Historical development of
Reinsurance
The concept of reinsurance dates back to the middle Ages. The oldest known
contract with reinsurance characteristics was concluded in 1370 in Genoa and
dealt with marine risks. But it was not until the 19th century that the
foundations of the modern industry developed, with the introduction of
whole-portfolio reinsurance and the emergence of specialized reinsurance
companies. The joint-stock primary insurers that were developing at this time
were small and locally based, and thus vulnerable to catastrophic losses. It
became clear that — by introducing an additional layer of diversification —
reinsurance was the solution to this problem.
The first specialized reinsurer, Cologne Re, was founded in 1846 in Cologne,
mainly in response to a devastating fire in Hamburg some four years previously.
Similar institutions followed, mainly in Germany, France, Belgium, Austria, and
Switzerland, among them Swiss Re (1863) and Munich Re (1880). Widespread
reinsurance was slower to develop in the United Kingdom and United States, in
the former case partly reflecting statutory restrictions and partly because of
the unique nature of the Lloyd’s market. This was initially confined largely to
marine risks and organized as a co-insurance market. Large risks were from the
beginning spread among a number of syndicates, which in turn were backed by
wealthy individual merchants in the City of London.
The development of the reinsurance industry in the 20th century was linked to
the world economic cycle, together with world wars and political and economic
crises. Major natural disasters, such as the earthquakes in San Francisco
(1906) and Tokyo (1923), demonstrated the resilience of the industry and its
crucial role in extreme loss events. This resilience, and the stabilizing
effect of the industry, were underscored by the very small number of
insolvencies of reinsurance companies, notwithstanding the voluntary market exits
and entries that occur in any industry.
More
recently, the industry’s development has been closely aligned with economic and
technological progress, leading to the emergence of new classes of business, such
as satellite insurance in the 1960s. Increasing trade liberalization toward the
end of the century has allowed the industry to sustain and promote greater
global risk diversification and has encouraged the development of new
reinsurance centers, Bermuda being the most prominent among them.
TRADE BARRIERS FACING REINSURANCE INDUSTRY
|
A number of broader
factors have combined in recent years to heighten interest in the functioning
of reinsurance companies.
First, the
importance of a robust and innovative reinsurance industry has been highlighted by several major natural
disasters and terrorist actions. These have imposed significant
stresses on the reinsurance sector. Although they have been handled without substantial
financial disruption, they have nevertheless focused increased attention on the
capital resources and risk management capability of the insurance industry
generally and, as a crucial part of that, the reinsurance sector. Indeed,
reinsurers are increasingly recognized to be crucial to primary insurers and to
global insurance capacity.
Second, in recent
years, the boundaries between different kinds of financial activity have been
progressively eroded (destroy gradually). This erosion has been most evident in
relation to banking and capital markets but extends across the financial sector
more widely, including insurance and reinsurance. It has been reflected in the
growth of financial conglomerates active in a range of different markets, in
contractual risk transfers of various kinds, and in the broader range of
counterparties that firms take on. Reinsurance companies have participated in
this evolution, for example through their involvement in credit risk markets
and in the development of structured products of various kinds, providing
attractive sources of efficient risk transfer to the capital markets. These
increasing interconnections have in turn served to raise the level of interest
in the reinsurance sector.
Third, the
reinsurance industry has become increasingly concentrated. Some ten firms now
account for over 60% of global reinsurance premiums, as against some 40% ten
years ago. Although there are good reasons why increased size may deliver
competitive advantage, as in other parts of the financial sector, the increased
concentration nevertheless means that primary insurance companies, and through
them many other economic agents, are dependent on the performance of a
relatively small number of reinsurance firms.
Against this
background, the reinsurance industry faces a number of important challenges and
constraints, which are being addressed but which will need to be decisively
resolved in coming years if the industry is to maintain its role in an
expanding global economy.
There is a growing
demand for risk cover in segments about which the primary insurance industry
has become wary: health, longevity, disability, and a variety of casualty and
liability risks. A key issue is whether the reinsurance industry can provide
cover for risks that might otherwise be avoided by the primary sector, even
though they are some of the fastest growing areas of demand — particularly
longevity risk in the developed world and health cover in the developing world.
■ The shift
to a lower interest rate and inflation environment means that the reinsurance
industry is unlikely to be able to generate the kind of investment returns it
has enjoyed in the past. It follows that improving underwriting performance
will be necessary to deliver the overall returns on equity necessary to attract
new capital from investors searching for yield.
■ Regulation of the
reinsurance industry is increasing, but in piecemeal fashion and without
agreement on key techniques and parameters. Well-conceived, internationally
consistent regulation, together with consistent legal, accounting, auditing,
and actuarial practices, would provide a foundation for a healthy, growing,
adequately capitalized industry and reduce the risk of financial instability.
The challenge is to ensure that practice in these areas, both by the industry
and by regulators, evolves in the right direction.
■ There is a
widespread perception that publicly available information about both the
financial state and the risk profile of reinsurance companies is in many cases
inadequate. Although several companies have taken significant initiatives in
this area, there are serious technical issues to be resolved before meaningful
disclosure processes are available for the industry as a whole. But the higher
profile of the reinsurance industry has reinforced demand for increased and
timelier public disclosure, especially concerning information about the
consolidated position of reinsurance groups.
■ Likewise, rating
agency capital models, on the basis of which the capital and financial strength
of individual insurers and reinsurers are evaluated, are not always transparent
or clearly based on sound risk-based analytics.
MARKET
STRUCTURE OF REINSURANCE
|
The worldwide
reinsurance industry consists of about 150 active providers of reinsurance, who
received total premiums of nearly $168 billion in 2004. Non-life premiums
accounted for 80% of the total (over $134 billion) and life premiums for the
remaining 20% (over $33 billion) (see Table 1). Non-life reinsurance premiums
(“ceded” premiums) represented over 11% of the premiums received by the primary
non-life sector, against 2% for the life sector.
Approximately half
(51%) of total reinsurance premiums arise in North America. Western Europe
accounts for about a third (31%) and the remaining 18% come from other regions.
The reinsurance business is dominated by specialized reinsurance companies
concentrated in a small number of financial centers. Table 2 lists the top 35
global reinsurance groups. Just over 92% of reinsurance premiums are ceded to
reinsurers in eight countries: Bermuda, France, Germany, Ireland, Japan,
Switzerland, the United Kingdom, and the United States. In less developed
markets there are often just one or two reinsurers, which typically focus
mainly or exclusively on their local markets.
Six of the eight
countries — including Bermuda, Germany, and Switzerland — are net exporters of
reinsurance services, while the United States has traditionally been the
largest net importer of reinsurance. The Japanese insurance industry also cedes
more business overseas than it assumes from other markets, but most Japanese
reinsurance business is domestic: more than 70% stems from pooling arrangements
for the country’s compulsory automobile liability insurance.
Traditionally the
international reinsurers based in Germany, Switzerland, and France have
accounted for a large proportion of global reinsurance capacity. Their share
totalled 44% in 2003, according to statistics collected by the International
Association of Insurance Supervisors (IAIS). Despite some difficulties at the
start of the 1990s, the London market is also still a very important trading
center for reinsurance coverage. While together Lloyd’s and the London-based
(re)insurers make up only 8% of worldwide capacity, this relatively low figure
belies the importance of the London market because much of the business placed
with insurers and reinsurers in Europe and elsewhere is transacted by London
market brokers and intermediaries. In the last decade, Bermuda and Ireland have
emerged as important offshore centers for all kinds of reinsurance cover.
Moreover, it is likely that Bermuda’s net export position is understated, given
that only two Bermudian companies are captured within the IAIS statistic
The Use of
Quantitative Tools to Measure Catastrophic Risk
|
With
advances in computer technology, new quantitative
tools have been developed to help manage
catastrophic risk. Geographic information systems
have allowed companies to resurrect the “pin
maps”, with significant additional abilities.
But, well beyond merely looking at exposures,
catastrophe simulation models have given us
the ability to estimate potential losses in a way
that truly reflects the long term
frequency and severity distributions. As actuaries,
we knowthatexpected catastrophic losses and
reinsurance decisions should not be based
upon past catastrophic losses. Insured loss
data from catastrophes has been captured for
roughly the last 45 years. Severe hurricanes
and earthquakes are so relatively infrequent that
this body of experience cannot hope to
represent the scope of potential occurrences.
Also, the distribution of insured properties
has changed dramatically over time with the
population movement towards the Atlantic and Gulf
Coasts and earthquake-prone areas of California. Clark
[1] and Friedman [2] have shown us alternative
methods for determining catastrophe losses through
the use of simulation modeling. This involves simulating
the physical characteristics of a specific catastrophe,
determining the damage to exposures, and
calculating the potential insured losses from
these damages. While specific catastrophe
simulation models are different, they all
operate within a simple framework. These
three steps, which we named the Science
Module, the Engineering Module, and the
Insurance Coverage Module, will be discussed
after we discuss the most important
component of catastrophe modeling: The Exposure. 85 The Exposure
All discussions of catastrophic exposure management
must begin with the accuracy and
availability of exposure data. The most sophisticated,
complex catastrophe modeling systems cannot estimate
an insurer’s losses if the insurer cannot
identify what insurance coverages have been
written and where those risks are located.
Company exposure databases vary considerably.
The decisions to retain exposure information may
be based on statistical agency, rate
filing, or management information requirements. Budget
restraints have also contributed to the
designs of some exposure databases. Catastrophe
exposure management considerations are almost
always of secondary importance. Exposure
information can be separated into two
categories: physical characteristics and insurance
coverage.
Physical
characteristics may include:
· type of risk
· location
· Construction
· Number of stones
· age of
risk
· Number of risks
86 The
type of risk can be described in insurance
terms through the line of business,
classification and type of policy codes. The
line of business codes can distinguish
between personal property, commercial property,
personal automobile, commercial automobile,
personal inland marine, commercial inland marine,
business owner, or farm owner policies.
Classification
codes can distinguish the type of risks such
as signs, boats, livestock, inventories, etc.
The type
of policy code can distinguish between
different types of commercial policies (mercantile,
contracting, motel, office, apartment, etc.). The
quality of location data available from
companies varies substantially. Often, the location
recorded is the billing location, rather
than the location of the property insured.
While this may be only a moderate problem for
personal lines, it can cause major
distortions when modeling commercial lines. For a
more complex commercial policy, many of the
locations will not be identified. This may
cause a false measure of concentrations at the
billing location, while understating other areas.
Some companies cannot provide location
detail at zipcode or street address. Location on
a county or state detail can be spread to
finer detail using population densities or
census data, but this can lead to severe
distortions in measuring the concentrations for a
specific insurance company. Insurance companies
must be encouraged to retain fine location
detail. Future exposure location identification
could use the latest satelite technology
(global positioning systems) to determine exposure
locations within a few feet.
Insured
coverage data may include:
· coverage type
· coverage
amounts
· Replacement cost
provisions
· Insurance-to-value
provisions
· deductibles
· co-insurance
· reinsurance
Coverage type
distinguishes the type of insured exposure such
as buildings, contents, appurtenant structures,
vehicles, business interruption, etc.
Replacement
cost and insurance- to-value provisions identify
those provisions where the insurance
coverage may be greater than the
specified coverage amount.
Deductibles,
co-insurance, and reinsurance provisions can
reduce the insured loss to the company.
Once
exposure data is deemed to be reasonable, the
modeling process can begin.
Now briefly
discuss the three modules in any
catastrophe simulation model.
The Science Module
The first
module simulates the natural phenomenon
(i.e., hurricanes, storm surge, earthquakes, fire
following earthquake, tornadoes, hail, winter
storms, etc.). The events can usually be described
through a series of scientific equations
and parameters that determine the resulting
force that causes damage. For hurricanes, numerous
models exist to estimate wind speeds at risk
locations caused by specific storms. A sample of a
simplistic hurricane function might look like this:
Wz =
f(dp, r, s, I, a, t)
Where
Wz = Wtnd speed at location z,
dp = Ambient
pressure minus central pressure
r = Radius of
maximum winds
s = Forward
speed of the storm
I = Land fall
location (longitude, latitude)
a = Angle of
incidence at landfall
t = Terrain or
roughness coefficient at location z .
The Engineering Module
The
engineering module is used to determine the
exposure damage resulting from the wind speeds or
shaking intensities. Wind and earthquake engineering provide the research
to
determine
these relationships. We can express these functions as
follows:
Pz,c,a,s,v =
f( Wz, c. a, s, v), for hurricane or
Pz, c, a, s, v = f(
Iz, c, a, s, v ), for earthquake
where
Pz,c,a,s,v = Percent damage at location z for
risk characterized by c, a, s and v
· c = Construction of
building
· a = Age of building
· s = Number of stones
· v = Coverage,
i.e.. building, contents, time element
If we
apply these damage percentages to the
exposed properties from an insurance company’s
database, the result will be an
estimate of the total damage to those properties
caused by the catastrophe being simulated.
Dz,c,a,s,v =
Ez,c,a,s,v x f( Wz, c, a, s, v) for hurricane
= Ez,c,a,s,v x f( lz, c, a, s, v) for earthquake
where
Dz,c,a,s,v = Damage at location z for risk
characterized by c, a, s, v
Ez,c,a,s,v =
$ exposure at location z for risks characterized by c,
a, s, v
Damages
can vary by more than just construction
type, number of stories, age of building, and
type of coverage (e.g., regional construction
practices, building code and building code
enforcement, occupancy use, surrounding terrain,
etc. ).
The insurance Coverage Module
The
last module translates the damaged exposure
into insured damaged exposure. This includes
reflection of limits, replacement cost
provisions, and insurance-to-value provisions. This
module also includes loss reduction provisions
such as deductibles, co-insurance, and reinsurance.
IDz,c,a,s,v =
f( ( Dz,c,a,s,v ), r, d, I )
where
IDz,c,a.s,v = Insured damage at location z for risk
characterized by c, a, s, v
Dz,c,a,s,v
= Damage at location z for risk characterized
by c, a, s, v
r =
Guaranteed replacement cost multiplier
d = Deductible
I =
Reinsurance limit
a = ALAE
percentage
--------
