Escrow White Paper. - reconciling stability and surplus. Alternative Finance. Think Pensions Stability Think Aon Hewitt

Transcription

1 Escrow White Paper - reconciling stability and surplus Contributors of Stability Alternative Finance Think Pensions Stability Think Aon Hewitt September 2014

2 Contents Page 01 - Synopsis Page 02 - Introduction Page 03 - What is a trapped surplus and why are we talking about it? Page 05 - Reducing the likelihood of trapped surplus Page 07 - How does an escrow work? Page 09 - Can an escrow be more efficient than paying into the scheme? Page 11 - When should schemes start to consider escrows? Page 12 - Tax position of the sponsoring employer Page 13 - Does the modelling tell the whole story? Page 14 - Implementation considerations Page 16 - Conclusions Page 17 - Appendix 1 More details of the modelling we have carried out Page 20 - Appendix 2 Summary of Pensions Stability White Paper turning theory into reality

3 Synopsis Escrow White Paper reconciling stability and surplus This paper summarises our research into the use and effectiveness of escrows within the wider context of Pensions Stability. In June 2014 we released our analysis of the finances of low risk pension plans. This was set out in Pensions Stability White Paper turning theory into reality, which explored what defined benefit (DB) schemes may look like in an environment of Pensions Stability, and how this may affect the decisions which schemes make now. It also showed how dealing with an over-funded scheme can become an issue, and could prevent schemes from achieving the stability that they seek. If you would be interested to find out more, your usual Aon Hewitt consultant will be happy to send you a copy of the white paper. Alternatively, a synopsis of the Pensions Stability White Paper is set out in Appendix 2. In this paper, we consider how pension schemes concerns over surplus can be addressed. Concerns around surplus - and specifically the idea of a trapped surplus - may seem a dim and distant issue for many pension schemes, but the modelling we have undertaken shows that it is an issue that may need to be addressed today as schemes become better funded and get closer to their desired long term solution. We recognise trustees would prefer cash paid into the scheme, but sponsoring employers will become increasingly reluctant to commit contributions to well-funded schemes, given the restrictions involved in the return of surplus funds, and may therefore be attracted to alternative financing approaches. For this paper we have focused on escrows as an example of such alternative financing methods, although in practice there are a range of other vehicles that can be used, such as charges over assets, letters of credit or surety bonds. Depending on how the rules of the escrow triggers are set, the use of an escrow for some or all future pension contributions may present an achievable balanced outcome for many pension scheme sponsors and trustees. For the scheme sponsor it may be more financially efficient than paying contributions directly into a pension scheme. Our modelling has shown that, for our specimen scheme, an escrow becomes an attractive financial solution once the scheme funding position reaches around 85% on a gilts + 0.5% p.a. basis and assuming a corporation tax rate of 20%. Where the sponsoring employer does not pay corporation tax, escrows will be an attractive solution regardless of the scheme funding position. For trustees who agree to escrow solutions, the payback could be that the sponsoring employer is more comfortable funding to higher levels in aggregate than they would be if paying contributions only into the scheme. This achieves greater security for members benefits and gives greater stability to overall funding. Scheme members would gain from the higher likelihood that their full benefits will be paid. Escrows could therefore bridge the gap and enable all parties to benefit from greater stability. 1

4 Introduction We can reasonably expect the funding level of defined benefit (DB) pension schemes to improve. Deficit contributions continue to be paid, while low gilt yields mean that almost any positive performance on growth assets will see assets increasing more quickly than liabilities. When that happens, at some stage, contributions to a wellfunded scheme are unlikely to make financial sense for the sponsoring employer and will require explanation to hard pressed shareholders. For assets in a pension scheme, the normal position is that a refund of surplus is only possible when the scheme is wound up and its liabilities are fully secured which could be many years in the future. There is also a significant possibility that the existence of a surplus may encourage demands for benefit improvements. For this reason, sponsoring employers are faced with the risk that any surplus which has built up in their scheme becomes trapped. If that was not enough, there is now a proposal that in some circumstances accounting surplus may not be able to be recognised under IAS19 in future for some sponsoring employers. This will bring forward the point at which some sponsoring employers are concerned about surplus, noting that 25% of the FTSE350 already have an accounting surplus. The reductions in corporation tax rates in recent years have also significantly reduced the tax efficiency of pension contributions. For all of these reasons, some sponsoring employers may be unwilling to put money into their pension schemes at the level required to achieve Pensions Stability entirely within the scheme itself. Therefore, while trustees may want more funding in the scheme in order to have greater confidence that they can meet members benefits, they may well find that receiving additional contributions after full funding is reached will become increasingly difficult. In this paper: We set out why trapped surpluses are an important consideration for schemes today We consider the likelihood of trapped surpluses occurring in future We explore how an escrow can work as a buffer to aid pensions stability We consider when an escrow is likely to be particularly effective We discuss some of the practical implementation issues of escrows Terminology When we discuss escrow in this paper we are looking at the broad economic concept. In the early parts of this paper, if you are familiar with charged accounts, funding trusts or special purpose vehicles we are grouping together all of these structural forms in the term escrow. If you are not familiar with these legal structures we will introduce you to them in the Implementation considerations section on page 14. We hope that you find this paper interesting and informative. If you have any questions or would like to explore any of the issues raised in further detail, then please contact your usual Aon Hewitt consultant or one of the authors. David Bush / david.bush@aonhewitt.com Lynda Whitney lynda.whitney@aonhewitt.com Kevin Wesbroom kevin.wesbroom@aonhewitt.com 2

5 What is a trapped surplus and why are we talking about it? What is a trapped surplus? Put simply, trapped surplus arises where sponsoring employers are unable to recover surplus funds from overfunded pension arrangements. In the UK, for example, sponsoring employers are not generally able to recover surplus assets from their pension arrangements until a scheme is wound up and its liabilities are secured in full. Even in this situation, a refund of surplus assets to the sponsoring employer may not occur because surplus funds can also be used to augment member benefits. If sponsoring employers do take a refund of surplus assets, the refund is also subject to a significant tax charge of 35%, whereas the tax relief they would have received on the way in is more likely to have been 30%, gradually falling to 20% more recently (if indeed any tax relief has been received). Trapped surpluses are therefore something that sponsoring employers are generally keen to avoid. Throughout the last decade, the pensions landscape has changed dramatically in the UK. For the majority of DB pension schemes, the funding surpluses which were commonplace throughout the 1990s and early 2000s disappeared, and were instead replaced by deficits. Scheme deficits have arisen for a variety of reasons, most notably rising life expectancies and falling bond yields which have made pension scheme financing more expensive for sponsoring employers. As a consequence of this, sponsoring employers have sought to address the deficits by paying significant extra contributions into their schemes. Why are we talking about trapped surpluses now? Although the risk of trapped surpluses may not have been a key focus in trustee and sponsor discussions in recent years, it is very important for all parties involved to consider the risks of them arising at some point in the future, as this may affect the decisions which are taken today. Chart 1 below shows how the Technical Provisions and IAS19 accounting positions of a typical scheme have varied in recent years. Chart 1 Financial development of typical scheme Funding position 105% 100% 95% 90% 85% 80% 75% 70% 65% 60% 01-Jan Jan Jan Jan Jan Jan-14 Funding Accounting As can be seen from Chart 1, the funding position of a typical scheme has improved considerably over the last couple of years. The IAS19 position has generally been less volatile than the funding position and the positions on both measures have converged somewhat in recent years. In total, roughly 25% of schemes in the FTSE 350 currently have an IAS19 surplus. Is this upward trend in funding levels likely to continue? If everything happens as the actuaries and financial markets expect, funding levels should increase. However, as we know, things do not always turn out as expected and it can be helpful to consider what future funding levels might look like if things follow a different path. Factors contributing to the increasing risk of trapped surpluses include: Pension scheme funding levels have actually improved in recent years since the global financial crisis in 2008/9. If economic conditions continue to improve as some expect, there could be significant future improvements in funding levels in the coming years. Schemes closing to new entrants and future accrual, together with the move to contract-based Defined Contribution (DC) provision, have taken away the ability to use up surpluses by funding future service benefits. 3

6 Specimen Scheme For the modelling presented in this paper, we have looked at the same specimen scheme used in our Pensions Stability White Paper. This is a private sector scheme, closed to future accrual, with fairly typical characteristics for example, having a roughly even split between pensioner and non-pensioner liabilities and with the majority of its benefits linked to inflation. Chart 2 Stochastic development of specimen scheme In our modelling, we have assumed that: The scheme holds 100M in assets The scheme s assets will target an investment return of 0.5% p.a. in excess of gilt yields (net of manager fees) once the stable position is reached. Interest rate and inflation risks have been fully hedged, using derivatives where needed. The funding target is set to be 0.5% p.a. in excess of gilt yields. Appendix 1 gives more details on how we have carried out our modelling and about the scheme involved. Chart 2 opposite shows a stochastic projection of future funding levels for our specimen scheme over the next 20 years, in this case assuming that the scheme is already fully funded and that mortality risks have been fully hedged. Our modelling involves a large number of projections of the future under varying economic and market conditions: The line at the top of the blue area is the 95th percentile outcome (i.e. only 5% of outcomes would be expected to be better than this); The line at the bottom of the red area is the 5th percentile (i.e. only 5% of outcomes would be expected to be worse than this); The chart also has lines marking the 25th, 50th (i.e. median) and 75th percentiles; The red line at the bottom is the first percentile. So even when assessed against a conservative funding target and with a relatively low risk investment strategy which is fully hedged against interest rate, inflation and mortality risks, significant surpluses can arise. If schemes want to be well funded, so the downside events are protected against, then one consequence is that outcomes giving rise to a surplus are more likely to occur than not. As outlined earlier, for a refund of surplus the normal position is that a refund is only possible, and even then not guaranteed, once all the scheme s liabilities have been fully secured. An alternative way to consider this type of output instead of running stochastic projections is to look at the specific economic scenarios that might give rise to such a range of outcomes. If you would like to see stoachstic projections such as Chart 2 or specific economic scenarios for your scheme circumstances, please contact your usual Aon Hewitt consultant or the authors of this White Paper. This chart shows a very large range of funding outcomes over a 20 year period from what would commonly be regarded as being a reasonably low-risk investment portfolio of gilts +0.5% p.a. there is a 90% likelihood that the funding position will be between a deficit of 12m and a surplus of 45m after 20 years (i.e. these are the deficits shown at the 5th and 95th percentiles on the chart). 4

7 Reducing the likelihood of trapped surplus Many trustees would argue that the risk of trapped surplus is low, because they can de-risk their investment policy and so use up any surplus which arises on their Technical Provisions funding basis. That is true in the short term, but we have shown that a significant risk of overfunding still remains, even when investing to target a relatively low risk return of gilts +0.5% p.a. Many sponsoring employers would not want to de-risk much further than that. In our Pensions Stability White Paper we also found that this was the case for even more conservative investment strategies such as those targeting a gilts + 0% p.a. return. DB pension schemes that are open to new entrants have a natural safety valve for dealing with the risks of overfunding and corresponding concerns regarding trapped surplus any surplus that arises can be used to fund future service benefits. However, with about 90% of DB schemes closed to new entrants and over 50% frozen to all future DB accrual (with more of such freezes expected to occur in coming years), this safety valve no longer exists for most schemes. In a similar vein, DB schemes with a DC section in the same trust may also have a safety valve. For these schemes, where the scheme rules allow, it is possible for any DB surplus to be used to fund future DC contributions. However, with the recent trend being towards setting up contract-based or mastertrust DC schemes, rather than own trust-based DC schemes, such pension arrangements are becoming less common. One way of dealing with these concerns is to hold a buffer outside of the pension scheme that addresses the downsides of trapped surpluses, but still provides additional security for members benefits if needed. The concept of a buffer is very common in the insurance world where the insurer holds reserves and expects cashflows to and from those reserves. This is demonstrated in Chart 3. Chart 3 Buffer in an insurance environment Capital injected at outset Premium Insurer Reserves Expected cost of benefits Profit Expenses Flow of cash in and out of reserves Benefits 5

8 Chart 4 Buffer in a pensions environment Capital injected at outset Contributions Sponsor Buffer Pension scheme Surplus Expenses Flow of cash in and out of buffer Benefits For a sponsoring employer to run its pension scheme in a similar way to an insurance company is not straightforward. In particular, once assets are put into the pension scheme they can only be taken out in very restricted circumstances. However, by using a buffer, which is in some way pledged to the scheme and its beneficiaries, that difficulty can be overcome. The equivalent diagram for a pension scheme with separate buffer might be as shown in Chart 4. This has contributions paid to and from a buffer, and only paid into the pension scheme when it is efficient to do so. By way of illustration, we have looked at our specimen scheme which holds assets of 100m. Assuming the scheme is fully funded on a gilts + 0.5% p.a. basis at 31 December 2013, we have considered the impact of holding an additional 6m buffer either inside or outside the scheme. When simply paid into the scheme as a normal contribution, such a buffer can: reduce the likelihood of further sponsoring employer contributions being required in the next 20 years from 50% to just 10% increase the probability of reaching full funding on the buy-out basis over the next 20 years to over 70%. More information on this can be found in our Pensions Stability White Paper. When paid into a buffer outside of the scheme, the additional 6m can have a similar impact to that outlined above, but with benefits to the sponsoring employer if the buffer turns out to not be required. For example in the 70% of cases where the scheme has been able to reach full funding on a buy-out basis, the surplus will definitely revert to the sponsoring employer. There are many alternative financing tools that can perform the role of the buffer. They are also used to solve a variety of other problems, for example the management of PPF levies, cashflow management, scheme mergers or covenant changes during M&A activity, etc. The range of tools includes charges over assets (sometimes wrapped in an asset-backed contribution vehicle), surety bonds (which leaves the buffer in the sponsoring employer but insures the risk that the sponsoring employer cannot pay) and parent company guarantees (where the trustees are comfortable that the parent company has the scale to manage the buffer within the business). However, the simplest solution that solves the trapped surplus issue is an escrow. In the rest of this paper we will focus on the escrow as a buffer but other alternative financing solutions should also be considered if there are secondary objectives such as management of cashflow. Please speak to your usual Aon Hewitt consultant or the authors of this White Paper if you would like to discuss your other objectives and the type of buffer most suitable in your particular circumstances. Use of escrows in the public sector Although primarily seen in schemes in the private sector, escrows may also have a role for sponsoring employers participating in the Local Government Pension Scheme for a finite period, such as those providing outsourced services to local government. They could do so for much the same reason of avoiding trapped surplus. 6

9 How does an escrow work? The principle of an escrow is that assets are set aside by the sponsoring employer and held by a third party on behalf of both the sponsoring employer and the scheme. The assets in the escrow are then released to the scheme or sponsoring employer based on pre-determined triggers. This bridges the gap between trustees who prefer more cash in the scheme and sponsoring employers who are concerned about a trapped surplus. In this section we consider some of the advantages and disadvantages of using escrows in this way. Advantage 1 An escrow avoids overpaying into the scheme Consider our sample pension scheme. If the 6m buffer illustrated on the previous page was instead held as additional assets within the scheme, there is over a 70% chance that the scheme would be overfunded relative to a buyout basis in 20 years time. In these situations there may be limited opportunities for the excess funds to be paid back to the sponsoring employer. Holding the buffer within a separate escrow gets round this problem by returning the money to the sponsoring employer if it is no longer required. Advantage 2 An escrow provides security for trustees and members If, by using an escrow, the sponsoring employer can be encouraged to target a higher funding level in total than if paying contributions only into the scheme, this can provide better security for trustees and members. Sponsoring employers may be willing to do this because there are circumstances where the monies held within the escrow can be recovered efficiently, whereas this is not always the case in a pension scheme. Advantage 3 An escrow allows assets to be retained on the sponsoring employer s balance sheet The treatment of pension deficit varies between rating agencies, banks, industry regulators, and investors. Retaining the asset on the balance sheet of the sponsoring employer rather than in the pension scheme can be advantageous in some of these assessments - and will at the very least be neutral. From perhaps January 2017, some companies may no longer be able to recognise accounting surpluses and may need to recognise recovery plan payments. Therefore the advantage of having an escrow on the sponsoring employer balance sheet rather than cash in the scheme may increase dramatically for some sponsoring employers. Starting an escrow now will ensure the sponsoring employer does not get closer to the accounting problem in See Accounting treatment of pension schemes on page 14 for more details. Advantage 4 An escrow can give the sponsoring employer some control over investment strategy The sponsoring employer can agree with the trustees that it retains control of the investment strategy of the escrow (or has joint power with the trustees). In addition, by considering the overall investment strategy of the escrow and the pension scheme, the sponsoring employer can adjust the total investment strategy by flexing the strategy used for the escrow to equal the overall risk they wish to have. They do not therefore need the trustees to change the investment strategy of the pension scheme in order to make changes to the overall strategy. However, they need to remember that in most cases the trustees can change the pension scheme s investment strategy if they so decide. 7

10 Disadvantage 1 An escrow means that tax relief on contributions is deferred until the contributions are paid from the escrow to the pension scheme For contributions paid into a pension scheme, getting tax relief at the time the contribution is paid can seem attractive compared to payments into an escrow, which do not benefit from this tax relief. However, tax relief on the contributions made to the pension scheme are at the prevailing corporation tax rate, which is falling to 20% in April 2015, and which can be lower if the sponsoring employer is not currently paying corporation tax. In contrast, the tax charge for a refund of surplus is 35%. Disadvantage 2 An escrow incurs tax payable on investment returns The investment returns earned within an escrow are taxable, unlike the returns earned on pension scheme contributions which are tax free. However, taking into account the 35% tax charge applied to a refund of surplus pension contributions, one may need to wait many years in order to get the same amount back via pension scheme contributions rather than via escrow contributions. This concept is discussed in more detail in the section Tax position of the sponsoring employer on page 12. Myth Busting There also a number of myths about escrows and this section aims to dispel those myths. Myth 1 An escrow can only be invested in cash An escrow can be invested in a wide range of investment classes. The myth has grown up because some escrow agents only want to be responsible for a narrow range of investments. By getting the right structure to your escrow at the start, you can invest in the full range of investment classes used by pension schemes. Myth 2 An escrow will be disliked by the Pensions Regulator The Pensions Regulator recognises that there is a wide range of alternative financing solutions available. There are certain types of contingent assets that the Pensions Regulator has more concerns about, in particular those where the asset is currently used by the company. An escrow therefore causes no concerns as long as it is invested in normal pension asset classes. Disadvantage 3 An escrow needs costs and resources to be set up and run To put an escrow in place, appropriate legal documentation is needed, along with a vehicle to hold investments outside of a tax-free pension wrapper and a custodian to protect the assets. While setting up an escrow is relatively straightforward and cheap to do (compared to more complex solutions such as asset backed contribution vehicles), it can still be time consuming and has fees attaching. We believe costs can be managed if the key triggers are agreed as part of the integrated funding strategy, and if the sponsoring employer and trustees are willing to have a joint working party to resolve other issues pragmatically. It may also be possible for the sponsoring employer and trustees to appoint a single lead lawyer to manage the drafting rather than seeking separate legal advice. Packaged products are under development which will greatly reduce the cost and increase the flexibility of the assets than can be invested in. 8

11 Can an escrow be more efficient than paying into the scheme? We have undertaken stochastic projections to compare the value of future sponsor contributions paid into an escrow versus those paid directly into a pension scheme. These projections allow for the different tax treatment of contributions in both cases, and they also include escrow triggers for the payment of contributions both back to the sponsoring employer and into the scheme. When assessing value in this instance we have calculated the Net Present Value (NPV) of contributions payable into an escrow and into a pension scheme, where NPV refers to the value today placed on all future contribution payments, adjusted for tax relief and any subsequent refunds of surplus. For this reason, a lower NPV is a better outcome for sponsoring employers because it represents a lower cost of future contribution outgo. A key consideration when setting up an escrow is to decide the terms of the triggers for when the funds in the escrow are paid into the scheme, or back to the sponsoring employer. A must have trigger for trustees is one which triggers the payment into the scheme in the event of sponsoring employer insolvency. Another relatively common trigger for payments to the scheme is if there is still a scheme deficit at the end of a specified timeframe. Other less common triggers for payments into the scheme include a credit down-grading of the sponsoring employer, or if any future deficit recovery periods exceed an agreed length. From the sponsoring employer s perspective, the must have trigger is repayment of funds to the sponsoring employer if there is a surplus on wind-up. Other potential triggers for payments back to the sponsoring employer include an agreed funding level being exceeded (possibly over a period of time), or if the trustees adopt an investment strategy that is not agreed with the sponsoring employer. Our research to date into the use of escrows in pension scheme financing has indicated that the rules for how the escrow triggers are set up are very important and can have a significant impact on the modelling results. In this paper we have modelled a particular set of possible escrow triggers, as set out below, but there will be several different ways in which these triggers can be structured that could give rise to materially different outcomes. In practice the trigger points agreed are likely to take account of the specific circumstances of the pension scheme and the sponsoring employer, including the interaction between the funding and investment position of the scheme and the covenant strength of the sponsoring employer. For our sample scheme, with liabilities of 100m on a gilts + 0.5% p.a. basic funding target, we have carried out stochastic projections to model the following over a 20 year projection period: An initial funding level of 100% within the scheme. If the funding level drops below 100% at any time, further sponsor contributions are assumed to begin to take the funding level up to 100% within the next three years. In addition, further contributions are payable in subsequent years with the aim of targeting a funding level of 105%. In each projection we have compared the outcome of paying these contributions either directly into the scheme or into an escrow. If contributions are paid into escrow they are added to scheme assets when measuring the funding level. For the projection using the escrow, if the funding level exceeds 110% during the projection period, any excess over 110% is paid back from the escrow to the sponsoring employer immediately. At the end of projection period, if the funding level is 106% or above, the scheme is assumed to buy out. At this point, if there is a deficit, escrow funds are transferred into the scheme to the extent required to pay for the buy-out. Any escrow surplus is returned to the sponsoring employer. Any surplus funds in the scheme are refunded to the sponsoring employer allowing for the 35% tax charge. There is a gradual decrease in the level of risk in the asset strategy as the funding level improves. The asset strategy assumes a typical mix of growth assets (such as equities) and matching assets (such as bonds or swaps). An assumption that the sponsoring employer is paying corporation tax at 20%. 9

12 The NPV is then determined by discounting the future contribution payments by the sponsor s cost of capital, here assumed to be 10% p.a. More details of the modelling can be found in Appendix 1. The results of our modelling are shown in Chart 5 below. This shows that paying future contributions into the escrow results in a lower NPV of future contributions than paying into a pension scheme in 84% of the projections modelled. The average NPV amount that the escrow outperforms by is 0.8m, which is equivalent to around 15-20% of the future contributions paid by the sponsoring employer - albeit it is worth noting that the level of future contributions paid by the sponsoring employer can fluctuate significantly across the different projections. Chart 5 NPV escrow comparison, assuming 100% funding level and 20% corporation tax Gain/Loss to the sponsor 8.0m 6.0m 4.0m 2.0m 0.0m - 2.0m - 4.0m - 6.0m Buffer with 100% initial funding, 20% corporation tax In this example, therefore, paying contributions into an escrow may be more attractive to the sponsoring employer than paying contributions directly into the scheme. It can be seen from Chart 5 that the escrow generally provides better value for the sponsoring employer as the average asset returns increase. The extra value of the escrow comes from various sources, with the two most common sources being the more effective tax treatment on contributions payable (i.e. where it is better to lose out on the 20% tax relief in order to avoid having to pay the 35% tax charge on a refund of surplus), and the ability of the sponsoring employer to take a refund of contributions earlier than eventual scheme wind up (i.e. given the sponsor s cost of capital, the NPV of a refund of 5m in five years time is just over 3m, whereas the NPV of the refund after 20 years is only 0.7m). The relative importance of these two sources will vary depending on the specifics of each scheme, including how well funded a scheme is to start off with and the tax position of the sponsoring employer. We expand on both of these areas in more detail in the next sections m -20% -10% 0% 10% 20% 30% 40% Average asset return p.a. 10

13 When should schemes start to consider escrows? To indicate at what point a sponsoring employer might want to start to divert future contributions into an escrow, we have compared the NPV of future contributions into an escrow to those paid directly into a pension scheme for a number of assumed starting funding levels. In practice, the funding measures which drive this decision will differ between sponsoring employers some sponsoring employers may use their corporate accounting measure (such as IAS19) while others may use the Technical Provisions or alternative measures. In our modelling, we have used a gilts + 0.5% p.a. measure for ease of comparison with the results shown on page 10. We have made this comparison assuming that the sponsoring employer is subject to corporation tax (see Table 1 below). There are more details on the impact of the marginal corporation tax rate in the next section. Table 1 NPV escrow comparison for different funding levels for tax paying employer Starting funding level on a gilts + 0.5% p.a. measure (level of assets) 80% ( 80m) 90% ( 90m) 100% ( 100m) Escrow NPV is better 48% of the time Escrow NPV is better 70% of the time Escrow NPV is better 84% of the time This shows that for a sponsoring employer which is paying corporation tax, the escrow becomes more attractive as the funding level improves; in the scenario we have modelled for our specimen scheme, the escrow starts to become an attractive option for initial funding levels as low as 85% or more on a gilts + 0.5% p.a. measure. In Appendix 1, we show the results of the above modelling using scatter charts in the same format as Chart 5. For regulated financial services entities, escrows may be even more attractive. Contributions into well-funded pension schemes can represent lost capital reserves because any accounting surplus in the pension scheme is usually written off from core capital ratios. A contribution to an escrow, however, may be more capital efficient as it may not need to be fully written off. The exact treatment of an escrow for capital purposes will depend on how the particular regulated entity treats such accounts more generally (and how its regulator expects the accounts to be treated). However, the escrow treatment should not be worse than a cash contribution to a pension fund, and could be much better. Further detail of the accounting treatment of escrows is included in the Implementation considerations section on page 15 of this paper. If you would like to see the results of similar modelling comparing an escrow to putting money into the scheme for your circumstances, please contact your usual Aon Hewitt consultant or the authors of this White Paper. 11

14 Tax position of the sponsoring employer The tax position of the sponsoring employer has a significant impact on when it may make sense to stop paying contributions directly into the pension scheme and divert them to an escrow. In the analysis in Table 2 we compare the same scenarios as those on the previous page but consider what happens if the sponsoring employer is or is not paying corporation tax. Table 2 NPV escrow comparisons for different funding levels and tax status Sponsoring employer is paying tax Sponsoring employer is not paying tax Starting funding level on a gilts + 0.5% p.a. measure (level of assets) 80% ( 80m) 90% ( 90m) 100% ( 100m) Escrow NPV is better 48% of the time Escrow NPV is better 91% of the time Escrow NPV is better 70% of the time Escrow NPV is better 97% of the time Escrow NPV is better 84% of the time Escrow NPV is better 99% of the time For a sponsoring employer paying tax the circumstances where the NPV is worse for the escrow are those where most or all of the buffer ends up being paid into the scheme, but the sponsoring employer has to wait to receive the tax relief. To illustrate this, imagine a sponsoring employer puts 100 aside either in a pension scheme or escrow, and generates a gross (before tax) investment return of 5% p.a. The money is expected to be repaid at some stage when the scheme buys out. How long would it take before the tax advantages of the investment returns in the scheme outweigh the 35% tax charge on the way out? At current corporation tax rates the payback period is 22 years. Therefore if a refund of surplus is likely in the future, sponsoring employers subject to 20% tax relief on pension contributions could potentially have some very long payback periods. The payback period becomes even greater when investment returns are assumed to be lower. Therefore the current tax position of the sponsoring employer, as well as how this tax position may change in the future, has a material impact on the relative attractiveness of using an escrow. And for sponsoring employers subject to corporation tax at 20%, it may actually be tax inefficient to continue to pay additional pension contributions directly into the scheme if there is a reasonable chance that those contributions will end up not being required. For a sponsoring employer that is not paying corporation tax, or is not paying enough to get full tax relief on the contributions, the impact of that deferred tax relief could be much lower, or even nil. The Pensions Regulator estimates that approximately a quarter of DB scheme sponsors have a tax position such that they cannot claim tax relief on some of the contributions it pays, and for those sponsors an escrow will be even more attractive. Even when a sponsoring employer is subject to corporation tax, reductions in tax rates in recent years have increased the attractiveness of escrows. The tax charge of 35% for a refund of surplus was put in place as part of the Finance Act At that time corporation tax (and hence tax relief on any pension contributions) was 30%, so the penalty was just 5% extra. However, since 2007 the Government has been gradually reducing corporation tax. From next April, corporation tax will hit an all-time low of 20% (which compares very favourably to the 52% peak hit in the late 70s and early 80s). 12

15 Does the modelling tell the whole story? Even where our modelling shows only a marginal advantage, other factors may still make an escrow attractive including: funding level improvements may occur in the future as a result of liability management techniques such as trivial commutation, pension increase exchange or enhanced transfer value exercises, especially after the 2014 Budget changes. legislative changes may occur that lead to funding level improvements, and in these cases the escrow structure provides additional flexibility. A relatively recent unforeseen example of such a legislative change was the move from pension increases linked to the Retail Prices Index to increases linked to the Consumer Prices Index. Sponsoring employers will not want an unknown unknown to cause a trapped surplus. Key factors for escrows Trapped surplus measured on which basis? What triggers? Current Funding Level Corporation tax paid? 13

16 Implementation considerations As well as the design of the trigger rules, there are a number of other practical considerations to take into account when determining how to set up an escrow, or indeed any other type of buffer. We have highlighted some of the main ones here. Investment choice When considering what assets to place in the escrow, in general we recommend that the trustees and sponsoring employer consider the assets of the scheme and escrow as one, when setting an overall balance between return seeking assets and matching assets. The risk and return associated with this combined asset portfolio can be set in light of the overall financing objectives for the scheme. The choice of which assets to invest in the escrow can then be driven by other factors. For example: As the investment returns on assets held in escrows are subject to tax, there is an argument for biasing the escrow towards relatively low risk and low return assets, with the risk and return on the assets in the pension scheme set slightly higher in order to achieve the desired overall risk and return objective. Of course if the sponsoring employer is not paying tax (and does not expect to do so in the near future), returns on assets held in escrow are effectively tax free, and this becomes less of a factor. If gains/losses in the escrow have to go through the sponsoring employer s profit and loss account, there may be a desire to hold less volatile assets in the escrow. Our understanding is that at present escrow gains/losses can be put through a sponsoring employer s accounts as Other Comprehensive Income instead of the profit and loss account. However, this could change in the near future see the Accounting treatment of pension schemes section opposite for further details. The recognition for PPF levy purposes requires the assets to be cash, UK property or securities. Legal wrapper The exact legal wrapper for the escrow will also be important to the implementation. There are a number of possible legal wrappers including: contract law - for example a charged account trust law - for example a settlor trust special purpose vehicle - for example a subsidary company established for the sole purpose of operating the escrow The legal wrapper chosen will depend on factors such as whether there is an overseas company involved, detailed tax considerations and the views regarding the investment strategy. Accounting treatment of pension schemes At meetings in July and September 2014 the IFRS Interpretations Committee of the IASB agreed to develop amendments to the provisions in IFRIC14 which control whether a company can recognise an asset in its accounts in relation to any pension plan surplus under IAS19. Their tentative conclusion was that Where Trustees have the power to use up a surplus by augmenting members benefits, the surplus should not recognised as an asset unless the entity has an unconditional right to an immediate refund of this surplus. Where Trustees have the power to wind up a plan an entity cannot realise an economic benefit from a refund of surplus through a gradual settlement of the plan, i.e. once the last pensioner member died. That is an asset that can only be recognised to the extent that a refund of surplus would be available immediately or to the extent that the surplus exceeds the cost of a complete wind-up of the plan. The Trustees power to make investment decisions including buying annuities while the plan is ongoing does not affect the economic benefit available from a refund of surplus, consistent with BC 10 of IFRIC 14 ( The existence of the asset at that date is not affected by possible future changes to the amount of the surplus. If future events occur that change the amount of the surplus, their effects are recognised when they occur. ) 14

17 This means that some sponsoring employers will not be able to recognise an accounting surplus if it occurs under IAS19. About 25% of the FTSE350 already have accounting surpluses and many more are heading that way. This proposal would therefore have negative impacts on both the balance sheet and P&L account for some employers. The Committee recognised this proposal as a significant change and so it seems unlikely that any amendments will take effect before 1 January However we expect it may change funding discussions now and lead to more sponsoring employers being concerned about trapped surplus earlier. This will lead to the need to use alternative financing solutions such as escrow to bridge the gap between the trustees and sponsoring employer s viewpoints. Points to consider include: The tax and accounting advantages of having the escrow under one legal entity, rather than another. Whether there are advantages from an insolvency protection perspective of having the escrow set up under a particular legal entity. Whether there is any reason for the contributions that are paid into the escrow to come from a different legal entity to the one that will receive any potential return of funds from the escrow. If so, how do these legal entities relate to the pension scheme s participating employers, and does this create any issues for the directors of legal entity making the contribution? Accounting treatment of the escrow As indicated above, our understanding is that at present gains/losses on escrow investments can go through a sponsoring employer s accounts as Other Comprehensive Income, provided the escrow investment is designated as available for sale (a term defined in IAS 39). If the escrow is not treated in this way, asset volatility will go through the profit and loss account. Depending on the size of the escrow compared to the size of the sponsoring employer, this could be a concern for sponsoring employers. However, we understand that the available for sale category is not included in IFRS 9 (the standard that replaces IAS 39, available for accounting periods beginning on or after 1 January 2015 and mandatory from 1 January 2018). Therefore putting gains/losses on escrow investments through the profit and loss account may be the only option from 1 January Where on the balance sheet does the escrow sit? The escrow is generally shown in the sponsoring employer s accounts as a company asset, rather than a pension scheme asset (unless the triggers are overly biased towards the scheme, i.e. where there is little chance of the sponsoring employer having escrow funds returned to it). Careful consideration will be required on exactly where in the corporate structure the escrow should sit. Accounting, legal and taxation issues will drive this decision. 15

18 Conclusions In this report we have covered a number of areas in relation to escrows. In summary: The risk of trapped surpluses can still be very material even in pension schemes with conservative funding measures and relatively low risk investment strategies. A buffer outside of the scheme can help to mitigate this, generating stability in a manner which balances the needs of sponsoring employers, trustees and members. There are a number of ways in which that buffer can be arranged. From a member perspective, an escrow may achieve greater security of benefits and give greater stability to overall funding. In addition, members could gain from the higher likelihood their full benefits will be paid. We therefore believe that trapped surplus vehicles, whether escrows or something different, could enable all parties to meet their needs, and create an environment of Pensions Stability. Escrows are a relatively straightforward and easy way to create a buffer, and can be an effective way to deal with concerns that a sponsoring employer may have about trapped surplus. From a corporate perspective, at some point during a scheme s flight plan to its destination, stopping paying contributions directly into the pension scheme and paying them into an escrow is likely to lead to a lower net present value of future contributions. The design of the escrow triggers will be crucial in this regard, but our modelling has shown that, for our specimen scheme, this point is reached when the scheme s funding level reaches around 85% or more on a gilts + 0.5% measure, assuming that the sponsoring employer is subject to 20% corporation tax. If the sponsoring employer is not paying corporation tax then escrows are an even more favourable solution and appear attractive regardless of a scheme s funding position. In addition, escrows may also be a way in which sponsoring employers can recognise a pension surplus on their balance sheets. From a trustee perspective, they would prefer cash in the scheme. But if they can achieve an increase in the security of members benefits by agreeing to implement a trapped surplus solution such as an escrow, with the sponsoring employer then prepared to fund the scheme and the escrow up to a higher level than the scheme alone, this can also be a successful outcome. 16

19 Appendix 1 More details of the modelling we have carried out For our modelling, we have used a scheme with the following characteristics: Current split of liabilities: 50% pensioners, 50% non-pensioners The scheme has been closed to future accrual The duration of the scheme is currently 17 years. This reduces to around 12 years by Pension increases in payment are a mix of typical DB scheme increases with the majority of pensions linked to RPI with a cap on annual increases of 5%. Deferred pension increases are linked to CPI. The initial funding liabilities are calculated using a discount rate of gilts + 0.5% p.a. and are equal to 100m. In order to carry out our modelling, we have Focused on a 20 year timeframe. Although a scheme s run off period will typically be around 70 years, it is likely to be at its largest during the next 20 years. Assumed that the cost of capital is 10% and the corporation tax rate paid (unless otherwise stated) is 20%. In this paper we have referred to investment strategies that target a median return of gilts + X% p.a. There are many ways to construct asset portfolios, however the broad characteristics of our portfolios are: A typical diversified portfolio of growth assets is used to generate the outperformance over gilts. The outperformance determines the amount to be held in growth assets. Interest rate and inflation exposure are hedged using gilt and swap market instruments. Typically a small amount of leverage is required to cover the liability exposures. Table 3 Modelling assumptions for different funding levels Initial funding level Initial Deficit contributions Initial Investment strategy In all cases we have assumed 80% 90% 100% 1.5m p.a. for 10 years Gilts + 2.0% p.a. 1.5m p.a. for 5 years Gilts + 1.5% p.a. n/a Gilts + 1.0% p.a. If the scheme plus escrow funding level reaches 105% on the gilts + 0.5% basis before the end of the period in which the initial deficit contributions are assumed payable, these initial deficit contributions cease to be paid and do not restart. If the funding level drops below 100% at any time, further sponsor contributions are assumed to begin to take the funding level up to 100% within the next three years. In addition, further contributions are payable in subsequent years with the aim of targeting a funding level of 105%. In each projection we have compared the outcome of paying these contributions either directly into the scheme or into an escrow. If contributions are paid into an escrow they are added to scheme assets when measuring the funding level. During the projection period if the funding level exceeds 110%, any excess over 110% is paid back from the escrow to the sponsoring employer immediately. At the end of projection period, if the funding level is 106% or above, the scheme is assumed to buy out at this point with any escrow funds transferred into the scheme if required in order to pay for the buy-out, and any excess scheme funds refunded to the sponsoring employer allowing for the 35% tax charge. There is a gradual decrease in the level of risk in the asset strategy as the funding level improves. For the triggers for when money flows into and out of the escrow, we have tried to be fairly realistic, and so we have slightly different approaches depending on what the initial funding level of the scheme is, measured relative to a gilts + 0.5% p.a. funding target. This is set out in Table 3 opposite. 17