Insurance: Limiting the Impact of Natural
Catastrophes on the Balance Sheet
Dr. Oliver Kübler
Dr. Matthias Schaub
“This has only been possible by insurers. They are the ones who really built this
city. With no insurance, there would be no skyscrapers. No investor would finance
buildings that one cigarette butt could burn to the ground” (Henry Ford)
Engineering
Insurance
Finance
Insurance enables technological progress
2
Losses in USD bn
Costs of Natural Catastrophes & the Protection Gap
Protection
gap
3
Global Insurance Industry Losses
in USD mn
Man-Made1)
Nat Cat
15 yr average
annual loss burden
6,406
43,565
15 yr highest
annual loss burden
9,432
127,595
1)
excl. terror
• Variance or volatility of Nat Cat loss is much higher than man-made, so
called long-tail distribution for cat events
• Because of high volatility and sometimes very low frequency of events,
sophisticated models are require to better understand the solvency events
4
(Re)insurance makes Society more Resilient
Christchurch, New Zealand 2011
“Extensive offshore reinsurance will fund a substantial share of the
rebuild costs and thus has helped reduce the financial impact on New
Zealand.”
“Comparisons with other major natural disasters suggest that the widespread
coverage of insurance, particularly accompanied by reinsurance overseas,
has helped to mitigate the longer-term adverse economic effects of the
earthquakes.”
Reserve Bank of New Zealand
5
(Re)insurance makes Society more Resilient
Country
Chile 2010, M8.8
Haiti 2010, M7.0
Shake map
maps shown in same scale
Fatalities
562
222,570
Economic loss:
USD 30 bn (12% of GDP)
USD 10 bn (136% of GDP)
GDP (2011):
USD 249 bn
USD 7.3 bn
6
Nat Cat Modelling and
Control of Losses
7
NatCat risk assessment: Premium income vs. losses
Fire:
Natural catastrophes:
8
The 4-Box Principle - Earthquake
Hazard
How often?
How strong?
Vulnerability
What degree
of damage?
Value
Distribution
What is
covered where?
Insurance
Conditions
What is
covered how?
9
Hazard
Vulnerability
Value
Distribution
Insurance
Conditions
Historic earthquakes in Japan
• Magnitude > 5.0
• Depth < 70km
• Years 0 to 2013
Visualized in SwissRe CatNet
Hazard
Vulnerability
Value
Distribution
Insurance
Conditions
Generating a Probabilistic Event Set
Fault
• 240 fault models
• Model parameters (geometry,
max magnitude, frequency)
consistent with JSHIS
Interface
• M9-class event
incorporated
Background
• Stochastic background events
being consistent with historical
catalogue
New probabilistic event set combines latest knowledge on Japanese
Earthquake hazard with latest earthquake modeling technology.
11
Hazard
Vulnerability
Value
Distribution
Insurance
Conditions
Hazard Uncertainty
Ground motion equations include hazard uncertainties
(inter and intra event)
ln 𝐼 = 𝑔 𝑀, 𝑅 + 𝜀𝑖𝑛𝑡𝑒𝑟 + 𝜀𝑖𝑛𝑡𝑟𝑎 , 𝜀…normally distributed
Expected loss is a function of the hazard and the
vulnerability/loss function
𝐸𝐿 =
𝐿 𝑖𝑚 𝑓𝐼𝑀 𝑖𝑚 𝑑𝑖𝑚
𝑙𝑜𝑠𝑠
Ω
12
Hazard
Vulnerability
Value
Distribution
Insurance
Conditions
Hazard Comparison to JSHIS
JSHIS
Swiss Re
JMA Intensity, at 475 year return period on amplified soil
13
Hazard
Vulnerability
Value
Distribution
Insurance
Conditions
figure:
Weatherill , Silva, Crowley, Bazzurro (2015)
• The vulnerability curves define the relationship of hazard and damage/loss
• The vulnerability is defined by:
• Hazard intensity
• Coverage
• Risk category (Occupancy)
• Quality/age
•
•
•
•
•
(Protection)/Preparedness
Construction type
Building height
Vulnerability Region
etc.
14
Hazard
Vulnerability
Value
Distribution
Insurance
Conditions
Structure types perform differently during earthquakes, compare e.g. steel and
unreinforced masonry
Source: Kam et al. 2011, Data source: Christchurch City Council
15
Hazard
Vulnerability
Value
Distribution
Insurance
Conditions
• Value distribution is the driver
for
– correlation between assets
– accumulation for insurer
– relevance of protection
measures
– ….
• Accuracy of provided
geographic information
(latitude/longitude) is key
16
Hazard
Vulnerability
Value
Distribution
Insurance
Conditions
• Insurance conditions (deductible, • This will drive the shape of the loss
limits, sub-limits) will be applied to
distribution function and hence the
each simulated event
credit on premium
17
Known Limitations of Earthquake Models
All models have limitations,
– Many limitations are known (by model developers and users)
– Until recently none of the vendor models considered Tsunami!
– Most models neglect or do not consider aftershocks resulting in underestimation
of risk
– Earthquake models, in general, produce more reliable results in assessing risk to a
population of structures and/or with a simpler coverage (e.g. physical damage)
compared to a single risk with a complex coverage (e.g. business interruption)
Exchange between model developers and underwriters is key!
18
Event set based Accumulation Control
Event 1
• We use event-based
simulations
• All insurance conditions can
be applied where appropriate
(coverage, location, policy,
…).
• Output is a huge loss table
called the event loss table
• By doing so correlation can
be considered!
Risk 1
Risk 2
345
BU l
Desk m
Acceptance n
479
BU o
Desk p
Acceptance q
43
76
…
… (large treaty)
…
Event 1'000'000
278
… (multi-location)
Risk 100
Event …
BU i
Desk j
Acceptance k
…
…
Event 2
BU r
Desk s
Acceptance t
Risk 1'000'000
120
..
Tsunami and Volcano Hazard
Tsunami
Global volcano hazard layer
First-in-Industry

Calculated inundation for Tohoku earthquake reflects observations.

New Tsunami model is fully-coupled with earthquake shock model.

Model was available for 2012 renewable and especially helped clients with
coastal exposure

Global Volcano ash fall hazard layer, 2016
20
CatNet – Swiss Re’s Natural Hazard Platform
Mw 7.0, Earthquake, Kumamoto, Kyushu, Japan, 2016-04-16
source: CatNet, USGS
21
Flood Risk & Risk Awareness
22
The 10 largest Flood Events (Insured Losses 1970 – 2015)
In Billion USD (in 2015 prices)
Insured
Loss
Total economic
loss
15.8
48.4
2011
Thailand
2013
Germany & Czech Republic
4.2
16.8
2002
Germany & Czech Republic
3.0
13.8
2007
UK
2.8
4.3
2005
Switzerland
2.6
4.2
2011
Australia
2.4
6.5
1997
Poland & Czech Republic
2.4
7.4
2007
UK
2.3
3.4
2010
Australia
2.2
5.6
1973
US
2.0
5.5
Source: Swiss Re Economic Research and Consulting
23
Don Muang Airport in northern Bangkok, Nov. 3, 2011
Duration:
July - November 2011
Economic loss:
USD 48.4bn (Swiss Re: Sigma-explorer 2016)
Insured loss:
USD 15.8bn (Swiss Re: Sigma-explorer 2016)
24
Flood Central Europe, Germany, June 2013
Duration:
27 May 2013 till 20 June 2013
Economic loss:
USD 16.8bn (Swiss Re: Sigma-explorer 2016)
Insured loss:
USD 4.2bn (Swiss Re: Sigma-explorer 2016)
25
River / Flash Flood: Forecast on precipitation extremes
Increase in 1-in-5 year return intensities for 1 day events
Based on 10 regional climate models (including KNMI's RACMO2.1)
Results imply a rising probability of more intense extreme precipitation events (in autumn, winter
and spring) and extended dry spells in the summer months
Source: Rajczak, J., P. Pall, and C. Schär (2013), Projections of extreme precipitation events in regional climate simulations for Europe and
the Alpine Region, J. Geophys. Res. Atmos., 118, 3610–3626, doi:10.1002/jgrd.50297.
26
River / Flash Flood: Representative client's exposure by
Flood Zone
Sum insured by Zone
3%
• Limited assets within highly
exposed flood risk zones
• However, 3% of the clients
assets will be affected with high
frequency and will drive the
overall expected annual loss
• Suppliers often face the same
issue and create a supply chain
risk for your production process
7%
8%
50 year
100 year
500 year
82%
outside
Expected Loss by Zone
25%
44%
50 year
100 year
25%
500 year
outside
6%
27
River / Flash Flood: Economics of flood protection and risk
management
Expected flood
loss
(no measures)
Expected flood loss
(various measures)
-22%
-10% to -15%
Establish proper
on-site risk
management
incl. BCP
Increase flood
protection to 10% of
assets to 250 yr
level
Increase flood protection of
3% of assets to 100yr level
-16%
Insurance
Self-insured
• Economical flood risk management entails investments into protection and
includes risk management culture and insurance
28
Japan Earthquake BCE Cover
(Business Continuity expense)
29
Commercial EQ Insurance – 3 Standard Components
Claims payable for loss
derived from Physical
Damages caused by EQ,
volcanic eruption and/or
Tsunami
Business
Continuity
Expense
Physical
EXISTING
COVERAGE
Damage
(BCE)
Loss of
Claims payable for Loss
of Profit caused by the
business interruption or
suspension due to predetermined events
Profit
Claims payable for
various expenses
required for business
recovery after an EQ event
REMARKS:
 Basis Risk: miss match
between actual loss and
amount paid
 Declaration of
expected expenses for
policy limit before
inception of coverage
30
Introduction to BCE Insurance
Examples for business
continuity expenses
Pre-Event Conditions (to be agreed)
1
Declaration
Determination
•Expected
expenses
(=policy
limit)
Contractual
•JMA
stations &
parametric
payout
•Define
application
& Premium
Post-Event Conditions (to trigger the ‘immediate’ payout)
Proof of Loss
•Physical damage to
property exceeds
the "franchise
amount"
EQ Intensity Trigger
•JMA stations
recorded the preagreed SHINDO
Intensity
Debris removal and cleanup
Temp. closing & reopening
2 costs, emergency adver-
tisments, announcements
Temporary building
3 instalment, temporary
furnishing
Relocation to alternative
4 facilities, evacuation of
stocks
5
Travel and accommodation
of supporting staffs
6 Damage evaluation
7 Alternative transit, logistics
8
Emergency response to
customers / supply chains
…. and more
31
JMA Station Selection Criteria
largest exposure
(90% profit earnings)
50%
100%
45%
90%
40%
80%
35%
70%
30%
60%
25%
50%
20%
15%
Top
90%
40%
30%
10%
20%
5%
10%
0%
0%
2. Select JMA stations
near the largest profit
earning markets
Tokyo
Osaka
Aichi
Fukuoka
Kanagawa
Hyogo
Hokkaido
Kyoto
Chiba
Shizuoka
Miyagi
Hiroshima
Saitama
Okayama
Tochigi
Ishikawa
% of Total Annual Gross Profit
stations all over
Japan
1. Determine areas of
Cumulative Gross Profit %
0. Approx. 600 JMA
32
Non-Physical Damage Business
Interruption (NDBI) Multiperil cover –
Swiss Federal Railways (SBB)
Swiss Insurance Innovation Award 2015
by ‘Schweizer Versicherung’
NDBI Cover – An Overview
Client
•Swiss Federal Railways (SBB) – major Swiss national railway company
Client’s Needs
•The client identified several 'black swan' type scenarios based on real life past
incidents or as potential risks that could materialize
•Clients internal risk assessment indicated substantial exposure to core businesses
being interrupted for several days or even weeks.
Insurance Solution
•Multiperil Non-Damage Business Interruption Cover
•Coverage: Business Interruption due to
•Imminent Natural Catastrophes
•Non-Performance of Information Systems (Cyber)
•Regulatory Abandonment of Rolling Stock
•3 year policy with an aggregate term limit
34
NDBI Coverage – Covers Imminent Natural Catastrophes
Perils covered
•Windstorm, Strong Precipitation, Avalanches, Mudflow, Rockfall, Landslide
Risk assessment
•Nat Cat risk assessment is based on analytical loss models (frequency and severity), i.e.
comprehensive exposure and scenario analysis as well as actuarial modelling
Example of scenarios to be considered:
•Small flood may trigger imminent natural catastrophe coverage through this policy.
(large flood expected to trigger traditional PD coverage - covered through separate
policy)
•Risk for Rockfall / Landslide / Avalanche
•Impact on Power Generation and Distribution
•and others
35
Wrap up
(Re)insurance and engineering make society
resilient
The world sustains significant Nat Cat losses
every year
Nat Cat risks can be modelled, assessed and
controlled
Taylor made insurance solutions
36
37
Legal notice
©2015 Swiss Re. All rights reserved. You are not permitted to create any modifications
or derivative works of this presentation or to use it for commercial or other public purposes
without the prior written permission of Swiss Re.
The information and opinions contained in the presentation are provided as at the date of
the presentation and are subject to change without notice. Although the information used
was taken from reliable sources, Swiss Re does not accept any responsibility for the accuracy
or comprehensiveness of the details given. All liability for the accuracy and completeness
thereof or for any damage or loss 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 or consequential loss relating to this presentation.
38