Transcript 204 kb PowerPoint presentation

Allocating Reinsurance Capacity to
Emerging Markets:
A View from Bermuda
Financing the Risks of Natural Disasters
Washington DC, June 3rd, 2003
Sean Ringsted
Phone 441-298-9519
E-Mail [email protected]
ACE Global Reinsurance
The Bermuda Market
Company Landscape
GPW
12,000
6,000
8,000
4,000
6,000
3,000
4,000
2,000
2,000
1,000
Renaissance Re
PXRe
Partner Re
Montpelier
IPCRe
Endurance
Axis
-
XL Capital
5,000
$ Millions
Capital & Surplus
10,000
AWAC

7,000
Arch

14,000
ACE

The Bermuda (re)insurance
market has a significant
presence in the world-wide
insurance market.
Approximately, >$26bn in
capital & surplus (excluding
captives and life operations)
and >$31bn of premiums
written (end of 2002).
Companies established postAndrew (1993) and post-WTC
(2001).
Market characterized by large
risk appetite, “clean” balance
sheets and technical
approach.
$ Millions

-
* Data from Deloitte and Touche Bermuda Insurance Survey 2003
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
2
ACE Global Reinsurance
The Bermuda Market
ACE Tempest Re Background

Formed in September 1993, following capacity crunch caused
by Hurricane Andrew, with $500m capital.
 Bought by ACE July 1996 for total consideration around $950m.
 In 1998, ACE acquired Cat Ltd and merged with Tempest.
 Commenced diversification plans in 2002; developing into
multi-line reinsurance operation in Bermuda, Stamford, London
and Dublin.
 Tempest Bermuda is the property catastrophe carrier (~$470 of
gross premiums written in 2003).
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
3
ACE Global Reinsurance
The Bermuda Market
ACE Tempest Re Business Approach
 Only accept risks that can be modeled and quantified:
 Requires detailed exposure data
 Understanding of client’s business plan
 Retrocession on an indemnity basis is less transparent (= less
attractive)
 Consistent, pricing (and capacity) through use of
probabilistic catastrophe simulations models to provide
objective risk assessment.
 Risks must be written to provide profitable return on capital
deployed.
Significant
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
capacity for risks that meet return hurdle.
4
ACE Global Reinsurance
The Bermuda Market
ACE Tempest Re Business Approach
 Flexible product structuring:
 Indemnity versus derivative
 Parametric (for example; wind speed, spectral acceleration)
 Catastrophe Bonds
 Swaps
 Multi-Year
 Support for “State”-sponsored catastrophe protection
initiatives:

California Earthquake Authority
 Turkish Catastrophe Insurance Pool

Norwegian Natural Perils Pool
 New Zealand Earthquake Commission
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
5
ACE Global Reinsurance
Risk Transfer and Management
Sources of Catastrophe
 Natural catastrophes

unlikely to cause surprise by cause

uncertainty around return period and severity

earthquake, hurricane, flood, tornado

meteors, tsunami, volcano
 Man-made catastrophes
Sean Ringsted
World Bank Conference, DC
June 3rd 2003

potential to cause surprise by source

uncertainty around most parameters

fire / explosion (Bhopal, Chernobyl) , aviation, terrorism

asbestos, asset-bubbles, cyber-risk, global warming
6
ACE Global Reinsurance
Risk Transfer and Management
Catastrophe Modeling
 Computer
catastrophe simulation software market developed in
late 1980s and early 1990s.
 These
models are event-based (not statistical) and exposure-
driven:

Allows for correlation within a portfolio

Allows portfolio weighted marginal capital allocation
Multiple models give better answers, but……..
Better understanding of the models leads to better risk assessment


 Large portion of reinsurance capacity determined by answers
from cat models, so it is imperative that models are available for
covered risks and that the results are understood.
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
7
ACE Global Reinsurance
Risk Transfer and Management
The Insurance Market and Model Introduction
US HU & TH
35
US EQ
EU WS
JP TY & EQ
UK FL
16
14
30
10
25
8
20
6
4
15
2
10
0
Global Non-Life Premium Growth (%)
Nat. Cat. Insured Losses (USDb, 2002)
12
-2
5
-4
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
-6
1980
0
Data from Swiss Re sigma reports, model release dates from Applied Insurance Research
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
8
ACE Global Reinsurance
Emerging Market Risk Transfer
Inequality of Risk and Insurance
Annualized Average Nat. Cat. Fatalities (per 1m pop.)
1000
Bangladesh
Fmr. Soviet Union
China
100
Honduras
Hong Kong
Venezuela
Canada
Turkey
10
Taiwan
Jamaica
UK & US
1
0.1
0.01
0
1
2
3
4
5
6
Non-life Premium (% of GDP) [2001]
Premium data from Swiss Re sigma reports, GDP data from US government sources, population data from US
Census Bureau, catastrophe data from the Center for Research on the Epidemiology of Disasters.
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
9
ACE Global Reinsurance
Emerging Market Risk Transfer
Inequality of Risk and Insurance
Annualized Average Nat. Cat. Fatalities (per 1m pop.)
1000
Bangladesh
Fmr. Soviet Union
China
100
Honduras
Hong Kong
Venezuela
Canada
Turkey
10
Taiwan
Jamaica
UK & US
1
0.1
0.01
0
1
2
3
4
5
6
Non-life Premium (% of GDP) [2001]
Premium data from Swiss Re sigma reports, GDP data from US government sources, population data from US
Census Bureau, catastrophe data from the Center for Research on the Epidemiology of Disasters.
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
10
ACE Global Reinsurance
Emerging Market Risk Transfer
Catastrophe Modeling

G7 and major western nations are well covered by natural catastrophe
models.

Aside from economic wealth to pay for the models, it is assumed that
these models can be built because these nations have good temporal
and engineering data on catastrophes and their impact.

However, several of the current cat models are based on surprisingly
short datasets. For example, records of tornado events in the US only
stretch back to the 1950’s, as do records of river flooding in the UK.

In the developing world, it has been assumed that models are difficult
to build because data on historical catastrophes is scarce or nonexistent. Recent studies challenge this assumption.
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
11
ACE Global Reinsurance
Emerging Market Risk Transfer
Catastrophe Modeling

Documentary evidence – indigenous and colonial
administrations contain a vast store of information, some of
which can be used to ‘re-construct’ historical catastrophes

Some recent examples of applied research:

Records from the Dutch East India Company (Verenidge Oostindische
Compagnie) have been used to reconstruct the effects of the Krakatoa eruption in
1883 (Winchester, 2003).

Spanish colonial and religious records were used to reconstruct historical
typhoons in the Philippines (Udías, 1996; García et al, 1999).
The frequency of land-falling typhoons of the last 500 years in Guangdong
Province, China, have been reconstructed from provincial, county and city
records (Qiao and Tang, 1993).



Sean Ringsted
World Bank Conference, DC
June 3rd 2003
Windstorm events in the North Sea stretching back 500-600 years have been
reconstructed by Lamb (1991).
The US hurricane records has been enhanced with the use of documentary
evidence to push the official record back to 1850 (Landsea et al, 2003)
12
ACE Global Reinsurance
Emerging Market Risk Transfer
Catastrophe Modeling

These and many other sources of data still await further
examination. Such data sources may include British Admiralty
records, British colonial records from India and Africa and
company records (such as the East India Company and Lloyds),
and Chinese Imperial Court Records, amongst others.

In addition to historical records, catastrophes leave physical
evidence of their passing:


Tropical cyclones can leave sedimentary evidence in particular locations as
storm surge washes sand from beaches over land (Liu and Fearn, 2000;
Donnelly et al, 2001).

Volcanoes can leave sedimentary evidence of their eruption (Winchester,
2003).
As well as this historical data, the Center for Research on the
Epidemiology of Disasters maintains an ongoing database of
catastrophes stretching back before 1900.
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
13
ACE Global Reinsurance
Emerging Market Risk Transfer
Catastrophe Modeling
 In addition to peril information, catastrophe models also require
complex engineering and insurance modules to provide a
complete assessment of loss potential.
 Whilst there is a rich and largely untapped source of historical
data for emerging market countries, data to support the other
modules is less researched.
 Even with “mature” US models:
Considerable calibration
and parameter risk issues
 Flood poorly modeled

Peril B - View 3
Peril B - View 2
Peril B - View 1
Peril A - View 3
500 Year Return Period
Peril A - View 2
250 Year Return Period
 These problems severely
100 Year Return Period
compounded for “emerging
market” regions.
Peril A - View 1
-
100
200
300
400
500
Loss (USD Ms)
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
14
ACE Global Reinsurance
Conclusions
 The Bermuda (re)insurance market offers:

Significant risk-appetite and capital to support risk tolerance

Proven track-record of structuring and product innovation

Regulatory environment and underwriting flexibility to support various risk
transfer structures (indemnity, parametric, multi-year, swaps)
 Emerging Market Risk should be attractive:

Developed / primary markets have relatively mature risk transfer
mechanisms so emerging markets offer growth opportunities

Risk diversification away from core US / European / Japan perils
 However, challenges with:

Objective and transparent risk assessment

Peril Modeling (particularly flood) and Data (insurance, engineering)

Moral Hazard (building-code enforcement, policyholder)

Lack of insurance penetration to pay for risk transfer alone
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
15
ACE Global Reinsurance
References
Winchester, S; Krakatoa : The Day the World Exploded: August 27, 1883; Harper-Collins, 2003.
Udías, A; Jesuits’ contribution to Meteorology; Bull. Am. Met. Soc., 77, 2307-2315, 1996.
García, R.R., et al; Atmospheric circulation changes in the tropical Pacific inferred from the
voyages of the Manila galleons in the 16th-18th centuries; Bull. Am. Met. Soc., 82, 2435-2455,
2001.
Qiao, S.-X. and W.-Y. Tang; Collection and Research of Climate Data from Historical Records in
the Guangzhou Area; Guangdong People’s Press, Guangdong, China, 1993.
Lamb, H; Historic Storms of the North Sea; Cambridge University Press, Cambridge, UK, 1991.
Landsea, C.W. et al; The Atlantic hurricane database re-analysis project : Documentation for
the 1851-1910 alterations and additions to the HURDAT database; Hurricanes and Typhoons :
Past, Present, and Future, R.J.Murnane and K.-B. Liu Editors, Columbia University Press
Accepted and revised for publication, 2003.
Liu, K.-L. and M.L. Fearn; Reconstruction of prehistoric landfall frequencies of catastrophic
hurricanes in Northwestern Florida from lake sediment records; Quaternary Research, 54, 238245, 2000
Donnelly, J.P., S. Smith Bryant, J. Butler, J. Dowling, L. Fan, N. Hausmann, P. Newby, B.
Shuman, J. Stern, K. Westover, and T. Webb III; 700 yr sedimentary record of intense
hurricane landfalls in southern New England; GSA Bulletin, 113 (6), 714-727, 2001
Sean Ringsted
World Bank Conference, DC
June 3rd 2003
16
ACE Global Reinsurance