Emerging risks:Risk perception at Swiss Re

Sean Russell

OCCA Meeting

July 4, 2007

Page 2

Disclaimer

Please note that this document, or any part of it, may not be disclosed or otherwise made use of without Swiss Re's prior written consent. Whilst any third party information contained in this document has been obtained from sources believed to be reliable, no representations are made as to its accuracy or completeness. Further, Swiss Re expressly disclaims any responsibility for liability or loss arising from use of such information or from any of Swiss Re's comments upon it.

Page 3

Overview

What are emerging Risks?

Why are emerging risks of concern for re/insurance?

How does Swiss Re detect emerging Risks?

Some examples of current emerging risks

How to manage emerging risks

Page 4

Emerging risks - definition

“Emerging risks are developing or changing risks which are difficult to quantify and may have a major impact on insurers book of business.”

Often they are already (unintentionally) present in an insurers existing book of business.

Page 5

Risk of change

René OefeliThe Emerging Risks LandscapeFinland, 9 May 2006

Technological advances & scientific knowledge discovery of new threats, proof of cause-effect relations, improved methods of measuring/detectingemissions

Economic circumstances higher concentration of values,stability of economics (local andglobal), intensity of insurance

Legal standardstransition of liability from ‘fault’ to ‘strict liability’

Social aspectsenhanced awareness, deep-pocket philosophy, loyalty of the employee vs employer, public trust, collapse of social security systems

Page 6

Common denominator of emerging risks

High uncertainty

Little information available

Frequency and severity difficult to assess

Difficult to quantify

Risk transfer questionable

No industry position

No one wants to make the first move

Difficulties for risk

communication

Danger of turning into phantom risks

Regulatory involvement necessary

Page 7

Group ER management framework Rationale (Illustration100 years of asbestos premonition)

2006 US Insurance loss as of 2004: USD 55 bn *Total estimated future losses: USD 275 bn **

Source: David Gee and Morris Greenberg, “Asbestos: from ‘Magic’ to Malevolent Mineral”.

Conclusion

Asbestos exclusions were introduced in the US in 1918 – but somehow they got “lost” in the following years

Data on harmful health effects of asbestos were reported as early as 1898 , but the insurance industry did not integrate this knowledge into its risk management

First worker's comp asbestos claims filed

French factory report of 50 deaths in female asbestos textile workers and recommendation of controls.

US insurers refuse cover to asbestos workers due to assumptions about injurious conditions in the industry.

UK Merewether Report cites 66% of long-term workers in Rochdale factory with asbestosis.

UK Asbestos Regulations specify dust control in manu-facturing only and compensation for asbestosis, but this is poorly implemented.

1906

1918

1930

1931

1950s

1960s First GL cases filed and first payments madeFirst claims paid by Swiss Re, Dec 1966

* according to AM Best

** according to Millimans

Page 8

Risk Management Process

Risk Policy

LegislationStrategy

Risk Assessment

PerceptionIdentificationAnalysisQuantification

Risk TreatmentAvoidancePreventionReductionFinancing/Transfer

Risk Control

EvaluationSteeringReporting

Risk Management

Risk informationObjectives & GoalsRisk communication

Page 9

P&C SONAR’s history

Risk perception network established to promptly detect faint signals

Purpose:

– Identification and evaluation of emerging and future risks

– Systematic screening of the risk landscape

– Centralising observations associated with risk from different countries, cultures and enterprises

Systematic

Observations of

Notions

Associated with

Risk

1995 2000 2002

Systematic Risk Perception started in 1995

SONAR Patent filed

Page 10

SONAR perceives early indicators

Early warning

Claims, losses

Reaction time

Early indicators Events

Page 11

Landscape of emerging risks

Invasionof privacy

Bogus parts

Powersystembreak

Organisedcrime

CO2 trading

Off-shore &internetmarkets

Spaceweather

ElectrosmogResistance to

antibiotics

Drinkingwater quality

Loss of reputation

Businessethics

Intercontinentaldata transmission

Customised drugs

NanotechnologyCaldera

erruption

RSI

Cyberrisks

Dirtybombs

Implants

Indoorpollution

Toxic mold

Foodcontaminants

Stress atwork

EndocrinedisruptorsMedia

risks

Ageinginfrastructures

Tele-medicine

CloningDeteriorating

safetystandards

Alcohol

Contingent Business

Interruption

MegaTsunami

Pervasivecomputing

Privatisation

Botox

Spread ofdiseases

Page 12

Silicosis

Silica is the second most common mineral on earth

Numerous industrial uses

Disabling, nonreversible and sometimes fatal lung disease

Disease cannot be cured but is highly preventable

Studies indicate evidence of carcinogenic properties of inhaled crystalline silica

Page 13

Obesity

Center for Disease Control and Prevention (CDC) has declared obesity an epidemic

1 out of 3 Americans is obese

400 000 obesity-related deaths annually in the US

Increases risk of heart disease, type 2 diabetes and colon, breast and endometrial cancers

Page 14

Critical parts

Reuse of single-use medical devices

Hospitals and external medical providers are reprocessing expensive endoscopic tools used for minimally invasive surgery. Reprocessed medical instruments can expose patients to increased risk of infection.

Page 15

Botox

Botox (Botulinum toxin) is a toxin injected into wrinkles to permanently paralyze certain muscles and help eliminate wrinkles. It has proven to reduce the appearance of moderate to severe frown lines in 18 to 65 year-old patients.

Massive use in the North America and Europe.

No long-term studies available. Ageing characteristics of treated facial tissues unknown – 'Michael Jackson facial syndrome' possible.

Page 16

Drugs and Cosmetics

Violent reactions to antibody therapy

In a London clinical trial six healthy volunteers suffered multiple organ failure.

Specialists agree that results of T cell activation are notoriously hard to predict.

Past experience with “small synthetic molecules” can no longer automatically be expanded to such molecules.

Potential impact on insurance: product liability can in general be triggered more frequently for drugs stimulating rather than neutralizing components of the immune system.

Page 17

Food

Pet food contamination in US

The major manufacturer of cat and dog food for many brands found that their food had been contaminated with a substance used overseas to poison rats.

The contamination comes from ingredients imported from a supplier.

“...pets are basically considered property at this point...”. If this changes this “...could bring further liability actions into play.”

Potential impact on insurance: increase of products liability exposure for pet food manufacturers. Largest exposure could be in veterinarian medicine and pharma for pets.

Page 18

Food

Bypass approval process for pharmaceutical

Food additives are not subject under the approval process for pharmaceuticals. For some products manufacturers can avoid the lengthy process for pharmaceutical approvals by declaring them as food additives or nutraceuticals

Nutraceutical refers to foods claimed to have a medicinal effect on human health.

Page 19

Complex or critical infrastructure

Critical software

Stable software has become crucial for the flawless function of aviation, automobiles, rails or infrastructure

MS Crown Princess came close to capsizing due to a faulty steering mechanism.

Reasons for such failures are often to be found in cost pressures, faulty time schedules or lack of risk/project management.

Potential impact on insurance: casualty of shipping company. If software was at fault the PI cover of the software company would be called on. Highlights the problems with safety relevant software. In cars a failure could trigger PL and/or Recall. Property damage (eg. ship on fire, defective cargo), business interruption.

Page 20

Environment

Mysterious, massive death of bees in the US

In the US, bee keepers are experiencing unprecedented die offs of bees some losing as much as 80% of their colonies. Some additional facts:

In 2000, the total U.S. crop value that was wholly dependent on the honey bee pollination was estimated to exceed $US 15 billion.

Potential impact on insurance:

– Parts of agriculture relying on pollination by bees will have loss of yield.

– If chemicals are causing the death of the bees the question who is guilty will be raised. If this is the result of a natural cause, it has no impact on insurance.

Page 21

Spreading of infectious diseases

Warning over global bird flu plan

A third of countries failed to set out how they would distribute medical treatment

Individual countries have not consistently prioritised population groups for vaccines and antivirals.

No countries prioritised population groups to receive ventilators, face masks and other critical resources

Half of the countries studied had prioritised children, despite a WHO recommendation against it.

Page 22

Public health

Builders sunburn

The risk of sunburn has been highlighted as “the new asbestos”, causing cancer which could become the next big threat to the building industry.

Contractors across the sector could be facing potentially huge employers’ liability claims amounting to millions of pounds in the coming decades if they do not warn their employees to slap on the suntan lotion.

Clearly it would be difficult to determine which sun caused the skin cancer, the one taken at work or the one taken during a vacation.

Page 23

Terrorism

Building weapons of mass destruction

Genetically engineered bio weapons developed by small teams could become reality within the next two to four years.

Virologist and state funding is no longer needed. With the help of a DNA synthesizer it would take just over $200 000 to build smallpox genome.

Virologist question if it would be that easy.

Potential impact on insurance: Product liability for labs synthesizing harmful sequences on demand, accidental releases from laboratories, pandemic with ten thousands getting sick/dead.

Page 24

Pervasive Computing

Overview

Miniaturization of electric circuits makes them more susceptible to cosmic radiation

Loss of sensitive personal data

Playing video games may result in hand injuries

Addictive gambling

Googlization

Internet creating conflagration

Robot safety

Vicious bloggers

Page 25

Nanotechnology

Nanos (means ‘dwarf’ in Greek)

Technology to manipulate matter down to 1 – 100 nm

Small size

– high surface to volume ratio

– unique properties (material strength and weight reduction, conductivity, new optical properties)

– new entry paths (high mobility in human body and environment)

1000 000 nm

1000 nm

100 nm

<100 nm

Page 26

Nanoparticles -Ubiquitous in industrial production

Materials Pharmaceuticals Aerospace

Chemicals MachinesElectronics

Page 27

Worldwide revenues in the year 2015:More than USD 1 trillion (1000 bn)

Nanotechnology revenues worldwide by 2015 (USD bn)

Materials 340

Electronics 300

Pharmaceuticals 180

Chemicals 100

Aerospace 70

Tools 20

Healthcare 30

Sustainability 45Source: National Science Foundation

Page 28

How much of proof do we need to react?

Challenge: early signals vs. impact to insurance

Do we need to adjust the risk capital?

How much statistical data are required to trigger action? Is a proactive behaviour possible? first mover disadvantage?

No loss no exposure? How can we adjust premiums without loss experience?

Page 29

How to set priorities?

Understand:

Time to impact

Probability

Severity (accumulation, cross sector impact)

Measures / down side protection / opportunities

Premium adjustments

Capacity steering, limitations

Page 30

Balance sheet protection:Risk mitigation levels

Measures to protect balance sheet encompass: 1 Risk communication

to influence the business environ-ment favourably,

2 Developing (under-writing) best

practices and 3 Developing rules

and adjusting pricing and risk models.

raising awareness (internally & externally)to influence the business environment

application of underwriting bestpractice

rules & modeladjust-ments

individual riskassessment by Risk Engineering Services

limitationof exposure throughwordings (e.g. triggerdefinition, exclusions,product type)

exclusions

limits

underwriting principles(e.g. market standards)

pricing and risk model adjustments

deviation possible,butmust be

documented in auditable form

referral process mandatory forany deviations

Level 1

Level 2

Level 3

risk communication to raise internal & external awareness with the objective to influence the business environment (e.g. through legislative measures)

Level 1

Raising awareness (internally & externally)to influence the business environment

(e.g. Nanotechnology)

-

individual riskassessment by Risk Engineering Services

limitationof exposure throughwordings (e.g. triggerdefinition, exclusions,product type)

exclusions

limits

underwriting principles(e.g. market standards)

pricing and risk model adjustments

deviation possible,butmust be

documented in auditable form

referral process mandatory forany deviations

Level 2

Application of underwriting best practice

(e.g. EMF, GMO, SilicaTSE, Chemicals)

Level 3Rules and model

adjustments (Asbestos)

risk communication to raise internal & external awareness with the objective to influence the business environment (e.g. through legislative measures)

SONAR: Keep/monitor on watch list, no immediateactions necessary (short term); permanent

observation

SONAR: No immediate actions necessary (mid term)

Page 31

Possible underwriting measures -‘Best practice’ in underwriting

Higher premiums offset the uncertain higher risk. No structural solution. Insureds pay risks from the past …

Higher retentions stimulate risk-adverse behaviour and prevention. Not a solution to large loss amounts

Lower insured limits: better assessment of the severity of occurrence, but not ‘full’ compensation of an insured’s damage (via insurance).

Page 32

Possible underwriting measures -‘Best practice’ in underwriting

Better risk differentiation stimulates risk-adverse behaviour and prevention fairer solidarity groups

Exclusions leave the insured’s damage uncompensated. Sometimes this ‘solution’ is inevitable

The liability insurance trigger should be changed from ‘act committed’ to ‘loss occurring’ or even - in some countries - from ‘loss occurring’ to ‘claims made’

Page 33

Managing emerging risksConclusions

High demand for further systematic observation of the ‘risk of change’ for current and future emerging risks

Promoting risk awareness or risk perception internally and externally is vital

Define necessary underwriting measures (‘best practice’ in underwriting) in good time

meaning:

impact analysis and

define specific underwriting guidelines

Page 34

Additional information

Swiss Re Publications:

“www.swissre.com”

Prion infection on the rise? Hospitals in need of modern risk management

Emerging risks – A challenge for liability underwriters

Liability and liability insurance: Yesterday - today - tomorrow

Nanotechnology – Small matter, many unknowns

The insurability of ecological damage