Bulk Annuity Annual Report Overview of the current bulk annuity market and recent developments RISK PENSIONS INVESTMENT INSURANCE

Transcription

1 RISK PENSIONS INVESTMENT INSURANCE 2017 Bulk Annuity Annual Report 2017 Overview of the current bulk annuity market and recent developments Bulk Annuity Annual Report

2 2 Bulk Annuity Annual Report 2017

3 Welcome to Barnett Waddingham s 2017 annual report on the bulk annuity market. In addition to providing an overview of the market activity in 2016 and the first half of 2017, we provide some insights into the future and focus on some of the key areas for scheme trustees and sponsoring employers to consider when looking at a potential buy-in, buy-out or longevity transaction. Market activity The bulk annuity market completed just under 20bn of transactions in Broadly half of these were with pension schemes; the remaining 9bn arose from the sale of Aegon s retail annuity back-book to Rothesay Life and Legal & General. Therefore, whilst the total volume of transactions with pension schemes was slightly down on the previous couple of years, overall this represented a further significant increase in the market capacity and demonstrated the insurers ability to transact at higher levels. This activity also comes under the backdrop of the introduction of Solvency II at the start of 2016 and the ongoing uncertainty resulting from the EU Referendum. Insurers have been able to adapt to the new regulatory regime, refining their strategies over the first 18 months of its application, optimising their capital efficiency in order to continue to support the competitive level of pricing available to pension schemes. The market has also shown a robust level of business over the first half of 2017, with over 5bn of transactions completed, and an increase in activity anticipated in the second half, in line with the pattern of previous years. Recently, pensioner buy-ins have been responsible for the majority share (c. 80%) of the total market. This reflects the highly competitive pricing available relative to gilts and schemes being able to exchange available low risk protection investments for a fully matched asset with minimal, or even positive funding impact - removing both the financial and demographic risks. We are typically seeing schemes reduce their financial risks as they continue to mature. With the longevity risk becoming relatively more important, insurance transactions can represent a practical and cost effective way of managing this. The increasing number of schemes carrying out a sequence of phased or tranched buy-in transactions over a period of time is a prevalent feature of the market allowing schemes to remove risk in a way, and at a time, that best fits with their specific circumstances and objectives. Buy-outs have been driven by those schemes which have historically been in a well-hedged position against the adverse movements in interest rates and inflation expectations, as well as those instigated by corporate activity or where de-risking has become more affordable for an overseas parent following the weakening of sterling. Medically underwritten transactions represented a smaller proportion of the overall market in 2016, in part due to the reduced level of direct competition following the merger of the two specialists (Just Retirement and Partnership) to form Just. However, medical underwriting continues to represent a meaningful share (especially for transactions below 100m) and provides a valuable approach within the range of bulk annuity options, with tailored deal structuring being possible to generate the desired level of competitive tension between insurers. Bulk Annuity Annual Report

4 Attractive pricing Pensioner buy-in pricing (on a non-medically underwritten basis) has been at the most competitive levels relative to gilts for a number of years - pricing with an implied investment return of the order of gilts plus 0.3%-0.4% p.a. having been achievable. From our experience, this has not been the sole domain of medium to large transactions, but also possible for more modest deal sizes. This attractive pricing has been supported by the insurers success in sourcing alternative cashflow matching assets with higher anticipated investment returns (e.g. benefitting from the associated illiquidity premium with these long term assets) including infrastructure, social housing and equity release mortgages. For schemes looking to take advantage of these pricing opportunities, it is important to be in an appropriately well-prepared position - especially for smaller schemes where favourable positioning can play a greater role in gaining the necessary insurer appetite. Constraints around the availability of suitable assets for the insurer also means that schemes may need to have some flexibility over their transaction timing to optimise the position. As highlighted in this report, for many schemes liability management exercises can make a material difference in enhancing the potential affordability of a transaction. For example, a number of our clients have successfully completed member transfer exercises or pension increase exchange exercises to reduce the cost of completing a bulk annuity transaction. Future outlook Last year saw some consolidation amongst the bulk annuity providers, with the departure of Prudential and the merger of two existing providers to form Just. However, Phoenix Life have now joined the market on a selective basis boosting the number of active bulk annuity insurers to eight, with some other potential new entrants also progressing their preparations to formally join the market. The introduction of new entrants will provide additional competition and capacity, benefiting pension schemes as a result. The successful capital raising activities carried out by some of the existing active providers over the last year or so have also been a positive indication for the supply side of the market. In addition to pension schemes, some very significant annuity back-books remain with insurers who are not seeking new business in this area. In particular, this includes Prudential and Standard Life; any partial sale of these back-books could distract the bulk annuity providers attention at some stage in the future. Conversely, this could also mean there are potential opportunities in respect of any insurers who are unsuccessful in acquiring these back-books having earmarked the necessary assets - which then become available for an alternative transaction. Whilst pricing continues to prove attractive, the flow of pension schemes looking to de-risk in the market is likely to increase, whether that is in the form of full pensioner buyins, or further phased transactions from the larger pension schemes. General market uncertainty and potential volatility is likely to persist as the negotiation around the UK s exit from the EU progresses. Again, this may provide opportunities for some schemes who are able to take advantage of any beneficial pricing movements as a result. For schemes with longer term self-sufficiency or buy-out targets, there may also be the emergence of insurer offerings which look to manage selective risks, or provide access to the insurer s experience of managing the pension risks. 4 Bulk Annuity Annual Report 2017

5 Our specialist bulk annuity team have extensive experience of advising both trustees and sponsoring employers on insurance transactions and are more than happy to discuss any of the areas raised in this report with you. Contents Page 1. Market update 6 2. Buy-in appeal Preparation steps FTSE350 analysis Liability management options 29 Investment options and journey planning 34 Longevity risk 37 GAVIN MARKHAM Partner, Head of Bulk Annuity Consulting gavin.markham@barnett-waddingham.co.uk Bulk Annuity Annual Report

6 1. Buy-outs and buy-ins: market update 2016 activity For UK bulk annuity providers, 2016 was another strong year, with total transaction volumes with UK pension schemes exceeding 10bn for the third consecutive year. This figure was slightly below the total business levels achieved in 2014 and 2015 which were both over 12bn. However, it is probably not surprising that 2016 saw a slight dip in pension scheme transactions as Solvency II, the new regulatory regime for insurers, became effective at the start of the year. This helped drive a flurry of transactions at the end of 2015, and consequently a slower start to It also took time for some of the insurers to adjust to the new regime and for the level of pricing to find its naturally competitive position. In addition, there were a couple of very significant transactions totalling 9bn completed by insurers in respect of the sale of the Aegon retail annuity back-book - providing a diversion from transactions with pension schemes. If these direct transactions between insurers were to be included, 2016 was another record-breaking year, with the total volume of business transacted by bulk annuity insurers pushing 20bn. 6 Bulk Annuity Annual Report 2017

7 The breakdown of transactions completed by the UK bulk annuity insurers with UK pension schemes over 2016, and the first half of 2017, was as follows: Number of transactions Value of transactions ( m) Year Year H1 H2 Total H1 Total H1 H2 Total H1 Total Aviva Canada Life Just Legal & General Pension Insurance Corporation Phoenix Rothesay Life Scottish Widows Total Aviva Canada Life Just Legal & General 641 2,698 3,339 1,504 Pension Insurance Corporation 897 1,632 2,529 1,875 Phoenix - 1,164 1,164 - Rothesay Life Scottish Widows , Total 2,692 7,521 10,213 5,086 Source: Insurers Just s figures include business completed by Just Retirement and Partnership pre-merger. Phoenix Life transaction is in respect of its own pension scheme. Whilst Rothesay Life did not carry out a transaction with a UK pension scheme during 2016, they did complete a 6bn transaction for the majority of Aegon s annuity back-book. Legal & General completed a 3bn deal for the remainder of Aegon s liabilities. Phoenix Life s transaction was a buy-in with their own pension scheme, the PGL Pension Scheme, and included the conversion of an existing longevity swap. Bulk Annuity Annual Report

8 The following charts show how the value of business for pension schemes has been split between the insurers for 2016 and for the period between 2012 to 2016 inclusive. While some insurers have been a significant presence for several years, the charts demonstrate how Scottish Widows have made their mark since joining the market in Insurer market share (Pension Scheme Transactions) Aviva Canada Life Just Legal & General Others Phoenix Pension Insurance Corporation Rothesay Life Scottish Widows 15% 6% 1% 9% 9% 6% 0.4% 6% 4% 14% 25% % 33% 11% 30% 2% Source: Insurers Just s figures include the transactions completed by Partnership and Just Retirement who merged to form Just in Bulk Annuity Annual Report 2017

9 As in previous years, the total market volume for 2016 was heavily influenced by a relatively small number of very large transactions, including the 1.1bn buy-out of the Vickers Group Pension Scheme and further buy-in tranches for the ICI Pension Fund. Schemes carrying out a sequence of transactions over a period of time has become an increasingly common feature of the market, demonstrating the value of building on prior successful experience and an efficient transaction framework. The actual number of transactions showed a fall in 2016, with just over a hundred deals being completed compared to nearer 160 to 170 in the previous two years, as shown in the graph below. To some extent, this reflects the shift in focus of some insurers to larger transaction sizes, and the application of both their financial capital and human capital in terms of transaction resourcing. Pension scheme transactions by UK insurers between 2012 and ,000 12, Value of transactions Number of transactions 10, Value of transactions ( m) 8,000 6, , , H Source: Insurers and Barnett Waddingham research Bulk Annuity Annual Report

10 First half of 2017 The total value of pension scheme transactions in the first half of 2017, just exceeding 5bn, suggests that this will be another strong year for the UK bulk annuity insurers; the value is nearly double the corresponding figure for the first half of This has been helped by attractive insurer pricing and continued demand by some schemes to carry out repeat buy-in transactions. Therefore, 2017 is well set to comfortably exceed 10bn, making it the fourth consecutive year this mark has been surpassed. After their leading market shares in 2016, Legal & General and Pension Insurance Corporation have continued strongly into the first half of 2017, completing over 3bn of business between them. The common theme of schemes looking to de-risk by securing a tranche of their liability through a buy-in is expected to continue for the remainder of the year and into 2018, supported by some highly competitive pricing levels and the healthy capacity of both the insurers and reinsurers. Market providers During 2016 there were several significant movements amongst the providers - the highest profile being the departure of Prudential from the bulk annuity market. Whilst relatively selective about their transaction appetite, Prudential were one of the longest-standing players in the market. Reasons for their exit were cited around challenges for the additional capital requirements under Solvency II, potentially preferring less long-term capital intensive business lines or overseas market opportunities. The anticipated merger of Just Retirement and Partnership also completed in the first half of 2016, driven by the benefits of scale and synergy, as well as the pressures on the individual annuity market following the introduction of the pension flexibilities. Following the completion of the buy-in with their own pension scheme in 2016, Phoenix Life have now joined the market, with their focus on selected pensioner transactions at this stage. These changes mean there are eight insurers actively quoting in the market at the current time, with differing appetites in terms of transaction sizes and business targets. Buy-in appeal The following chart shows the distribution of business completed over 2016, split by buy-outs and buy-ins and transaction size. VALUE OF TRANSACTIONS ( M) 4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 This distribution shows the balance of transaction sizes and demonstrates how buy-ins have tended to dominate the market over the year and across each of the transaction size groups. Around 80% of the total business volumes with pension schemes in 2016 relate to buy-ins. Source: Insurers and Barnett Waddingham research BUY-OUTS BUY-INS BUY-OUTS BUY-INS BUY-OUTS BUY-INS UNDER 100m 100m - 500m OVER 500m 10 Bulk Annuity Annual Report 2017

11 So what influences this strong appeal of buy-ins? Affordability For the majority of schemes, a full buy-out of all the scheme s benefits is likely to involve a large additional contribution from the sponsoring employer. In part, this reflects the relatively high cost of insuring deferred members. However, recent attractive buy-in pricing for pensioners over the last 12 to 18 months has been around, or potentially below, the corresponding liabilities calculated on a scheme s valuation basis. Therefore, insuring part, or all, of the pensioners through a buy-in can often be achieved without reducing the scheme s funding level and so requiring no additional contribution from the employer. Typically, this has been achieved by exchanging some of the scheme s gilt or bond holdings to fund a buy-in (which assumes these holdings are not required to support any wider LDI or hedging strategy for the scheme). Matching A buy-in generally provides a full match for the scheme s liabilities, with the payment to the scheme from the insurer being equal to the members required pension payments. Unlike other alternative asset classes, this removes all the investment risk and demographic risks associated with the liabilities insured. For example, if a buy-in is completed, the insurer will be required to continue to pay benefits if members live longer than expected or benefits become payable to spouses, financial dependants or children. Phasing An increasing number of schemes have insured their liabilities through a sequence of phased buy-ins, with each transaction representing progression towards full buy-out, and a tangible step along their de-risking journey. This multiple stage transaction approach can ensure the transaction timing and size fits with their wider objectives of the scheme. For larger schemes, splitting the pensioner liability into more manageable tranches for the insurers may increase the level of competitive tension and allow greater flexibility to take advantage of market pricing opportunities. Conversion Under the policy terms of most buy-ins they can be simply converted to a buy-out (i.e. individual policies are issued to members) at a later stage - generally at the time the remaining scheme liabilities are being secured and the scheme is looking to move to wind-up. This simple conversion process will typically be an option at no additional cost. Buy-in contracts also normally allow for some flexibility to adjust the benefits payable under the policy at the point of buy-out if required. Whilst not every transaction is publicised, our estimate is that in excess of 20bn of bulk annuity business (over the period since 2008) has been completed by schemes that have carried out more than one transaction. Bulk Annuity Annual Report

12 Smaller transactions One notable trend is the increase in the average size of transactions over the last few years. The average deal size in 2016 was just under 100m whereas five years ago, it was around 30m. This primarily reflects the increasing emergence of very large deals over this period, but also to some extent a shift in the market appetite for smaller transactions. However, contrary to some perceptions, schemes are still able to successfully complete transactions at the smaller end of the spectrum. For schemes that are attempting to complete transactions under 10m, it can sometimes be challenging to generate a competitive process involving quotations from a range of insurers. However, it is certainly not impossible to complete transactions at this size; Barnett Waddingham has advised a number of clients on the completion of transactions in this size range over the last year or so, down to deals of just over a million. In the main, these smaller transactions are full scheme buy-outs on route to winding-up. Once the transaction size exceeds 15-20m, the range of insurers increases - with three or four insurers potentially willing to participate. For the smaller sub- 10m transactions, it is more important than ever that the scheme is well-prepared when formally approaching the market - with the trustees and sponsoring employer in a position to respond quickly and transact efficiently. At very small transaction sizes, it may also be necessary to be patient and be in a position to take advantage of any windows in the insurer s appetite. Insurers clearly have competing demands; the focus will be on cases where affordability is demonstrated and the transaction can be achieved with limited negotiation and within reasonable timescales. Medical underwriting The use of medical underwriting in the bulk annuity market continues to be a factor. However, this market has consolidated following the merger of Just Retirement and Partnership in April 2016, forming the combined entity now known as Just. This has impacted the level of direct competition, given Just Retirement and Partnership were responsible for the vast majority of the medically underwritten business (in excess of 2bn) written prior to the merger. Despite Legal & General having carried out the largest medically underwritten transaction at the end of 2015, their appetite - and that of Aviva, the other participant - has been very limited. All business completed in 2016 through a medically underwritten process was completed either by Just or its predecessor companies - a similar pattern to the first half of The following chart shows the change in the percentage of the total bulk annuity market that is attributable to medically underwritten transactions (note that this covers all transactions completed by Just, including a limited number which were completed on a traditional basis). This shows a high in 2015 of around 12% falling back to just under 5% in In part, this reflects some diversion created by the consolidation activity of the business following the merger. Fully understanding the affordability, providing clean data in the insurer s preferred format and a clear of description of the schemes benefits that has been reviewed by the scheme s lawyers can all help with insurer engagement for smaller cases. We have discussed this in more detail in our Preparation section on page Bulk Annuity Annual Report 2017

13 Value of business ( bn) Medically underwritten Traditional Source: Insurers and Barnett Waddingham research We believe, for the right transactions, the use of medical underwriting can continue to provide value relative to a traditional non-underwritten approach. The medical underwriting process, where health and lifestyle information is gathered about the particular members being insured, can help insurers set their assumptions around life expectancy with greater confidence and so help reduce pricing generally. This is in addition to the more direct impact arising from the specific health characteristics and conditions of the members which are obtained by the process. In the absence of any genuine direct market competition on an underwritten basis, insurer tender processes are now often structured on a traditional basis initially, with a view of potentially moving to an underwritten approach if the underlying pricing of the underwritten specialist proves to be sufficiently competitive. In particular, this can be done on a post-deal underwriting approach where the pricing is locked in on transaction and the underwriting process is carried out afterwards. The impact of the underwriting can then be reflected using a pre-agreed adjustment to the pricing. For example, the impact could be reflected in full (positive or negative depending on the outcome) or on a downward only basis where any positive impact (such as reduced pricing) is shared between the scheme and the insurer. However, if the impact is negative, then the pricing remains unchanged and the insurer effectively absorbs the impact. Bulk Annuity Annual Report

14 Looking forward What could affect the supply and demand balance in the bulk annuity market over the next few years? We consider three possible factors below: Pension scheme demand The demand for de-risking by pension schemes using a bulk annuity transaction is only likely to increase as they continue to mature, with further employer funding and a greater proportion of the financial risk removed via their investment strategies. Whilst the existing market has already demonstrated the ability to absorb annual business of the order of 20bn, capacity could come under pressure, either as a result of the required capital or human resourcing side. For example, if financial conditions were to ease the feasibility of an insurance transaction may be accelerated for a significant number of schemes (see our commentary on FTSE350 schemes in section 4). Ensuring early visibility with insurers, and completing the necessary preparatory steps in advance, will tend to serve schemes well in this environment. New entrants into the market In addition to Phoenix Life, we are also aware of other parties who have been potentially looking to join the market. Any further new entrants would help to increase the level of competition and potential capacity in the market, providing benefits in terms of pricing and future innovation. The appetite of the potential new entrants may span a range of transaction sizes and as previously highlighted, increased competition for smaller transactions would be a welcome addition. Entering the bulk annuity market and becoming an established provider is not a straightforward proposition however. As well as being able to source suitable assets to support a competitive level of pricing, new providers also need to comply with all the regulatory approvals, have access to an adequate source of capital and achieve both the awareness and trust from schemes, sponsoring employers and their advisers. Insurer-to-insurer transactions During 2016, there were two significant transactions between insurers as Aegon s retail annuity book was split and portions transferred to Rothesay Life and Legal & General respectively. There are some other larger annuity back-books remaining in the market which could possibly transfer in a similar manner. For example, the annuity book of Prudential who are no longer seeking new business, and also Standard Life who have previously stated that they would consider selling their annuity book if a suitable price could be achieved. Any further transactions could potentially be attractive to the bulk annuity insurers. The completion of other multi-billion pound insurerto-insurer transactions would clearly have some knock-on effects for the shorter-term market capacity available for pension scheme transactions. 14 Bulk Annuity Annual Report 2017

15 Insurer pricing Pensioner buy-in Pensioner pricing has been extremely attractive over the last year or so, with pricing relative to gilts moving to historically low levels. This is despite the implementation of Solvency II, the ongoing uncertainty of Brexit and a surprise general election. Whilst pricing in absolute terms may have tended to increase, schemes assess their liabilities using a market-related liability measure. For example, it has been possible to achieve pricing with an implied investment return of the order of gilts plus % p.a. with the additional benefits of having a perfect match to the liabilities and removing the longevity risk. Pensioner pricing 140 Insurer Pricing Gilts PRICE OF TYPICAL PENSIONER PROFILE MORE EXPENSIVE LESS EXPENSIVE 90 30/06 30/09 31/12 31/03 30/06 30/09 31/12 31/03 30/06 30/09 31/12 31/03 30/ Source: Barnett Waddingham Bulk Annuity Pricing model One of the main impacts on absolute pricing was the outcome of the Brexit vote in July 2016, with an immediate rise in pricing as long-term interest rates fell sharply. However, this has also raised some opportunities. For example, where schemes had a significant level of interest rate hedging in place, the widening of corporate bond credit spreads offered attractive pricing. In addition, the fall in sterling has also made the cost of a de-risking transaction look more affordable from the perspective of an overseas sponsoring employer. Bulk Annuity Annual Report

16 Solvency II Insurers have now had a little time to adapt to the capital requirements under Solvency II, the regulatory regime for insurers which was introduced at the beginning of Whilst considerable work was carried out by insurers in the lead up, they have now had the opportunity to refine their approach through the experience gained and completion of new business under the regime. For example: the efficient use of longevity reinsurance; the diversification of asset strategies to optimise capital requirements; and the effective management of the Matching Adjustment requirements. Scheme buy-out Although the longer duration liabilities for deferred members, relative to pensioners, have provided greater challenges for insurers under Solvency II - the insurers have been able to optimise their approach to help mitigate any adverse impact on pricing from the more stringent capital regime. The full buy-out position for specific schemes will vary widely depending on their previous funding and investment strategies. Those schemes that have been relatively wellhedged in relation to interest rate movements will have tended to fair relatively well, especially taking into account any positive growth asset performance over the last year or so. The asset sourcing ability of the insurers can be more significant in the case of full buy-out, and we have seen pricing move materially (by the order of a few percent or so) where specific assets opportunities have arisen. As highlighted in section 5, liability management exercises can also play a significant role in helping schemes close the gap to buy-out affordability. 16 Bulk Annuity Annual Report 2017

17 2. Buy-in: when might it make sense? De-risking journey As more and more defined benefit schemes have closed to both new entrants and future accrual, the focus has firmly shifted to the run-off of the liabilities already on the balance sheet. In particular, to the efficient financial management of these legacy benefits. The ultimate aim of schemes is generally to pass the obligation on to an insurance company at an appropriate point in the future there may also be an interim objective of reaching some form of self-sufficiency (i.e. lower risk) position. However, for a large proportion of schemes, despite having paid significant deficit contributions over the last few years, being sufficiently well-funded on a solvency buy-out basis is still likely to be some time away. In the meantime, some schemes may be running a sizeable level of risk in order to support their funding plans. Schemes in general will continue to be exposed to a number of risks: 2.0 DEFICIT INCREASE RISK 1 IN 20 ( m pa) INTEREST RATES INFLATION CREDIT TARGET RETURN LONGEVITY DIVERSIFICATION TOTAL Source: Barnett Waddingham model BWARM Bulk Annuity Annual Report

18 With the greater range of de-risking options available to schemes. increasing care is needed in considering which approach, or approaches, best fits with the specific objectives and circumstances of the scheme. The use of a buy-in (hedging both the financial risks and longevity risk) represents an important part of the potential de-risking toolkit. As highlighted earlier, during the first part of 2017, insurer pricing for pensioners has continued at historically attractive levels relative to gilts. This is shown in the following chart calculated using the Barnett Waddingham Bulk Annuity Pricing Model - which reflects current pricing from a range of leading insurers. The highly competitive pricing has allowed schemes to insure pensioners at a cost where the implied return underlying the buy-in policy has been significantly in excess of gilts. Whilst pricing is clearly a key consideration, any potential buy-in needs to be assessed in the context of the scheme s overall objectives, and provide an attractive risk reduction/cost balance relative to the alternative strategies. For example, a buy-in may represent an increasingly effective option, as the scheme s longevity risk becomes more meaningful relative to the financial risks. This will tend to happen naturally over time as schemes continue to mature, and take action to reduce their investment, interest and inflation rate risks. Pensioner pricing 125 PRICE OF TYPICAL PENSIONER PROFILE /06 31/07 31/08 30/09 31/10 30/11 31/12 31/01 28/02 31/03 30/04 31/05 30/ Insurer pricing range Insurer pensioner pricing Gilts Assessing potential buy-in In assessing the potential merits of a buy-in relative to the alternative de-risking options, there are a number of key areas that should be considered as illustrated overleaf. Their relative impact will vary from one scheme to another depending on the characteristics of each scheme, including the risk appetites of the trustees and sponsoring employer. 18 Bulk Annuity Annual Report 2017

19 Market pricing levels Factors Financial/ accounting cost Risk profile Hedging levels Income / liquidity Future investment flexibility Financial/accounting cost How do you measure the cost of de-risking? The impact on the future expected returns from the scheme s assets (and so the potential funding or cost implications) will depend on the assets being used to support the transaction. For example, an exchange of gilts (or other low yielding assets) for a buy-in may have a minimal, or even positive, impact on scheme financing whilst using growth type assets is likely to increase the anticipated cost. Any impact should be considered in the context of the scheme s longer-term objectives and the likelihood of achieving them, allowing for any additional funding from the employer for the transaction. The general competitiveness of the insurer pricing, say relative to recent historical market levels, is also likely play a role in most cases. The corporate accounting implications (such as FRS102 or IAS19) of carrying out the transaction will also need to be acceptable. Risk profile It is important to consider the holistic impact of the de-risking exercise - both financial and longevity risks. Stochastic modelling allows the reduction in longevity risk to be quantified alongside the financial risks to provide a meaningful comparison. As part of the risk analysis, it is also possible to assess the impact of insuring different groups (e.g. highest liability pensioners, random cross sections of members, young/ old pensioners etc.) where a partial buy-in is considered. Hedging levels The impact on the overall level of interest rate and inflation rate hedging for the scheme will depend on the particular assets used to meet the buy-in premium. In practice, a minimum criteria for the buy-in transaction may be that it has a neutral (or better) overall impact on the hedging of the interest and inflation rate risks. For some schemes it may be possible to introduce leveraged Liability Driven Investment (or increase the extent of any existing leverage) to help release assets to support a buy-in. Where this is an option, any leveraging should be within acceptable limits allowing for the potential collateral requirements. Income/liquidity Whilst a buy-in does provide a source of future cashflow based on the underlying pension payments insured, the general cashflow impact needs to be considered and sufficient liquidity retained within the scheme s asset portfolio to support anticipated requirements such as future retirements or transfers. Looking at the before and after comparison of all risks can show the potential impact of the buy-in transaction (or alternative de-risking strategy) on the overall profile of the scheme. Bulk Annuity Annual Report

20 As part of the assessment process, the above framework also helps shape the size of the potential buy-in transaction. For example, the analysis may indicate that the originally intended size of transaction is too large given the financial constraints and a smaller transaction size provides a better fit with the wider funding and investment strategy. This can be coupled with an intention to carry out further transactions as and when the timing is right. Although slightly less tangible in the immediate term, a buy-in transaction has other additional benefits. Completion of a transaction demonstrates to the market the scheme s intent and ability to de-risk successfully putting them in a better position for any follow-up transactions. Phasing transactions (subject to maintaining suitable partial transaction sizes) can also provide some element of market smoothing, both in terms of the relative pricing achieved over time and any potential changes in the competitive nature of the bulk annuity market. Further considerations There are also a number of other considerations including: The effect on future buy-ins and buy-outs De-risking strategy facilitating buy-in transaction The previously outlined approach can help assess the derisking journey, and the potential role of any partial buy-in transactions within it. In a number of cases, the buy-in may be facilitated by adjustments to the wider investment strategy. The following example (based broadly on a recent client case) illustrates this: 30 % DIVERSIFIED GROWTH FUNDS 35 % CORPORATE BONDS 60 % INTEREST RATE COVERAGE Take care to avoid any selection bias; use random sampling of members or an objective justification. Consider implications following any medically underwritten approach. 35 % INDEX LINKED GILTS 80 % INFLATION COVERAGE Anticipated timescales for future buy-in tranches or buy-out? Insurer preferences for transaction sizing and liability profile? 1. The Scheme was already invested in a reasonably defensive manner using 70% of its assets (mixture of indexlinked government bonds (gilts) and corporate bonds) to hedge interest rate and inflation rate risks, with the remaining 30% held in growth assets split between two diversified growth funds (DGF) 20 Bulk Annuity Annual Report 2017

21 30 % DIVERSIFIED GROWTH FUNDS 48 % FREED UP ASSETS 22 % LDI ASSETS 60 % INTEREST RATE COVERAGE 80 % INFLATION COVERAGE 2. The trustees and sponsoring employer agreed a move to a leveraged Liability Driven Investment (LDI) approach to provide a more efficient way of hedging the interest and inflation rate risks. In effect, this freed up a proportion of the scheme s protection assets not required to support the LDI investment. These could be used to increase hedging further (including the previously unhedged longevity risk via a buy-in) or target a higher investment return. Both the trustees and the employer had limited desire to increase the growth asset risks. Partial buy-in of 50% of pensioners 36 % DIVERSIFIED GROWTH FUNDS 24 % LDI ASSETS 40 % BUY-IN 75 % INTEREST RATE COVERAGE 75 % INFLATION COVERAGE 3. Using all of the freed up assets to support a buy-in of the Scheme s current pensioners led to a much higher return required from the remaining assets (based on the current funding basis). To achieve this additional return with the remaining assets, the trustees would have had to move assets away from LDI towards growth assets, reducing the hedging. By using a partial buy-in policy to cover around 50% of the current pensioners, the balance of the freed up assets were sufficient to hedge the interest rate and inflation risks of the remaining liabilities and still provide sufficient return to support the current funding basis. Bulk Annuity Annual Report

22 3. Being prepared: what can schemes do now? Even where a potential transaction may realistically be some way into the future, trustees and sponsoring employers can carry out valuable steps in preparation. These preparatory actions not only ensure the scheme is better positioned when the relevant time comes, but can also help demonstrate the necessary commitment to a transaction helping achieve the desired level of engagement from insurers. Early consideration of the issues and detailed planning is key in order to deliver the overall de-risking exercise in an efficient and effective manner. For example, as highlighted in section 5, implementing member option exercises can deliver some material savings against the cost of insuring the benefits with an insurer. However, a reasonable period of time (e.g. 6 to 9 months or more) can be required to deliver the more complex member exercises successfully - recognising the member communication and IFA advice process involved. Depending on the circumstances, it can be possible to integrate an exercise more directly into the transaction process. 22 Bulk Annuity Annual Report 2017

23 Planning initial stakeholder engagement bespoke education and training identify key issues and process steps detailed project planning Benefits Financials affordability/risk reduction analysis development of pricing metrics funding and accounting impacts Membership data data assessment and gap analysis member and spouse information GMP recon/equalisation prioritise data cleanse actions identify historical issues or uncertainties consider any discretionary practices member option terms legal review Liability Management feasibility assessment transfer value/pie exercise? trivial commutation? Stakeholders trustee/sponsor engagement appropriate governance structure clear decision-making process Bulk Annuity Annual Report

24 Bespoke training at the outset of the project covering all of the relevant stakeholders can provide a number of benefits. In particular, this can help: Identify main transaction issues at early stage Facilitate agreement of stakeholder objectives Support development of effective governance/decision making process Ensure transaction fit with wider scheme funding/investment strategy Manage stakeholder expectations on process and time scales Consider merits of alternative transaction structures Good governance and clear decision-making structure including both the trustees and sponsoring employer is fundamental for any efficient de-risking exercise. Membership data and scheme benefits Two areas of information that are key to the transaction, and ultimately form an integral part of the policy documentation, are the membership data and benefit specification. Work carried out in advance to make sure these areas are transaction ready (i.e. appropriately accurate and complete) can pay dividends in terms of removing uncertainties and lead to better pricing terms. It is important to identify any gaps in the priority data areas and focus any data cleansing activity accordingly. This approach can help strike the right balance between having information that is fit for transaction purposes, without spending too long refining data such that the transaction is unnecessarily delayed, and even the danger of a de-risking opportunity being missed. For the detailed benefit specification, legal review in advance of provision to the insurers can avoid any subsequent delays or issues arising. Review by the scheme actuary and administrators can also help ensure that the benefit specification is consistent with the way the benefits are actually being delivered in practice, for example in those areas which are not explicitly defined under the rules of the scheme. Trustees will also need to consider any implications for conversion terms which relate to member options, including cash commutation at retirement and transfer values. Solvency II limits the potential flexibility for insurers in these areas, especially for schemes where terms may be hard-coded into the rules, or are set at relatively generous levels. 24 Bulk Annuity Annual Report 2017

25 Related to membership data and scheme benefits is the ongoing saga of reconciliation, rectification and eventually equalisation of Guaranteed Minimum Pension (GMP) benefits. With virtually all schemes looking to complete the reconciliation exercise before the 31 December 2018 cut off, demand for NICO s service is very high and turnaround times have been increased. Barnett Waddingham have a specialist team experienced in carrying out GMP equalisation exercises and there could be some pricing opportunities in this area for schemes over the next 12 months with revised regulations on GMP equalisation and conversion due to be published. Once GMP benefits have been reconciled and rectified, it will also be necessary to consider benefit changes for sex equalisation requirements. Monitoring Increasingly our clients are using online actuarial tools to monitor their funding position, and as an early indicator of potential bulk annuity opportunities for their scheme. Monitoring scheme specific metrics (e.g. insurer pricing levels, asset availability, hedging requirements etc.) can form part of a scheme s trigger strategy, set in the overall context of their Integrated Risk Management framework. For more information on Illuminate visit: illum nate Bulk Annuity Annual Report

26 4. FTSE350 analysis: the current market position? One of the key factors that will influence pension scheme demand for bulk annuity transactions is the strength of their funding position, particularly in the case of any possible move to scheme buy-out where the sponsoring employer is likely to be required to make a shortfall payment to meet the premium. The total defined benefit liabilities for UK occupational pension schemes are just over 2trn, with roughly half of this associated with the FTSE350 companies. Not all of these liabilities will ultimately find their way to the insurance market (or indeed be able to be insured), but the figure clearly demonstrates the scale of the potential demand from pension schemes in the future. A scheme s buy-out funding position will typically be expected to improve over time as more members retire, the duration of the liabilities reduces, the employer continues to pay deficit contributions and any anticipated growth asset performance is potentially realised. The position could also be accelerated over a significantly shorter timeframe if financial conditions improve, for example if long-term interest rates were to rise materially beyond market expectations to the extent that schemes are not already hedged. 26 Bulk Annuity Annual Report 2017

27 The following chart illustrates the estimated buy-out funding position for the FTSE350 companies with defined benefit obligations, and also the projected position in The charts show the potential impact of an increase in long-term interest rates (not already factored into the market) on the position, assuming an increase in interest rate expectations of 0.5% pa and 1.0% pa respectively. FTSE350 - Estimated Buy-out Funding Level 250 Number of companies BASE SCENARIO 0.5% pa INTEREST RATE INCREASE 1.0% pa INTEREST RATE INCREASE Source: Barnett Waddingham analysis based on published financial information. Less than 50% 50% to 75% 75% to 90% Over 90% For these FTSE350 companies, around 5% have buy-out funding levels (i.e. ratio of assets to buy-out liability) of 90% or over, which increases to the order of 10% and 20% for the 0.5% pa and 1.0% pa interest rate increase scenarios respectively. There is also a significant improvement in the projected funding position in 2019, allowing for the estimated employer contributions over this period. This illustrates the potential improvement in funding positions and so possible increase in demand moving forward, as well as the sensitivity to changes in financial market conditions. Bulk Annuity Annual Report

28 The chart below shows the number of FTSE350 companies who may be able to fund their buy-out shortfall by using part, or all, of their existing cash holdings. As per the previous chart, we have shown the projected position in 2019 and also the sensitivity to possible changes in long-term interest rates. FTSE350 - Buy-out affordability using existing cash holding Number of companies BASE SCENARIO 0.5% pa INTEREST RATE INCREASE 1.0% pa INTEREST RATE INCREASE able to buy-out - 75% to 100% cash holdings able to buy-out - 50% to 75% cash holdings able to buy-out - 25% to 50% cash holdings able to buy-out - less than 25% of cash holdings Source: Barnett Waddingham analysis based on published financial information. In practice, companies are likely to utilise their cash holdings for other business purposes rather than necessarily seeking to fully de-risk their pension scheme through buy-out. Any buy-out has to be supported by the appropriate business case and justified from a cost/risk perspective. However, based on this cash affordability measure, it illustrates how demand could significantly rise over a fairly short period. 28 Bulk Annuity Annual Report 2017

29 5. Liability management options: could you be closer than you think? Member option exercises are increasingly being used by schemes and sponsoring employers to help manage the scale of their defined benefit liabilities and reduce the associated risks. In the context of a bulk annuity transaction, it may be possible to significantly improve the affordability of a deal by carrying out appropriate member exercises in the run up to a transaction. In some cases, this could ultimately make the difference between a transaction being a feasible proposition or not. For example, member option exercises can be used to help reshape complex member benefits in order to reduce the cost of a pensioner buy-in, or improve the overall solvency positon of the scheme to enable the sponsoring employer to bridge the gap to a full scheme buy-out. For any exercise, clear objectives, a well-designed proposition, detailed planning and effective member communications are essential in achieving a successful outcome. In particular, wellreceived member communications will help drive a greater level of member engagement, which in turn increases the likelihood of higher option take-up rates. Delivered well with the appropriate level of support - offering additional choices to members that enable them to receive their benefits in a form which may better suit their own individual needs has the potential to be a mutual win for the members, trustees and sponsoring employer. Bulk Annuity Annual Report

30 Possible liability management exercises LIABILITY 3,500,000 3,000,000 2,500,000 2,000,000 1,500,000 Trivial Deferreds? Deferreds - ETV? Deferreds - FRO? Trivial Pensioners? Pensioners - PIE? Pensioners - Top Slicing? 1,000, , Source: Barnett Waddingham research and scheme analysis Transfer value exercise The introduction of the pension flexibilities in April 2005 under freedom and choice provides members of defined contribution arrangements with a much wider range of ways to access their benefits. This, combined with relatively high transfer values resulting from the historically low level of gilt yields, means that transfer values from defined benefit schemes to defined contribution schemes have looked attractive to an increasing number of deferred members evidenced by the general increase in the volume of transfers being paid from pension schemes. Percentage of transfer value 150% 145% 140% 135% 130% 125% 120% 115% 110% 105% 100% 95% Typical funding range Buy-out Transfer value For the vast majority of schemes, when a member transfers out of the scheme there will be a saving compared to the potential cost of securing their benefits with an insurance company (at that point in time). To some extent, this reflects the higher cost of insuring deferred members relative to pensioners. Therefore, member transfers will generally improve the solvency position of the scheme and so reduce the potential cost of buying-out. For a typical scheme, the following chart illustrates the cost of buying-out the member benefits broadly compared to their transfer value from the scheme (and funding reserve) for members over age 55. Member s age Source: Barnett Waddingham analysis As a result, many schemes are looking to carry out some form of exercise to communicate the transfer option to their non-retired scheme members. This exercise may be targeted at members over age 55 who are eligible to take their benefits immediately, or simply positioned as part of the scheme s ongoing retirement process. 30 Bulk Annuity Annual Report 2017

31 It is also possible to provide an enhancement on top of the standard transfer value as an added incentive for the member to transfer out of the scheme, or to look at some form of partial transfer option. Compliance with the Incentives Code of Practice is key for these exercises, and in any case, represents good practice more generally. Even where transfer enhancements are applied, an exercise can still result in some material savings against the solvency deficit. Based on our own experience and anecdotal evidence from Independent Financial adviser firms (who are involved in the individual member advice process), assuming a reasonably modest take-up rate and potential enhancements to transfers, a well-run exercise could result in a reduction to the solvency deficit of potentially up to 20%. In reality, the specific impact will depend on the individual circumstances of the scheme and its members. Pension increase exchange Even where the exercise is carried out to pass on the full best estimate value of the increases being exchanged, savings can still be made relative to the cost of insuring the original scheme pension. Where the scheme pension increases are complex, for example linked to a combination of inflationary caps and collars, these can be relatively expensive to insure due to the additional challenges for the insurer in looking to hedge or reserve for these payments. Therefore, exchanging these complex increases for non-increasing pension can result in a lower insurance cost. An exchange can reduce the impact of any deflationary floor (i.e. where inflation falls below zero but the pension in payment cannot be reduced). In addition, whilst insurer pricing for CPI-linked pension increases has shown some improvement, it can still be relatively expensive to insure when compared to the typical scheme funding assumptions. Therefore, a PIE may be particularly beneficial in reducing the overall proportion of CPI-linked benefits. Whilst transfer options can be an effective way to reduce the solvency deficit in respect of non-retired members, they are not possible for pensioner members. For pensioner members a pension increase exchange (PIE) can be considered, and could also be implemented for deferred members, for example at retirement. Under a PIE, members are offered the option of exchanging their future non-statutory increases in payment, typically for a higher non-increasing pension. For example, increases on Annual pension ( ) 40,000 35,000 30,000 25,000 15,000 10,000 Difference in annual pension Annual scheme pension Annual re-shaped pension pensions in excess of Guaranteed Minimum Pension built 5,000 up before April 1997 may be exchanged. In recognition of the reduced level of future pension increases, the member receives an immediate one-off uplift to the pension in Age payment on execution of the exchange. In particular, some members may prefer the higher level of pension whilst they Source: Barnett Waddingham research, <100% Balance Deal Percentage are younger and potentially have more use for the income. Bulk Annuity Annual Report

32 Other liability management options Further options which could be used include: Trivial commutation - trivial commutation exercises can be used to discharge liabilities for members with the smallest benefits in the scheme. The ongoing costs of administering these small benefits can be disproportionately large, and so removing them in exchange for a cash lump sum paid to the member can be an attractive option. Where the exercise is carried out as part of a full buy-out and scheme winding-up process, then under the regulations a greater range of members can be offered the option. Early retirement simply reminding members that they have the option of retiring early (generally those members over age 55) can reduce the number of deferred members in the scheme and so have a positive impact on the potential solvency deficit. Particularly if retiring members exercise a cash commutation option. Combining member options Appropriately combining member option exercises schemes can optimise the overall outcome, resulting in a material reduction in the buy-out deficit. These exercises can be used if the focus is on transacting in the immediate term, or equally can form part of the longer-term de-risking strategy. Where there is a potential transaction in the short term then positioning them as part of a one-off bulk exercise is likely to be more appropriate (either in the lead up, or directly linked, to the transaction). Alternatively, they could be structured as a business as usual option forming part of the standard retirement process, in which case any savings will emerge over time as members exercise the options on retirement. Potential buyout deficit 120, ,000 80, S 60,000 40,000 20,000 0 Starting Solvency deficit TV exercise for under 55s TV exercise for over 55s Bulk PIE Exercise Trivial commutation and early retirement Ending Solvency deficit MEMBER EXERCISE Note: Notional Scheme based on Barnett Waddingham illustrative calculations with reasonable assumed take up rates and enhancements. 32 Bulk Annuity Annual Report 2017

33 Closing the gap to buy-out further? Even after allowing for potential liability management options, the buy-out deficit is still likely to be material in most cases. But for certain schemes, there may be other factors which could be considered to close the gap in developing the wider affordability picture. These can include: the value of the remaining future deficit contributions agreed under the recovery plan; Bringing all of these factors together may mean that the shortfall can be considered affordable by the sponsoring employer. Ultimately, in deciding whether to move to buy-out at a particular point in time, the business case should consider the alternatives and seek to demonstrate it is the desirable route for the scheme in the context of the pension risks being removed, both financial and demographic. savings from the removal of the ongoing operational expenses of the scheme on buy-out (net of transaction and wind-up costs); and potential for favourable movements in insurer pricing (including sterling exchange rate movements for overseas parents where applicable). Potential buyout deficit 120, ,000 80, S 60,000 40,000 20,000 0 Starting Solvency deficit Liability management exercises Present value of deficit contributions Present value of future expenses (net) Favourable annuity pricing Ending solvency deficit Note: Notional Scheme based on Barnett Waddingham illustrative calculations. Bulk Annuity Annual Report

34 6. Investment options and journey plan: how might the investment strategy evolve? For schemes heading towards buy-out, it is important to develop an appropriate investment strategy to reflect this underlying objective. Investing in a more consistent way to an insurer allows schemes to better match changes in insurer pricing and protect their aim of buy-out. This approach allows schemes to lock down investment risk, whilst progressively plugging the funding gap and moving towards their goal of fully securing member benefits. For illustration the charts show the broad investment strategy of two bulk annuity insurers although different, they demonstrate the role of liability matching, credit assets and potential return enhancing assets. We consider three investment areas that could be of interest to schemes with this in mind. 34 Bulk Annuity Annual Report 2017

35 AAA BBB Loans Current Assets Other Financial Assets Debt Assets gilts INSURER 1 INSURER 2 Debt Assets A Source: Insurer annual returns and Barnett Waddingham research BBB Debt Assets AAA A AA gilts Buy and Maintain corporate bonds Why would you buy? Insurers typically have a significant holding in corporate bonds (credit) as part of their investment portfolio. Credit is attractive to insurers as it provides regular income (like the annuities they back), as well as an additional premium over gilts (to cover the risk of default and of lower liquidity etc.). As very long-term investors, insurers are able to absorb a lot of the liquidity risk (i.e. significant costs associated with selling credit), as well as some of the market value changes associated with holding them. In particular, provided the corporate bond issuer does not default, the insurer receives the expected cashflows from the bond in full. Insurers hold a well-diversified credit portfolio, across sectors and geographies, such that they are far better diversified than a standard sterling bond index. They also expect to have very low turnover within the bonds these are assets held to provide income, which meets the insured pension payments. This type of management can be viewed as a middle ground between a pure passive approach and an active approach, known as Buy and Maintain. The distinguishing feature of the Buy and Maintain approach is low turnover. Bonds are purchased (subject to various maximum allocations to enhance diversification) and held to maturity, unless there are fundamental changes to the issue. Although there are many influencing factors, a key driver of insurer pricing is the asset portfolio used to match the liabilities, and therefore changes in market conditions such as corporate bond pricing. Pension schemes that include a similar sized holding in corporate bonds are better able to match changes in insurer annuity pricing. Better still is a portfolio of credit that is more aligned with the credit assets the average insurer is likely to hold. This not only provides a closer match to pricing changes, but also potentially reduces investment transaction costs, as insurers are more likely to accept these assets in-specie, avoiding the material costs of the scheme s assets being sold and the insurer having to buy the assets they wish to hold going forward. How do you buy? Larger pension schemes can use a segregated basis to follow a Buy and Maintain approach to corporate bond investing, while smaller schemes can access the approach through pooled Buy and Maintain credit funds. Bulk Annuity Annual Report

36 LDI funds Why would you buy? As schemes approach full funding on a low-risk basis and move closer to buying-out with an insurer, they typically want to have hedged the majority of their interest rate and inflation risk. Without doing so, there is a significant risk of an adverse movement in yields increasing the cost of buyout beyond reach. To do this, whilst maintaining desired exposure elsewhere (for example to growth assets), schemes have looked to leverage their hedging assets. In practice, this Liability Driven Investment (LDI) approach involves entering into derivative contracts, which synthetically make the assets held a close match for the liabilities by changing the level of duration or inflation exposure. The range of fund choices available now includes retention of growth asset exposure whilst also providing interest rate protection. How do you buy? In addition to the more bespoke LDI strategies for larger schemes, pooled LDI funds also allow smaller schemes with a lower governance budget to gain leveraged exposure to interest rate changes. Many schemes have now gone down this route, at the very least allowing them to broadly mirror a typical insurer s portfolio of gilts and credit, whilst maintaining hedging. Secured-income assets Why would you buy? Insurers do not hold just bond investments. They naturally also look to achieve higher risk adjusted returns elsewhere, but in doing so, are seeking assets with a contractual income, given the need for regular income to match the pension benefits and provide an acceptable fit within the regulatory capital regime. Pension schemes can also consider investing in similar assets, like infrastructure and long-lease property to source these higher expected returns, whilst maintaining a cashflow-matching approach. These assets can help bridge the gap between self-sufficiency and buy-out. How do you buy? This depends on size. Very large schemes are able to buy these assets directly, or invest significant amounts into certain vehicles that provide access. However, this comes with a higher governance burden and might be a barrier to entry to many schemes. Smaller schemes do not have the option of direct investment given the size of these assets but pooled vehicles are emerging, which give access to a diversified basket of illiquid alternatives. That said, individual insurers may be quite selective as to what assets they are willing to take in specie. This is an area where the investment and insurance market is likely to develop further. 36 Bulk Annuity Annual Report 2017

37 7. Longevity Risk: quantifying the known unknown? Longevity market over 2016/17 Pension scheme trustees attending Barnett Waddingham s 2017 pensions conference voted longevity as the second most important risk facing pension schemes today after the low yield environment. We see increasing appetite from pension schemes to secure or reduce longevity risk. Key drivers are: the increasing maturity of pension schemes; diminishing returns from additional interest rate or inflation rate hedging; the successful implementation of past longevity transactions; and potential concerns over long-term capacity of the longevity transfer markets. However, for the moment the main users of the longevity transfer market remain bulk annuity insurers. The Solvency II supervisory regime which came into force at the start of 2016 means that reinsuring longevity risk continues to provide capital efficiency for insurers writing bulk annuity business, demonstrated by the 1bn reinsurance transaction taken out by Rothesay Life on their back book during the year. To satisfy pension schemes desire to manage their longevity risk reduction, reinsurers will need to maintain their appetite for UK longevity risk and providers will need to become more innovative. The recent introduction of uncollateralised swaps with a simplified benefit structure goes some way towards meeting this demand. The reduced cost in relation to legal terms, collateral management and data transfer mean that longevity can be transferred cost effectively for smaller schemes. We have been working with life insurers interested in entering the market for similar transactions and more competition in this area will be welcome. Some pension schemes have adopted a wait and see policy from concern that reinsurance pricing was slow to reflect the recently observed slowdown in mortality improvements. During 2017 the new CMI_2016 projection model was published, offering additional data to schemes and reinsurers. We are watching with interest to see how this impacts pricing. For pension schemes, notable transactions during the past year included: 1bn swap carried out by the Manweb Group of the Electricity Supply Pension Scheme with Abbey Life Assurance Company Limited (since taken over by Phoenix Group) conversion of a longevity swap for the Phoenix Group s Pension Scheme into an internal buy-in contract a series of swaps between 50million and 600million carried out on an uncollateralised basis with Zurich Bulk Annuity Annual Report

38 Longevity modelling and risk management Longevity modelling The longevity risk picture looks different for every pension scheme. It can depend on a number of factors including: Age of the members Concentration of liability whether a large part of the liability depends on a small number of senior or longservice employees Whether scheme characteristics have changed over time for instance, many pension schemes have transitioned from a largely blue collar occupational profile to more white collar or management in recent years Benefit structure higher pension increases mean that longevity risk is carried for longer and amplified over time Spouse benefits schemes with generous spouse benefits, or where a large proportion of members are married, may also have a higher level of longer-term longevity risk than other schemes Holistic risk analysis Many pension schemes have already implemented a significant level of interest rate and inflation hedging. Our approach is to consider pension risk holistically and our longevity model works in conjunction with our financial risk model to give an overall picture of the risks a scheme is exposed to. This approach can help trustees and sponsors judge where their governance time and budget is best spent. For example in the case of one 500m scheme, we were able to demonstrate that increasing the size of the LDI hedge further would provide minimal risk reduction, due to the uncertainty from longevity in the underlying funding basis. To illustrate the impact of longevity risk and the mitigation that schemes can take in different circumstances we have provided a brief longevity analysis for three different scheme profiles (mature, young and significant concentration risk). Pension schemes and their sponsors should have a keen understanding of their longevity risk profile and assess their appetite before approaching the de-risking market. This approach will help to identify what type of transaction is most affordable and whether a transaction will be cost effective relative to their funding reserve. In order to help schemes better understand the risk posed by longevity, we have developed a stochastic longevity model which analyses the expected cashflow profile and the underlying components of mortality risk inherent in a scheme. 38 Bulk Annuity Annual Report 2017

39 1. Mature scheme Longevity risk projected benefit cash-flows 6,000,000 Longevity VAR (scheme lifetime): 1 IN 20 RISK EXPERIENCE 10,000,000 5,000,000 8,000,000 4,000,000 6,000,000 4,000,000 3,000,000 2,000,000 1,000,000 50% Scenarios 90% Scenarios 95% Scenarios 95.7% Scenarios 2,000,000 0 Process Risk Level Risk Trend Risk Diver sification of 3 components of Longevity Risk All - Longevity Risk Source: Barnett Waddingham longevity model The chart above shows the expected future cashflows payable by a mature pension scheme projected over the next 50 years. Our model runs several thousand simulations of how the scheme s mortality experience might turn out. In half of all scenarios, the cashflows paid will be within the green area. This is a fairly narrow band but still means that the annual cashflow might be 5% higher or lower than the scheme funding valuation has forecast. The colours orange, green and blue show successively more extreme scenarios. The risk is fairly well balanced in this case with a broadly even opportunity for the scheme to gain or lose from mortality. We also show a sample Value at Risk (VaR) measure at the 95% level (i.e. 1 in 20 chance) for the cost a mature scheme (c. 100m) might experience from longevity over a 3-year horizon. There is broadly equal contribution to the longevity risk from three underlying components: Process Risk the random chance that more members than expected survive in any given year Level Risk uncertainty associated with the current assumed level of mortality Trend Risk the risk that longevity improves quicker than anticipated over time Mature schemes have significant longevity risk exposure in some cases it can be a more significant risk than interest rate, inflation rate or investment risks. The most appropriate route for a scheme of this nature will depend on its overall funding plan: As a starting point, they should consider whether their long-term goal is to reach buy-out or self-sufficiency and their likely horizon and current risk appetite. Depending on size the scheme may be able to secure its risk entirely either in one transaction or by a series of buy-in transactions. If buy-out is not the long-term target or appears unaffordable, then a longevity swap may be the most effective way to manage this risk. Bulk Annuity Annual Report

40 2. Young scheme Longevity risk projected benefit cash-flows Longevity VAR (scheme lifetime): 1 in 20 risk experience 6,000,000 5,000,000 4,000,000 50% Scenarios 90% Scenarios 95% Scenarios 95.7% Scenarios 10,000,000 8,000,000 6,000,000 4,000,000 3,000,000 2,000, ,000,000 Process Risk Level Risk Trend Risk Diver sification of 3 components of Longevity Risk All - Longevity Risk 1,000, Source: Barnett Waddingham longevity model Here we show comparable analysis for a younger pension scheme (i.e. a scheme where the majority of members are not yet drawing their pensions). In this case the most significant cashflows are a number of years in the future. The overall cashflows from the scheme are more uncertain (the green band shows a +/- 10% error margin around the scheme funding assumption). This is illustrated in the VaR analysis where the largest exposure is to Trend Risk the risk that in future longevity improves much faster than expected. Pension schemes with large numbers of deferreds find that there are limited options available to help them de-risk; bulk annuity pricing looks relatively expensive compared with scheme funding reserves and longevity swap providers have preferred to focus on pensioner transactions. Pension schemes with younger populations looking to reduce risk will often put in place interest rate and inflation hedging strategies. They should monitor the cost of bulk annuities relative to their funding strategy and consider how they can be ready to take pricing opportunities that arise. Liability management through member option can also play an important role. The scheme may be able to adopt a split strategy with liabilities for pensioner members protected through a buy-in and assets held in respect of non-pensioners applied more efficiently to generate returns. 40 Bulk Annuity Annual Report 2017

41 3. Scheme with significant concentration risk Longevity risk projected benefit cash-flows Longevity VAR (scheme lifetime): 1 in 20 risk experience 6,000,000 10,000,000 8,000,000 5,000,000 6,000,000 4,000,000 4,000,000 3,000,000 2,000, ,000,000 Process Risk Level Risk Trend Risk Diver sification of 3 components of Longevity Risk All - Longevity Risk 1,000, Source: Barnett Waddingham longevity model Finally, we look at a mature scheme where a small number of large liability members, represent broadly half of the overall liability. Here, the uncertainty in the cashflow projection is much wider and the overall VaR is higher. The cashflows are sensitive to these large liability individuals living just a few years longer than expected. Although the three schemes we have looked at have equal funding liabilities, this scheme has the largest exposure to longevity risk due to the concentration of risk as shown by the larger Process Risk. A scheme with a concentration of individuals with large pensions may be able to achieve risk reduction by entering a buy-in. We have been able to demonstrate that in some cases a scheme can use just 20% of their assets to remove 25% of their risk. Bulk Annuity Annual Report

42 Assessing the transaction When we advise clients considering a de-risking transaction, there are two fundamental questions: How much risk is removed? What is the cost of the transaction? This could be measured on the funding basis, the accounting basis or another financial measure the costs of implementing the transaction should also be factored in. The outcome of this analysis depends on the insurer s view of the scheme members longevity, and in particular relative to any margins that may be held by the scheme in their funding basis. For some schemes in 2016, pricing was sufficiently attractive that the transaction would have been risk-reducing at little or no cost on a funding basis. Illustration of reduction in longevity risk through full pensioner buy-in % Distribution 2.5 % After pensioner buy-in Before pensioner buy-in 2.0 % 95th percentile 16.4m risk reduction 1.5 % 1.0 % 0.5 % 0.0 % Source: Barnett Waddingham longevity model Change in liability ( m) 42 Bulk Annuity Annual Report 2017

43 Longevity analysis and CMI_2016 Longevity analysis The large majority of all UK pension schemes set their funding assumptions using projections of future longevity published by the Actuarial Profession s Continuous Mortality Investigation (CMI). It is important for pension schemes to have a clear view of their likely mortality statistics (based on the best information available to them). Whilst often the scheme s funding basis will form part of the first point of reference for considering whether a de-risking transaction represents value for money, it is also important to consider the best estimate view of mortality. Using qbw, one of our specialist longevity tools, we can perform analyses of pension scheme data to derive schemespecific mortality assumptions. Two main types of analysis can be performed in qbw to estimate the recent mortality level in a scheme: Estimate based on socio-economic profile - For example by considering the income, job titles of members, or using a postcode model. Estimate based on experience analysis - Looking at the actual deaths that occurred in a scheme compared with how many were expected. Summary of exposure and deaths - Example scheme 700 Exposure Actual Deaths Expected Deaths 600 Exposure (life years) Source: Barnett Waddingham longevity model Age band We also combine both a socio-economic and an experience analysis for a pension scheme, using statistical methods, to determine a credibility-weighted, single best estimate adjustment to a chosen base table. The resulting assumption will be more reliable, as it is based on more information and potentially allowing for the removal of unnecessary margins for prudence. Bulk Annuity Annual Report

44 Recent mortality and CMI_2016 The chart below shows a simplified index which illustrates how longevity has improved over the past 35 years. For every 100 deaths in the year 2000 in a standardised pensioner population, the graphs shows how many deaths were seen in earlier or in later years. Mortality rates fell steadily over the 1980s and 1990s (illustrated by fewer deaths over time), and longevity improvements appeared to accelerate between 1998 and 2011 with longevity improvements of around 2.5% pa, but longevity improvements have been close to zero since Improvements in the first half of 2017 (not shown on the chart) have also been low, leading to a growing consensus that recent experience could represent a change in future longevity trends, rather than just a short-term blip. Illustration of historical improvement in longevity Improvement: 1.3% p.a 2.5% p.a 0.3% p.a Source: Barnett Waddingham calculations, based on ONS data. A scheme adopting CMI_2016 will see its liability value fall due to reduced life expectancy when compared with earlier projection models. It is important that schemes consider their specific membership profile and characteristics when setting their longevity assumptions. The CMI s projections reflect longevity improvements at a national level as they are calibrated to the general population of England & Wales. There is no guarantee that pension scheme members will have the same experience as the England & Wales population particularly for schemes of white collar workers and individuals with large pensions. Indeed, the CMI s analysis of its own data suggests that pension scheme members generally have had higher longevity improvements in recent years compared to the general population. It remains to be seen how insurer pricing may or may not change in the future to reflect the most recent mortality experience. If schemes begin to adopt lower assumed rates of future improvement as a result of the new data, but this is not reflected in the pricing of reinsurance (and consequently bulk annuity insurers) this may lead to a perceived reduction in the attractiveness of insurer pricing. 44 Bulk Annuity Annual Report 2017

45 If you would like to contact us to discuss bulk annuities, please contact: GAVIN MARKHAM FIA Partner and Head of Bulk Annuities CHRIS HAWLEY FIA Associate, Actuary MARK PAXTON Senior Bulk Annuity Consultant For information about longevity risk transactions, please contact: RICHARD GIBSON FIA Partner and Head of Longevity Risk Transactions For information about financial strength reviews of insurers, please contact: KIM DURNIAT FIA Partner and Head of Life Consulting Bulk Annuity Annual Report