General Insurance Spring Seminar Facing up to Risk

1
General Insurance Spring SeminarFacing up to
Risk
abcd

• 13-14 May 2002
• Scarman House, University of Warwick, Coventry

2
The Worst Event
abcd

• Richard Winter
• K P M G

3
The Worst Event

• Suggest a title for a workshop to Julian Leigh,
  asking him to find a presenter
• Get an email by return saying That sounds
  interesting I look forward to hearing you do
  it.
• Then discover that the workshop has become one of
  the main sessions

4
The easy way to assess the worst possible event

• Assume that every policy pays out to its maximum
  exposure
• That is all the capital you need to hold just in
  case
• Impossible even if you look at just one class
  of business within one company
• Hold a reasonable level of capital
• At some point, need to introduce judgement into
  process

5
The Worst Event

• Prior to 2001 - Biggest UK motor claim 9m
• Severe injury / very high assumed future earnings
• Selby estimated cost 40m - 50m
• WTC step change in property claims
• Toulouse Explosion estimated insurance bill
  2bn, with properties severely damaged 1km from
  site of explosion

6
Aim of Presentation

• Consider how to manage extreme catastrophe risk
• Overlaps with other presentations at this event
• Manage extreme risks within product lines and
  individual risks
• Most senior management focus on major risks (what
  could put us out of business / cause insolvency)
• Product manager / Underwriting / Pricing focus
  risk management at a localised level

7
How do you manage risk?
8
How do you manage risk?
Control
Identify
Administer
Finance
9
Identify Risk

• This is meant to be what insurers are good at
• Sinister Risk
• From an unusual or unexpected source
• One claim to maximum cover (and massively
  beyond)(WTC)
• Unexpectedly large consequential loss (Selby)
• Large number of relatively small claims
  (Asbestos)

10
Why were events so unexpected

• WTC nothing would happen that could take both
  towers out simultaneously
• Selby Road Traffic Accident is going to be on
  the road there will not be multiple severe
  bodily injuries

11
What would it take to.

• Get a claim that exhausts the cover on this
  policy?
• What is the maximum damage that the entity we are
  insuring could cause?
• Is there any way that every policyholder could
  claim under a policy?

12
What would it take to.

• Get a claim that exhausts the cover on this
  policy?
• WTC
• What is the maximum damage that the entity we are
  insuring could cause?
• Dog
• Toddler
• Roast Potato
• Is there any way that every policyholder could
  claim under a policy?
• Cashback

13
We have a problem

• This is the second talk this morning
• You had a heavy time last night
• Its audience participation time

14
Practical Exercise

• You are the insurer of 500,000 private vehicles
  under current UK legislation
• How did the first 100m individual claim happen?
• How did you lose 15m by 1 in 10 of your
  policyholders making a 300 claim.

15
Identification Realistic Disaster Scenarios

• Lloyds Market Risk Unit
• USA Windstorm 50bn loss
• European Storm / Flood 10bn
• Japanese Earthquake 15bn
• Marine Collision (Ultra Large Crude Carrier and
  Cruise Liner)
• North Sea loss of complex
• Aviation collision (over major US city)
• Loss to largest single risk including PML failure
• Space storm destroys 6 in orbit satellites
• Political Risk

16
Unrealistic Disaster Scenarios

• Expect the unexpected
• If something bad can happen, it probably will
• Identify key assumptions in thinking
• Since May 1984 there has been no significant
  flood risk to London because of introduction of
  Thames Barrier

17
Unrealistic Disaster Scenarios
18
Control

• Limit exposure
• Dont sell as much cover
• Ensure exposure to region / one risk / one event
  is understood
• Restrict cover
• Terrorism exclusions
• More risk bands
• Enforcement of policy conditions
• Encourage Mitigation
• Withdraw Cover
• Government Insurer of last resort

19
Control

• Needs a long (corporate) memory
• Not conducive to building market share
• Easier in a hard market
• Temptation follow the market practice

20
Finance

• Ignore it, and hope it doesnt happen

21
Finance

• Ignore it, and hope it doesnt happen
• Valid approach
• Acceptable company failure rate
• 1 in 1000 year events do happen
• How would you perform relative to the rest of the
  market?
• If all insurance companies went insolvent
  together, would the Government bail them out
• No one is being singled out over WTC because all
  companies have taken a big hit

22
Finance

• Ignore it and hope it doesnt happen
• Payback
• The purpose of commercial insurance is to smooth
  declared profit, not to alter the total amount
  of profit over time
• Reinsurance is similarly to smooth profit for the
  direct insurer, not to alter the total amount of
  profit over time
• Personal lines insurance is a means of spreading
  the costs of a natural catastrophe amongst the
  insured population over time, not a transfer of
  risk to the insurer

23
Does Payback work?

• If make too explicit, ceases to be a contract of
  insurance
• Assumes barriers to entry and exit
• In practice, relatively easy to introduce or
  withdraw capacity
• At extreme, assumes purchaser is willing to
  respond quickly enough to save insurer that gets
  into trouble

24
Finance

• Ignore it and hope it doesnt happen
• Payback
• Reinsurance
• Limits loss to an acceptable level
• Can cover be placed?
• Assumes reinsurer able to pay up
• What about reinsurers aggregation of risk?

25
Finance

• Ignore it and hope it doesnt happen
• Payback
• Reinsurance
• Diversify
• Ideally take on a risk with the opposite
  financial effect
• Normally not obvious for an insurance product

26
Control / Finance What has happened in practice?

• Property PML has been reassessed (e.g. whole
  buildings destroyed not just floors)
• Terrorism cover withdrawn (or not as easy to
  obtain)
• Some companies reducing exposure
• Lot of new entrants to market limits upside for
  companies that made heavy losses

27
Administer

• Aggregation of risk
• Within class of business
• Across multiple classes
• Demonstrate monitoring exposure properly (not
  just have the capability to do so)
• Review continually
• Only a fool learns by his own mistakes

28
Key issues

• It is important to work out the worst that can
  happen both at a macro and micro level
• Appropriate action may be to do nothing but it
  should be a conscious (and documented) decision
• If you dont identify a risk you cant control it
• Is the concept of payback dead and if so, are
  companies acting in a way consistent with this?
• How short is the memory of the insurance
  industry?

29
(No Transcript)