Monetary Review - 4th Quarter 2007

29

Longevity Bonds – a Financial Market

Instrument to Manage Longevity Risk

Governor Jens Thomsen and Jens Verner Andersen, Financial Markets

INTRODUCTION

Life expectancy has been increasing over the last centuries. Naturally, it
is positive that people live longer, but this also means that they must
reconsider their savings in order to maintain a satisfactory standard of
living when they retire. If people outlive their reserves, they will at some
point have to reduce their standard of living.
Since the exact time of death is not known, there will always be some
uncertainty regarding the savings required. Many pension schemes
include life annuities, so that the pension companies make fixed
monthly payments during the remaining part of the policyholder's life1.
This transfers the longevity risk2 from the individual policyholder to the
pension company. If life expectancy increases more than expected, pen-
sion companies will have to pay out more than projected, resulting in a
loss to those companies.
With a view to managing the uncertainty related to future life
expectancy, various players have sought to develop financial instruments
that are indexed to the longevity of the population. These new instru-
ment types, known as longevity bonds, transfer the risk in connection
with higher life expectancy to investors in the financial markets.
The market for the issuance of longevity bonds has received consid-
erable attention in the financial press. However, neither issuers nor in-
vestors have so far fully embraced the instrument. Against that back-
ground, this article discusses the prospects for future use of such instru-
ments. The underlying considerations behind the products are outlined
in the first section, followed by a status of the current usage in the se-
cond section. The third section analyses supply and demand factors gov-
erning the limited use, and finally, the fourth section examines the po-
tential for establishing a market for longevity bonds and discusses why
there is no basis for government involvement.

1
In Denmark, life annuity schemes constitute a relatively large share of the pension schemes, but the
2
ratio is declining. See ATP Faktum No. 27 (in Danish only), December 2005.
In the following, risk is reviewed in relation to management of longevity risk in the pension sector.
 Monetary Review - 4th Quarter 2007

30

IMPROVEMENT PER DECADE IN LIFE EXPECTANCY IN OECD COUNTRIES

1960-2000 Table 1

Country At birth At 65

Canada ........................................................... 1.9 1.0

Denmark.......................................................... 1.2 0.6
France ............................................................. 2.2 1.3
Netherlands .................................................... 1.3 0.7
Italy ............................................................... 2.6 1.2
Spain ............................................................... 2.5 1.1
Sweden ........................................................... 1.7 0.9
UK .............................................................. 1.7 1.0
USA ............................................................... 1.8 0.9

Note: Life expectancy is measured as a weighted average of life expectancy at birth and at 65, respectively, calculated
on the basis of estimated mortality rates. The improvement in life expectancy is the growth rate between 1960
and 2000.
Source: Human Mortality Database, OECD and Antolin (2007).

BACKGROUND

In many OECD countries, life expectancy has increased by up to 2.6 years

per decade over the last half century, cf. Table 1. The higher life expect-
ancy has been achieved following substantial declines in the mortality
rates of both the young and the old.1
Despite of the positive message contained in this information, it is a
challenge for the population, pension companies and policymakers to
implement mechanisms that can alleviate the implications of higher lon-
gevity. At the centre of the discussion is the possibility that the life
expectancy of e.g. younger generations may far exceed the present pro-
jections, which may have social implications2. If people outlive the cap-
ital that has been saved they will have to reduce their standard of living
sooner or later.
Life expectancy has been increasing during the last centuries and pro-
jections show that this will continue for future generations in the west-
ern world, supported by new, improved healthcare treatments, among
other things. Many governments have this issue on the agenda and con-
tingent planning has been initiated in order to prepare for the change
in demographics. Some of the measures implemented in OECD countries
include later retirement and incentives for higher savings. In many
countries the possibility of postponing the retirement age of future gen-
erations has been considered3.
Thus, at the centre of the discussion is the speed and magnitude of
these changes, and not the direction. In other words, the main issue is

1
2
Gillian Tett and Joanna Chung, Death and the salesmen, FT Magazine 24/25 February 2007.
3
See, for example, Mervyn King (2004).
Danish Welfare Commission, Future welfare – what are other countries doing? (in Danish only), 2005.
 Monetary Review - 4th Quarter 2007

31

HIGHER LIFE EXPECTANCY – UNEXPECTED VERSUS EXPECTED Box 1

Model projections show that life expectancy in Denmark will follow an upward trend
1
in the period until 2100. From the risk management perspective of the pension com-
panies, the risks associated with higher life expectancy stem from the uncertainty re-
lated to the prediction of the precise number of years the individual policyholder will
live. If policyholders outlive the projections, pension companies will realise losses.
Chart 1 illustrates life expectancy for women, as well as the related uncertainty, il-
lustrated by the 5th and 95th percentiles. In other words, in 2100 life expectancy will
be around 87 years, while the uncertainty, expressed as a 90 per cent confidence
interval, will be ±2 years.
The pension sector manages the two effects differently; higher life expectancy is ad-
dressed by reducing payments to policyholders, while deviations are handled through
specific risk premiums which are part of the general provisions of pension funds.

EXPECTED AND UNEXPECTED DEVELOPMENT IN LIFE EXPECTANCY FOR

WOMEN IN DENMARK Chart 1
Age

90

85

80

75

70

65

60

55

50
1850 1900 1950 2000 2050 2100 2150
Life expectancy including projections 95th percentile
5th percentile

1
Haldrup (2004). For men, an upward trend is also seen in the period until 2100.

that the life expectancy of present generations may deviate substantially

from the course projected today, cf. Box 1.
Life expectancy in Denmark has not increased at a constant rate, cf.
Chart 2. There have been substantial fluctuations during the period. For
example, the pandemic influenza in the 1920s has been viewed as one
of the factors that improved the life expectancy of subsequent gener-
ations. Conversely, obesity is seen as a factor which may reduce the life
expectancy of present generations1.

1
Jacob S. Siegel, The Great Debate on the Outlook for Human Longevity: Exposition and Evaluation of
Two Diverging Views, Society of Actuaries, 2005
 Monetary Review - 4th Quarter 2007

32

DEVELOPMENT IN LIFE EXPECTANCY IN DENMARK 1835-2006 Chart 2

Years Years

80 8

6
70
4

2
60

50
-2

-4
40
-6

30 -8
1835 1855 1875 1895 1915 1935 1955 1975 1995
Deviation from trend (right-hand axis) Development in life expectancy (left-hand axis)

Note: Life expectancy is measured as a weighted average of life expectancy at birth, calculated on the basis of
estimated mortality rates. Over the above period, life expectancy has increased by an average of 3 months per
year, and the deviation from the trend is measured on the basis of this average growth rate.
Source: Human Mortality Database.

Implications for the pension sector

In general, past projections have underestimated the fact that people
live longer. In practice this implies that pension companies need to make
additional provisions in order to address the shortfall. Given that pen-
sion companies have made long-term commitments, they need to man-
age the longevity risk1.
Before 2000, there was awareness of this issue, but high returns had
helped to alleviate the problem. Due to poor equity market perform-
ance and low interest rates in the following years, it became evident
that decades of improvements in life expectancy had become a chal-
lenge for the pension industry2.
Moreover, many countries have introduced new pension regulation,
requiring pension fund managers to update their pension plan assump-
tions and apply a mark-to-market approach in the valuation of the risks
embedded in the pension plans3. Finally, many pension companies have
implemented new risk management techniques, known as asset liability

1
2
Ted Clarke, Living with the Risks – Is your pension deficit what it seems?, Prudential, 2007.
3
Hermes: BT Pension Scheme: A decade of outperformance, 2007.
Since 1 January 2005, listed insurance companies have been subject to the International Financial
Reporting Standards (IFRS).
 Monetary Review - 4th Quarter 2007

33

management (ALM), which imply a closer linkage between pension fund

assets and liabilities.
Against this background there is an increasing interest in analysing
new instruments that can be used to hedge against longevity risk. The
pension industry has for some time struggled with other risk factors such
as the management of interest rate risk, and in many defined benefit
pension schemes inflation is also an embedded risk element. In both
cases there exist a range of financial instruments, such as cash products
(bonds with maturities of up to 50 years, and inflation-linked bonds) and
derivatives (interest-rate swaps and inflation-linked swaps) that offer
the pension industry options for managing such risk factors.
In relation to longevity risk, the application of capital market instru-
ments has been limited. So far the pension industry has used other types
of hedging tools which can basically be divided into three categories:

1) Self-insurance and implicit coverage

Typically pension funds have managed liabilities related to longevity risk
without specific hedging instruments. Thus, the exposure to longevity
risk has been part of their overall risk management. However, life insur-
ance companies that enter into contracts with both pension and insur-
ance elements have been able to obtain some degree of hedging.
After an insurance event such as pandemic influenza there will nor-
mally be a negative correlation between insurance and pension claims.
The claim from insurance policies increases, but at the same time future
payments on pension policies are reduced because the lives of the pol-
icyholders were shorter than initially forecasted1.

2) Reinsurance
Traditionally, reinsurance companies have provided capital to life insur-
ance companies and pension funds seeking hedging opportunities. How-
ever, longevity risk is so specific that reinsurance companies have so far
been reluctant to undertake business unless it was part of an existing
client relationship. Nevertheless, factors such as increasing demand from
insurers and pension funds plus better modelling of longevity risk may
increase reinsurers' appetite for this business2.

3) Sale to external buyout funds

In recent years, companies with defined benefit schemes (primarily in
the UK) have been looking for alternative risk transfer solutions, and

1
2
Richards (2004).
Richards (2004).
 Monetary Review - 4th Quarter 2007

34

more are expected to follow suit in the coming years. According to a

survey by PwC, 11 per cent of UK companies with defined benefit
schemes are considering the option of selling off their pension obliga-
tions within the next 5 years1. The set-up may take various forms; how-
ever, the key underlying model is that companies transfer the portfolio
of pension contracts to an external investor that will assume the con-
tractual obligations. In the last couple of years, the UK has seen the es-
tablishment of around 20 new investment companies specialising in this
area.2

STATUS FOR ISSUANCE OF LONGEVITY BONDS

The evolution of financial market instruments has brought a number of

new risk-diversification options to the attention of investors. Most not-
able has been the development in credit markets, where a multitude of
products have been introduced during the last decades. This reflects,
among other things, an interest in developing products that can help
financial institutions to hedge against various types of risk. Experience
from other markets has generated an interest in developing capital mar-
ket products that can also help diversify longevity risk.
Nevertheless, issuance of longevity bonds has so far been limited. A
few reinsurance companies have used capital market instruments to
transfer risk from pension and life insurance companies to investors in
financial markets. The most developed market is based on mortality-
linked derivatives, which in some respects resemble longevity bonds, but
are nevertheless significantly different in that they are based on mortal-
ity rates, while longevity bonds are based on survivor rates, cf. the Ap-
pendix.
Since 2003 some companies have issued mortality-linked securities, cf.
Box 2. In addition, anecdotal evidence suggests that a number of invest-
ment banks have been active in setting up similar structures, but these
transactions have been established on the basis of private placements3.
Issuance of longevity bonds has not taken place so far, even though
there has been an indication of strong interest from the pension sector.
In 2004, the EIB aimed to launch a longevity bond with a value of GBP
540 million which was intended for UK life insurance companies and
pension funds. The structure was initiated in partnership with BNP
Paribas and Partner Re, which ultimately were the final assumers of the

1
PricewaterhouseCoopers, PwC pensions survey: Tracking the views of key decision makers, Autumn
2
2007.
Phillip Inman, More firms seek to sell off their pension schemes as costs rise, The Guardian, 19
3
October 2007.
IFR, UK longevity risk hot spot, 10 June 2006.
 Monetary Review - 4th Quarter 2007

35

ISSUANCE OF MORTALITY-LINKED SECURITIES Box 2

In 2003, Swiss Re established a special purpose vehicle (Vita Capital) that issued USD
400 million in 3-year notes. In 2005, Vita Capital issued a second note with an
outstanding amount of USD 362 million and 5-year maturity. In December 2006, Vita III
was launched with an outstanding volume of USD 700 million in tranches with 4- and
5-year maturities. The key objective was to replace Vita I, which expired at the end of
1
2006.
2
In 2006, Scottish Re raised USD 155 million via Tartan Capital in 3-year notes. Osiris
Capital was arranged by Swiss Re, but on behalf of the AXA Group, which was the
ultimate buyer of protection in 2006. The outstanding volume was EUR 345 million
3
and the maturity was 4 years.

1
Swiss Re.
2
Scottish Re.
3
IFR, Taking a view on mortality, 21 October 2006.

risk. The launch received much attention in the financial press because
the deal was considered to be the first of its kind and had an innovative
set-up. However, it never reached the market. Besides technical issues
such as design problems, anecdotal evidence indicates that pension
funds and life insurance companies did not subscribe to the deal, primar-
ily because coverage was considered to be too expensive, see Box 3.
The above-mentioned capital market transactions have all been
focused on hedging the pension sector's longevity risk. In addition, a
number of bonds have been issued, predominantly in the USA, with the
purpose of funding life insurance business or funding regulatory capital

EIB LONGEVITY BOND PROJECT Box 3

In November 2004, the European Investment Bank announced plans to issue the first
longevity bond that would offer coverage for UK pension schemes and life insurers
with exposure to longevity risk for the male population of England and Wales. The
initial size of the note was GBP 540 million and the maturity of the bond would be 25
years.
Although the bond was launched by the EIB, issuance was arranged and managed by
BNP Paribas. Under this structure BNP would effectively bear the investment risk and
the longevity risk would be covered by Partner Re that had concluded an agreement
with BNP. The bond would still be rated AAA, equivalent to the EIB rating, and thus
purchasers of coverage would have an AAA counterparty risk.
The issue was withdrawn in late 2005 without being issued, primarily because the
pension industry found the price of coverage on longevity risk too high. In addition,
other factors were mentioned such as missing mandates in the pension industry and
concerns about basis risk between the index embedded in the bond and the longevity
risk faced by insurance and pension funds.

Source: T. Cox (BNP Paribas) and F. Blumberg (Partner Re), Longevity bonds, 2005 Life Convention, Barnett
Waddingham LLP, Longevity bond to be issued by the EIB, 2005 and Financial Times, Changing attitudes
mean there is more life left in longevity bonds, 22 November 2006.
 Monetary Review - 4th Quarter 2007

36

requirements. These transactions have used a securitisation set-up in

which asset-backed securities have been structured on the basis of a
pool of life insurance policies. In general, these transactions are aimed at
financing traditional life insurance business without having a specific
focus on hedging longevity risk. The bulk of the outstanding volume has
been linked to such structures, cf. Chart 3.

INTEREST IN LONGEVITY BONDS AMONG INVESTORS AND ISSUERS

Notwithstanding the appealing characteristics of longevity bonds, the

market has not evolved. A number of factors have contributed to the
slow growth in longevity bonds; factors related that are applicable to
both buyers and sellers. In general, the market is short of investors in
longevity risk, and capital needs to be attracted by offering alluring risk-
adjusted returns. In addition, some externalities may hinder the develop-
ment of markets. A factor such as the outstanding volume is often essen-
tial for a market to develop, but once the volume has passed a critical
level, growth may potentially be very high. In the following the various
factors behind the lacklustre evolution will be illustrated.

General capital market considerations

Viewed in retrospect, the most successful launches of new capital market
instruments during the past decade have been undertaken in an envir-

INSURANCE-LINKED SECURITIES AND MORTALITY NOTES Chart 3

USD billion
16

14

12

10

0
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006*

Insurance-linked securities Mortality notes

Note: The data for insurance-linked securities is based on information until the end of August 2006. The data includes
accumulated issuances since 1996.
Source: Swiss Re and IFR.
 Monetary Review - 4th Quarter 2007

37

onment where initially there was a fair balance between participants

with buying and selling interests. If markets are relatively biased to-
wards one side, it may be difficult to launch a product due to lack of
interest in and understanding of business characteristics. Consequently,
the price will typically be established at a level that is not deemed at-
tractive.
In the case of longevity bonds, the most obvious challenge is the lack
of natural investors who would benefit from an unexpected rise in life
expectancy. Pharmaceutical companies and care providers are often
mentioned as examples of investors with a natural interest in assuming
this risk1. These sectors would be exposed to losses if life expectancy
decreases, and thus the companies would be able to hedge their own
exposures by issuing longevity bonds.
In spite of the theoretical arguments, it is more than doubtful whether
the natural investors would actually enter into these transactions, due to
corporate governance considerations, among other things. For example,
it is debatable whether a company would be able to explain to its share-
holders how it can benefit from entering into transactions that may
influence profits in a distant future. If this obstacle can be overcome, it is
also uncertain whether the potential volume would be more than a
drop in the ocean when compared with the overall demand in the pen-
sion industry.
Another challenge in relation to establishing a market for longevity
bonds is that the underlying characteristics differ significantly from
those of other types of financial instrument, such as mortgage-credit
bonds, where the risk premium is determined on the basis of the devel-
opment in the credit quality in the mortgage-credit market. The process
related to development in life expectancy is characterised by a high dur-
ation and low volatility, particularly in recent years, cf. Chart 4.
If a comparison is made with the development in credit quality in the
mortgage-credit market (illustrated by growth in GDP at factor cost), the
duration is lower and the volatility substantially greater. The reason is
that the economy has historically gone through complete business cycles
in less than 10 years. Consequently, an investor in mortgage-credit
bonds does not have to wait many years to realise whether the invest-
ment was profitable. In addition, economic indicators are frequently
published, giving an idea of future developments.

1
European Pensions and Investment News: Looking for a buyer in face of longer lives, 4 December,
2006.
 Monetary Review - 4th Quarter 2007

38

GROWTH IN DANISH LIFE EXPECTANCY AND ECONOMIC GROWTH IN

DENMARK ILLUSTRATED BY THE CHANGE IN GDP AT FACTOR COST Chart 4
Per cent
10

-2

-4

-6

-8
1875 1885 1895 1905 1915 1925 1935 1945 1955 1965 1975 1985 1995 2005
Annual growth in life expectancy Annual growth in GDP at factor cost

Note: Life expectancy is measured as a weighted average of life expectancy at birth, calculated on the basis of
estimated mortality rates.
Source: K. Abildgren, Monetary Trends and Business Cycles in Denmark Since 1875, Working Paper 43, Danmarks
Nationalbank, 2006, and Human mortality database.

Perspectives for investors in longevity bonds (hedging sellers)

Besides companies with a natural exposure to longevity risk, a number
of other investors have also been mentioned in the debate about the
issuance of longevity bonds. These are primarily investors that are al-
ready active in the asset management sector, including hedge funds, in-
vestment funds, etc. The low correlation between an unexpected in-
crease in life expectancy and the yield on other financial instruments is
typically cited as the major reason why longevity bonds would be
attractive. The instrument would contribute to reducing the risk on the
investors' aggregate investment portfolio, which means that the invest-
ors would require only limited risk premiums when investing in this in-
strument.
In contrast, the financial instruments presented so far have had a num-
ber of disadvantages. Primarily, they have been highly complex instru-
ments with maturities in the range of 25 years or more. So far, it has
been difficult to sell products that combine high complexity and long
maturities. It has, at the same time, been difficult to reach agreement on
some measure of standardisation, including how to estimate the future
development in life expectancy and the related uncertainty1.

1
JP Morgan (2007).
 Monetary Review - 4th Quarter 2007

39

Against that background it has been difficult to build up a market since

investors have had to accept that they were buying an instrument that
could prove to be practically impossible to resell. In view of the long
maturity of the product, investors would have limited opportunities to
exit the transaction for many years. The lack of liquidity has led investors
to demand a liquidity premium, so that the resulting price of hedging
has been unattractive for the pension sector, despite the low correl-
ation.

Perspectives for issuers of longevity bonds (hedging buyers)

So far, the pension sector has predominantly seen longevity risk as a
factor that has been managed via internal risk-management systems.
Capital provisions have been made as a general buffer against fluctu-
ations in the value of pension commitments.
However, as the pension sector has increased its focus on risk manage-
ment, it has also turned its attention to instruments for explicit manage-
ment of the primary risk factors that a pension company is exposed to,
including longevity risk.
In spite of this, the pension sector has not welcomed the launch of
longevity bond. The reason may be that the instruments presented so
far, such as the EIB's longevity bond, allow only partial hedging of the
pension sector's exposure to unexpected increases in life expectancy. For
example, the development in life expectancy may differ considerably
from one socio-economic group to another, and between geographical
areas, etc. Against this background, a Danish pension fund investing in
the EIB's longevity bonds would not be able to fully hedge its exposure
owing to the different developments in life expectancy in Eng-
land/Wales and Denmark, cf. Box 3.
These issues reflect a dilemma, in that on the one hand pension funds
want the closest possible correlation between hedging and exposure in
order to avoid a basis risk, and on the other hand investors want instru-
ments that are as standardised as possible.
Furthermore, pension funds are typically not yet empowered to enter
into transactions that may involve less exposure to longevity risk. This is
attributable to factors such as the uncertainty concerning the regulatory
treatment of longevity bonds.

PERSPECTIVES FOR FUTURE DEVELOPMENT OF LONGEVITY BONDS

The lacklustre market for longevity bonds has led to considerations as to

whether government issuers should intervene in this market. Tradition-
ally, government issuers have an objective of covering the central gov-
 Monetary Review - 4th Quarter 2007

40

ernment's financing requirement at the lowest possible long-term costs,

subject to a prudent degree of risk. Furthermore, their aim is to facilitate
the central government's access to the financial markets in the longer
term and to support a well-functioning domestic financial market.
Consequently, it might be considered whether government issuers
should get actively involved in the development of a market for longev-
ity bonds. In some countries government issuers have in this way con-
tributed to the development of bond segments with long maturities as
well as inflation-linked instruments, cf. the 50-year issuances in France
and the UK. This has provided a benchmark for private issuers of finan-
cial instruments in the same segments1.
Likewise, it could be argued that the central government should issue
longevity bonds in order to provide a benchmark that could support the
development of a market for longevity bonds. However, it is difficult to
imagine that sovereigns would be able to issue such longevity bonds on
attractive terms2. Furthermore, the central government is already highly
exposed to increasing life expectancy as extra government spending will
be required to tackle the implications of demographic changes3. By
issuing longevity bonds to pension funds, the government would there-
fore further increase their balance-sheet exposure.
Several pension funds have also indicated that they see the manage-
ment of unexpected increases in life expectancy as an inherent challenge
for the pension funds4. It could also be argued that the pension sector
knows its clients better than the external investors and is therefore bet-
ter equipped to manage the risk, rather than selling it off to investors in
the financial markets.
One issue that has been pointed out in relation to the new products
launched in the credit market has indeed been that it would, perhaps,
be best that financial institutions manage the risks in relation to their
commitments, instead of repackaging them and selling them off to fi-
nancial investors. The same argument could be put forward in respect of
longevity. On the other hand, by selling off, the pension sector disperses
the risk to an investor group which may be better equipped to manage
risk that is characterised by a low probability but a potentially high cost.
For investors in the financial markets there are, however, a number of
obstacles to be overcome before longevity bonds become a marketable
instrument, and it is very likely that the final product design has not yet
been developed. In any case, it is important to achieve a higher degree

1
2
Danish Government Borrowing and Debt 2005, chapter 9: Issuance of Long-Term Government Bonds.
3
Swedish National Debt Office, Comment on Longevity bonds: summary by Bo Lundgren, 2005.
4
The Danish Welfare Commission's main report 2006 (in Danish only).
European Pensions and Investment News: Looking for a buyer in face of longer lives, 4 December
2006.
 Monetary Review - 4th Quarter 2007

41

of standardisation within the area. Several investment banks, including

JPMorgan and Credit Suisse, have issued reports on this topic, thereby
generating debate and awareness about the issue.1 One of the critical
themes raised is the development of better models for predicting life
expectancy.

CONCLUDING REMARKS

Higher life expectancy will pose substantial challenges for pension funds
in the coming years. In general, increasing life expectancy needs to be
addressed via higher savings for the retirement age. Moreover, the un-
certainty about projections of the future development in life expectancy
implies that pension funds that have issued life annuities will have to
manage the risk on their liabilities. In itself this is nothing new; pension
funds have hedged their risks for many years. In recent years pension
funds have, however, to a large extent implemented asset-liability man-
agement techniques, and thereby focused on almost complete hedging
of risks not deemed to be part of their core business.
Viewed in that perspective, some pension funds, e.g. in the UK, have
actively sought opportunities to hedge the balance-sheet risk, including
the possibility of using longevity bonds. Nevertheless, the market for
longevity bonds has not really taken off yet, although there have been a
few experiments with products whereby investors assume the risk of an
unexpected increase in life expectancy against receipt of a risk premium.
The extensive focus on the pension sector by several investment banks
reflects investments in the development of various models. In view of
the innovation seen in other financial areas, it is not impossible that a
private market for longevity bonds will emerge, although not necessarily
based on the product structures known today.

1
JPMorgan (2007) and Credit Suisse Longevity index.
 Monetary Review - 4th Quarter 2007

42

REFERENCES

Antolin, Pablo (2007), Longevity Risk and Private Pensions, Financial

Market Trends, OECD, Volume 2007/1 No. 92, June.

Antolin, Pablo and Hans Blommestein (2007), Governments and the

Market for Longevity-indexed Bonds, Financial Market Trends, OECD,
Volume 2007/1 No. 92, June.

Blake, David and William Burrows (2001), Survivor Bonds: Helping to

Hedge Mortality Risk, The Journal of Risk and Insurance, Vol. 68, No. 2,
pp. 339-348, June.

Blake, David, Andrew Cairns and Kevin Dowd (2006), Living with
Mortality: Longevity Bonds and other Mortality-Linked Securities,
Presented to the Faculty of Actuaries, 16 January.

Blake, David, Andrew Cairns, Kevin Dowd and Richard MacMinn (2006),
Longevity Bonds: Financial Engineering, Valuation, and Hedging, The
Journal of Risk and Insurance, Vol. 73, No. 4, pp. 647-672.

Brown, Jeffrey R. and Peter R. Orszag (2006), The Political Economy of

Government Issued Longevity Bonds, Prepared for The Second Inter-
national Longevity Risk and Capital Market Solutions Conference, April.

G10 (2005), Ageing and pension system reform: implications for financial
markets and economic policies, September.

Gründl, Helmut, Thomas Post and Roman N. Schulze (2006), To Hedge or

Not to Hedge: Managing Demographic Risk in Life Insurance Companies,
The Journal of Risk and Insurance, Vol. 73, No. 1, pp. 19-41.

Haldrup, Niels (2004), Estimation of average longevity for men and

women in Denmark 2002-2100 based on the Lee-Carter method (in
Danish only), Working report 2004:3, Danish Welfare Commission, 15
May.

JPMorgan (2007), LifeMetrics. A toolkit for measuring and managing

longevity and mortality risks, Version 1.0, 13 March.

King, Mervyn (2004), What Fates Impose: Facing Up To Uncertainty, The

Eighth British Academy Annual Lecture.
 Monetary Review - 4th Quarter 2007

43

Loeys, Jan, Nikolaos Panigirtzoglou and Ruy M. Ribeiro (2007), Longev-

ity: a market in the making, J.P. Morgan Securities Ltd., July.

Richards, Stephen and Gavin Jones (2004), Financial aspects of longevity

risk, The Staple Inn Actuarial Society, 26 October.

Securitization – new opportunities for insurers and investors, Sigma,

Swiss Re, No. 7/2006.

Health Committee and Political and Economic Committee of the Danish

Parliament (2006), Significant increase in life expectancy – what is the
situation and what are the potential implications for the future retire-
ment age? (in Danish only),
www.folketinget.dk/samling/20061/almdel/PØU/Bilag/12/313205.PDF, 18
October.

Visco, Ignazio (2006), Longevity Risk and Financial Markets, Banca

d'Italia, October.
 Monetary Review - 4th Quarter 2007

44

APPENDIX

The following provides a description of the overall structure of the mor-

tality-linked instruments issued to date, as well as a comparison with the
EIB's proposal for issuance of longevity bonds1.

Swiss Re Vita I
The overall concept is that the issuer (Swiss Re) seeks to hedge any risk
arising from a substantial increase in the mortality rate. Viewed in isol-
ation, the issuer achieves a gain on the instrument if the mortality rate, Mt,
substantially exceeds a predefined index, M0. On the other hand, the
investor gains if the mortality rate, defined by Mt, is not extremely high.
The maturity of the instrument is 3 years. The Mt index is defined as a
basket of mortality rates in five countries with different weights – the USA
(70 per cent), the UK (15 per cent), France (7.5 per cent), Switzerland (5 per
cent) and Italy (2.5 per cent). The gender distribution is 35 per cent women
and 65 per cent men, and the instrument has a broad age distribution.
The instrument pays a variable coupon, equivalent to LIBOR plus 135
basis points, on an ongoing basis. The part of the principal that is lost is
determined using the following loss function:

0% Mt < 1.3 M0
Lt = (Mt – 1.3 M0)/(0.2 M0)*100 % if 1.3M0 ≤ Mt ≤ 1.5M0
100 % 1.5* M0 < Mt

Issuance takes place via Vita Capital, a special purpose vehicle. This
structure entails a number of advantages, partly due to the transaction's
status as an off-balance-sheet item, partly because it offers more favour-
able counterparty risks for the investor.

EIB longevity bond

The EIB's proposal for a longevity bond (that never reached the
market) is constructed with a maturity of 25 years. The instrument
hedges the pension fund against an unexpected increase in the survivor
index for people in the age group 65-90 years. The bond has an annuity
structure, with the annual coupon payments starting from the base of
GBP 50 million. Subsequently each coupon payment reflects the
percentage of the population of England and Wales aged 65 that is still
alive each year until the 25 years have passed. If the survivor rate
increases substantially, the pension fund will receive higher coupons.
The resulting overall coupon structure is as follows:

rt (St) = GBP 50 million * St for t = 1,2,….,25, where St is the survivor index

in the year t.
1
The description is based on Standard & Poors, Vita Capital Ltd.'s Principal-at-risk variable rate
mortality catastrophe-indexed note, 2003, and Blake, Cairns, Dowd and MacMinn (2006).