Systemic and Catastrophic Risk Are we hanging by a thread or securely tethered? David Alexander, Swiss Re ASHK Appointed Actuaries Symposium Symposium, November 2010
Agenda Insurers' contributions to systemic risk in our financial systems Systemic and catastrophe risk in the insurance industry Risks Mitigants Conclusions
2
Crisis hit insurers much less than banks Cumulative credit26% losses from 2007 2’000
USD bn
USD 1’715 BN 1’500
1’000 1 000
Cumulative credit losses from 2007, as % of 2006 Shareholders Equity 80%
60%
x6
40%
500
68%
x 2.6 26%
20%
USD 271 BN 0
0%
Insurers
Banks
Insurers
Banks
Source: Geneva Association Systemic Risk Report 2010 based on Marsh EMEA Insurance Reports 2007, 2008 and 2009, Bloomberg (as at 10 February 2010), DataStream, Oliver Wyman analysis
3
The iinsurance and Th d reinsurance i iindustries d i weathered the financial crisis fairly well
Business was conducted as usual: cover was always provided both in insurance and reinsurance, and claims were paid as usual throughout the crisis. i i
Prices and capacity remained stable in most lines of business.
There was no experience Th i off ““runs on iinsurers”” and d llapse rates iin lif life insurance remained stable.
Most capital was lost on assets, but (re)insurers remained solvent throughout the crisis
Problems were with monoline insurers involved in credit business and (re)insurers with FS operations
Very few insurers needed government support – and the support needed was minimal compared with banks
Approach for assessing systemic relevance of risk activities) Criteria considered for assessing the systemic relevance of risk activity
Risk activities
Transmission mechanisms Transmission mechanisms
Specific considerations when assessing the criteria •D Does a transmission t i i mechanism h i exist? i t? • What are the triggers? • What is the likelihood of these triggers being breached? • Can the risk activity cause significant losses? • Can the activity be substituted by someone else? • How big is a potential impact?
Size/Impact
Speed
Systemically relevant risk activities activities
• How fast does the loss materialise? • How fast does the loss transmit?
Risk activities are deemed systemically relevant if they pass the risk assessment
5 Source: Oliver Wyman analysis
Timing of transmission between insurers is slower than between banks Timing of World Trade Centre Insurance Claims Cumulative proportion of claims made
% of to otal net claims paid
100% 80% 60% 40% 20%
Quarters after 11 Sept 2001 Quarters after 11 Sept 2001
Gradual cash flows S Source: R Reinsurance i A Association i i off America, A i C Catastrophe h L Loss D Development l S Study d
6
Q2 2008
Q4 2007
Q2 2007
Q4 2006
Q2 2006
Q4 2005
Q2 2005
Q4 2004
Q2 2004
Q4 2003
Q2 2003
Q4 2002
Q2 2002
Q4 2001
0%
Insurers core Insurers’ core-activities activities are not systemically relevant Risk activities not deemed systemically relevant A t Liability Li bilit M Managementt A o Asset
& Strategic Asset Allocation
o Underwriting catastrophe risks i k B o Underwriting long term risks o Writing business with redemption p options p o Writing life business with embedded guarantees o Hedging with derivatives o Reinsurance and retrocession C o Insurance linked securities and derivatives
D o Long-term capital raising E o Credit insurance
Systemic relevance rejected for at least one of the following reasons
Size: Limited size o No disruptive effects on financial markets Interconnectedness o Level / intensity of interaction g does not create contagion Substitutability o Given ability of market to find a substitute Timing: Slow speed of impact o Allows insurers to absorb impact, through capital raising or orderly wind-up
7
In conclusion…
!
Size needs to distinguish between absolute concentration of exposure or dilution of risk through diversification
!
Interconnectedness needs to allow for the materiality of the connections compared to the overall balance sheet
!
Substitutability needs to distinguish between the uniqueness of any individual market participant, and the ability of a market as a whole to be re-capitalised
!
Timing of transmission between insurers is significantly slower than between banks, allowing mitigation measures that dampen systemic risk
Thus, we argue that the criteria need to be considered against specific risk activities, rather than institutions 8
Agenda Insurers' contributions to systemic risk in our financial systems Systemic and catastrophe risk in the insurance industry Risks Mitigants Conclusions
9
Significant risk is all around us
Pollution – environmental and health risks
Euthanasia – Acceleration of Claims Payouts
M k t Risk Market Ri k
Terrorism
N t lC Natural Catastrophe t t h
P d i Pandemic
Advancements in Medical Technology – Cost per Intervention
Asbestosis
10
Naturall C N Catastrophe h Large Loss Potential E h Earthquake k
T i l Cyclone Tropical C l Extra tropical Cyclone Flood
11
Insured Nat Cat vs. vs man man-made made losses 1970–2009 2005: Hurricanes Katrina, Rita, Wilma
USD bn, at 2009 prices
50
• Increasing values
• Concentration C i iin exposed d areas
40
• Insurance penetration
1992: Hurricane Andrew
g g hazard • Changing
30
2004: Hurricanes Ivan, Charley Charley, 1999: Frances Winter storm Lothar
2008: Hurricanes Ike, Gustav
- climate variability - climate change
1994: Northridge EQ
20
2001: Attack on WTC
10 0 1970
1975
1980
Earthquake/tsunami
1985
1990
1995
Weather-related Nat Cats
2000
2005
Man-made disasters Source: Swiss Re, sigma No 1/2010, Figure 3
12
The most costly Th l iinsurance llosses in 2009 Insured loss 1 (in USD m)
1 2
Victims
2
Date (start)
Event
Country
3 372
25
24.01.2009 Winter storm Klaus, winds up to 170 km/h, heavy rain France, Spain
1 350
15
10.02.2009 Thunderstorms, winds up to 145 km/h, hail
US
1 193
11
23.07.2009 Hail storm Wolfgang, winds up to 130 km/h
Switzerland, Austria, Poland et al
1 130
2
09.04.2009 Tornadoes, storms, winds up to 105 km/h, hail
US
1 079
173
07.02.2009 Victorian bush fires, winds up to 100 km/h
Australia
1 050
1
10.06.2009 Thunderstorms, winds up to 128 km/h
US
995
6
25.03.2009 Thunderstorms with hail
US
800
-
20.07.2009 Storms, heavy rain, hail
US
760
2
26.05.2009 Hail storm Felix, winds up p to 90 km/h
France, Germany, y Belgium g
615
5
08.10.2009 Typhoon Melor/No 18, winds up to 204 km/h
Japan
570
-
07.05.2009 Storms, thunderstorms, winds up to 145 km/h; floods
US
Property and business interruption, excluding liability and insurance losses D d and Dead d missing i i
Source: Swiss Re, sigma No 2/2010 table 8
13
Cat Risk - Natural Perils across Asia
14
Hazard H dM Modelling d lli : 4 Box Principle
Natural Hazard
Example H icane Hurricane “Charley” Aug 2004
Vulnerability
Ocean Drive Drive, FL FL, 1926
What can happen?
How do the exposed values react? Ocean Drive, FL, 2000
Distribution Di ib i off property values
Insurance conditions
What is affected?
Limit
Deductible
How is the exposure covered? 15
Case study Caribbean: C ibb Caribbean C Catastrophe h Ri Risk k Insurance Facility (CCRIF) Solution features The CCRIF offers parametric hurricane and earthquake insurance policies to 16 CARICOM governments The policies provide immediate liquidity to participating governments when affected by events with a probability of 1 in 15 years or over Member governments choose how much coverage they need up to an aggregate limit of USD100 million The mechanism will be triggered by the intensity of the event (e.g. (e g winds exceeding a certain speed) speed). The facility responded to events and made payments: – Dominica & St. Lucia after earthquake (2007) – Turks & Caicos after Hurricane Ike (2008) – Haiti (2010) Involved parties Reinsurers: Swiss Re and other overseas reinsurers Reinsurance program placed by Aon Benfield Ltd. Derivative placed by World Bank Treasury
Background information Caribbean states are highly susceptible to natural disasters and h have only l li limited i d options i available il bl to respond. With small economies and high debt levels, they often depend on donors to finance post-disaster needs, but do bu donor o resources esou ces o often e a arrive e late ae or not at all The CCRIF was launched in June 2007 on behalf of the Caribbean Community (CARICOM) heads of government under the guidance of the World Bank with financial support from international donors CCRIF participating governments are: Anguilla, Antigua & Barbuda, Bahamas, Barbados, Belize, Bermuda, Cayman Islands, Dominica, Grenada, Haiti, Jamaica, St Kitts & Nevis, St Lucia, St Vincent & the Grenadines, Trinidad & Tobago, Turks and Caicos Islands
16
CCRIF - Present structure The CCRIF is the first multi-country risk pool in the world, and is also the first insurance instrument to successfully develop parametric policies backed by both traditional and capital markets.
Currency: USD
90m
Losses of up to USD 12.5m is retained by CCRIF. USD 132.5m is paid by the reinsurance contributed by the international reinsurers -
Reinsurance
132.5m
Cat Swap
Next USD 30m of losses are reinsured with reinsurance companies.
30m
12.5m 12.5m
First, USD 12.5m of losses are reinsured with reinsurance companies. companies
R i Reinsurance
Two- third of the top p layer y i.e. USD 60m is paid p byy reinsurance companies,
Reinsurance
USD 30m (i.e. one-third) in the top layer of risk is placed into capital market via a catastrophe swap between CCRIF and the World Bank Bank.
CCRIF Source: CCRIF Annual Report 2008-09 2008 09
The group retains all the losses above USD 145m. CCRIF is backed by donor funds held by the World Bank as a Multi-Donor Multi Donor Trust Fund. Fund 17
Nat Cat underwriting – key take aways
Be mindful of loss accumulation potential due to natural catastrophe.
Come up with MPL scenarios. scenarios Control your exposure or buy adequate protection.
Use site specific risks factors to identify best risks from Nat Cat perspective
Identify and price for Nat Cat risks.
Introduce Nat Cat specific deductibles/sublimits in your policies. Extremely important for low frequency perils in any market
Financial Fi i l and dM Market k Ri Risk k 10-year Bond Yield History
Hong Kong
Japan
8.22% 10.177%
0.901%
2.091%
19
Interest rates STILL on the way down! ASIAN MARKETS COUNTRY
5YR GOVT YIELD LAST CLOSE
10YR GOVT YIELD YTD Δ (%)
LAST CLOSE
YTD Δ (%)
Taiwan
0.88
-13.51
1.22
-21.12
Hong Kong
1.12
-41.78
2.07
-19.66
Singapore
0 86 0.86
-32 32.81 81
2 00 2.00
-24 24.81 81
Vietnam
10.44
-10.67
11.03
-3.65
Malaysia
3.32
-12.40
3.81
-10.35
South Korea
3.75
-23.78
4.18
-22.45
Thailand
2.51
-30.83
2.98
-28.75
Indonesia
6.43
-28.34
7.16
-28.85
Philippines
5.11
-19.64
5.94
-26.74
India Australia
7.79 7 79 4.99
7.32 7 32 -3.49
8.14 8 14 5.16
7.26 7 26 -8.65
New Zealand
4.38
-20.15
5.10
-12.27
China
3.03
1.34
3.66
0.55
Japan
0.30
-37.00
0.90
-30.73
as of 22 Oct 10
Source: Bloomberg 20
Financial Market Issues
Falling interest rates cause significant strains on life companies – Matching is not perfect – Liabilities often long
Equity and bond market falls can deplete solvency capital and force more capital to be held – there are trade offs and compromises between regulatory and economic focus
Inability to raise capital may hamper growth for some products
Cost of guarantees is higher
Furthermore, poor economics can be linked to – poor claims eg disability, mortality (suicide) – higher lapsation as customers adjust their finances 21
Pandemics – are you ready?
Page 22
Influenza pandemics can cause one-time one time mortality shocks
Mortality has generally been improving for many decades
But life insurers still face the risk that an influenza p pandemic could cause a one-time mortality shock
Pandemics are infrequent and unpredictable
It is not easy to specify in advance the loss value from such an event …
… and therefore the amount of capital to hold
This makes it difficult for life insurers to quantify the risk and to manage p efficientlyy their capital
At the same time… – regulatory regimes are forcing insurers to consider this threat – there is a heightened general awareness of pandemic in the population – capital p markets are becoming g more aware of mortalityy risk
H5N1 was a near miss Emergence of a new virus sub-type
Global population has little or no immunity
New virus is capable of replication in humans and causes serious illness
New virus is capable of efficient human-to-human transmission
H5N1
x
Emerging and re-emerging infectious diseases
Influenza A H1N1
A Assess the h exposure Insurance penetration i 2008 14%
Real premium growth Emerging Asia Life
2009 6.7%
Non-life
Non-life p premiums, 2008 (USD10 bn)
Taiwan
12%
Life premiums, 2008 (USD10 bn)
10% Hong Kong
Japan
8%
14.2%
South Korea
6%
Singapore
Australia Australia
4%
J Japan
2% 0%
S Korea
Taiwan
Singapore
Malaysia Thailand
India Vietnam
Indonesia Philippines
-2% -10%
Malaysia
0%
Average life penetration India China
Average non-life penetration Vietnam
China
Indonesia Philippines
Hong Kong
Thailand
10%
20%
30%
40%
Average real premium growth (1999 (1999-2008) 2008) Sources: Oxford Economics, Swiss Re Economic Research & Consulting.
26
Modern p pandemic with characteristics of the 1918 influenza
The graph shows the cumulative effect ff on mortality rates of selected changes between 1918 (the “Base”) Base ) and 2006
27
Sensitivity of outcome to plausible changes h iin variables i bl 1 of 2
28
Sensitivity of outcome to plausible changes h iin variables i bl 2 of 2
29
Agenda Insurers' contributions to systemic risk in our financial systems Systemic and catastrophe risk in the insurance industry Risks Mitigants Conclusions
30
Mitigants •Asset and Liability Management Enhancing capital position to make insolvency less likely – •Enhancing consider reinsurance as an efficient source of such capital •Adjusting Adjusting dividend policy – dividends for both policyholders and shareholders •Writing g business with less risk or diversifying y g risk eg g investment linked products and protection products •Product Design & Surrender value Policy •Hedging and Reinsuring excess risks to counterparties or capital markets
31
Insurance regulatory developments in Asia means our exposure will ill become b more transparent Higher minimum capital requirements Adoption Low minimum of RBC and solvencyAdoption of RBC solvency systems solvency capital p systems t requirements in the past. Introduction of dynamic stress Higher tests and the use of scenarios statutory requirement push capital to Increased focus on consumer protection (eg the forefront forefront, policyholder post GFC. protection funds) Rating capital ((usuallyy higher) g ) will become Alignment of accounting standards to IFRS more important.
Regulatory uncertainty still high
Consolidation to accelerate in some markets
A driver of M&As
Impact on insurance portfolio as well as on asset management
Increase demand for reinsurance
Regional harmonisation
Conclusions
Like it or not our companies are exposed to significant catastrophic risks
We are g getting g better at p perceiving g and mitigating g g that risk…
…but are we yet good enough… – for our policyholders? – for our shareholders? – for our own high standards?
33
Thank you
Legal notice ©2010 Swiss Re. All rights reserved. You are not permitted to create any modifications or derivatives of this presentation or to use it for commercial or other th public bli purposes without ith t th the prior i written itt permission i i off S Swiss i R Re. Although all the information used was taken from reliable sources, Swiss Re d does nott acceptt any responsibility ibilit ffor th the accuracy or comprehensiveness h i off the details given. All liability for the accuracy and completeness thereof or for any damage resulting from the use of the information contained in this presentation is expressly excluded. Under no circumstances shall Swiss Re or its Group companies be liable for any financial and/or consequential loss g to this presentation. p relating
35
