures on the health service. This scenario, which could be considered reasonably likely given experience in the Southern Hemisphere, is considered below. Lessons to learn from Australian flu Due to different seasons in opposing hemispheres, the Australian flu season precedes the season in the UK and the rest of Europe. In Australia, the flu season runs broadly from April to October, with a peak in August. Conversely, in the UK, the season runs from October to April, typically peaking in January or February. This year the prevalence of flu in Australia (measured by laboratory confirmed notifications) has been much higher than is typical, and more than two and a half times higher than There are concerns that this may be replicated in the Northern hemisphere. In Australia, the H3N2 subtype of influenza A was the dominant strain over the majority of the season. Seasons with this dominant strain tend to be more severe the flu spike in was mainly attributable to the H3N2 variant (whereas the less severe season was characterised by the H1N1 subtype). Compared to other strains, this variant also tends to cause more deaths amongst the elderly, and this has been the case in Australia - whilst the overall number of deaths attributed to flu has been relatively low, the majority have been among the elderly. The World Health Organisation has determined the strains in the flu vaccination based on those circulating in the Southern Hemisphere the concern will be that the virus mutates so that the vaccination is ineffective (the vaccine in was less effective than the target average of around 50% whilst initial estimates were around 3%, this was later updated to around 30%). Even if death rates from flu are relatively low, if there are large numbers of cases in the coming season, it is likely to cause additional strain on the NHS in Australia, hospitals closed to new patients and were reporting unusually long waiting times. This experience serves as a potential guide to what might happen in England & Wales in the 2017/18 flu season. If we do see another spike in mortality in late 2017/early 2018, similar to the spike that was seen in early 2015 (which was attributed in part to high levels of flu and the ineffectiveness of the vaccine) this will affect the data going in to the models CMI_2017 and CMI_2018, and lead to lower projections of life expectancy. Users will need to consider the extent to which any short-term experience should be allowed for in the projections they use for valuation purposes. Higher mortality in pension schemes would also lead to higher base mortality rates and (all else being equal) lower liabilities again, it will be important to consider how much weight is given to heavier than expected mortality, if this is not expected to recur. Risk Settlement Bulletin 6

7 Contact Information John Baines +44 (0) Martin Bird +44 (0) Karen Gainsford +44 (0) Tiziana Perrella +44 (0) Steve Bale +44 (0) Hannah Cook +44 (0) Tim Gordon +44 (0) Tom Scott +44 (0) Paul Belok +44 (0) Phil Curtis +44 (0) Dominic Grimley +44 (0) Michael Walker +44 (0) About Aon Aon plc (NYSE:AON) is a leading global professional services firm providing a broad range of risk, retirement and health solutions. Our 50,000 colleagues in 120 countries empower results for clients by using proprietary data and analytics to deliver insights that reduce volatility and improve performance. For further information on our capabilities and to learn how we empower results for clients, please visit: Aon plc All rights reserved. The information contained herein and the statements expressed are of a general nature and are not intended to address the circumstances of any particular individual or entity. Although we endeavour to provide accurate and timely information and use sources we consider reliable, there can be no guarantee that such information is accurate as of the date it is received or that it will continue to be accurate in the future. No one should act on such information without appropriate professional advice after a thorough examination of the particular situation.