The End Game?
An analysis of the bulk buy-out market & other de-risking solutions
Richard Jones FIA, Martin Hunter & Oliver Herbert
June 2008
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
●
The buy-out market amounted to slightly under £3 billion in 2007 which is very
small in relation to the £50+ billion capacity of the market and the growth of the
£800 billion of UK private sector pension liabilities (estimated at £28 billion per
annum). It is exceedingly unlikely that buy-outs will increase to such a level as to
start to reduce the size of UK private sector pension liabilities.
●
Nearly all the business in 2007 was written by just two insurers (Legal & General
and Paternoster) despite the large number of new entrants in the market.
Some high profile new entrants, such as Synesis Life, have yet to write any
annuity business.
●
People considering buy-out need to understand that a bulk buy-out does not
provide a guarantee to the benefits, but reduces the risk of a shortfall. However,
there remains a not insubstantial risk of members not getting full benefits due to
the insolvency of the insurer.
●
All life insurers are regulated by the FSA and subject to a restrictive regulatory
regime. This regime, combined with insurers’ profit requirements, severely limits the
price that insurers can charge. The new entrants do not have a “magic bullet” that
allows them to change the pricing dynamic.
●
Differences in mortality assumptions between company accounting and insurers’
pricing only account for around 25% of the additional cost of securing annuities –
the majority of the difference is driven by insurers’ profit requirements.
●
We, along with many market participants, believe that high levels of price
competition are leading to insurers accepting very low returns on capital in order
to ensure new business. This cannot continue forever and therefore prices are likely
to rise in the future.
●
The key consideration in selecting an insurer should be price, as the long term
nature of the contracts makes extrapolating capital strength, strategy or service
levels impossible. Pensioners are often perceived to be a “cheap” buy-out option,
since the premium payable to insurers is lower than for deferred pensioners.
However, this simply reflects the significantly lower level of risk presented by
pensioners, and as such they are not genuinely cheap.
●
Whilst all companies have a price at which they would willingly transfer their
scheme to an insurer, the current pricing, even allowing for the discounts currently
available, is uneconomic for the majority of sponsors.
●
Other alternative solutions also exist, the most interesting of which is sponsor
covenant insurance from BrightonRock.
●
We think it is extremely unlikely that the current noise around buy-out solutions
will lead to the rapid growth of the market predicted by some market participants.
1
Key Findings
Key Findings
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Executive Summary
With a large number of players entering the UK pension bulk buy-out market in
recent times, there has been considerable press speculation surrounding this market,
with some commentators expecting to see rapid growth in this area. This report seeks
to set out the current state of the annuity business, including the details of the
environment in which it operates, the risks faced by insurers and a discussion of
current pricing levels. It also discusses each of the current players in the market and
describes the annuity purchase process, as well as looking at the buy-out decision
from the perspective of the sponsoring employer. We then consider alternatives to
buy-out offered by the market, as well as other de-risking solutions available to
pension schemes and their sponsoring employers.
The State of the Market
Slightly under £3 billion of bulk buy-out business was written by insurers over 2007,
and we estimate that over the first four months of 2008 around £2.5 billion of
business has been written. However, it is worth noting that these figures have
been influenced by a small number of very large transactions, which had some
unique characteristics.
These figures are a drop in the ocean compared to the potential market available,
since the UK’s private sector pension schemes currently have buy-out liabilities in
excess of £1 trillion. Although the new entrants have significantly increased the
capacity of the bulk buy-out market, the current volumes of business are having
negligible impact on the overall size of the UK private sector’s pension liabilities.
Indeed, we have estimated that the size of “UK plc’s” defined benefit pension
liabilities actually increased over 2007.
Insurance legislation is considerably more onerous than pensions legislation. In
particular, insurers must meet tighter reserving requirements, typically requiring them
to hold capital margins of around 8% of their “technical provisions”. Despite this,
insurers still face a material risk of insolvency, of something less than 5%, which
trustees and scheme sponsors should consider when evaluating the buy-out decision.
Pricing
We have constructed a model to estimate the impact of the FSA’s reserving
requirements on the pricing of insurers. In order to obtain their pre-tax return on
equity of around 15% per annum, insurers would generally need to price the pension
scheme’s liabilities using a discount rate of around 30 basis points below the longterm gilt yield. Coupled with a prudent allowance for future improvements in
longevity and expenses, this might lead to a buy-out price of around 135% of the
FRS 17 liabilities for a “Typical Scheme”, with a mix of deferred pensioners and
current pensioners.
The buy-out price quoted for a cohort of pensioners would be lower than this level.
Due to this pensioners are often perceived to be a “cheap” buy-out option, but this
lower price simply reflects the significantly lower level of risk associated with
pensioners, and as such they are not genuinely cheap.
The most significant risk faced by an insurer is in respect of longevity, since they take
minimal investment risk. If longevity improvements are greater than expected by the
insurer when writing annuity business, it could significantly reduce their investors’
return on equity, as well as create strains on their capital. The potential impact of this
2
This report seeks to set
out the current state of
the annuity business,
including the details
of the environment in
which it operates, the
risks faced by insurers
and a discussion of
current pricing levels.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
However, the allowance for longevity improvements only comprises around 25%
of the premium above the value of the FRS 17 liabilities payable to the insurer,
with most of the differential being due to the pricing yield used.
We speculate that insurers are currently offering discounts in their pricing to
entice trustees and scheme sponsors to write business with them, in order to boost
their profile and get up and running in a competitive market. There is no “magic
bullet” which insurers can use to make pricing more attractive and therefore these
discounts are being facilitated by investors being willing to accept lower returns on
their equity. Our model suggests that offering a 10% discount to the price (which
we estimate is around the level of discount currently being offered) might reduce
the investors’ pre-tax return on equity to as little as 6% per annum. This would be a
very low return on equity, particularly given the risks involved, particularly in relation
to longevity.
Our view is supported by decisions by the likes of Prudential to stop competing in the
bulk buy-out market. They currently view pricing levels to be uneconomic, and as
such will not engage in writing new annuity business until pricing levels revert to
more normal levels. A number of other market participants, such as Pension Insurance
Corporation, appear to share this view.
Since we do not consider the existing pricing levels to be sustainable, we think bulk
annuity prices are currently at a low. Therefore, if a pension scheme is looking to
buy-out now is the time to act to secure the lowest price possible.
The Annuity Players
The elder statesmen of the bulk buy-out market, Legal & General and Prudential, have
been joined by the likes of Paternoster, Pension Insurance Corporation, Lucida, Synesis
Life and Rothesay Life over the past couple of years. This increased level of
competition has led to the downward pressure on prices outlined above.
The increased number of players in the market has also led to product innovation.
There are now a number of variations available on the standard annuity buy-out,
including phased buy-outs, profit or risk sharing, and insurers being willing to take on
data risk. Some insurers also incorporate a bulk transfer prior to buy-out, to allow
them to alter the liabilities prior to buy-out and speed up the process.
A pension scheme can consider a number of factors when determining which insurer
to choose. These include the financial strength of the insurer, their administration and
implementation capabilities, the precise benefits which will be provided, the brand
name offered by the insurer, and the future direction of the insurer. However, in our
view, since all insurers are subject to the same FSA regulation, selection should be on
the grounds of price.
Although a number of the deals done recently have been conducted using an online
auction process, we do not consider this to be an optimum strategy. The online
auction process has a number of limitations, most notably that designing an
optimum auction is extremely complex, and the auction approach generally used is
overly simplistic. In addition, not all market players are willing to take part in auctions.
We believe that a better outcome will be achieved by undertaking a process similar to
that used in mergers and acquisitions.
3
Executive Summary
was illustrated when Legal & General announced that they were updating their
mortality assumption in their reserving basis in March 2008, leading to a hit to their
2007 profits of £269 million.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
The Buyer’s Perspective
The key stakeholder in the decision to purchase annuities is the sponsoring employer.
Given the increasingly onerous requirements associated with operating a defined
benefit pension scheme, it is understandable that some scheme sponsors are
considering the buy-out option.
However, on purely financial grounds, we do not consider buying-out the pension
liabilities with an insurer to make economic sense for most companies. Instead, the
sponsoring employer could retain a well-funded pension scheme on its balance sheet,
and by following a similar approach to that which would be adopted by an insurance
company, manage it so that surplus is gradually released to the company. By
transferring a pension scheme to an insurer, the Finance Director is effectively giving
up his claim to future profits which the well-funded pension scheme would be
expected to generate.
A number of the recent transactions involved unique circumstances. For every
sponsoring employer there is a price at which they would be pleased to offload their
pension liabilities to an insurance company. However, for most companies this price
is still below the current pricing levels in the buy-out market, even allowing for the
discounts being offered by insurers. We therefore think that the muted level of current
activity in the buy-out market does not bode well for the aggressive expansion plans
of some insurers.
Buy-Out Alternatives
A number of organisations have attempted to bridge the gap between the pricing
of the traditional buy-out option and the prices at which Finance Directors are willing
to trade.
Pension Corporation and Citigroup have done deals involving the acquisition of a
company in order to take on their pension liabilities, which allows them to continue
to run the pension scheme as a going concern, and therefore take advantage of the
greater flexibility of the pensions regulatory environment, rather than the insurance
environment. However, since this approach involves acquiring small companies with
very large, well-funded pension schemes, it has limited applicability.
Occupational Pensions Trusts have also designed a product which involves transferring
the pension scheme to a shell company, which OPT then buys for a nominal sum.
However, OPT does not provide any external capital, and the trustees are likely to
require the pension scheme to be very well-funded as compensation for the employer
covenant which they are being asked to give up. We do not consider this solution to
be an attractive option to trustees and sponsoring employers, and expect it to
generate extremely limited amounts of business.
Alternative De-Risking Solutions
A range of products have been launched in recent times which aim to reduce
significantly the risks associated with running a defined benefit pension scheme,
without severing the link between the sponsoring employer and the pension scheme.
PensionsFirst, Tactica and PensionsRisk have all designed products which aim to
protect against particular pension scheme risks for a defined period of time, such as
10 years. Although the products have technical differences in the way in which they
4
A number of
organisations have
attempted to bridge
the gap between the
pricing of the traditional
buy-out option and the
prices at which Finance
Directors are willing
to trade.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
operate, they typically involve taking on responsibility for meeting the scheme’s cash
flows over the period, and at the end of the period assets will be returned to the
scheme equal to the size of the liabilities. However, we think most scheme sponsors
and trustees will consider the cost of these products to outweigh their benefits.
BrightonRock also intend to launch a product offering protection against employer
insolvency later in 2008. On face value it appears that their innovative product could
appeal to a broad segment of the pensions industry, particularly those seeking to take
equity risk in order to bridge their funding gap, since with the policy in place trustees
should be more relaxed regarding both investment risk and the level of prudence
required in their funding valuation. However, further analysis of the BrightonRock
business model is likely to be required when their product is launched, as trustees
will need to have confidence that BrightonRock’s probability of insolvency is low.
A number of players in the market are also now developing a range of mortality
products. The development of this market is still in its infancy, but it could provide
trustees with the final piece of the risk management jigsaw. The key obstacle to the
growth of this market is the lack of obvious counter-parties for the longevity risk
faced by pension schemes.
Conclusion:
So is this the end game for defined benefit pension schemes?
Is UK plc likely to rush to buy-out their pension scheme liabilities
over the next few years?
Our answer is categorical – NO!
5
However, further
analysis of the
BrightonRock business
model is likely to be
required when their
product is launched,
as trustees will need
to have confidence
that BrightonRock’s
probability of
insolvency is low.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Table of Contents
6
Key Findings
1
Executive Summary
2
1. Introduction
7
2. Background
8
3. The Annuity Business
11
4. The Annuity Players
22
5. The Purchase Process
27
6. The Buyer’s Perspective
32
7. Buy-Out Alternatives
41
8. Alternative De-Risking Solutions
46
9. Mortality Hedging
55
10. Conclusion
59
Appendix – Improvements in Longevity
60
About the Authors
62
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Once upon a time, the UK pension buy-out market was a fairly dull and boring place,
with relatively few press-worthy events. However, over recent years this has started
to change, with news of developments in the market appearing on an almost daily
basis. This report seeks to explain some of these developments, some of which may
appear extremely complex at first glance, and discuss why they have occurred. We
also describe in detail the key current players in the market.
The big unknown is whether the market will take off as some participants, such as
Mark Wood, Chief Executive of Paternoster, expect or whether the number of
transactions will remain relatively small in the context of the whole UK defined
benefit market.
“Growth looks likely to accelerate: the pipeline of business on which it is
quoting stands at more than £30 billion and, while not all of that will turn
into buyouts, let alone be won by Paternoster, Wood is predicting that total
pension buyouts this year could be as high as £12 billion, almost five times
last year's £2.5 billion.
Even that, says Wood, will be 'small potatoes' in the context of the
£1.2 trillion total value of defined benefit schemes. While he thinks that
most FTSE 100 companies – which account for about a third of the total –
will stick with their schemes, within five years or so it will be 'anachronistic
for a medium-sized company to still have a defined benefit scheme in its
balance sheet'.”
The Observer, 2nd March 2008
Some of this business has started to filter through in 2008. We estimate that in the
first quarter of this year annuity business worth slightly over £2 billion was written,
and by the end of April business written in the year to date was around £2.5 billion.
Players in the insurance market are therefore confident that the 2007 figure of nearly
£3 billion will be surpassed by the middle of 2008.
Due to the rapidly evolving nature of the market, we anticipate some of the content
of this report will become obsolete in the near future, as new products are launched
and more deals are done. We therefore anticipate producing further updates in the
future; however, please contact us if you have any queries on developments in
the market.
7
Introduction
1. Introduction
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
2. Background
Historically the only way for the obligations of a UK defined benefit pension scheme
to be settled, other than running it onwards until the last pensioner has died (which
may not be for eighty or more years into the future), was for the trustees to purchase
an annuity contract for every member in the scheme with an EU regulated insurance
company.
Where a scheme purchases such contracts for all its members this is called a “bulk
buy-out” of the liabilities. Traditionally this option was usually taken when the
sponsor of the scheme had become insolvent and the scheme was winding-up,
however, in recent times this solution has been increasingly applied where the
sponsor is still solvent but no longer wishes to run the risks implicit in having an
ongoing defined benefit pension scheme.
2.1 The Current State of the Market
Until only a few years ago, insurers predominantly wrote bulk buy-out business for
pension schemes going into wind-up due to the insolvency of their sponsoring
employer. Schemes were looking to use their remaining assets to secure the best
possible arrangements for their members through an insurance company, and in most
cases members received poorer benefits than they had been promised through their
pension scheme.
However, the introduction of the Pension Protection Fund (“PPF”) in April 2005
deprived UK insurers of this type of business, since these schemes will now
automatically fall into the PPF. Over the past couple of years the insurers have
instead been required to shift their focus towards solvent companies looking to
offload their pension liabilities. Since 11th June 2003 the debt on solvent employers
looking to wind-up their pension schemes has been the buy-out debt, whereas prior
to this date the amount payable was typically a lower amount. Therefore, employers
seeking to relieve themselves of their defined benefit pension scheme obligation are
now required to do so in the buy-out market.
The table below sets out our understanding of the levels of new annuities written
by insurers in the UK market over 2007, based on press coverage and discussions
we have had with some of the insurers. It should be noted that some of these figures
are estimated, since not all of the insurers disclose full details of the deals they are
involved in.
Table 2.1 – Value of new annuities written by insurer over 2007
Insurer
8
Value of New Annuities Written
over 2007 (£ billions)
Paternoster
Legal & General
Norwich Union
Aegon
Prudential
Pension Insurance Corporation
Synesis Life
Lucida
Rothesay Life
1.4
1.2
0.1
0.1
0.1
-
Total
2.9
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
These increased levels of activity have continued in 2008. We estimate that the first
quarter saw a volume of business of slightly over £2 billion (of which around £700
million was represented by a single transaction by Rothesay Life, discussed further in
Section 4), and by the end of April the value of bulk annuity business written in the
year to date was around £2.5 billion. This has led insurers to predict that if the current
levels of business are sustained, the 2007 figure of nearly £3 billion will be surpassed
by the middle of 2008.
Whilst nearly £3 billion is clearly a very significant amount of money, it is a drop in
the ocean compared to the potential market available. The combined liabilities of
private sector UK pension schemes are currently estimated to be around £800 billion
(on an FRS 17 basis), and an approximate split of these liabilities by size of pension
scheme is set out in the table below.
Table 2.2 – Split of the UK’s private sector pension scheme liabilities by scheme size
Size of scheme
Less than £20 million
£20 million - £500 million
More than £500 million
Liabilities (£ billions)
100
200
500
This gives some indication of the huge potential market which could be targeted by
the pension buy-out players, and why there have been so many organisations trying
to get involved in the market over the last few years. Since the FRS 17 liabilities are
around £800 billion, this would suggest that the buy-out liabilities exceed £1 trillion,
as stated by the likes of Mark Wood. However, the success of the buy-out market will
depend on the players involved being able to persuade pension schemes and their
sponsors that the risks faced by a pension scheme are significant enough to consider
paying a premium to remove.
A number of commentators have claimed that the acceleration in bulk buy-outs,
coupled with the existence of the PPF, will remove the majority of defined benefit
liabilities from the UK market within a matter of a few years. Indeed, Mark Wood
believes that within five years or so it will be “anachronistic for a medium-sized
company to still have a defined benefit scheme in its balance sheet”. Given the
current state of the UK bulk buy-out market, we do not consider this to be likely.
The number of pension schemes falling into the PPF will depend on rates of employer
insolvency. However, let us assume that the PPF will take on around £2 billion of
liabilities each year, which is broadly in line with their long term risk model. Assuming
bulk buy-outs account for £10 billion each year, which would be considerably higher
than was the case in 2007, this would suggest defined benefit liabilities of around
£12 billion are being removed each year.
However, we also need to allow for pensions being paid to members, interest on
liabilities, and the accrual of benefits. The table overleaf includes some very rough
estimates of what these values might be for the universe of UK defined benefit
9
Background
As shown, a number of the new entrants to the market did not write any business
during 2007. It is also worth noting that a significant proportion of the business
written related to a small number of large deals, completed in the last quarter of
2007. For example, Paternoster’s £800 million transfer of pensioner liabilities from
P&O accounted for over half of their new business, whilst Legal & General’s £240
million transfer from the Weir Group also boosted their figures substantially.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
pension schemes, and illustrates how the liabilities might develop from the £800
billion estimate shown above over the course of a year.
Table 2.3 – Estimate of change in liabilities for universe of UK defined benefit pension
schemes over a year
Liabilities (£ billions)
Liabilities at start of year
Liabilities discharged through PPF
Liabilities discharged through bulk buy-out market
Benefits paid1
Interest on liabilities2
Benefits accrued3
800
(2)
(10)
(20)
40
20
Liabilities at end of year
828
This corresponds to growth in the size of “UK plc’s” defined benefit pension liabilities
over the year of around 3-4%.
Although the numbers used here are clearly very rough, this indicates that the current
levels of the bulk buy-out market are not leading to the rapid contraction of the UK
defined benefit pensions market. On the contrary, we would estimate that the market
is currently growing modestly in size. To cease this expansion in the market, the UK
bulk buy-out market would need to write around £35-£40 billion of bulk annuities
per annum. Even the most optimistic estimates from insurers are that the volume
of business written in 2008 might reach £15 billion, and this is at a time when profit
margins of insurers have been squeezed to try to make buy-out prices more
attractive. Therefore, it seems unlikely that buy-out levels will grow to the extent
required to lead to the UK defined benefit pension market contracting in size.
The levels of business written should also be viewed in the context of the newly
expanded capacity of the market for such transactions. We estimate that the crop
of recent entrants alone have capacity to write in excess of £50 billion of annuity
business, without even considering the long standing insurance companies who are
active or have expressed an interest in the market. Unless the demand for these
products does start to grow substantially, there are going to be some very
disappointed investors.
10
PSTS Comment:
The traditional source
of business for insurers,
insolvent wind-ups, is
now drying up due to
the presence of the
Pension Protection Fund
and insurers are relying
on solvent employers
deciding to offload
their pension liabilities.
Despite the very large
amount of capacity
within the market, the
current volumes of
business being written
are having little or no
impact on the overall
size of the UK private
sector pension
liabilities. There remains
a massive potential
market for annuity
companies. The key
question over the
coming years will be
whether this potential
will convert into actual
business.
1
The Purple Book 2007 (prepared
by the Pensions Regulator and
the PPF) shows that in 2007
only 33% of pension scheme
members in the UK are currently
receiving a pension from their
scheme, so benefit outgo is
probably relatively modest.
2
Estimated based on an interest
rate of 5% per annum.
3
The Purple Book 2007 shows
that 25% of members of UK
pension schemes are currently
accruing benefits.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
The traditional buy-out market offers pension schemes the opportunity to pass the
responsibility for meeting pension scheme liabilities to an insurance company, in
exchange for a premium. These insurance companies are regulated by the Financial
Services Authority (“FSA”), rather than occupational pension scheme legislation, and
must therefore meet tighter reserving requirements. This leads to the cost of
transferring occupational pension scheme liabilities to an insurer exceeding the
expected cost of providing these benefits through the pension scheme itself.
3.1 A Brief Recent History
Until a few years ago the traditional UK buy-out market comprised only two major
and long established players; Legal & General and Prudential. However, there have
been a significant number of new entrants to the market in recent times, as other
insurers see this as a part of the insurance market which could offer significant
growth opportunities. Since the increased supply of insurers in the marketplace does
not yet appear to have been matched by pension scheme demand, buy-out prices
have fallen slightly. It is fairly difficult to quantify this, since no two pension schemes
are the same and there is not an abundance of quotes in the public domain, but it
would appear that during the course of 2007 prices may have fallen by around 10%
due to the increased level of competition.
Historically there has been little difference between the pricing approaches of Legal
& General and Prudential. The market was fairly evenly divided between them and
there was relatively little downward pressure on prices. However, as discussed in
Section 2, the circumstances in which buy-out solutions might be appropriate have
changed significantly over recent years, primarily due to the introduction of the
Pension Protection Fund (“PPF”). There have also been a large number of players
attempting to enter the market, since they see potential for significant growth in
the buy-out area.
3.2 The FSA Regime
Insurance companies can, and do, become insolvent and therefore there is significant
regulation on their activities, mainly deriving from the various European Union
Insurance Directives. The regulation of the insurance industry is carried out in the
UK by the Financial Services Authority.
Legislation exists which provides protection to policyholders should their insurance
company default. This was first introduced by the Policyholder Protection Act 1975,
where the Policyholder Protection Board (the “PPB”) acted as an industry-funded
safety net when UK insurance companies became insolvent. The costs of providing
protection and the PPB’s administration costs were provided for out of levies raised
against insurance companies operating in the UK. From midnight on 30th November
2001 the Financial Services Compensation Scheme (the “FSCS”) took over
responsibility for providing policyholder protection in the UK, under the terms of the
Financial Services & Markets Act 2000. The FSCS covers annuities written, and it aims
to ensure that policyholders receive 90% of the future benefits from the annuity in
the case of insurance company insolvency.
Since the compensation arrangements were established, two small life insurers have
defaulted, leading to claims on the FSCS. However, as these insurers were very small,
the protection which would be offered by the FSCS in the event of the default of a
large life insurer is relatively untested. As the FSCS is unfunded, it would be reliant on
11
The Annuity Business
3. The Annuity Business
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
being able to claim levies from solvent insurance companies in the case of a large life
insurer defaulting.
In order to try to reduce the chance of an insurer becoming insolvent, new rules
regarding the capital requirements of insurers were introduced in 2005. In particular,
the Individual Capital Assessment process was brought in, where an insurer must
conduct its own comprehensive assessment of the risks faced in its business, and
determine how much capital (and in what form) it needs to hold to mitigate these
risks. This is typically defined as enough capital to ensure the insurer is at least 99.5%
likely to remain solvent over the next year, or in other words their capital is sufficient
to ensure they can survive all adverse circumstances, other than a 1-in-200 year
event, or worse.
Typically, this capital requirement necessitates insurers to hold capital margins of
around 8% of their “technical provisions” in respect of their annuity book. This means
that if an insurer takes on annuities of £100 million, they must typically hold total
assets to back these annuities of around £108 million.
The FSA’s capital requirements also depend on the level of “matching” between the
insurer’s assets and their liabilities. The matched position for annuities is generally
considered to be a diversified portfolio of very secure bonds, such as government
bonds and AAA or AA corporate bonds, with a similar duration (typically long-term)
to the liabilities. Due to the scarcity of very long-dated bonds some insurers also
use swaps to improve their level of matching. The bonds should typically match the
nature of the liabilities, for example index-linked bonds should be held to match
inflation-linked pensions, whilst fixed interest bonds could be held to match pensions
with fixed increases in payment. If an insurer opts to move away from this matched
position, they would be required to increase significantly their level of capital held.
As such, insurers tend to take minimal investment risk. This leaves longevity as the
most significant risk faced by an insurer.
Section 9 provides further discussion of the current longevity risks faced by insurers
and pension schemes. As this section describes, there is considerable uncertainty
regarding future mortality levels. The FSA returns for 2006 set out the mortality
allowance which various UK insurers were making in their reserving basis. The
allowances for some of the larger insurers are set out in the table below.
Table 3.1 – Level of improvements allowed for in mortality assumption for males
in 2006 reserving basis, split by insurer
Level of improvements allowed for in mortality
assumption for males in 2006 reserving basis
Aegon
Medium cohort projection, with 1.00% floor
Prudential
Medium cohort projection, with 1.25% floor
Standard Life
Medium cohort projection, with 1.50% floor
Norwich Union
Medium cohort projection, with 2.00% floor
Friends Provident Average of medium/long cohort projection, with 0.50% floor
Legal & General Average of medium/long cohort projection, with 0.80% floor
It should be noted that there is not a significant difference between these allowances
and the allowances which most pension schemes are now being encouraged by the
Pensions Regulator to make in their ongoing funding valuations. Under the scheme
specific funding regime, which was introduced for actuarial valuations carried out
12
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
from September 2005 onwards, trustees have been encouraged to be increasingly
prudent when setting their mortality assumptions.
In February 2008 the Pensions Regulator published a consultation document which
included an analysis of the assumptions used by the 1,138 schemes which had
submitted their recovery plans covering valuation dates between September 2005
and April 2006. This revealed that the majority of pension schemes (55%) made
allowance for improvements in line with the medium cohort effect, whilst 11%
allowed for the short cohort effect and 33% made no allowance for the cohort effect.
As an indication of the impact of these differences in mortality projections, using
Legal & General’s allowance in Table 3.1 above might increase the liability in respect
of a 65-year old male pensioner by around 3% compared to simply using the medium
cohort effect.
Although the gap is not vast based on the data above, the FSA has recently been
undertaking a similar campaign on mortality issues to the Pensions Regulator and has
been in discussions with life insurers to remind them of the requirement to hold
prudent reserves and to ensure that their mortality tables remain completely up to
date. This is likely to lead to insurers strengthening their mortality tables from the
above rates in the coming years. (See Section 9 for further commentary on the
impact of strengthening mortality assumptions.)
In the current FSA-regulated environment, the capital requirements of insurers are
not fixed in a formulaic way. Instead they are based on models built by the insurer,
approved by the FSA, to test the insurer’s requirements in various ways. As described
above, these models are intended to protect against 1-in-200 year events over a oneyear time horizon. However, the FSA recognises that this time horizon may not be
appropriate for testing life company capital requirements. Translating a one-year
probability rule into a ‘run-off’ ruin probability is not straightforward and, for each
entity, will depend on many things. The FSA has stated that it looks for run-off ruin
probabilities of 5% or less and will expect capital to be increased until this type of
target is reached. However, this is not an exact science and it is to be expected that
there will be a wide variety of answers for any given insurance entity.
In summary, annuities secured through an insurance company are exposed to a very
small risk of default arising from the provider’s insolvency, and if insolvency does
occur members are guaranteed to receive 90% of their benefits. This could be
considered to be fairly similar to the position which members of uninsured defined
benefit pension schemes face, albeit that the risk of insolvency is somewhat larger for
schemes whose sponsoring employer does not have a AAA or AA credit rating. Again
compensation benefits of broadly 90% would be payable through a government
established compensation arrangement; in this case the PPF.
It is also worth noting that the level of cover provided by the PPF could be
significantly different from that provided under the FSA’s compensation agreement,
depending on the member. For example, although the impact of the compensation
cap would significantly reduce the benefits payable to a deferred pensioner with a
large accrued pension under the PPF, there would be no reduction in benefits for a
member who has already reached their Normal Pension Age and is entitled to a
pension which does not increase in payment.
3.3 Pricing Implications
The FSA regime has significant implications for insurance company pricing of annuity
contracts. It is worth considering the pricing of these annuity contracts, since this has
direct implications for the return on capital that the insurers can achieve. In order to
13
PSTS Comment:
It is extremely
important for trustees
to understand that
whilst securing annuities
may extinguish their
own obligations under
the Trust it does not
guarantee that
members will receive
their benefits in full.
Whilst it is likely that the
chances of them
receiving their benefits
in full will be much
improved, there is a risk,
of something less than
5%, that the insurer will
fail and the members
will receive a reduction
in benefits through the
Financial Services
Compensation Scheme.
In nearly all cases
purchasing annuity
contracts will enhance
benefit security,
although some
members may have
better “safety net”
coverage under the
Pension Protection Fund
than the Financial
Services Compensation
Scheme.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
do so we will construct a very simplified model of a life insurance company writing
annuity business to consider the general concept. Note this is not meant to be
accurate except in the sense that it is designed to be indicative of the key drivers of
annuity pricing.
Firstly, in order to construct our model, we need to make a few simple assumptions:
●
●
●
●
●
●
●
The insurer wishes to make a pre-tax return on equity of 15% per annum, which
seems neither excessively high nor low given the very large risks (particularly on
longevity) that insurers run.
The insurer will invest all their assets in a mixed portfolio of government and
high quality corporate bonds (which for all intents and purposes life insurers do
actually do).
The FSA requires the insurer to hold an 8% capital margin against its “technical
provisions”.
The insurer will price their annuities equal to the “technical provisions”, so they
are only providing the capital margin and not subsidising or being subsidised
by customers.
In essence, as they will typically invest shareholder equity equal to 8% of the
“technical provisions” to back 100% of their “technical provisions”, shareholder
equity capital is effectively 12.5 times leveraged.
IAS 19 and FRS 17 require that the discount rate used to value the pension
liabilities on the sponsoring employer’s balance sheet is set with reference to the
yield available on AA rated corporate bonds. As at 1st March 2008 this would lead
to a discount rate of around 6.4% per annum, which is almost 200 basis points
above the yield available on equivalent government bonds, of 4.5% at this date.
This spread has recently increased due to the “credit crunch”, and over the past few
years a more typical spread has been between 75 and 100 basis points. For the
purpose of our example let us assume that spreads revert to this level in the future,
so that the AA corporate bond yield is 5.5% per annum, or 100 basis points above
the 4.5% per annum yield available on long-term government bonds.
As discussed the insurer will hold a diversified portfolio of government and high
quality corporate bonds to match their annuities. After allowing for management
costs, a few defaults on corporate bonds and some active management, such a
portfolio might return 50 basis points above the government bond yield over the
long term (or 5% per annum in this example).
Taking these assumptions together we can calculate that, in order to obtain their 15%
return on equity, the insurer must value the liabilities using a government bond yield
less 30 basis points when pricing (or 4.2% per annum in this example). This is since
they are 12.5 times leveraged, such that:
Return on Equity =
=
=
Return on Assets + 12.5 x (Return on Assets – Pricing Yield)
5% + 12.5 x ( 5% – 4.2% )
15% per annum
We can now calculate the premium above FRS 17 or IAS 19 liabilities that the insurer
must charge in order to achieve their required return on equity for different types
of liabilities.
The duration for a fairly typical pension scheme, with a mix of deferred pensioners
and current pensioners, might be somewhere between 15 and 20 years. For the
purpose of our example we will assume the duration is 17 years. Since we are
14
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
assuming that the FRS 17 or IAS 19 liabilities are discounted using a corporate bond
yield of 100 basis points above the gilt yield (i.e. 5.5% per annum), and this insurance
company needs to price at 30 basis points below the gilt yield (i.e. 4.2% per annum),
this leads to a difference in pricing between the company’s accounting liabilities and
the insurer’s pricing of around 23%, since:
17
( 11 ++ 5.5%
4.2% )
– 1 = 23%
In addition to this, the insurer will also need to make an allowance for the future
administration expenses associated with paying the benefits to members. Insurers
might be able to reduce these costs slightly compared to a stand-alone pension
scheme due to economies of scale, so a typical allowance might be somewhere
around 3% of the FRS 17 or IAS 19 liabilities. There is currently no requirement for
a capital value of future administration expenses to be included on a company’s
balance sheet under FRS 17 or IAS 19 and thus these must be added on to the
FRS 17 or IAS 19 liability figure.
As described in Section 3.2 above, insurance companies will also need to use a more
prudent allowance for future mortality rates than the trustees of the pension scheme.
We stated above that this might only produce liabilities around 3% higher than the
scheme might use in their ongoing funding valuations. However, the assumptions
used in the company’s accounts under FRS 17 or IAS 19 are set by the directors of
the company. Since the valuation assumptions must include a margin for prudence,
directors often opt to use slightly weaker mortality assumptions, which might be
considered to be closer to a ‘best estimate’ assessment. Let us assume that the
insurer’s mortality assumption adds around 9% to the FRS 17 or IAS 19 liabilities for
a typical scheme.
Therefore, the price offered by the insurer might be somewhere in the region of 135%
of the FRS 17 or IAS 19 liabilities. This is typical of the sort of levels that insurance
companies have been pricing at over the past few years (i.e. since the advent of
significant competition in the annuity market).
It is interesting to note that in this example only around 25% of the price differential
arises as a result of the mortality assumption adopted. The vast majority of the
difference between the value of the liabilities on an accounting basis and the value
placed on the liabilities by an insurer is due to the insurer’s enforced investment
strategy consisting predominantly of high quality bonds, and the margin required in
order to generate a 15% per annum pre-tax return on shareholder equity.
Insurers may wish to claim that they need to aim to generate this high return on
equity in order to compensate their shareholders for the longevity risk that they must
bear in order to enter this business. This is true but it is still the case that the change
in expected future mortality rates between a pension scheme and an insurer only
makes up a modest portion of the difference between the FRS 17 or IAS 19 liability
figure and the cost of buy-out.
In the example above we considered a fairly typical pension scheme, consisting of a
mix of deferred pensioners and current pensioners (the “Typical Scheme”). Let us
now consider two further examples:
1) A mature pension scheme, consisting only of current pensioners,
with a duration of 10 years (“Pensioner Scheme”); and
2) A relatively immature scheme, consisting only of deferred pensioners,
with a duration of 25 years (“Deferred Scheme”).
15
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
The mortality difference for Pensioner Scheme will be considerably less than in the
example above. This is since the uncertainty relating to mortality mainly relates to
the rates of future improvements, and the shorter the duration of the liabilities is,
the less time there is for mortality rates to diverge from expectations. Conversely,
Deferred Scheme will need a larger mortality adjustment.
The possible buy-out prices for these two schemes, as well as the Typical Scheme
described above, are set out in the table below.
Table 3.2 – Estimated buy-out prices as proportion of FRS 17 or IAS 19 liabilities
for pension schemes of varying duration
Addition to FRS 17 or IAS 19 liabilities
Pensioner Scheme Typical Scheme
Deferred Scheme
FRS 17 Liabilities
Yield
Mortality
Expenses
100%
13%
5%
3%
100%
23%
9%
3%
100%
36%1
15%
3%
Buy-out Liabilities
(as proportion of
FRS 17 Liabilities)
121%
135%
154%
Therefore, the buy-out price for our Pensioner Scheme might be around 121% of the
liabilities on the FRS 17 or IAS 19 basis, and for our Deferred Scheme the price might
be approaching 160%. Again, these prices are fairly consistent with figures which have
been quoted in recent buy-out transactions.
It is worth noting that current pensioners as a membership class are the cheapest to
insure through annuity contracts, relative to the FRS 17 or IAS 19 liability valuation,
and this has led many to believe that they are the best value. However, we need to be
clear that the pensioners represent the lowest risk part of the liability where the
current mortality trends can be most easily discerned and the tail risk is limited. Thus
we should expect that the insurance premium payable to cover such risks is lower
than for the deferred pensioners. Deferred pensioners are considerably more risky as
the longevity risk for such members extends over much longer and more uncertain
timeframes.
PSTS Comment:
Only a minority of the
premium required above
FRS 17 or IAS 19 is
explicitly driven by
differences in views
between insurers and
pension schemes of the
underlying mortality
trends. The majority
of the pricing
differential is driven by
the need for the life
insurer to make a return
on its shareholders’
equity, in return for
taking on the liability
and putting up capital
as required by the FSA.
Pensioners are often
perceived to be a
“cheap” buy-out
option, since the
premium payable to
insurers is lower than
for deferred pensioners.
However, this simply
reflects the significantly
lower level of risk
presented by
pensioners, and as
such they are not
genuinely cheap.
It is worth noting that for all three theoretical pension schemes the impact of the
mortality differential only represents around 25% of the overall price difference.
Clearly these examples are somewhat simplistic, and the actual pricing techniques
adopted by life insurers will be slightly more sophisticated than this. However, these
examples serve to illustrate the key pricing issues for an annuity buy-out, and in
particular the key pricing differential is due to:
a) The capital penalty imposed by the FSA if an insurer moves away from a
‘matched’ investment strategy, which severely restricts the strategy which can
be followed by insurers; and
b) The insurer requires a relatively high return on equity (typically around 15%
per annum before tax) as compensation for the significant risks that it faces
when writing annuity business.
16
1
Note: it is often the case for
deferred pensioners that a larger
yield adjustment is observed due
to the downward sloping nature
of the government yield curve
which begins to have a very
significant impact at longer
liability durations.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
3.4 Discounts Available?
PSTS Comment:
Some of the buy-out deals announced towards the end of 2007 and at the start of
2008 appear, on face value, to have been written on significantly lower premiums to
FRS 17 or IAS 19 than suggested by the analysis above. It could be suggested that
since some of the newer insurers are very keen to write business, they are offering a
lower premium in an attempt to boost their profile and get up and running in a
competitive market.
The current competitive
pricing environment
suggests, through our
simple model of annuity
business, that some
insurers are writing
business at a discount.
In effect these insurers
are willing to accept a
low return on
shareholders’ equity in
order to establish
themselves in this highly
competitive market.
Since the FSA still requires the insurer to have assets to back the “technical
provisions” plus the capital margin of around 8%, it is likely that insurers will need to
take a hit on their return on equity to write business at these levels.
In order to analyse the impact of offering such a discount on the return on equity, it is
necessary to use a slightly more sophisticated model than the simplistic one outlined
above. The above model is inadequate when pricing is not assumed to be equal to the
“technical provisions”, as allowance also needs to be made for the different pace at
which capital is returned to the investors under this scenario. If a discount is offered a
greater proportion of the capital must be used to pay the benefits, eroding the return
on equity. We note that even this more complex model is not designed to be totally
reflective of a life insurance business but more to illustrate the general impacts.
We have run a model showing the impact on the return on equity of offering a range
of pricing discounts. The model gives the same output as our simplistic method above
in the case where no discount to the “technical provisions” is offered. The results for
our Typical Scheme outlined above are set out in the table below.
Table 3.3 – Estimated pre-tax returns on equity available for various discounts to
“technical provisions” offered in price
Discount to “Technical Provisions”
Offered in Price
Pre-tax Return on Equity
(per annum)
0%
5%
10%
15%
20%
15%
9%
6%
4%
3%
Therefore, if we were to conclude that some of the current deals were being written
at around a 10% discount to the value of the “technical provisions”, this would
suggest that investors are accepting a pre-tax return on equity of around 6% per
annum. This would reduce the premium above FRS 17 or IAS 19 to around 25% for
our Typical Scheme or to around 11% for our Pensioner Scheme above.
Although this might still be considered to be an acceptable level of return for some
investors, allowance should be made for the risks involved, particularly relating to
longevity. It is difficult to quantify a risk-adjusted rate of return, due to the
uncertainties associated with mortality projections, but this illustration does serve to
illustrate that it is only possible to write bulk annuity business at the current levels by
significantly reducing investors’ return on equity. The ability of insurers to continue
writing business at the current levels therefore very much depends on whether
investors are happy to continue to accept these levels of returns.
17
This view is confirmed
by a number of market
participants
commenting on the
pricing available in the
market. However, in
the long term, such
pricing is unsustainable
and we therefore expect
prices in the market to
increase in the near
future.
Therefore, if a scheme
is looking to buy-out,
then swift action is
required in order to
take advantage of the
current pricing before
pricing reverts to more
normal levels.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
It could be the case that a number of players in the market are attempting to price
competitors out of the market. They may not intend to adopt their current pricing
policy in the long-term, but in the short-term are attempting to take market share,
with the view that in the long-term they will be able to increase prices back towards
the levels illustrated in our examples above.
Section 2.1 illustrated that the vast majority of the bulk buy-out business is currently
being written by Paternoster and Legal & General. Meanwhile, some large insurers
(such as Prudential) have indicated that they currently consider pricing levels in the
bulk buy-out market to be uneconomic, and to not provide sufficient return on their
capital for the level of risk involved, and as such are writing limited amounts of new
business. Since the FSA regime provides a level playing field for all insurers, this
suggests that the insurers writing business must be accepting a lower return on
equity to do so.
Therefore, we conclude that if a pension scheme is considering purchasing annuities
to extinguish their liabilities, now is an ideal time to do so. We expect that buy-out
prices have bottomed out, and in the long-term the differential between FRS 17 or
IAS 19 prices and buy-out is likely to revert to the long-term levels outlined above.
3.5 Longevity Risk
It is also worth briefly mentioning the significant impact which mortality could have
on insurance companies, particularly in the current environment of competitive buyout pricing.
As is discussed in Section 9 there is considerable uncertainty surrounding
expectations of future mortality rates. As such, if rates of improvement in longevity
prove to be greater than are currently expected by insurance companies in their
pricing, this could significantly reduce the return on equity provided by bulk buy-outs.
As described in Section 3.2 above, the FSA requires insurers to hold sufficient capital
to meet their “technical provisions” plus a margin of typically around 8%. Insurers
face a major risk that at some point in the future they are required to update the
mortality tables they use to calculate their “technical provisions” to a more prudent
table, and are therefore required to inject more capital into their annuity fund. Thus
prudent reserving for mortality is important for insurers in order for them to avoid
having calls on capital and lower overall returns on their annuity books.
Table 3.4 below shows the potential impact that unanticipated improvements in
longevity could have on the investment return achieved by an insurer. If longevity
improvements are greater than expected, this will increase the value of the liabilities.
Similarly, reducing the discount rate used to determine the present value of the
liabilities would increase the liabilities. It is therefore possible to estimate the impact
of improvements in longevity by applying a reduced discount rate to value the
liabilities, which in turn reduces the return achieved by the insurer.
The assumptions used here for the return on assets and the pricing yield are the same
as used in Section 3.3 above, whilst the reduction in the investment return of 0.5%
per annum caused by potential unanticipated longevity improvements is taken from
data from the Government Actuary’s Department.
18
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Table 3.4 – Potential impact of unanticipated improvements in longevity
to investment margin
Insurance company
Rate of return achieved on assets (per annum)
Pricing yield (discount rate applied to calculate liabilities)
5.0%1
(4.2%)2
Investment margin
Potential unanticipated impact of longevity improvements
0.8%
(0.5%)3
Revised investment margin
Effective reduction
0.3%
63%
If unanticipated longevity improvements increase the effective discount rate on the
liabilities from 4.2% to 4.7% per annum this would also reduce the pre-tax return on
shareholder equity.
To show the impact on the return on equity figure, let us return to our simplistic
example for the Typical Scheme outlined in the pricing discussion in Section 3.3
above. Assuming that the insurer is 12.5 times leveraged, we previously had a pre-tax
return on equity of 15%, calculated as:
Return on Equity =
=
=
Return on Assets + 12.5 x (Return on Assets – Pricing Yield)
5% + 12.5 x ( 5% – 4.2% )
15% per annum
Adjusting the underlying discount rate on the “technical provisions” to equal 4.7%
per annum would alter the calculation as follows:
Return on Equity =
=
=
PSTS Comment:
Longevity risk is the key
risk for a life insurer and
thus life companies
need to be relatively
prudent in setting future
mortality rates in order
to avoid mis-pricing
leading to demands for
further capital and
reductions in returns on
shareholders’ equity.
Even taking this prudent
view into account it is
clear that mortality
differences are a minor
contributor to pricing
differentials between
annuity costs and FRS
17 or IAS 19 liability
figures. This is
confirmed by the
mortality tables used
by insurance companies
for their reserving
disclosed in annual
returns to the FSA.
Return on Assets + 12.5 x (Return on Assets – Pricing Yield)
5% + 12.5 x ( 5% – 4.7% )
8.8% per annum
Therefore, if unanticipated longevity improvements increase the effective discount
rate underlying the calculation of the “technical provisions” by 0.5% per annum, it
would reduce the investors’ pre-tax return on equity from around 15% to 8.8% per
annum. More importantly perhaps this reduction eliminates most of the excess return
above the return on the bonds in which the shareholders’ equity is invested.
Although this illustration is clearly simplistic, it shows the significant impact which
improvements in longevity rates could have on the return on equity of insurance
companies.
The other key issue for insurers around mortality is that, not only will they make a
lower return on shareholders’ equity if they get mortality wrong in their pricing, there
can be significant strains on their capital at future points. If, for example, at some
point in the future they were required to update the calculation of their “technical
provisions” to include some enhanced allowance for mortality then they would need
to find capital to plug the gap immediately in order to maintain their capital buffer.
The results announced by Legal & General in March 2008 illustrate the potential
impact of mortality. They enhanced their reserving basis to allow for a 65-year old
male to live for 25.1 years rather than the previous assumption of 23.8 years. This led
to a hit to 2007 profits of £269 million.
19
1
The rate of return on an
investment strategy of investing
in government and investmentgrade bonds.
2
Estimated using a long-term gilt
yield of 4.5% less 30 bps.
3
Estimated impact of
unanticipated mortality
improvement 1963 – 2003
from Government Actuary’s
Department data.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Therefore, for an insurance company, the major risk that they run in writing annuity
business is in longevity risk. This is because they take little or no risk with investments
and thus have no source of excess returns to fund the gradual improvements seen in
mortality rates over time.
3.6 Variations on a Theme
Given the significant amount of new entrants to the market, insurers have
increasingly started to offer variations on the standard annuity buy-out. Due to the
competitive pricing environment they aim to differentiate their offering and to make
their solutions more attractive to trustees and sponsors.
Examples of the types of products offered by some or all of the annuity players
include:
●
Phased buy-outs – the pension scheme’s liabilities are transferred in stages over
time, according to agreed milestones with future pricing determined at the outset.
In the interim, active management of the scheme’s assets through a Liability
Driven Investment (“LDI”) approach is used to try to generate strong returns to
fund the buy-out process.
●
Staged payments – the liability is transferred in full when the deal is struck, but
the insurer allows the sponsoring employer to spread the premium payments over
a period of time. There is a cost to such an approach as the insurer will wish to
ensure that they are protected from credit default.
●
Profit sharing – actuarial profits which arise after the liabilities have been
transferred to the insurer can be shared with the sponsoring employer. Such profits
might arise from options taken by the members, for example choosing to commute
pension for tax-free cash at retirement, which may reduce the value of the
member’s liability depending on the level of the cash commutation factors used.
Alternatively the member might choose to take a transfer value to an alternative
pension arrangement, the value of which might be less than the insurer’s expected
cost of providing the pension benefits. It is also possible to share profits arising
from more common experience items like mortality or investment returns.
●
Risk sharing – similar to profit sharing, only losses can be shared as well as profits.
Whilst these arrangements will reduce the cost of securing the liabilities with an
insurer, it also leaves some risk relating to the pension scheme with the sponsoring
employer, which would not be the case otherwise. For example, it might be possible
to leave the extreme longevity tail risk with the sponsoring employer which would
significantly reduce the buy-out price due to a reduction in capital required by
the FSA.
●
Tranching – different classes of members can be bought-out at different times. For
example, pensioners could be bought-out at the outset of the contract, with the
older deferred pensioners following a few years after, then the younger deferred
pensioners secured as they approach retirement age. Again, it might be possible to
fix the buy-out prices in advance.
●
Data and exception risks – traditionally insurers only ever took on promises to pay
precisely defined benefits, but with the increasingly cut-throat competition in the
market they have become willing to take on the obligations of a scheme lock, stock
and barrel even if the data cannot be completely verified. This has greatly improved
execution speeds, as trustees used to have to dedicate considerable time and effort
to auditing their data in order to ensure that it was completely accurate before the
insurer would take on the liabilities. Another variant on this theme is insurers taking
20
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
on the financial risk of any future requirements to equalise Guaranteed Minimum
Pensions (“GMP”).
●
Bulk transfers – the pension scheme’s liabilities are transferred to a new pension
arrangement operated by the insurer prior to being bought-out. This approach
allows the more flexible environment of UK occupational pensions legislation to be
used as an interim stage to improve the efficiency of the outcome for all parties
whilst the insurer immediately comes “on risk” to protect the liabilities at the point
of transfer. Uses of the bulk transfer stage are as follows:
– It can be used to speed up the buy-out process, and allows the trustees to
pass the risk of data errors to the insurer, rather than retaining this risk as
was traditionally the case. This approach was used by Paternoster in their
£170 million transaction with Emap at the end of 2007.
– It can also allow the insurer to alter the liabilities before securing them with
their life assurance arm, through liability reduction exercises (most typically
offering deferred pensioners a cash lump sum to take to a personal pension
in lieu of the annuity that would otherwise be secured for them by the
insurer). This is the approach which Rothesay Life has been promoting as
their unique selling point and allows the insurer to trim the cost of buyingout deferred pensioners marginally.
These variations have allowed insurers to tailor deals to the requirements of the
trustees or sponsor looking to buy-out their pension liabilities. However, we should
note that these solutions are all variations on the same theme, since all of the
approaches will culminate in the pensions being bought-out with an FSA-regulated
insurance company. These methods cannot get around the capital constraints
imposed by the FSA and therefore pricing is very much driven by the regime.
The two most significant innovations seem to be the willingness of insurers to accept
less than perfect data (which has always been a problem for trustees looking to
purchase annuities) and the use of bulk transfers to pension schemes operated by
the insurers, to increase speed and efficiency of the execution. Other insurers are
increasingly likely to offer these options if they become popular in the market as
they are relatively simple to understand and manage.
21
PSTS Comment:
The competitive
environment has led
insurers to attempt to
differentiate and drive
new business by
providing innovative
solutions. Most of these
solutions are, however,
not making the annuity
product more attractive
given that the pricing
is constrained by the
FSA regime.
Two areas of significant
advancement have been
the willingness of
insurers to take on
unknown risks (such as
data errors and GMP
equalisation) and the
use of liability reduction
strategies to manage
down the costs of a
buy-out.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
4. The Annuity Players
There are now a huge number of providers in the bulk annuity market turning the
duopoly of Legal & General and Prudential into a thriving competitive market. As well
as long established life insurance companies re-entering the market or entering the
market for the first time a significant number of new “mono-line” insurers have raised
funds and entered the market to specialise solely in annuity buy-outs.
Legal & General
Legal & General (“L&G”) have been involved in the UK pension buy-out market for
many years, writing business through its AA+ rated subsidiary. L&G is the third largest
insurer in the UK and bulk buy-outs of UK pension schemes represents a relatively
small proportion of their overall business.
Over 2006 L&G wrote just over £1 billion of new bulk buy-out business, which
increased to £1.2 billion over 2007. Their most high profile transaction was the
£240 million buy-out of 4,500 pensioners from the Weir Group which was announced
in December 2007, but L&G also took on pension liabilities from Electricity
Association Services and Queens Moat House Hotel Group over the year.
L&G have, alongside Paternoster, been the most aggressive on pricing annuities
over the past twelve months.
Prudential
Prudential have also been involved in the UK pension buy-out market for many years
and writes business through a AA+ rated subsidiary, in common with L&G.
In recent years Prudential have focused on buying whole annuity books from other
insurers. For example, in December 2007 Prudential confirmed that it had completed
the transfer of Equitable Life’s with-profits annuities book, covering approximately
62,000 policies and with estimated assets of around £1.7 billion as at 31st December
2007. In previous years Prudential have acquired £1.5 billion of with-profit annuities
from Phoenix Life and Pensions and £1.1 billion of annuities from Royal London.
However, Prudential did not actively participate in the bulk buy-out market during
2007, instead operating solely in the insurer back-book market. Their 2007 new
business results explain their rationale:
“Prevailing pricing levels for bulk annuities were insufficient to meet
Prudential’s return on capital requirements, based on its view of future
longevity improvements, and Prudential UK chose not to write business
at those un-economic levels in 2007.”
Our analysis in Section 3.4 supports this view and Prudential are not the only insurer
to have made such comments and a number of participants have expressed similar
views in informal discussions with PSTS.
Paternoster
Of all the new entrants in the UK bulk buy-out market, none have been more high
profile than Paternoster. Over the course of 2007 their founder Mark Wood has
regularly appeared in the press stating that he expects the bulk buy-out market to
dramatically increase in size imminently. It remains to be seen whether his forecasts
will materialise.
22
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Paternoster currently have around 110 employees, 70 of whom are based in Mumbai,
making Paternoster the largest employer of actuaries in India. They claim this gives
them the opportunity to reduce costs compared to the traditional players in the
market, and reduce their quotation turnaround time.
Paternoster wrote very little business in 2006, but over the course of 2007 wrote £1.4
billion of new bulk buy-out business, predominantly in the final quarter. Their largest
deal was the £800 million transfer of pensioner liabilities from P&O, although this
was only a partial buy-out and the responsibility for paying the pensions remains
with the scheme. It also appears that this transaction was not instigated by the
sponsoring employer; rather it was the trustees of the scheme who considered
securing annuities with an insurance company to be an appropriate use of the
scheme’s assets.
Other significant deals included Eni Lasmo for £150 million and Emap for £170
million. Eni Lasmo was of interest since this was the largest buy-out done to date
using an online auction process (discussed in Section 5.2). The Emap transaction was
the first where the insurer also took on the risk for verifying the data involved in the
buy-out. Traditionally it could take many months to verify the member data being
used, to ensure that the benefits being insured correctly match those promised to
members. In order to speed up the buy-out process, Paternoster agreed to accept the
liabilities before this verification was complete, so that they assumed the risk of data
errors (as discussed in Section 3.6).
In order to secure this new business it would appear that Paternoster have priced
lower than the two traditional bulk buy-out players. This view is backed-up by the
general belief that buy-out prices have come down slightly over the course of 2007,
as insurers seem willing to accept lower returns on their capital than was previously
the case, as we discussed in Section 3.4.
However, the transactions completed to date represent a very small proportion of
what Paternoster believes is potentially a huge market. Therefore, Paternoster may
have been using extremely competitive pricing in order to stimulate the market and
gain a leading market position, in the hope that high levels of business written over
future years on more prudent terms will be more profitable. It is also worth noting
that the majority of the business written by Paternoster has been in respect of
current pensioners rather than deferred pensioners, where buy-out prices are lower,
primarily because the insurer takes on less reinvestment risk due to the shorter
duration of the liabilities.
Pension Insurance Corporation
Pension Insurance Corporation (“PIC”) is the FSA-approved life insurance company
of Pension Corporation.
Pension Corporation was founded in 2006 by Edmund Truell, who was formerly one
of the founders of the private equity firm Duke Street Capital. It has secured around
£1 billion of equity commitments, primarily to provide bulk buy-out annuity solutions
through PIC, which they expect would allow them to write up to £20 billion
of business.
23
The Annuity Players
Mark Wood was formerly the Chief Executive at Prudential (UK and Europe) before he
established Paternoster in December 2005. After receiving £500 million of equity
financing from a variety of investors including Deutsche Bank and the hedge fund
firm Eton Park International, Paternoster gained FSA approval in June 2006.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
However, PIC has found the price competition in the market extremely difficult to
overcome, and have written very little business to date. They have only recently
announced details of their first deal, which was a £72 million buy-out with Swan Hill
Pension Scheme at the end of May. Indeed PIC has made similar comments to
Prudential regarding the current levels of pricing being uneconomic, and they appear
to have instead focussed on their other solutions for the pension market.
Pension Corporation’s alternative solutions for the pension market are discussed in
further detail in Section 6.1.
Lucida
Jonathan Bloomer established Lucida in 2007, having previously been Group Chief
Executive at Prudential. Unlike most of the other new entrants to the bulk buy-out
market, Lucida have a solitary investor; the US private equity firm Cerberus Capital
Management. Cerberus has committed initial capital of £1 billion, which we estimate
would enable them to write around £10 billion of new business. Jonathan Bloomer is
a Partner of Cerberus European Capital Advisors, which should allow Lucida to raise
further funds fairly easily if required.
Lucida employ around 25 actuaries (both life and pensions) in the UK, and they
intend to target the larger, “more complex” pension schemes, with liabilities of at
least £50 million. They received FSA authorisation in November 2007.
Lucida wrote their first deal in January 2008 and this was slightly unusual in that it
was a reinsurance transaction with another insurer. Lucida agreed to reinsure over
€100 million (£75 million) of annuities held by Bank of Ireland Life and they will also
reinsure the majority of Bank of Ireland Life’s future annuity business, estimated to be
worth another €40 million per year. In turn, Lucida have reinsured some of their
longevity risk through JPMorgan, using JPMorgan’s LifeMetrics Longevity Index
(discussed further in Section 9). Therefore, some of the longevity risk is retained by
Bank of Ireland Life, the majority is passed on to Lucida, and the extreme tail risk is
covered by JPMorgan.
We understand Lucida did not set out to make reinsurance a key part of their
business plan, but thought that this deal was attractive to them. Although they intend
to write most of their business directly with pension schemes in the future, they
would consider doing more reinsurance deals as well as providing mortality
protection for pension schemes in isolation.
In March 2008 they announced their first transaction with the trustees of two
schemes sponsored by Morgan Crucible. The deal was in respect of the current
pensioners, worth around £160 million, who were secured with annuities, whilst the
schemes retained the liabilities for current active members and deferred pensioners.
Lucida consider current pricing levels to be extremely competitive and have indicated
that they will not reduce their return on equity for Cerberus below double digits.
However, they do expect to be able to compete on price with the likes of Paternoster.
Since they have a single investor, Lucida may be able to remain in the buy-out market
without writing significant levels of business for longer than some other new entrants,
whose investors might require higher levels of activity. We would therefore not expect
them to drop out of the market, at least in the short term. In common with the other
annuity players they are optimistic about the number of pension schemes who will
look to buy-out their pension benefits in the near future.
24
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Synesis Life
Synesis Life was set up in 2006 by Isabel Hudson and Jay Shah, former Senior
Executives at Prudential. At the time of writing they have yet to write any business,
and although capital has been pledged to Synesis Life, they have not yet capitalised
the insurance entity. Due to this they have not yet secured authorisation from the
FSA, but they intend to capitalise immediately prior to their first transaction, which
they expect will lead to FSA authorisation.
Synesis Life have indicated that they will only be interested in taking on liabilities of
£100 million or more, and they believe they have capacity to write up to £10 billion
of business.
As yet there has been no news regarding when Synesis Life expects to carry out their
first transaction.
Rothesay Life
Goldman Sachs have also launched a bulk buy-out product through their AA-rated
and FSA regulated subsidiary Rothesay Life. Their buy-out team was set up in 2006 by
Addy Loudiadis, who was formerly head of their European investment banking
franchise. Goldman Sachs hopes to target medium to large pension schemes with
liabilities of around £150 million or more, and they expect to be able to write up to
£20 billion of new business initially.
Their approach involves a bulk transfer of the pension liabilities to a new scheme
(sponsored by Goldman Sachs and effectively guaranteed by Rothesay Life) prior to
buy-out, which allows the trustees to be offered an indemnity package to protect
them from all future benefit related claims, such as those which may result due to
data errors at the time of the deal. Before buying-out the liabilities through Rothesay
Life, Goldman Sachs would carry out liability reduction exercises, such as offering
transfer values to deferred pensioners and running early retirement programmes.
By running liability reduction exercises Goldman Sachs hope to be able to reduce the
buy-out cost for the scheme, which in turn would allow them to offer more
competitive pricing. Since there is more scope for liability reduction exercises to be
carried out with deferred pensioners than with current pensioners, it is likely that
their pricing will be more competitive for non-pensioners. However, since the level of
member uptake for liability reduction exercises can be notoriously difficult to predict,
this may present a risk to Goldman Sachs.
However, there will be nothing to stop other insurers in the bulk buy-out market also
running liability reduction exercises in an attempt to reduce the buy-out liabilities, if
this approach proves successful. Indeed, pension scheme sponsors might consider it to
be more cost-effective for them to run the liability reduction exercises themselves,
before buying-out the pension scheme liabilities with an insurer. This would allow
them to gain from all of the cost savings which a liability reduction exercise might
lead to. It could also lead to a higher member take-up rate, since if the liability
reduction exercise was offered by Goldman Sachs members might place a higher
value on the benefits they were being asked to give up, in the knowledge that if they
did not take up the offer the alternative would be to have their benefits secured
through an annuity with Rothesay Life.
Rothesay Life announced details of their first deal at the end of February 2008. The
transaction was with the UK casino, bingo and online betting company Rank, and is
reported to be worth around £700 million. The pension scheme was fairly well
25
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
funded, but despite this the company had committed to pay future contributions of
£30.8 million. Following the deal this requirement was removed, and Rank also
expects to receive a payment of “at least £20 million” due to the funding position of
the pension scheme.
The circumstances surrounding this deal were quite unique. Since Rank were
committed to paying a significant level of contributions into the pension scheme over
the next few years, the buy-out option was particularly attractive to them. Goldman
Sachs had also provided corporate advice to Rank prior to them deciding to buy-out.
Other Market Entrants
There are a number of other insurance providers who have added their names to the
list of insurers who now claim to be operating in the UK pension bulk buy-out market.
However, information regarding deals done by these insurers is currently relatively
sparse. Whilst it is not possible to list every player in the UK pension bulk buy-out
market, some of the insurers include:
●
●
●
●
●
●
Aegon, one of the world’s largest insurance companies, who entered the market at
the start of 2007 with the aim of targeting small to medium sized pension
schemes. They have also joined forces with the investment bank UBS to offer a
product aimed at large pension schemes, where UBS invests the scheme’s assets
with a view to gradually buying-out the pension scheme liabilities as various
investment performance triggers are hit.
AIG Life, who are also one of the world’s largest insurers, entered the market in
2006 having previously offered bulk annuity contracts in the US.
Norwich Union, the largest insurer in the UK, entered the bulk buy-out market
in 2006.
Canada Life, who took on Equitable Life's £4.6 billion non-profit annuity book in
2006, comprising of around 130,000 annuities. This took their total number of
annuities in payment to around 280,000.
Pearl Insurance, which was established by Hugh Osmond, who was previously
connected with Pizza Express and the pub chain Punch Taverns. The buy-out would
be provided through Axial, an investment management business owned by Pearl,
who would place assets with external managers rather than managing the funds
themselves. However, at the time of writing there has been no news regarding
Pearl’s first deal.
MetLife, another global insurance company looking to enter the UK bulk
buy-out market.
Whilst most of these large insurers have not done significant amounts of business in
the UK bulk buy-out market, due to their size they are well placed to become more
active players if they see an opportunity in the market.
26
PSTS Comment:
There are a lot more
options available now
for schemes looking to
purchase annuity
contracts. However, a
number of players in the
market do not believe
that the current pricing
levels are sustainable.
PSTS believe that
anyone looking to
secure annuity contracts
should move quickly as
there is definitely some
“loss leader” pricing
going on in the market
at the moment and this
cannot be sustained in
the longer term.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
After a decision has been made by the trustees and/or sponsor to secure the benefits
with annuity contracts they must decide how to go about implementing this decision.
5.1 How to Choose an Insurer
Given the abundance of insurers who are now in the market, pension scheme
sponsors or trustees who are considering buy-out could be forgiven for not knowing
where to start. In particular, they may struggle to differentiate between the different
insurers given that they are all offering practically the same product; an FSA regulated
annuity contract for each member to match the benefits promised under the scheme.
Whilst price is always a key consideration, there are a number of softer factors which
are often considered when selecting an insurer and upon which insurers attempt to
differentiate their offering in the market. These softer factors are considered below.
Financial Strength
Since insurers are regulated by the FSA, they are all required to meet prescribed
minimum capital requirements. The FSA has stated that it looks for run-off ruin
probabilities to be 5% or less, and will expect capital to be increased until this target
is reached. Trustees and sponsoring employers transferring their liabilities to an insurer
might well be surprised to discover that the minimum requirements allow for as high
as a 1 in 20 probability of the insurer defaulting. However, the case of Equitable Life
illustrates that it is possible for insurance companies to get into severe financial
difficulties and, more recently, the case of Northern Rock has also demonstrated that
major financial institutions regulated by the FSA are not immune to financial distress.
The extent to which insurers exceed this minimum requirement may therefore be
considered by the trustees to be a significant factor in selecting an insurer and is
often used by insurers to differentiate their offering.
The majority of the new players in the bulk buy-out market write (or propose to
write) their business through a subsidiary with a AA- credit rating, as determined by
the likes of Moody’s and Standard & Poor’s. The long established players such as L&G
and Prudential offer slightly more financial strength, writing business through a AA+
rated subsidiary.
Administration & Implementation Capabilities
Although a pension scheme buy-out might end the involvement of the trustees and
the sponsoring employer with the pension scheme, it does not mark the end of the
members’ interest. The insurer might be taking on responsibility to pay pensions to
members for the best part of a century, so the trustees and sponsoring employer
might be keen to ensure that the insurer looks after the members well.
It is therefore worthwhile checking the adequacy of the systems used by the insurer
to administer the pensions. This should include analysis of the payroll system, as well
as ensuring that data is stored securely. Ultimately, if former members of the pension
scheme are not satisfied with the service which they receive, they may complain to
the sponsoring employer.
Benefits Provided
Depending on the complexity of the benefits offered by the pension scheme, the
insurance company might not guarantee to match the precise benefits currently
27
The Purchase Process
5. The Purchase Process
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
offered to members. This could lead to different insurance companies providing
different prices, due to varying levels of benefits covered.
For example, some schemes provide fairly complex pension increases, incorporating
various caps and floors which often vary depending on when the pension was
accrued. Due to administrative complexity, insurers may be unwilling to match these
pension increases precisely.
Some insurance companies might not be willing to use the same definitions as
currently used in the scheme rules, preferring to use their own standard definitions
instead. This could affect the circumstances under which ill health early retirement
benefits and death benefits are paid.
An insurer may also restrict the options available to deferred pensioners when they
retire. For example, some pension schemes provide members with the option to
forfeit some of their initial pension in exchange for a higher spouse’s pension at
retirement. Some insurers might be unwilling to offer such an option, since it could
leave them vulnerable to selection.
Brand Name
Given the substantial cost of a pension scheme buy-out, the trustees and the
sponsoring employer will want the buy-out decision to be appreciated by members.
This could be helped by using a large, reputable insurer for the buy-out, which most
members will have heard of. Using a household name might give members the
impression that their benefits are more secure than using a relatively unknown
insurer, even if this is not actually the case in practice.
Future Direction of Insurer
The trustees and sponsoring employer will want to ensure that they are securing
members’ benefits for the full term of the liabilities. They might therefore be
interested in considering the future direction of the insurer’s business.
As described in other areas of this report, there has been a raft of new entrants to
the UK pension buy-out market in anticipation of a surge in the number of pension
schemes looking to buy-out. However, if this explosion in demand does not
materialise, the buy-out market could quickly become over-saturated, which may
lead to some providers leaving the market. Such insurers may seek to remove their
exposure to the market completely, by transferring any liabilities to another insurer.
Trustees and sponsoring employers may prefer their former pension scheme liabilities
to not be moved around different insurers, to give more stability to former members.
Therefore, they might favour insurance companies who appear to be committed to
the market for the long-term.
Other Factors
The theory of behavioural finance explores a wide range of factors which affect
financial decisions. Trustees and sponsoring employers could be influenced by press
commentary on the buy-out market, favouring insurers who have recently done
high-profile deals. They may also favour providers who give them a better quality
of service during the quotation process. Finally, personal views and experiences with
the alternative insurers may influence the final decision.
28
PSTS Comment:
As the purchase of an
annuity is an extremely
long-term contract, the
current position of the
insurer provides little
long-term information
about security. In
addition, all insurers
are subject to the same
FSA solvency regime,
so there is little to be
gained by observing
the current financial
position of the insurer
or other “soft” factors,
except to mitigate any
discomfort felt by the
members regarding
the transfer.
Therefore, it is our
belief that selection
should be on the
grounds of price alone.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
5.2 Creating Price Competition
The traditional approach to a pricing exercise for a block of annuities would have
been relatively straightforward. A full quotation would have been sourced by the
advising pension consultancy from all the interested parties in the market and the
cheapest quotation used.
With the substantial increase in the number of competitors and their perceived
willingness to compete on price grounds, increasingly complex processes are being
undertaken to try to achieve the cheapest price. Much more complex approaches are
justifiable when looking at extremely large transactions, where even small percentage
changes in the final price can equate to many millions of pounds in savings.
It is worth noting that in a buy-out transaction the pension scheme buying the
annuities has a significant amount of negotiating power over the insurer selling them.
Transactions are voluntary, since annuities do not have to be purchased if the price is
not considered to be attractive, and many of the new players in the market are keen
to be seen to be writing bulk annuity business. This contrasts to the situation before
the days of the PPF, when purchasers had no alternative to buying bulk annuities
through an insurer in the case of company insolvency.
Auctions
Some consultancy firms have started to offer an auction process with the aim of
helping the decision-makers in a buy-out process select an insurer. Paternoster’s
£150 million buy-out of Lasmo’s pension liabilities in the last quarter of 2007 was
conducted through an online auction process.
The auction process is intended to simplify the buy-out for the company or pension
scheme looking to offload their pension scheme liabilities. The firm managing the
auction will liaise with the insurers to obtain the quotes, negotiating with them
where necessary, then provide their client with the lowest quote at the end of the
process. If the two lowest bids are fairly close, the auctioneer may provide the client
with details of both bids to allow them to take account of non-price factors, such as
those outlined in Section 5.1 above.
Companies offering this service claim that advantages of the auction process include:
●
●
●
●
A clear audit trail is produced, allowing the trustees to justify their choice
of insurer.
A reduction in the time taken to buy-out.
The insurance companies are reassured that the client has a genuine desire
to transact, so competitive quotes are obtained from the outset.
A reduction in management time required in handling the buy-out process.
However, the auction process has a fair number of critics. Edmund Truell of Pension
Corporation has stated that he will not take part in an auction process, since he
considers them to be fundamentally flawed. An auction approach could be considered
overly simplistic; there are a significant number of factors which should be considered
when choosing an insurer, other than price, as described in Section 5.1 above. Critics
of the auction process believe that it is the trustees who should be considering these
qualitative factors, rather than the auctioneer. To enable the trustees and the
company to make a fully informed choice of insurer they will need more information
than simply the lowest price quoted.
29
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
As discussed previously, we are of the view that the softer issues are not of particular
relevance as all insurers are regulated by the FSA and observing the current
positioning of the insurer in relation to a very long term contract is unlikely to be
indicative of the position in five, ten or even fifty years time. Price is, for us, the
determining factor. Auctions simplistically therefore seem like the best approach to
get a decent pricing result.
Our concern around auctions is that designing an optimum auction is extremely
complex and has been the subject of significant academic research. One of the
potential issues arising is that an auction process, if not correctly designed, can lead
to a worse outcome than a more simplified or usual process. This is particularly the
case if certain players drop out of the auction process altogether (as suggested by
Pension Corporation) which would reduce competitive tension.
Perhaps the most famous example of an optimum auction design leading to a very
good outcome was the auction of the third generation (“3G”) mobile telephone
licences by the UK Government in 2000. The auction process was designed after
many years of study (the original consulting on the project started in 1997) and
simulation by a team of economists from University College London and Oxford
University. Other countries’ 3G auctions were less well designed and generated less
favourable outcomes for their governments.
A detailed paper setting out the considerations and the process of designing the 3G
auction entitled “The Biggest Auction Ever: the Sale of the British 3G Telecom
Licences” by Binmore and Klemperer gives an intriguing insight into the complexities
of designing an efficient auction process.
Whilst an auction for a book of annuity business is not directly comparable with the
3G auction, the principles of designing an auction to maximise the efficiency of the
outcome are highly relevant. We question whether suitable economic research into
the market and the design process has been undertaken to ascertain whether the
online auctions that have taken place are maximising the pricing benefits.
Dissecting the Transaction
One approach that has yet to be seen in the annuity market is parcelling up the
contract into various homogenous chunks (such as very old pensioners and younger
pensioners) and selling these chunks individually to the cheapest bidder. The idea here
is both to allow the different views of insurers on where the most longevity risk lies to
be captured in the bidding process and, similar to the approach taken in the UK
Government 3G auction (discussed above), to create an imperative amongst the
various competitors that they must win one of the chunks of annuities available.
Designing the parcels and the subsequent auction or selling process required to
ensure that the optimum price is achieved is not straightforward. As such it is likely
that many organisations will not be willing to commit the necessary resources to the
design phase of the process to ensure the best economic outcome is achieved.
Such approaches are going to become increasingly necessary if the market for large
scale annuity buy-outs continues to develop.
Long List, Short List, Exclusivity
For very large and complex transactions a straight auction process may not be
appropriate. Alternatively, the trustees or the sponsor could obtain quotes from all
the available insurers, then whittle these down on price grounds to a smaller subset
(perhaps only two or three). Further detailed price negotiations would then be
30
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
undertaken, before finally selecting one party with whom to have final exclusive
discussions to finalise the purchase of the annuities.
This process is very similar to the processes undertaken in the mergers and
acquisition market, where very high value, complex disposals are being made by a
voluntary seller. A number of the large transactions that have recently completed
have been undertaken through a long list, short list, exclusivity approach.
Such processes may be better suited than an online auction to achieving the best
result for the seller.
Economic Research
If the market for large scale annuity buy-outs continues to grow then some form
of detailed research into the design of the selling process is likely to be required.
An issue which is likely to arise is who will pay for such research. The annuity firms are
unlikely to fund research that leads to lower prices for themselves, whilst sponsoring
employers interested in selling off their liabilities are unlikely to consider such
investment as useful (given that there is a “free rider” problem present in that most
of the benefits will accrue to future sellers of annuity books rather than the party
that has paid for the research).
PSTS Comment:
Creating pricing tension
between the insurers is
essential to get the
cheapest price for a
particular block of
business. However,
auction processes are
subject to significant
behavioural
considerations and only
a very carefully
designed auction can be
guaranteed to provide
the best outcome.
For proof of this
comment consider the
considerable effort
expended by the UK
Government to design
correctly the third
generation mobile
telephone auctions
in 2000.
We believe that a better
outcome will be
achieved by undertaking
a process similar to that
used in mergers and
acquisitions, except in
very small transactions,
but academic research
into the most efficient
structure for the sale of
annuity business would
be extremely helpful.
31
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
6. The Buyer’s Perspective
The bulk annuity market has seen two key drivers of annuity business in recent years:
●
●
Trustees of pension schemes, where they can afford the cost of securing benefits
from their own assets, have been key in driving forward partial annuity buy-outs
(particularly of current pensioners) in cases such as P&O and Morgan Crucible.
Sponsoring employers have been the key drivers of wholesale annuity buy-outs
such as in the cases of Eni Lasmo, Emap and Rank.
Regardless of who is driving the buy-out process, the key stakeholder in the decision
to purchase annuities is the sponsoring employer, as they have to meet the ultimate
cost of running the pension scheme. In a large number of cases it is very difficult for
the trustees to purchase annuity contracts without the consent of the sponsoring
employer, and in any case they may prefer to have the employer’s support when
making decisions of such significance to the pension scheme.
In this section of the report we therefore focus on why sponsoring employers may
wish to implement a bulk annuity purchase.
6.1 An Increasing Burden
Over recent years the defined benefit promise has become an increasingly onerous
requirement. Most of the reasons for this are well-documented, including:
●
●
●
●
●
●
●
Increased legislative requirements, such as those arising as a result of the Pensions
Act 2004.
Improvements in longevity already experienced and the considerable uncertainty
relating to future improvements.
The low inflationary environment experienced over recent years, increasing the
value of fixed or capped pension increases.
The removal in 1997 of the tax credit available on share dividends for pension
schemes.
The requirement under the accounting standards FRS 17 and IAS 19 to disclose
the funding position of the pension scheme on the sponsoring employer’s balance
sheet, including any deficit as a debt item, which can increase the volatility of
company balance sheet and profit and loss account due to the market valuations
required.
The creation of the Pensions Regulator whose influence on the valuation process
and corporate activity through the “clearance” process is viewed as increasingly
burdensome.
The increasing and volatile levies required to be paid to the recently established
Pension Protection Fund, which has increased the administrative costs of pension
schemes.
Operating a defined benefit pension scheme involves a great deal of uncertainty.
The trustees of the scheme are required to carry out a formal valuation at least
every three years, and the funding position revealed in this valuation depends on the
assumptions chosen by the trustees (albeit after consultation with the sponsoring
employer). This may lead to a deficit being revealed, which the trustees will require
the sponsoring employer to fund, usually over a relatively short period of time, such
as five years. Unlike most other items in a company’s cash flow projections, the level
of contributions likely to be requested by the trustees can be very difficult to predict.
32
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
The FRS 17 or IAS 19 funding position, as disclosed in the company accounts, is also
volatile. This is particularly the case for pension schemes with large equity holdings,
since these investments do not match the AA corporate bond index which is used to
determine the discount rate and hence value the liabilities under the accounting
standards. This can have a significant effect on the company’s balance sheet,
particularly for companies for whom the pension scheme is large compared to the
equity value of the sponsoring employer.
This has led to many sponsors of defined benefit pension schemes being keen to
look for possible exit strategies to remove the obligations connected with their
pension scheme.
In Section 6.2 we consider the buy-out decision from the economic viewpoint of
the sponsor, before going on to consider the softer reasons in Section 6.3.
6.2 A Sponsor Deciding
In order to consider the position of the pension scheme sponsor let us consider a
sponsoring employer with a “Typical Scheme” (as defined in Section 3.3) with assets
of £1.35 billion that has, through a mixture of good investment returns and
experience gains, become funded to 135% of FRS 17 (so that the FRS 17 liabilities are
equal to £1 billion). The sponsoring employer has a number of quotations for annuity
buy-out that all represent 135% of the FRS 17 liabilities. The sponsoring employer
needs to consider the merits of winding-up the scheme and buying-out the liabilities
with an insurer.
Before looking at the softer issues in Section 6.3, let us consider the economics. If the
sponsor decides to wind-up the scheme then the assets will be passed to the selected
insurer who will then purchase a portfolio of corporate and government bonds to
match the expected benefit outflows. They will put up a small amount of capital
against this (circa 8% of assets transferred) and will carry only a very small risk of this
amount of capital being eroded (a 1-in-200 chance over one year).
In return for putting up the capital the life insurer will aim to run off the liabilities
over a long time period at a net present value lower than the assets received, in order
to generate profits. In Table 3.2 in Section 3.3 we set out the how the buy-out price
might be made up from the value of the FRS 17 liabilities for our “Typical Scheme”.
In this example, the yield differential was calculated to be 23%, corresponding to the
difference between the corporate bond yield (of 5.5%) and the pricing yield (of 4.2%)
over the 17 year duration, since:
(
1 + 5.5%
1 + 4.2%
17
)
– 1 = 23%
However, the insurer’s expected return on assets, based on their diversified portfolio
of government and high quality corporate bonds, was assumed to be 5.0% per
annum. If this rate were used, rather than the pricing yield of 4.2%, the yield
differential would only be 8%, calculated as:
17
5.5%
)
( 11 ++ 5.0%
33
– 1 = 8%
The Buyer’s Perspective
The Pensions Regulator has been putting pressure on trustees to increase the
degree of prudence used in their actuarial valuations, particularly around mortality
assumptions and discount rates, as well as looking for trustees to fund any deficits
over as short a period as practicable.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
The table below shows the difference to the overall pricing level to which using the
insurer’s expected return on assets, rather than their pricing yield, would lead.
Table 6.1 – Impact of yield differential between FRS 17 or IAS 19 yield and expected
return on assets
Addition to FRS 17 or IAS 19 liabilities
Based on Pricing Yield Based on Expected Return on Assets
FRS 17 Liabilities
Yield
Mortality
Expenses
100%
23%
9%
3%
100%
8%
9%
3%
Proportion of
FRS 17 Liabilities
135%
120%
The difference between the 135% at which the insurer prices its annuity contracts
and the 120% at which it expects to be able to run off the liabilities is the net present
value of the profits that it hopes to make from the contract over its lifetime. This 15%
of assets may seem rather high, particularly since only 8% capital is put at risk, but
this profit margin has to sustain a return on capital of around 15% per annum for the
whole lifetime of the contract, which may exceed 50 years.
However, this illustrates that the sponsoring employer is passing over a net present
value of profits to the insurer of 15% of the FRS 17 liabilities on the day they buy the
annuities. In this case, that corresponds to £150 million, since the liabilities are valued
at £1 billion on the FRS 17 basis.
As an alternative to buying-out the liabilities with an insurer, the sponsoring
employer could retain the scheme on its balance sheet. Since the pension liabilities
are valued on prudent assumptions, surplus would be expected to appear gradually
over time, as actual experience is more favourable than expected. This would generate
profits for the sponsoring employer at a similar level to that of the insurer, without
the need to put explicit capital at risk in the way the insurer does. This would be
achieved at a similar level of risk to that of the insurer, since the insurer only faces a
small risk of actually burning through the 8% capital buffer which they hold, and the
pension scheme trustees can choose to follow a similar investment strategy to that
followed by an insurance company.
In most cases the company can have control over the destiny of the pension scheme,
including when it winds-up. The company could continue to run the scheme onwards
until the last member has died and then wind-up the scheme and collect the surplus.
Alternatively it could wind-up at an earlier date, once a substantial proportion of the
profit has been achieved, and when the fixed expenses of running the scheme become
a significant proportion of the cash flows. (Most schemes would expect around 50%
of the net present value of the profits to be accumulated in the first 20 years.)
Alternatively, the sponsoring employer could seek agreement from its trustees to
release profit from time to time under the surplus regulations, subject to this being
allowed by the scheme’s trust deed and rules. This would be a particularly attractive
option if release of surplus regulations are amended in future to make it easier for a
sponsoring employer to remove surplus from their pension scheme.
When the insurer puts up capital, this is dedicated to their annuity book and invested
in their portfolio of bonds. This constraint does not apply to the sponsoring employer
34
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
of a pension scheme, since they do not need to put up explicit capital to back their
pension obligations in addition to the funds held within the pension scheme.
Therefore the capital that the sponsor (effectively) puts up can be used twice; once in
the business and secondly to support the pension scheme. As long as the sponsoring
employer is not going to be caused severe financial distress by any calls for cash from
the scheme then the additional risk being run by the sponsoring employer is small,
and the profits available substantial.
In addition, if the sponsoring employer chooses to retain the pension scheme on its
balance sheet, it does not necessarily need to fund the scheme up to 135% of the
FRS 17 liabilities using cash. Instead, contingent assets could be used to provide
security to the trustees, allowing the scheme sponsor to retain the cash for use in
their core business.
In summary, the sponsoring employer should be able to acquire the insurance
company’s profits at broadly the same level of risk, by taking a similar approach to
managing the pension scheme as the insurer would do. However, the sponsoring
employer maintains further advantages over an insurance company due to the
different regulatory and accounting regimes that apply to pension schemes and
insurance companies.
Firstly, short term volatility of returns (such as corporate bond spreads or default
rates rising) does not have a direct impact on the sponsoring employer’s balance
sheet. This is because, under FRS 17 or IAS 19, subject to auditor’s opinion, any
surplus above 100% of the liabilities that cannot be directly accessed is treated as
an off balance sheet item and its fluctuations taken off balance sheet (through the
STRGL or SORIE depending on which accounting standard is in use).
The following example illustrates the impact of gains and losses on the company
balance sheet for the company outlined above. Let us assume that over a year the
liabilities rise by 3%, due to experience losses on the liabilities, and assets fall by 3%,
due to credit losses. For simplicity let us also assume that total returns on bonds are
broadly zero over the year.
Table 6.2 – Impact of experience losses for a pension scheme with a large surplus
on the company’s accounting position
Assets
FRS 17 Liabilities
Surplus
Unrecognisable Surplus
Recorded in balance sheet
Opening
£ millions
Closing
£ millions
1,350
(1,000)
1,310
(1,030)
350
(350)
280
(280)
0
0
Although the true funding position of the scheme has deteriorated, there is no impact
on the balance sheet position. Profits generated by the pension scheme going forward
are likely to be lower, unless losses are reversed by similar gains in the future.
However, as this is all off balance sheet and only recognised in the profit and loss
account when they become actual profits, the changes have no impact on the
accounting for the pension scheme. As long as investors have not been banking on
profits from the pension scheme in advance there should be no impact on the stockmarket valuation of the business.
35
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
The change in the funding position also has no cash impact for the company.
The pension scheme remains well-funded by pension scheme standards, at 127%
of its FRS 17 liabilities. The trustees are therefore unlikely to require an immediate
restoration of their previous funding position. If we contrast this with the position of
an insurer experiencing the same experience losses, they will be required to restore
the solvency position of their annuity book. This would require them to add a further
£70 million in capital to their annuity book, with a commensurate impact on their
balance sheet and profits.
The less onerous regime for pension schemes presents the sponsoring employer with
a lower risk impact of running the same strategy as an insurer.
Therefore, in purely financial terms, the sponsoring employer may be better able to
generate profits from the pension scheme than an insurance company. This may lead
to a sponsoring employer preferring to retain their pension scheme, rather than
transferring it to an insurer. Whilst the mechanisms for extracting profits are
somewhat less clear cut for a sponsoring employer than for an insurance company,
due to the regulations regarding the release of surplus, the risks are almost identical.
The sponsoring employer can also effectively use its capital twice and has a stronger
position in terms of accounting and demands on cash flow.
6.3 Softer Issues
There are a number of non-financial factors which also influence the sponsor’s buyout decision. The key reasons sponsoring employers tend to give for looking to offload
their pension obligations can usually be distilled into the following:
A pension scheme is an increasing drain on senior management time and is a
distraction from running the underlying business. Pension schemes can also pose
particular difficulties in transactional or restructuring situations.
● The sponsoring employer, unlike a life insurer, has no competitive advantage in
managing pension scheme risks and therefore should not be in the business of
running defined benefit pension schemes.
● Pension schemes are too risky (particularly in terms of investment risk and
longevity risk) for the company to bear, and lead to a higher volatility of the
business with increased capital markets cost.
● Outsourcing the scheme to an insurer leads to lower administrative costs for the
arrangement due to the economies of scale that exist.
These arguments seem, on the face of things, to be relatively sound. However, we
discuss each of them in turn below.
●
●
A pension scheme is an increasing drain on senior management time and is
a distraction from running the underlying business. Pension schemes can also
pose particular difficulties in transactional or restructuring situations.
In order to buy-out the pension scheme liabilities it might be necessary to fund the
scheme to around 135% of its FRS 17 liabilities. With the pension scheme funded at
this level, and with an insurance company style low-risk investment strategy in place,
there is little need for senior management to become involved in the day-to-day
management of the pension scheme.
Generally, a greater amount of senior management time is required to manage
pension schemes where the FRS 17 deficiency is a significant financial burden to the
company, or the general size of the scheme is extremely large in relation to the
resources of the sponsoring employer. This is because in such cases investment risk
36
PSTS Comment:
In purely financial terms
a bulk annuity purchase
does not seem to make
sense for a sponsoring
employer as long as
any likely cash calls can
be absorbed.
If the sponsoring
employer considers
the annuity business to
make financial sense,
then why not effectively
run the pension scheme
as an annuity business
and reap the rewards?
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
needs to be taken in order to generate sufficient returns to run off the benefits
from the current level of assets.
Additionally, if the scheme retains reserves sufficient to buy-out its liabilities with
an insurance company then any problems caused by the pension scheme in a
transactional or restructuring situation can be solved by actually triggering the
wind-up and securing the liabilities with annuity contracts.
Therefore, for a pension scheme that is fully-funded on a buy-out basis, the senior
management distractions and limitations on activity are trivial.
●
The sponsoring employer, unlike a life insurer, has no competitive advantage in
managing pension scheme risks and therefore should not be in the business of
running defined benefit pension schemes.
It is true that most sponsoring employers do not have a competitive advantage in
managing pension scheme risks. However, there are many activities that a company
engages in where it does not have a competitive advantage (such as accounting, tax
and legal issues). In such cases it is normal to outsource the function to a third party
who is a specialist in this area. Running a pension scheme involves a similar process,
whereby services, such as actuarial, administration and investment of the assets, are
outsourced to third parties.
It is vital to note that it is not possible to manage the key risk of running an annuity
book; longevity. It is only possible to be better at pricing its risks. We can assume that
insurance companies are best placed to price the longevity risks, but that these will be
reflected in the annuity cost against which we are comparing. Once an insurer has
taken on a liability it cannot manage its mortality exposure, except through attempts
to hedge it or offload it; options which are increasingly becoming available to any
pension scheme, as discussed in Section 9.
Therefore, as long as the pension scheme is of sufficient size to justify outsourcing the
management of the scheme to third parties, it should be possible to run it efficiently.
●
Pension schemes are too risky (particularly in terms of investment risk and
longevity risk) for the company to bear, and lead to a higher volatility of the
business with increased capital markets cost.
As shown in Section 6.2, it should be possible to run a well-funded pension scheme
with minimal financial risk.
●
Outsourcing the scheme to an insurer leads to lower administrative costs for
the arrangement due to the economies of scale that exist.
For large schemes (worth at least £250 million, for example) administration and
management costs are relatively efficient. It is perhaps (rather cheekily) worth noting
that Paternoster employs over 100 staff to manage its current £2 billion annuity
book, which is rather more people than a pension scheme of a similar size would
provide a livelihood for. Whilst we recognise that Paternoster has recruited a
significant proportion of their staff in anticipation of establishing a bigger annuity
book in future, we should also note that due to the differences between insurance
and pensions regulation, insurers require more resources to manage their annuity
books than pension schemes.
An area where insurance companies do have an advantage is that it does not have
to pay for levies to the Pension Protection Fund. However, it could be called upon to
fund the Financial Services Compensation Scheme in the event that other insurers
become insolvent (which is a similar concept to the funding of the PPF).
37
PSTS Comment:
Most of the softer
reasons for outsourcing
the pension scheme to
an insurer seem to fall
away when considered
in further detail. A
sponsoring employer
with a well-funded
pension scheme does
not suffer many
inefficiencies or
management costs
compared to an
insurance company.
Managing mortality risk
is made more difficult
since murder is currently
illegal in the UK!
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
6.4 Why the Recent Activity?
Given the above commentary it would seem somewhat bizarre that almost
£3 billion of annuity business was written in 2007, with a significant proportion
of this business being driven by the sponsoring employer.
Some of the transactions involved unique circumstances, and in most cases price was
the key driver. If the price demanded by the insurer is relatively low (possibly since
the insurer is offering a discount in order to win the business), and the net present
value of the profit stream that is available to the sponsoring employer in retaining the
scheme is low, then the inefficiencies and distractions of running the scheme come
more into focus.
For example, consider the company with the “Typical Scheme” funded at 135% of its
FRS 17 liabilities, as described above. In Section 6.2 we illustrated that a pension
scheme this well-funded might be expected to generate profits over time of around
15% of the FRS 17 liabilities, which equated to around £150 million in our example.
If an insurer offered a 10% discount, so the price offered was 125% of the FRS 17
liabilities, then the dynamics of the situation change dramatically. The sponsoring
employer can immediately crystallise £100 million of the profits (although tax will be
payable on this when the surplus is refunded after winding-up) and only sees £50
million in net present value of profits pass to the insurer. Given that there are some
negatives and friction costs to running a pension scheme (even though these fade
somewhat when the financial benefit is high), the company may prefer a guaranteed
£100 million now compared to a probable, but uncertain, £150 million in the future.
Our discussions with Finance Directors reveal that there is always a price at which
the Finance Director would be willing to offload their pension obligations, even if this
means undertaking significant capital raising in order to do so. However, this price
varies depending on the particular circumstances of the company and scheme
involved, with few Finance Directors willing to trade at a very significant premium
to FRS 17 or IAS 19. The graph below represents a stylised view of the maximum
pricing levels at which the Finance Directors would be willing to transfer their pension
liabilities to an insurer based on surveys we have undertaken and discussions we
have had with Finance Directors.
Graph 6.3 – probability distribution of price Finance Directors are willing
to pay to remove pension liabilities with an insurer
90%
38
100%
110%
120%
130%
140%
Maximum price Finance Directors willing to pay as a proportion of FRS 17 or IAS 19 liabilities
150%
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Our research suggests that the price at which most Finance Directors would be
willing to transfer their liabilities to an insurer would be around 110% of their FRS 17
or IAS 19 liabilities. As shown, very few Finance Directors were interested in insuring
their pension liabilities when the price was around 135% of their FRS 17 or IAS 19
liabilities, which was supported by the very low levels of annuity business written
until mid-2007. However, since annuity prices have recently incorporated a significant
discount and have therefore been reduced, to around 125% of the FRS 17 or IAS 19
liabilities for a “Typical Scheme”, the level of business being written has increased, as
more Finance Directors are interested in trading at this lower price. It is our
contention that it is the discounted pricing that has been observed in the market in
recent times that has driven the higher activity levels and not an overall shift in the
view of Finance Directors.
However, it is worth noting that there are still a limited number of Finance Directors
willing to trade even at these discounted prices, and those that are willing to do so
are likely to be currently in the process of buying-out. Therefore, even if bulk annuity
prices remain at their current levels, we would not expect to see the current levels of
business being written to continue indefinitely.
Nevertheless, the price at which Finance Directors are willing to trade is, at least in
part, a function of the funding position of their scheme, and thus the amount of
capital that the company needs to inject into the scheme (if any) to buy-out.
Therefore, if funding levels of pension schemes improve in future this could alter the
distribution above, changing Finance Directors’ views as to the maximum they would
pay, and allowing further trades to take place.
A number of the deals done recently have involved unique circumstances. We
consider some of the higher profile solvent sponsor pension buy-outs undertaken in
recent times below, in order to consider whether there were special circumstances
involved in the companies or schemes such that the Finance Directors were willing
to trade at a higher price than other companies, in more normal circumstances, may
be willing to consider.
39
●
Rank plc undertook a full buy-out with Rothesay Life in early 2008 for its
£700 million pension scheme. The interesting piece of the jigsaw for Rank is
that they were committed, under a prior agreement with the trustees, to
paying £30 million per annum in cash contributions to the scheme even
though the scheme was now in significant surplus. By completing the buyout now, at the attractive price available, they could free themselves from
this contribution burden and also receive “at least £20 million” in surplus.
There was no additional funding cost required to complete the buy-out and
significant immediate financial gains for the company. Following significant
falls in the company share price over 2007 the scheme was also larger than
its market capitalisation, indicating that the company would find it difficult
to bear any volatility from the presence of the pension scheme. All in all
Rank seems to have faced a fairly unique set of circumstances and therefore
was willing to trade at a relatively high price given the immediate benefits
of being able to cease their cash contributions, release surplus and
remove risk.
●
Eni Lasmo was a £150 million buy-out completed in late 2007. The scheme
previously belonged to Lasmo plc and was supported by a UK company with
equity value of £2.5 billion and this company was owned by ENI, an Italian oil
major with a market cap of €80 billion. The company had been funding the
scheme for many years and the premium payable (and the effective profit
PSTS Comment:
For every sponsoring
employer there is a price
at which they would be
pleased to offload their
liabilities to an
insurance company.
However, this price is
unlikely to be coincident
with the long term
prices that insurers will
be able to provide
unless the sponsoring
employer and its
scheme have specific
and unique dynamics
that raises the
“clearing price”.
Given the relatively
limited market activity
at this time of very low
prices, it is questionable
whether any business
will be done at all if
prices revert to more
normal levels.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
transferred to the insurer) was therefore extremely trivial
from the company’s perspective. We would suggest that the
company undertook this buy-out as a tidying up exercise
more than anything else and was spurred into action by the
attractive pricing that became available.
●
Emap plc undertook a full buy-out of its liabilities towards the end of 2007.
Whilst the scheme was relatively small compared to the company as a whole,
Emap plc was in the process of breaking itself up into constituent parts. This
presented an issue in that the scheme would either have to be left with one
of the businesses or split across the businesses in some ratio. In either
scenario the trustees would want recompense for the weakened employer
covenant. In addition, if the large scheme was left with one of the smaller
businesses the price received for the sale of the businesses would be
depressed due to the scale of the pension liabilities in relation to the
business. Resolving this complex issue would have led to significant costs
arising, in particular in dealing with the trustees’ concerns on employer
covenant. The premium paid to remove this issue was relatively small,
particularly in relation to the size of Emap plc at that time.
In conclusion, all sponsoring employers have a price at which they are willing and able
to trade and at which they would happily move their pension scheme to an insurer.
Where there are specific reasons for a trade or the premium is trivial in relation to the
size of the business the price at which the company may be willing to trade will be
higher. However, for most Finance Directors this clearing price is still some way below
the buy-out prices available at the current time, even allowing for the discounts being
offered by insurers.
Clearly the premium that Finance Directors are willing to trade at may change over
time and specific circumstances of each company and scheme will drive the price at
which they are willing to trade. When prices of insured solutions fall, as they have
done recently, then activity will clearly rise (as has also been seen in recent times).
However, as we suggest in Section 3.4, since prices are depressed below long term
sustainable levels and activity is still fairly muted this does not bode well for the
aggressive expansion plans of some insurance solution providers.
40
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
A number of organisations have attempted to bridge the gap between the pricing of
the traditional buy-out option and the prices at which Finance Directors are willing to
trade. Clearly if some mechanism was created that bridged this gap then the demand
for the product could be overwhelming.
However, despite the activity in the market there has yet to be a solution created
with wide applicability that competes effectively with insurance company buy-outs
on both pricing and structure.
7.1 Pension Corporation
With the exception of Paternoster, few new entrants to the market have received as
much commentary in the pensions press over recent months as Pension Corporation.
Indeed, they have even been mentioned in a recent debate in the House of Commons
regarding the security of pension provision, in relation to their takeover of telent plc.
Pension Corporation (“PC”) consists of three businesses: Pension Corporation
Investments (“PCI”) is essentially a vehicle that can be the sponsor of a pension
scheme; Pension Insurance Corporation (“PIC”) is an FSA-approved insurance
company; and Pension Security Insurance Corporation (“PSIC”) is an approved
reinsurer based in Guernsey.
PC was founded in 2006 by Edmund Truell, who was one of the founders of the
private equity firm Duke Street Capital. It has secured around £1 billion of equity
commitments, primarily to provide bulk buy-out annuity solutions through PIC, which
they expect would allow them to write up to £20 billion of business. However, as
discussed in Section 4, PIC has written very little bulk annuity business due to the
current pricing environment in the market, since they consider premium levels being
offered by some of the other insurers to be uneconomic.
Instead PC has been involved in alternatives to traditional annuity buy-out. Over the
past three years they have been involved in three high profile deals, which have
involved them taking over a sponsoring employer with the primary aim of securing
the attaching pension scheme. Most recently they have been involved in a £400
million takeover bid for telent plc, and there has been considerable speculation over
the implications of such a deal for the company’s pension scheme, particularly since
there is an escrow account associated with the scheme. The Pensions Regulator’s
decision to appoint independent trustees to the pension scheme’s board in order to
safeguard members’ benefits has increased the profile of the deal but despite some
jitters on behalf of PC they proceeded with the transaction. The Pensions Bill 2008
will give the Pensions Regulator stronger powers to appoint independent trustees as
a result of this transaction.
In June 2007 PC acquired Threshers (First Quench) and Thorn and effectively sold on
the underlying businesses to another private equity firm leaving them with economic
exposure to the £93 million Threshers scheme and £1.17 billion Thorn scheme.
In both cases PC are continuing to run the pension scheme as a going concern, with
the aim of improving the funding levels over time, which might ultimately allow PC
to take a refund of any surplus which emerges above the buy-out level. The key
advantage of running the schemes as going concerns rather than buying them out is
that they are governed by pensions legislation rather than the FSA. In the case of
Thorn they have appointed PCI trustees to run the scheme, but the Threshers pension
scheme has retained its own trustees. In both cases investment advice is now
provided through PCI, generating additional income for PC.
41
Buy-Out Alternatives
7. Buy-Out Alternatives
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
It is likely that future deals will be done as and when PC identify an opportunity,
rather than through a competitive pricing process. Indeed PC have now acquired a
significant stake in the listed company Aga Foodservices which has a well-funded
pension scheme which is very large compared to the size of the business.
We can summarise the PC strategy relatively simply. They look to acquire relatively
small businesses which have a large, well-funded, pension scheme. They aim to
dispose of the underlying business (although due to their expertise in private equity
they could continue to run it themselves if they wished) and effectively lock down
the funding position by hedging, whilst generating a small level of investment outperformance using their in-house expertise. At some point in the future (perhaps even
ten or twenty years away) they will have accumulated a significant surplus in the
scheme above the cost of buy-out and then they will trigger a winding-up and collect
the surplus. At this point we would assume that the annuities would be purchased
with their FSA-regulated insurer, PIC, but they would not have to be.
From the sponsoring employer’s perspective, this option is very similar to a traditional
buy-out with an insurance company.
7.2 Citigroup
The acquisition of the Thomson Regional Newspapers’ pension scheme by the multinational bank Citigroup in the summer of 2007 was regarded by many commentators
as a significant development in the market for buying-out pension liabilities. Did
Citigroup really find an innovative new solution which could help companies with
closed pension schemes dispose of their pension liabilities?
To acquire the pension scheme, Citigroup bought the shell company Thomson
Regional Newspapers (“TRN”). The TRN accounts for the year ended 31st December
2006 revealed that the company was a holding company, and did not trade. The
company’s only significant asset was an amount of £46.7 million owed to it by other
companies within the Thomson Corporation, and the only significant liability was also
in respect of an inter-company loan, this time for an amount of £23.3 million. These
loans led to a net asset figure of £23.4 million.
The company is also responsible for a defined benefit pension scheme. This does not
appear directly on the balance sheet since the scheme is not in deficit. The notes to
the accounts reveal that the pension scheme had assets of £218 million and liabilities
of £194 million, leading to an “unrecoverable surplus” disclosed in the accounts of
£24 million at 31st December 2006. This corresponds to a fairly healthy FRS 17
funding level of 112%.
However, the headline FRS 17 figures do not tell the full story. The pension scheme
held annuity policies which were valued at £114 million at 31st December 2006,
corresponding to over half the total asset value. These match the liabilities of some of
the current pensioners, and as such the scheme is not exposed to any investment or
longevity risk in respect of these members. Removing these pensioners and the
corresponding annuity policies from both the assets and the liabilities suggests that
the scheme had assets of £104 million and liabilities of £80 million. Therefore, in
effect the £24 million surplus actually corresponds to a funding level of 130%. This is
probably getting close to the funding level which would be sufficient to secure the
remaining benefits of the scheme through an insurance company buy-out.
By purchasing TRN, Citigroup have assumed full responsibility for the operation of the
pension scheme and, since the Pensions Regulator has granted clearance for the
transaction, the company’s former owners have effectively washed their hands of the
pension scheme, removing all risk associated with it.
42
PSTS Comment:
The Pension
Corporation model is
interesting as it relies
on using the pension
regulatory environment,
which is more flexible
and less onerous than
the FSA regulated
insurance environment.
However, the PC model
as it stands only applies
to an elite band of
sponsoring employers
and schemes – small
companies with very
large, well-funded
pension schemes. It
therefore has limited
applicability.
Additionally PC have yet
to design a mechanism
that allows them to own
the pension scheme
assets without taking
over the company – an
approach which will not
appeal to many.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
The sponsor of the scheme will continue to be TRN. Therefore, it will be TRN that the
trustees will turn to if they require any additional contributions in the future. The only
change due to the transaction is that TRN will now be a subsidiary of Citigroup, rather
than a subsidiary of the Thomson Corporation. We understand from regulatory filings
made in the US that Citigroup has not provided any guarantees to TRN and hence to
the pension scheme.
PSTS Comment:
The pension scheme itself is unchanged by the transaction. The trust deed and rules
governing the pension scheme will not change, and the trustees will still have the
same level of responsibility, whereas in the case of a buy-out with an insurance
company the trustees obtain a “clean-break”.
However, in the case of
TRN, Citigroup acquired
only a shell company
rather than an operating
business which would
severely restrict the
amount of cases where
this would be
applicable.
The trustees themselves will also be the same, except that Citigroup have the power
to add a company-appointed trustee to the board. Therefore, the trustees will
continue to have control over the operation of the scheme, including the investment
strategy, and they will only be required to consult with TRN. Although the companyappointed trustee will be able to make suggestions to the other trustees, there will
be no requirement for the trustees to heed this advice.
This deal differs from a traditional insurance company buy-out since Citigroup will be
transferring the investment and longevity risk onto its own balance sheet, so will be
governed by pensions rather than insurance regulation. This would allow Citigroup to
adopt a more adventurous investment strategy than an insurance company would be
allowed to, although it is likely that the investment strategy will still be fairly prudent.
However, they will require trustee support before making any changes to the
investment strategy.
Citigroup has indicated that it plans to use a liability-driven investment approach
to reduce the scheme’s exposure to investment risk. Currently, around half of the
scheme’s assets are held as annuity policies, probably to match the liabilities of
current pensioners. Most of the remaining assets are held in gilts, with around 13%
of the assets being held in equities. This appears to be a fairly prudent investment
strategy, since the value of the annuities and the gilts should correspond closely to
the value of the liabilities as calculated on the FRS 17 basis. The equity holding is
broadly comparable to the amount of the surplus, so could be considered to be
unmatched, and used to maximise expected investment returns.
The appeal of the deal to Citigroup relies upon their ability to maintain the surplus
within the pension scheme, and eventually tap into it. Although it is fairly difficult to
get access to a surplus within a pension scheme at the current time, since the
benefits of all members must be secured with an insurance company before any
surplus can be reclaimed by the sponsor, Citigroup should be able to take a long-term
view on the release of surplus. The recent deregulatory review also proposed making
it slightly easier to remove surplus funds from pension schemes, so it is possible that
it may become easier for Citigroup to remove the surplus in the future.
This is a near identical strategy to that being followed by PC, as discussed in the
previous section.
Although there has been considerable press speculation suggesting that Citigroup
may be about to become a major player in the UK pension buy-out market, this
seems to be fairly unlikely. It appears that they will only be interested in closed
pension schemes which have a very mature membership, since they are likely to want
to limit their exposure to longevity risk. Indeed, they have since stated that they do
not intend to start getting involved in tender processes, instead preferring to take
advantage of opportunities in the market as and when they appear. As such, they will
43
The Citigroup/TRN
transaction seems very
similar to the Pension
Corporation model.
In any event Citigroup
have yet to undertake
any further transactions
and are indicating that
they will only undertake
transactions on an
opportunistic basis
rather than attempting
to build a large scale
business in the pensions
market.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
probably be fairly similar to Pension Corporation, who use a very similar strategy in
their deals, although to date Citigroup have been less active.
7.3 Occupational Pensions Trusts
Occupational Pensions Trusts Ltd (“OPT”) also aims to provide a cheaper alternative
to companies or trustees seeking to extinguish their pension liabilities. Launched in
September 2007, OPT has been established by a combination of individuals from an
investment and pensions background. They include the former Chairman of the
National Association of Pension Funds, Robin Ellison, as well as a number of directors
from Telereal, a company established in 2001 that provides property services to BT
and O2, amongst others.
Under the OPT approach a shell company is established by the current sponsoring
employer and a new occupational scheme created within this to take over the
pension responsibilities. The scheme is then bulk transferred using GN16: Retirement
Benefit Schemes – Transfers Without Consent into this new scheme, with a new Trust
Deed and Rules. As part of the transfer an injection of funds is paid into the scheme
by the current sponsoring employer, to compensate the trustees for the weakened
employer covenant provided by the shell company. The new company is then sold to
OPT for a nominal sum (£1).
The OPT model requires no external capital; all of the additional cash required to
bring the funding of the scheme up to the level required by the trustees comes from
the original sponsoring employer. OPT suggest that the post-transaction funding
level will be circa 120-130% of the FRS 17 liabilities. In practice, however, this
will be highly scheme specific, depending on the covenant of the original
sponsoring employer.
In terms of management of assets, OPT plans to adopt the “Yale Model” of
investments. They intend to take a multi-asset investment approach similar to that
undertaken by large US endowment funds such as Yale and Harvard, where there are
a number of asset classes (usually in excess of 5). Each asset class is intended to be
highly diversified and exhibit little positive correlation with the other asset classes.
The resulting level of risk adopted is greater than if the funds were invested in gilts,
but lower than for equity investments. The expected investment returns exceed that
expected from corporate bonds, but with a similar standard deviation of returns.
In transferring the pension scheme from the original company to OPT, the trustees
are exchanging one employer covenant for another. They are exchanging the
covenant of a going concern company for that of a shell company with no assets.
The assets held in excess of 100% of the FRS 17 liabilities are broadly the value of
the “promise” that the sponsoring employer will pay its obligations over the lifetime
of the liabilities, likely to be around 80 years for most pension schemes.
Companies with a strong covenant will therefore be required to make larger
payments to the pension scheme than weak companies, since transferring the
scheme to a shell company will lead to a larger weakening of the employer covenant.
However, it is precisely these companies that are best able to run off their pension
schemes and guarantee the payment of their pension liabilities.
The trustees might also have difficulty determining how much money the sponsoring
employer should contribute to the pension scheme as compensation for the
reduction in the employer covenant.
Determining the value of the initial employer covenant can be a very complex and
subjective task. However, given that the value of the employer covenant which will
44
PSTS Comment:
The OPT model is
completely different
from all other players
in the market in that it
does not bring any new
capital in to support the
pension obligations.
Effectively the trustees
of the scheme give up
their ability to call on
the company in the
future in return for a
lump sum payable
immediately – similar
to a compromise of a
“Debt on Employer”.
OPT provide a structure
to complete the
compromise and future
management expertise,
but no capital.
OPT is likely to be a
niche product, priced at
just under the cost of
buying-out, and reliant
on a compromise being
possible. Even where a
compromise is possible
trustees may well
consider some of the
other options available
to be more attractive,
and thus we expect
OPT to generate
extremely limited
amounts of business.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
be offered by the new shell company is zero, the trustees might require the cash
injection to be large enough to allow them to be self-sufficient, since they will not be
able to rely on the sponsoring employer being able to make good any deficits which
emerge in the future. (The Pensions Regulator is also likely to insist on self-sufficiency
in order to grant clearance to the transaction.) This is likely to lead to the trustees
de-risking their investment strategy, probably switching to gilts or other low-risk
investments, which would reduce the discount rate used to value the liabilities.
This might lead to the value placed on the liabilities being close to the buy-out level,
in which case the sponsoring employer would probably prefer to buy-out the
liabilities entirely.
Further, even if the trustees and the Pensions Regulator could be satisfied with
funding the pension scheme to a level significantly below the buy-out level, there
are alternative solutions that would be more attractive to the trustees. As has been
demonstrated by the recent acquisition by Citigroup of the Thomson Regional
Newspapers’ pension scheme (discussed in Section 7.2) and the Thorn transaction
with Pension Corporation (discussed in Section 7.1) a scheme funded at 120-130%
of its FRS 17 liabilities would be an attractive target for some players in the market.
Rather than receiving a nominal sum for a well-funded scheme the company may
be able to acquire a significant payment from one of these players.
Since the arrangement involves a company sale, the original company remains a
“connected party” to the scheme. This means that the company retains the risk of
being issued with Financial Support Directions for up to one year, and Contribution
Notices for up to six years, even if the Pensions Regulator grants clearance. This is
because the Pensions Regulator grants clearance based only on a specific set of
circumstances and does not provide blanket clearance to cover all and every future
circumstance. The Pensions Regulator may not look favourably on the deal if OPT
runs into difficulties in future, particularly since OPT do not put up any capital as
part of the deal. Therefore, if OPT is looking likely to default on their pension scheme
liabilities, and the original sponsoring employer is in fine health, the Pensions
Regulator might seek to get the original sponsoring employer on the hook for the
pension liabilities.
OPT are targeting companies with pension schemes which are closed to future
accrual and have liabilities between £10 million and £500 million. OPT have not been
involved in any deals to date, but have indicated that they have had preliminary
discussions with a number of clients.
7.4 Further Innovations
We are aware of a number of other innovative solutions that are currently seeking
funding and may add to the range of options available. However, the “killer
application” which offers a significantly cheaper alternative to annuity buy-out has
yet to appear. Due to the huge market potential and the Pensions Regulator’s
encouragement of the development of innovative products that help him meet his
statutory objectives and bring new capital to bolster the security of pension promises,
it is likely that more new products will continue to come to the market.
45
PSTS Comment:
The alternative
solutions for removing
liabilities from the
balance sheet show the
effort being undertaken
by market participants
to find innovative
solutions for the
perceived value gap
between the long-term
cost of pension
promises and the
annuity market pricing.
However, the solutions
available to date are
only applicable in niche
situations.
Market innovation may
at some point bring
forward a solution with
wide applicability that is
cheaper than annuity
buy-out, but at the
moment such a product
has yet to appear.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
8. Alternative De-Risking Solutions
There is also a range of products which have recently been launched in the UK
pensions market which aim to significantly reduce the risks associated with running
a defined benefit pension scheme, without severing the link between the sponsoring
employer and the pension scheme. This section describes some of those products.
8.1 PensionsFirst
PensionsFirst was established in 2007 by Amelia Fawcett, formerly of Morgan Stanley,
with the aim of bundling up certain pension scheme risks into capital market
products that can be traded by investors and could eventually develop into a new
asset class for hedge funds and other market participants to trade in. We understand
that although they have secured significant seed capital to get the business
established, they are currently fundraising to allow them to sell their first products
to pension schemes.
PensionsFirst think their capital market products could address all the risks involved in
managing a defined benefit pension scheme. Their range of products can be tailored
to be scheme specific, and are designed to protect against credit, market and
longevity risks. It is intended that schemes will be able to choose which risks to
protect against (ranging from a single risk, such as longevity, to complete protection
so that all future cash flows are matched) and what section of their scheme to
protect (for example, 50% of the scheme, or only the deferred pensioners).
PensionsFirst will provide pension schemes with a bond which is an asset of the
scheme. The bonds are named using different colours depending on the benefits they
offer. The most comprehensive level of cover is provided through the “Blue Bond”,
which covers all future scheme benefits, and PensionsFirst describes it as being
“economically equivalent to buy-out”.
In exchange for the scheme assets plus a premium, PensionsFirst set up a Jersey
based special purpose vehicle (“SPV”). This will issue a bond that becomes a scheme
asset, whose coupons will exactly match the benefit outgo of the scheme.
PensionsFirst will add capital to the SPV to ensure it attains a AAA credit rating,
although it is anticipated that this capital element will be relatively small (less than
5%). PensionsFirst aims to maintain the AAA rating throughout the term of the
bond, adding capital to the SPV or reducing investment risk as required. Each SPV
will be ring-fenced, so that it cannot be used to support investments from other
pension schemes.
The assets held will vary depending on the scheme, but are likely to be predominantly
fixed interest bonds and bank deposits. Increasing the holding of bank deposits could
be used to maintain the AAA rating. They are intended to be highly diversified.
The make-up of the bond will be fairly complex. One part of it will be a “longevity
master cell”. PensionsFirst hopes to be able to combine the longevity master cells of
all the schemes it takes on to spread the risk. They plan to repackage this longevity
risk so that they can sell it to the capital markets in tranches. They think one of the
reasons the longevity market has been slow to develop is that the capital market is
not interested in taking on long-term longevity risk, but may be interested in
products offering exposure to longevity risk over the short or medium term. They
could therefore sell these risks on, maintaining only the tail risk, although since the
most significant longevity risk is in the tail this may not reduce PensionsFirst’s
exposure significantly.
46
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
The sponsor will retain the scheme’s liabilities on its balance sheet, although
PensionsFirst hope it will be possible to disclose the asset value as exactly matching
the liabilities, so that no accounting or funding deficit is ever disclosed.
Benefits covered by the products offered are summarised briefly below. PensionsFirst
intends these products to show the versatility of the products available, and further
variations would be possible. They expect schemes may be interested in buying cover
for small elements of the scheme initially, incrementally increasing the cover over
time as it becomes more affordable.
●
●
●
●
●
●
●
●
Blue Bond – all scheme benefits
Term Blue Bond – all scheme benefits for an agreed term
Light Blue Bond – defined proportion of scheme benefits
(e.g. 50%, or pensioners only)
Geared Blue Bond – can cover same benefits as any of the above products, but
paid for over defined time period (so designed for schemes in deficit)
Green Bond – all scheme benefits for a defined term, then scheme returned to
sponsor at a pre-defined funding level, e.g. 110% of IAS 19 liabilities
White Bond – all scheme benefits with a fixed inflation assumption used, intended
for schemes that already have an inflation swap overlay in place
Dynamic investment bond – same as Blue Bond (or a variation thereof) but
amount paid to PensionsFirst dependent on a defined stock market index
Risk share solution – all scheme benefits, but sponsor can share in longevity risk,
for example by investing in the longevity notes produced through the “longevity
master cell”
For a hypothetical “average” scheme, PensionsFirst calculate the cost of the Blue Bond
to be circa 155% of the IAS 19 liabilities for deferred pensioners. They have been
advised that the buy-out cost for these members would be circa 160%, although this
is likely to have fallen slightly over the past few months due to increased
competitiveness in the buy-out market.
For pensioners, the price quoted was 123% of IAS 19 liabilities, although the
comparative buy-out price for this cohort of hypothetical members was not known.
Although these prices do not suggest that PensionsFirst will be significantly cheaper
than buy-out, PensionsFirst are keen to stress that they are not targeting the buy-out
market. They expect to target schemes who are interested in protecting a section of
their scheme for a specified period of time, and consider their premiums to be more
attractive for such deals. For example, to cover the benefits of current pensioners for
the next 25 years would cost 108% of the IAS 19 liabilities, or for deferred pensioners
the cost would be 127% of their IAS 19 liabilities to cover their benefit payments for
the next 40 years.
Since PensionsFirst describes their Blue Bond as being “economically equivalent to
buy-out”, it seems logical to compare the two options. A key disadvantage of the
Blue Bond is that the trustees of the pension scheme would maintain ultimate
responsibility for paying the benefits, unlike in a buy-out. This is likely to significantly
47
Alternative De-Risking Solutions
Ignoring the possibility of selling some of the longevity risk to the capital markets,
PensionsFirst’s model suggests that they could cope with future longevity being in
line with the “00” mortality tables, with allowance for the long cohort effect with a
minimum improvement of 2% per annum. They have not revealed the actual
assumption they will use when pricing.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
decrease its appeal to trustees. It also means that the management costs of operating
a defined benefit pension scheme are not removed, and regulatory risks remain.
It also means that the expenses of running the scheme would still be payable by the
sponsor, and, with the exception of investment expenses, these are unlikely to reduce
significantly, since administration and actuarial services (including triennial
valuations) will still be required. This may increase the cost of the Blue Bond by 3% or
more, effectively removing the price benefit relative to buy-out.
It might also be expected that a buy-out solution would provide more security, since
theoretically the Blue Bond’s rating could fall, or it could even default. However,
PensionsFirst have stated that whilst their bonds are unconditional obligations to pay,
insurance policies are conditional obligations to pay subject to there being no
defences available, e.g. failure by the insured to disclose any information considered
to be material. PensionsFirst therefore think that although insurance policies might
be appropriate for bulk buy-outs, so that the liability is to each individual member
(where the PR implications of refusing to pay a pensioner could be damaging), they
may be less appropriate as scheme assets.
It could therefore be concluded that PensionsFirst will not provide significant
competition to the buy-out market. However, the product range may be more
attractive to schemes looking to only cover a selection of their liabilities.
Unfortunately it also remains to be seen how the bond will be recorded in the
sponsor’s accounts. Although PensionsFirst state that the assets should be disclosed
as exactly matching the value of the liabilities, this will need to be agreed by
individual scheme auditors. For example, if the credit rating of the bond deteriorates
to less than AA, it seems likely that an accounting deficit would need to be disclosed
to make allowance for the risk of default on the bond. Even if the credit rating is still
AAA, auditors may not be comfortable simply matching the assets to the projected
liabilities, since the bond will be scheme specific and hence will not be a tradable
commodity, so that mark-to-market pricing will not be possible. Auditors may insist
on looking through to the surrender value (that is the bid-value) of the bond to value
the underlying assets.
There is also uncertainty regarding how the credit rating of the bond will be assessed.
Whilst the likes of Moody’s and S&P have experience of estimating the probability of
entities facing insolvency over the duration of bonds of up to 20 years, they have
little experience of rating complex longevity linked bonds which have 80+ year terms.
The recent criticism of credit rating agencies in relation to CDOs would be likely to
increase concern over this area, particularly given the complexity of the bond which
will be set up by PensionsFirst.
PensionsFirst are currently aiming to put £250 million of capital in place, which would
allow them to write around £15 billion of business. The issue of longevity notes would
allow them to raise more capital thereafter. They are still in the process of explaining
their product to advisers, and plan to launch their formal marketing at some stage in
2008. They are currently working with two pension schemes who have expressed an
interest in their product.
Although the PensionsFirst products are interesting, in our view there would be a
limited number of scenarios when their products might be attractive. Depending on
the objectives of the pension scheme and the sponsor, we would consider either
running the pension scheme as a “closed fund” or buying-out the benefits with an
insurer would be more attractive options in most cases.
48
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
8.2 Tactica
PSTS Comment:
Tactica are another company aiming to offer a product which they hope will appeal
to trustees and scheme sponsors looking to reduce their exposure to risk without
going as far as a bulk buy-out. They gained FSA authorisation in November 2007
and are based in Gibraltar, partly for tax purposes. They will outsource most of their
calculations work to Pension Capital Strategies (part of Jardine Lloyd Thompson).
We struggle to see the
appeal for Tactica’s
product. The only
difference between
investing the scheme
assets in a cash flow
matched portfolio of
bonds for ten years and
investing in Tactica’s
product is that Tactica
provides a guarantee on
the investment return
(but also caps this
return below what
might be reasonably
expected from a
portfolio of bonds over
this period) and hedges
the limited fluctuations
in experience that
would be observed
during the life of
the policy.
From the perspective of pension scheme trustees and sponsors, the product offered
by Tactica is very similar to the “Term Blue Bond” offered by PensionsFirst, but with
a few technical differences in the underlying structure. Rather than offering pension
schemes capital market products, they will instead provide them with an insurance
policy. They will underwrite an agreed proportion of the scheme’s liabilities for the
life of the policy, and guarantee to:
●
●
pay the cash flows arising from the insured liabilities; and
return assets at the end of the policy equal in value, on a defined basis,
to the remaining liabilities.
The basis used to value the liabilities at the end of the period will be agreed when
the contract is established. This basis will also be used to value the liabilities for
accounting purposes throughout the policy term. In exchange for this insurance the
trustees or the scheme sponsor will need to transfer the value of the liabilities, as
calculated on the agreed basis, to Tactica at the start of the policy. This can either
come through the pension scheme or be paid by the sponsor. There are also options
available to pay part of the premium over the lifetime of the policy if the pension
scheme is under-funded on the agreed valuation basis.
Tactica typically expect the term of the policy to be around 10 years, since they
consider this the period of concern to most pension schemes and sponsors, although
they would be happy to provide quotes for other timeframes. They would also be
happy to “roll over” the contract before expiry if the client was interested in this.
The funds received will be passed to Goldman Sachs, along with an additional 25% of
Tactica’s shareholders’ capital, and held in a protected cell structure. Goldman Sachs
will invest the funds in a diversified range of asset classes, which might include global
equity markets and possibly funds of hedge funds. The out-performance of the assets
above the level required to pay the cash flows throughout the period and the
liabilities at the end of the policy term will result in profits being generated for Tactica
and Goldman Sachs. In exchange, Goldman Sachs will offer a derivative product
guaranteeing 100% of the value of the liabilities. Tactica believe that because their
product is supported both by capital provided by Tactica and by the minimum
guarantee provided from Goldman Sachs’ balance sheet, pension scheme trustees
will view it as secure.
It will also be possible to include allowance for profit sharing in the contract, so
that the sponsoring employer can share in gains or losses resulting from experience
differing from expected. The client could also choose to participate in any
investment upside.
Tactica will not accept any selection risks as part of the policy. For example, if the
pension scheme offers generous early retirement terms on redundancy, Tactica would
not be prepared to cover such liabilities under the contract. Otherwise they would be
exposed to company actions which could increase the cash flows Tactica are required
to pay out, as well as the amount of assets transferred at the end of the policy.
49
There is no protection
provided for changes in
mortality expectations
during the life of the
policy, which is likely
to be a key concern
for trustees.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Tactica have indicated that the yield they would use to price the liabilities is likely
to be slightly above the long-term gilt yield, for example gilts plus 20 basis points.
However, they have also suggested the price will be dependent on the size of the
pension scheme. They do not consider the longevity risk over a period of 10 years to
be significant, so expect that in most cases they will be able to adopt the mortality
allowance used in the company’s accounts in the contract’s valuation basis. The main
difference between the contract’s valuation basis and the scheme’s ongoing valuation
basis is therefore likely to be in the discount rate used.
At a level of pricing of around gilts plus 20 basis points it is difficult to see the benefit
to the pension scheme of taking the insurance. Instead, the scheme could invest in a
diversified portfolio of government bonds and high quality corporate bonds (for
example AAA), which have very low levels of credit risk. With this portfolio it should
be possible to match the level of return allowed for in Tactica’s pricing, or possibly
exceed it. It should also be possible to structure the bonds so that they provide a
reasonable match for the pension liabilities, and pay coupons which could closely
match the scheme’s cash flows. Although Tactica’s product would offer exact
matching rather than close matching, we would expect the level of matching
provided by a cash flow matched bond portfolio to be sufficient for the needs of
most pension schemes.
It is also debatable whether trustees or scheme sponsors will be interested in
obtaining the limited level of protection provided by Tactica’s policy. The liabilities
paid out at the end of the policy will be calculated on the same basis as used at the
start of the policy. This means that, for example, the mortality assumption used
remains unchanged. Whilst the pension scheme is protected from mortality diverging
from expectations during the lifetime of the policy, they are not protected from
changes after the end of the period. This means that if mortality expectations change
significantly during the term of the contract, requiring the trustees to re-assess their
mortality assumption, this change will not be reflected by the payment which the
pension scheme receives at the end of the contract. Therefore, this could lead to a
deficit existing in the scheme when the assets are returned at the end of the policy.
When the trustees change their mortality assumption, under the Scheme Specific
Funding legislation the company may also be required to remove this theoretical
deficit, possibly resulting in the company paying premiums to Tactica and deficit
contributions simultaneously.
Tactica hope to raise initial capital of £250 million, which would allow them to write
up to around £1 billion of business, but they have not yet managed to raise this level
of committed funds. They hope to take their level of capital up to £1 billion over time
if they are successful. They intend to target a small number of large pension schemes,
ideally with liabilities exceeding £100 million, and due to the structure of their
product they do not expect to be directly competing with companies offering
buy-out.
8.3 PensionsRisk
PensionsRisk is a joint venture between the Beachcroft Regulatory Consulting and
a specialist firm of independent reinsurance advisers and arrangers and consulting
actuaries, Paterson Martin. It is headed up by Andrew Campbell-Hart, and they will
act as an insurance intermediary, regulated by the FSA. They would seek to place the
cover in the market on behalf of the scheme sponsor and trustees.
PensionsRisk will offer an insurance contract between the scheme sponsor, the
trustees and PensionsRisk. The non-accruing pension scheme liabilities covered by the
50
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
contract will be transferred to the insurer for the term of the contract (of 10 years,
say), and PensionsRisk promises to pay the cash flows for the term of the contract. If
a deficit exists within the pension scheme at the start of the contract, the sponsoring
employer commits to remove it. Scheme administration will also remain the
responsibility of the trustees.
At the end of the contract PensionsRisk guarantees to return assets equal in value to
the liabilities. As such, the contract is very similar to that offered by Tactica. However,
one key difference is that when calculating the value of the liabilities at the end of
the contract, PensionsRisk will update the mortality assumption “by reference to an
objective and independent benchmark”. The PensionsRisk insurance product therefore
provides some level of protection against the risk of mortality expectations changing
during the term of the contract, unlike the Tactica product, which only provides
protection against mortality altering the cash flows payable during the term of
the policy.
However, an objective and independent benchmark of future mortality expectations
does not currently exist, and it seems unlikely that such a benchmark will exist in
future. Therefore, there is a risk to trustees that the measure used by PensionsRisk to
assess their mortality assumption is not consistent with the assumption trustees will
adopt for use in their ongoing valuation.
In common with the Tactica product, PensionsRisk also intend to offer various options
for risk or profit sharing. The annual premium payable for the cover is expected to be
within a range of 0.5% - 1.0% of the value of the liabilities covered, although this will
depend on the scheme’s characteristics, as well as any risk or profit sharing.
PensionsRisk was launched in early 2007, and have yet to conduct their first deal.
8.4 BrightonRock
Although BrightonRock do not intend to formally launch their product to the UK
pensions market until the summer of 2008, it is of interest since it appears to offer
an alternative approach to managing the risks of operating a defined benefit
pension scheme.
BrightonRock will be based in Malta, although they will still meet the UK’s FSA
requirements. They expect to raise equity capital of £600 million, giving them a AAA
rating profile. They intend to target pension schemes of between £30 million and
£3.5 billion in size.
BrightonRock’s approach focuses on what they consider to be the fundamental risk of
running a defined benefit pension scheme: employer insolvency. They plan to provide
insurance to a pension scheme to protect against this risk. BrightonRock can therefore
be seen as a competitor to the PPF that insures the full benefits payable by the
pension scheme rather than the reduced level of benefits provided by the PPF.
BrightonRock considers that most of the risks of operating defined benefit pension
schemes arise as a consequence of the uncertainty surrounding the future of the
sponsoring employer. Pension schemes must allow for the possibility that they may
need to continue long after their sponsoring employer has become insolvent. As a
result they must target a 100% funding level on a prudent set of assumptions
(increasingly prudent as a result of the Statutory Funding Objective introduced under
Scheme Specific Funding), and any deficit must be removed over as short a time
period as possible, paying only limited regard to the effect these cash flows could
have on the sponsoring employer’s probability of insolvency. Deficits usually arise at a
51
PSTS Comment:
PensionsRisk offer a
very similar product to
Tactica. Although they
remove the risk of
mortality expectations
changing during the
term of the contract,
we still consider their
product to have a very
limited appeal.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
time when equity values have fallen, so that contributions are required when the cost
of capital for scheme sponsors is high.
Therefore, BrightonRock does not consider it to be optimal for pension schemes to
target a funding level of 100% on their ongoing basis. Instead, they believe that the
optimal funding level is usually somewhere between 85% and 95%, depending on
the specific characteristics of the scheme.
From the perspective of pension scheme trustees, as long as the solvency of the
employer is guaranteed they should not be concerned about any other risks, such as
longevity and investment risk, since there will always be an employer to support the
pension scheme and pay any unexpected costs which arise in future.
BrightonRock will guarantee to pay the scheme’s pensions through its EU-regulated
insurance company in full should the sponsor default. This provides a higher level of
cover than the PPF, which broadly speaking covers around 75-90% of benefits. If the
sponsor defaults, BrightonRock will top up the funding level of the pension scheme to
enable it to buy-out its liabilities through BrightonRock’s life assurance arm.
The cover will be perpetual, with the premium fixed (relative to the liabilities) for the
life of the scheme or sponsor (whichever ceases to exist first). The premium is
expected to be around 0.1-0.5% of the FRS 17 liabilities (as calculated for disclosure
in the company’s accounts) and BrightonRock will be unable to refuse to renew the
contract. It will also be possible to offset any elective bulk annuity purchase cost with
the premiums paid to date should the trustees or scheme sponsor want to wind-up
the scheme provided that such a buy-out is with BrightonRock’s annuity company.
On the other hand, the pension scheme can choose to cancel the cover at any time
by discharging the liabilities. For example, the trustees could transfer the liabilities
(under GN16) to a new pension scheme if they wished the cover to cease. This
ensures that BrightonRock should continue to maintain their AAA rating, since if this
falls their customers could walk away. This is not the case for some of the products
described elsewhere in this report.
BrightonRock are likely to require the pension scheme is funded to at least 85% of
their FRS 17 liabilities at inception. They will also impose various warranties and
covenants on both the sponsor and the trustees. For example, they may require the
scheme to target a specified funding level, which BrightonRock determines to be
optimal. For most schemes BrightonRock expects this to be 100%, but in some
circumstances they might consider a lower target to be appropriate, such as
somewhere between 85% and 95%, if the opportunity cost of scheme contributions
is high.
There are also likely to be covenants regarding the credit worthiness of the sponsoring
employer. However, a breach of a covenant will not automatically trigger cancellation
of the policy; instead it may trigger an additional payment being required from the
sponsoring employer to BrightonRock, or a renegotiation of the terms of the contract.
However, BrightonRock have made it clear that there are a number of areas they do
not want to get involved in. For example, they do not intend to impose restrictions on
the investment strategy of the pension scheme, or the ongoing management of the
scheme. (However, they may require disclosure of the investment strategy to allow
them to hedge their position in certain circumstances.) They also stress that they do
not want to get involved in the corporate affairs of the sponsor, other than to the
extent that covenants are breached.
The BrightonRock contract would be an asset of the pension scheme. It would be
valued under International Accounting Standard 39, and should be a contra-cyclical
52
PSTS Comment:
BrightonRock are offering
a truly innovative product
that covers the one risk
that trustees face which
could result in pensions
not being paid in full;
sponsor insolvency. As
BrightonRock will allow
the scheme to be funded
to less than 100% of their
FRS 17 liabilities at outset
and for the assets to
remain invested in equities,
this product could appeal
to the significant segment
of the pension industry
that still wishes to take
equity risk and use the
expected excess returns on
equities to bridge their
funding gap.
With a BrightonRock policy
in place the trustees
should be a lot more
relaxed on both
investment risk and the
degree of prudence
required in their funding
valuation.
However, the
BrightonRock policy is
a new approach and
trustees are likely to
be concerned as to
whether BrightonRock will
be around in (say) thirty
years’ time should their
sponsor become insolvent
in a time of economic
distress. Due diligence on
BrightonRock will be a
major exercise and trustees
will wish to have as much
comfort as possible on the
BrightonRock business
model.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
asset; increasing in value as the size of the scheme deficit increases or the sponsor’s
creditworthiness worsens. It should therefore help to reduce the volatility of the
sponsor’s balance sheet.
BrightonRock have run models to predict the probability of their own insolvency.
These models suggest that in the worst case scenario the probability of insolvency
is highest in their third year of operation, when it is 4 in 10,000.
BrightonRock hope that their product will be attractive to trustees since it will be
cheaper than the alternatives, and would still allow trustees significant investment
freedom. It would allow them to take a more risky investment strategy, hence
increasing expected returns and it would significantly reduce the volatility of the
pension scheme on the sponsoring employer’s balance sheet.
As discussed above, they will require that pension schemes have assets which cover
at least around 85% of their liabilities (as measured on the company’s FRS 17 basis)
at inception. However, this is a significantly lower level than would be required to
buy-out the liabilities with an insurer, and most pension schemes in the UK are likely
to already be funded to at least this level.
In order to be able to offer the cover for 0.1-0.5% of the FRS 17 liabilities,
BrightonRock have indicated that they will require “investment grade” sponsor
creditworthiness. This corresponds to a default likelihood of up to 1% per annum
(based on BrightonRock’s default assessment), but does not necessarily require a
published credit rating. BrightonRock considers around 70% of the pension schemes
in the UK to be associated with a sponsoring employer with at least this level of
creditworthiness. However, the pension schemes in the UK that do not have a
sponsor with this level of credit worthiness would be likely to see their quoted
premium rise above 0.5% of their FRS 17 liabilities. This could make the cover
prohibitively expensive for these schemes.
BrightonRock also think that their contract may be useful in corporate transactions,
particularly in situations where pension scheme trustees are concerned about the
impact on the security of benefits after a highly leveraged deal. For example, in
private equity transactions BrightonRock have suggested that they might take an
equity share in the business in exchange for a weakened employer covenant, rather
than significantly increasing the premium payable.
BrightonRock are currently in discussions with the Pensions Regulator regarding their
product, since they feel PPF levies, Section 179 requirements and clearance
applications should be unnecessary. However, the Minister of State for Pensions
Reform, Mike O’Brien, has recently denied BrightonRock’s request for exemption from
the PPF levies for schemes purchasing their product. Although BrightonRock hope to
be able to reduce the risk based levy payable to zero, by using their contract as a
contingent asset for the scheme, this means that the scheme based levy will still be
payable. Although this is likely to be relatively small (at the time of writing estimated
to be 0.0152% of the scheme’s Section 179 liabilities for the 2008/09 levy payment),
the trustees and scheme sponsor may resent feeling that they are double-paying for
cover. (However, on the other hand trustees may prefer to pay the scheme based levy,
as it would still leave the scheme eligible for PPF cover, to protect against the
insolvency of BrightonRock.)
On the face of it, the BrightonRock product does not appear to be particularly
complex or hard to replicate. However, BrightonRock have spent considerable time
collating the necessary data to analyse the insolvency probabilities of the sponsoring
employers, and as such they expect it to be around two years before other
competitors are able to replicate their product.
53
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
It is difficult to estimate the impact such a product will have on the UK pension buyout market until it is formally launched. On face value the product looks like it could
be attractive to a range of pension schemes. However, this will depend on:
●
●
●
BrightonRock being able to keep the premium affordable (and at the suggested
level of 0.1-0.5% of FRS 17 liabilities it probably would be);
the covenants imposed on the sponsoring employer not being too restrictive; and
BrightonRock being able to demonstrate to pension schemes that its own business
model is sound, so that the probability of their own insolvency is low enough that
it can be considered to be negligible.
This last point is likely to be the most significant. Schemes are effectively paying a
premium to BrightonRock to enable them to add BrightonRock’s employer covenant
to that of their own sponsoring employer. For this to be of significant value, they
must be convinced that BrightonRock can offer them long-term security.
Finally, even at the suggested premium levels it is unlikely that the BrightonRock
product will appeal to all. Some schemes and sponsors will consider the premiums
(which equate to a net present value of around 8% of the FRS 17 liabilities) to be
higher than they are prepared to pay.
54
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
A further de-risking alternative to the solutions described in Section 8 would be to
hedge out investment related risks through a so-called Liability Driven Investment
(“LDI”) solution and then hedge the only significant remaining liability risk, longevity
risk, through a mortality hedging product. Although mortality hedging could simply be
considered to be an alternative de-risking solution to buy-out, in common with the
products discussed in Section 8, we discuss it separately since longevity risk is one of
the most significant risks faced by the trustees of UK pension schemes. A discussion
of the recent improvements in longevity is included in the appendix.
The first UK Government bond with a longevity component were the “Tontines”,
issued in 1693 to help fund the war with France. However, these were of limited
success, since holders of the bonds soon realised that they could increase the size of
their annuity payments by killing off other bond holders! Fortunately mortality
products have developed significantly in the intervening 315 years.
9.1 Introduction
Some life insurance companies have a (limited) natural hedge against longevity
improvements, due to the assurances which they sell. This is because, if people live
longer than expected, they will pay out less on assurances, which might help to offset
the increased cost of paying pensions for longer. The extent to which these effects
offset each other will depend on the balance of assurance and annuity business
written by the insurer and also the mixture of people buying assurances and
annuities. However, this does not apply to pension schemes and therefore the
possibility of their pension scheme members living longer than expected provides
a significant risk.
Some players in the financial markets have recently developed mortality products in
an attempt to make longevity risk a tradable commodity. For example, JPMorgan (in
conjunction with Watson Wyatt and the Pensions Institute at the Cass Business
School) have developed the LifeMetrics Longevity Index, whilst Credit Suisse, Natixis,
Swiss Re and PensionsFirst have developed bespoke mortality products. These
initiatives are designed to parcel up mortality, so that it can be actively traded,
allowing pension schemes the opportunity to hedge their longevity risk.
As discussed in earlier sections, a scheme can remove longevity risk completely by
buying-out their liabilities with an insurer. However, recently companies have
attempted to create financial products which link mortality and cash flows out of the
scheme. In the past, bonds were issued whose coupon payments were matched to
mortality indices. A 3-year bond was issued by Swiss Re in 2003 and BNP followed a
year later with a 25-year bond, which aimed to produce cash flows designed to
replicate those of a pension scheme based on a pool of 65-year olds. The efforts over
the last year or so by the likes of JPMorgan and Credit Suisse reflect a renewed
attempt to hedge longevity risk in a more scheme specific manner.
Whilst the Office of National Statistics compiles annual records of mortality
experienced across the population, actuarial mortality tables – those used to calculate
the net present value of annuities – are produced far more infrequently. For these
figures to be accurate significant periods of time need to be examined, making them
costly and time-consuming to produce and update. For instance, even the data
underlying the Self Administered Pension Schemes (SAPS) tables, which is currently
deemed to be the most statistically credible dataset for calculating the “where are we
now” view of mortality, was collected between 2000 and 2006. It covers 200,000
55
Mortality Hedging
9. Mortality Hedging
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
male deaths and 150,000 female deaths but this is still not considered to be a long
enough period to estimate credible rates of improvement (the “where will we
be” view).
It is important to consider the limitations of the data when it is being used. For
example, it is unlikely to be appropriate just to apply the mortality rates derived from
the data to a sample population (such as a pension scheme), since the mortality rates
will vary depending on the characteristics of the lives in question. Therefore, the
bonds initially issued by Swiss Re and BNP, which were designed with an even more
specific mortality index (a pool of 65-year olds) in mind, were not a good hedge for
a pension scheme’s liabilities.
In order to improve the accuracy of the mortality assumptions used for a pension
scheme, it is necessary to analyse the mortality experience of the scheme in detail.
However, few schemes are sufficiently large (nor do they have the resources) to
perform a full mortality investigation.
Recent investigations into mortality have revealed links between mortality rates and
factors such as salary, occupation, post code, lifestyle and pension size. This refined
approach to mortality has enabled some banks and insurers to develop so-called
“mortality swaps”, which can allow pension schemes to hedge longevity risk more
effectively.
9.2 Mortality Swaps
A mortality swap requires the trustees of a scheme to pay a fixed or pre-determined
stream of cash flows to a counter-party (usually an insurance company or investment
bank) for a set period of time. This is called the “fixed leg” of the swap. In return the
counter-party pays the trustees the “floating leg”; the cash flows required to pay out
benefits to members. This would be payable for the actual lifetimes of the members,
which is unknown when the contract is written. The pension scheme, by paying the
fixed cash flows, is then able to protect itself against future unexpected longevity
improvements.
In most derivatives, at the date of entering into the contract the net present value of
the two cash flow streams are expected to be identical, so that the swap has an initial
value of zero. In the case of general financial swaps, such as interest rate or inflation
swaps, they are usually priced using “best estimate” assumptions. However, in the
case of mortality swaps, there is currently a lack of a developed market in such trades
and a limited pool of capital market participants who are seeking to take the opposite
exposure to longevity risk to that of a pension scheme (a swap which provides
positive returns if people live longer than expected). This lack of supply is likely to lead
to pension schemes that are looking to enter into such a swap being required to pay a
premium to do so, or basing the swap on prudent rather than best estimate mortality
assumptions. The extent of this premium payable compared to the value placed on
reducing the longevity risk will be a key factor in deciding whether to proceed with
the swap.
Mortality swaps are likely to have a fixed term, of say 50 years. However, a pension
scheme with relatively young deferred pensioners might expect to be paying pensions
for the next 80 years or so. Therefore, such a swap may not provide mortality
protection to the pension scheme over these last 30 years, leading to significant “tail
risk”. It is precisely this tail risk which usually causes the most concern to the scheme,
since the further into the future mortality rates are projected, the more uncertain
they become. To mitigate this risk, some issuers of mortality swaps include a “final
exchange”, so that when the swap matures a final payment is made to reflect the life
56
PSTS Comment:
Trustees are now able to
hedge effectively most
risks faced by a pension
scheme, such as
inflation and investment
risk. The ability to hedge
longevity risk could be
the final piece of the risk
management jigsaw.
The key obstacle to the
growth in this market is
the lack of obvious
counter-parties for this
risk. However, if this can
be overcome so that the
price of mortality
products falls, we would
expect these products
to prove very popular
with pension schemes.
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
expectancy of the remaining pensioners. The final exchange can be timed to occur
at the point when the swap arrangement becomes inefficient for both parties
to maintain.
There are two key risks which mortality swaps can be used to protect against:
the risk that ‘average’ people actually live longer than is currently expected in
that the actuarial table used is ‘wrong’; and/or
● the risk that a particular individual actually lives longer than expected due to
random variation, even if the actuarial table used is ‘correct’, which is a particular
risk for smaller schemes.
Index-linked hedging can be used to protect against the former risk, where payouts to
the pension scheme are based on a mortality index. Alternatively, named lives
hedging is a bespoke product which can be used to protect against the latter risk. The
difference between named lives hedging and buying an annuity with an insurer in the
name of the member is that under named lives hedging it is possible to only protect
against longevity risk if desired, leaving risks like investment and inflation with the
pension scheme.
●
There is also a significant link between inflation risk and longevity risk for pensions
which increase in payment in line with inflation. Mortality products can therefore be
tailored to incorporate an inflation rate swap to mitigate this risk.
There are a number of problems associated with index-linked hedging. The experience
of the scheme members could vary considerably from this index, particularly for
smaller pension schemes, which could still leave the scheme exposed to significant
longevity risk. For such a swap to be of genuine value to a scheme, it would need to
be based on the scheme’s actual mortality rates. However, it is unlikely that it will be
possible to find a willing counter-party to such a specific financial instrument, which
might make it more expensive and also reduce the liquidity of the swap. Reduced
liquidity would be an issue if the scheme wanted to close-out the contract before
expiry, although this is unlikely to be a significant issue since most pension schemes
would probably be “buy and hold” investors. It is primarily due to these difficulties
that the mortality swaps market has, as yet, been relatively slow to develop.
Some consideration must also be given to the application of derivatives to indices
where there is no physical market. Since it is not possible for a replicating portfolio to
be constructed, mortality swaps will not be priced on a “no arbitrage” basis in the
same way as other “classic” derivatives, such as interest rate or inflation swaps. In
addition, due to the lack of players currently seeking to acquire longevity risk,
investment banks are using a fairly prudent allowance for future longevity
improvements in their pricing (which is more prudent than the trustees are using in
their already prudent ongoing valuation assumptions), so that the price they are
willing to trade at is too expensive for the majority of pension schemes.
Furthermore, mortality tables take a considerable length of time to be produced, due
to the period of time which must be analysed to produce statistically credible tables.
There is therefore likely to be a substantial lag between the time period analysed and
the production of the relevant mortality table. Such lags are not seen in most other
derivatives where both the underlying physical asset and the respective derivatives
are continuously traded and updated. The result is that in purchasing such an
instrument pension schemes may be placing themselves at considerable market risk
in terms of the fixed payments they will have to pay.
As discussed in the appendix, there is considerable uncertainty regarding rates of
future improvement in longevity. Due to this uncertainty and the lack of an obviously
57
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
identifiable counter-party for this longevity risk, the banks may need to be prudent
when setting their benchmark mortality rates. By entering into the swap the pension
scheme is also tying itself into the current expectations of future mortality, so that if
mortality rates are heavier than expected the pension scheme will not benefit from
this. The trustees might therefore tie themselves into a mortality assumption which is
more prudent than that used in their most recent scheme specific funding valuation.
9.3 Other Mortality Derivatives
As well as swaps that match, either exactly or approximately, the mortality exposure
of the pension scheme, some market participants are beginning to offer other
mortality related derivatives, such as options. Such products will not provide a true
“hedge”, as a swap contract is designed to do, but will provide a degree of protection
for the trustees in the event that longevity improves significantly faster than
currently expected.
For example, a product which provides an option on mortality rates is offered by
some of the players in the mortality products market. The pension scheme could pay
a premium now, and in exchange would be able to receive a lump sum payment in
10, 15 or 20 years’ time, say, if longevity improvements in the general UK population
exceed a pre-determined level. The lump sum payable would be designed to provide
sufficient additional funding to cover the extra liabilities that would have been
created by the significant improvements in longevity.
This option could be refined to give a better match to the actual liabilities by
weighting the measurement and payout by various age bands and socio-economic
groupings. The mortality protection provided would always remain imperfect, but it
would cover the key risk for the trustees – longevity improving across the UK
population as a whole far faster than they have funded for – but they would remain
exposed to variations in their own pension scheme relative to the UK population.
Such an option would allow a pension scheme to fund its liabilities using a mortality
table with a reasonable degree of future longevity improvements and get effectively
be insured against the risk of actual experience being worse than expected.
Alternatively, a product offered by the French investment bank Natixis aims to
protect pension schemes from the possibility of adverse market conditions combined
with unexpected improvements in longevity. In exchange for this protection, the
pension scheme must forfeit some of their upside exposure to the possibility of
strong investment market performance and mortality rates being heavier than
expected. The product will be designed to have zero initial cost.
The product works by providing the pension scheme with put options driven by
longevity. These provide payments to the pension scheme if the investment index
chosen (usually an equity index) is below a pre-agreed strike price, but are only
triggered if longevity improvements are greater than expected. The amount of the
payment depends on the extent to which the experienced mortality rates have been
lighter than expected.
This product might appeal to pension schemes since the cost of the product is the
sacrifice of some of the upside potential. Pension schemes might be happy to forfeit
the possibility of large returns in exchange for this protection against extreme risk.
Products can also be designed to protect trustees against changes in mortality trends
over a pre-defined period, such as ten years. Under such an arrangement the trustees
would receive funds in ten years’ time which would be expected to be sufficient to
meet their pension obligations in the future, based on mortality expectations at
that time.
58
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Over the last few years the buy-out market has been a fast developing arena and has
generated a huge amount of interest in the press and across the pension industry.
Indeed a massive amount of new capital has been dedicated to the buy-out market
in recent years.
However, business has been slow and few players are generating significant volumes
of business. Major deals did not start to be transacted until the second half of 2007,
and most of this business seems to have been driven purely by the very attractive
pricing seen.
We should not forget that buy-out has always been available to sponsoring employers
and their trustees, since Legal & General and Prudential have been providing these
services for more than twenty years. The new capital has not greatly changed the
long term dynamics of the market as all buy-out insurers face a level playing field due
to the regulation by the FSA. There is no “magic bullet” that the new entrants can
bring to the industry to make pricing more attractive.
For most sponsoring employers we believe that buy-out will remain an unattractive
option for the foreseeable future as it involves passing substantial value to an
insurance company. However, the price at which each sponsoring employer is willing
to trade will vary and this trading price can rise significantly where there are
significant corporate benefits to the elimination of the pension scheme. This is
particularly the case when the pension scheme is large in relation to the employer
and the buy-out could be funded from the existing resources of the pension scheme
generating a cash benefit for the company (as was the case in the recent Rank
transaction). Buy-out might also become a more attractive option to some
companies if accounting standards and/or cash funding standards are further
strengthened in the future.
It is too early to tell if other innovations in the market will prove to be successful.
BrightonRock has a unique product that does address a real concern for a segment of
the market, although trustees will want significant amounts of due diligence to be
undertaken which may constrain its appeal. Mortality products are likely to be
popular if the pricing gap can be bridged by the application of more capital, since this
obstacle has led to only a limited number of transactions being completed in this
field to date.
Our model of the pensions market suggests that it is still growing in absolute terms
and that the current products available are unlikely to make a major impact on the
overall size of the market.
So is this the end game for defined benefit pension schemes? And will it, within five
years, be anachronistic for a medium-sized company to still have a defined benefit
scheme in its balance sheet?
Our answer is categorical – NO!
Richard Jones FIA
Martin Hunter
Oliver Herbert
Punter Southall Transaction Services
June 2008
59
Conclusion
10. Conclusion
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Appendix – Improvements in Longevity
There has been considerable press commentary in recent years surrounding increasing
life expectancies in the UK. This is a significant issue for pension schemes, since if
members live longer their pensions become more expensive to provide. This appendix
sets out a discussion of recent research into mortality levels.
Mortality statistics are split into two components:
a) Experience to date; and
b) Expectations of how mortality rates will change in future.
It is fairly straightforward to see that people’s life expectancies have increased based
on experience to date, due to factors such as improvements in healthcare and better
nutrition. However, there are differing views of how mortality rates will develop in
future. Some commentators believe that the improvements in longevity will continue
(and possibly accelerate), due to factors such as:
●
●
●
further medical advances;
reductions in smoking rates; and
improved levels of health education.
(Indeed, a geneticist at Cambridge University called Aubrey de Grey believes that the
first person to live to the age of 1,000 might already be 60-years old!)
Alternatively, other analysts think that the improvements in longevity have been
caused by one-off factors, such as the impact of healthy lifestyles during the interwar period, and future improvements will be limited by:
●
●
●
●
●
●
increased levels of obesity, partly caused by the increased consumption
of fast foods;
a slowdown in the level of medical advances, based on the assumption that the
significant medical developments of the last 20 years or so cannot be replicated
in future;
the possible emergence of global pandemics, such as bird flu or AIDS;
the possibility that the human body can only last so long, and therefore life
expectancy has a natural cap;
the impact of environmental issues, such as climate change; and
terrorism/war.
Therefore, there is considerable uncertainty regarding the rates of future longevity
improvements, which is reflected by the wide range of possible adjustments which
can be applied to current actuarial mortality tables. The graph opposite illustrates a
range of possible future life expectancies for males currently aged 65.
60
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Similarly, the graph below is taken from a lecture given by Mervyn King, and
illustrates the possible range of life expectancies from birth for females, based on
projections from 2004 onwards.
61
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
About the Authors
Richard Jones FIA
[email protected]
020 7533 6967
Richard has eleven years’ experience in advising corporate entities on pension and
investment issues. He is the head of Punter Southall Transaction Services, which is
the specialist transactions consulting division within Punter Southall.
Richard has spent the past eight years with Punter Southall, working on a variety
of UK and international pension issues from the corporate perspective. He has been
involved in a large number of international mergers and acquisitions, particularly
involving UK and US interests for both corporate buyers and private equity firms.
Amongst other clients, Richard is the lead consultant for an investment company,
for whom he provides actuarial and investment advice on a number of global
pension schemes with more than $3 billion in assets.
Whilst at the firm, Richard has worked for Punter Southall LLC in their Framingham,
USA office for eighteen months.
Richard has a degree in Business Economics and became a Fellow of the Institute
of Actuaries in 2002.
Martin Hunter
[email protected]
020 7533 1885
Martin joined Punter Southall in 2004, and initially worked on a team advising
pension scheme trustees. In 2006 he transferred to Punter Southall Transaction
Services, where he now advises private equity houses, investment companies
and other corporate entities on the acquisition and disposal of defined benefit
pension schemes.
Martin also advises corporate clients on the ongoing management of their pension
liabilities, including offering advice on liability reduction exercises and other
de-risking strategies. He is a member of the Punter Southall buy-out team.
Martin graduated from Keble College, Oxford, with a BA in Mathematics in 2003.
Oliver Herbert
[email protected]
020 7533 1820
Oliver joined Punter Southall Transaction Services in 2006 where he offers advice
to corporate entities involved in transactions, advising clients on both the buying
and selling side. He also works on the valuation of employee stock options, both
for accounting disclosures and expert witness cases, and is a member of a
Punter Southall team analysing international pension issues. He is also a member
of the Punter Southall buy-out team.
Oliver graduated from Brasenose College, Oxford, with a BA in Modern History
in 2004. Following this, he went on to obtain an MSc in Economics, with distinction,
at the University of Bristol.
62
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Contact Us
Punter Southall Transaction Services
126 Jermyn Street
London
SW1Y 4UJ
www.pstransactions.co.uk
Tel: 020 7533 6990
Fax: 020 7533 6978
Email: [email protected]
Further copies of the report can be requested by contacting Martin Hunter.
Every effort is made to ensure that the information is correct at the time of going to press
however Punter Southall Transaction Services accepts no responsibility for any errors or
omissions. The report should not be reproduced in whole or in part without the permission
of one of the authors. The report in itself does not constitute advice, and we would be
pleased to provide scheme specific advice on request.
© Copyright Punter Southall Transaction Services June 2008
63
The End Game? – An analysis of the bulk buy-out market & other de-risking solutions
Notes
64
For further information, visit our website at www.pstransactions.co.uk or email [email protected]
Pensions and Actuarial Due Diligence Provider of the Year
Punter Southall Transaction Services
Punter Southall Transaction Services is a division of Punter Southall Limited and is a member of The British Private Equity and Venture
Capital Association. Registered office: 126 Jermyn Street, London SW1Y 4UJ · Registered in England and Wales No. 3842603
A Punter Southall Group company