Marine Reinsurance Rating

Document Sample
Marine Reinsurance Rating Powered By Docstoc
					CARe 2006: Marine Reinsurance

Lee Tookey
Underwriter, Marine and Aerospace
Aspen Re

(and nearly an actuary)
Marine Reinsurance:

•   Introduction: review of the basics
•   Objectives: what the actuary can do
•   Some typical questions
•   Some observations on specific classes
    – Hull
    – Cargo
    – Energy
• Closing remarks
Marine Reinsurance: Introduction

• Significant Actuarial involvement for at
  least 15 years
• Slow improvement in data
  – Always has been plenty of it but little
    consistency and difficult to model.
• Little interest from commercial modelling
  – In house development and discussion papers
Introduction (2)

•   Proportional and Non-Proportional
•   Facultative and Treaty
•   Pure and Composite
•   “Marine” Classes include
    – Hull
    – Cargo & Specie
    – Offshore Energy
    – Liability (Marine and Energy)
Introduction (3): Non Proportional
Treaty (pre 2005)
• Risk and event common.

• Attachment point of programme a function
  of maximum risk line.

• Combined or “Whole Account” layers
  above specific class layers
Introduction (4): Rating Non-
Proportional Treaty
• Usually experience and exposure rating
  – Experience from actual and “as-if” losses
     • Attritional risk losses
     • Headline risk losses
     • Catastrophes (natural and man made)
  – Exposure from
     • Risk profile and size of loss curve for risk losses
     • Market share and market loss return periods
     • Possibly stochastic models
Introduction (5): Where is the
• Attachment point for excess of loss
  typically 10% of maximum line.
• Usually few losses at this level
• Money swap layers generally avoided
• Low level backup layers in softer market
• Frequency/severity approach would
  usually need to look at layer below
An aside on attachment points

• Of 68 marine programmes seen in 2006..
  – 12 had a ROL>50% for first layer
  – Of these, 9 were domiciled in Americas or led
    by US reinsurers
  – All attached below 10% of maximum line
• Possibly Lloyd’s influence
  – Underwriting capacity measured in terms of
    written premium

• Using information we have, provide
  guidance to underwriters on
  – Pure premium
  – Volatility of result
  – Changes to exposure over time
  – Claims inflation
  – Accumulations
Objectives (2)

• Understanding the business from an
  insurance perspective
• Understanding what information is
  collected and why
• Understanding effects of changes in the
• Understanding the results of our analysis
  in light of the actual experience.
Common Questions

• Why is the experience rate so different to
  the exposure rate?
• What has changed over the experience
  period and how can I quantify it?
• How homogeneous is the exposure data
  and does that cause a problem?
• What more is there to know and can the
  underwriter tell me?
Marine Hull Excess of Loss
• Typically loss experience given excess of
  50% of attachment point last year.
• Several years data usually available but
  what is the effect of…
  – Change in mix of vessel types
  – Change in lines size
  – Changes in policy conditions and coverage
  – Inflation
Marine Hull Excess of Loss
Information (2)
• Risk Profiles
  – Gross or Net of reinsurance
  – By type of vessel ?
  – In force or risks written, what period
     • 9 month written profiles not uncommon
  – Premium, sum insured and count
  – Losses by band as well?
Range of ocean going vessels

• Very large container
• Tankers
• Bulk carriers
• Passenger vessels
• Car Carriers
• Fishing vessels
• Service boats
Where is the value?

• Container Vessel
  – Mostly in hull and engines, some machinery
  – Little else of value (apart from cargo)

• Cruise ship
  – Hull and engines smaller percentage of value
  – Upper, accommodation decks high value
So what if there is a fire?
Underwriting approaches to book
building vary
• For larger vessels, percentage line, dollar
  line or both, smaller vessels 100% writing
  – “Normal” maximum line
• Use of proportional treaty
• Facultative reinsurance
  – Proportional or excess
• Territorial considerations
• Special acceptances
All hull insurance is not the same

• Standard coverage include
  – Hull and machinery (H&M)
  – Total Loss / Increased Value (IV)
  – Mortgagees Interests (MI)
  – Loss of Hire (LOH)
  – Collision Liability
• But we rarely see the claims or the
  exposure broken down like this
Hull Interest Example

• H&M $200m      • Maximum partial loss
• IV $50m           – $220m (H&M + LOH)
• LOH $20m       • Maximum total loss
                    – $250m (H&M + IV)

• In a risk profile, this may appear as three
  entries (200, 50, 20), one entry of $250m or
  one of $270m
Hull and Machinery versus Total
• In some markets, TLO coverage is limited
  to certain percentage of total insured value
• Premium rate significantly different
  – TLO may be 30% of all risk rate
• Size of loss curves different for H&M and
  – TLO pro rata

• Very unpredictable
  – Salvage
  – Labour
  – Steel
• Reflected to some extent by ship-owner
  revaluing vessel and we get rate on sum
What we should aim to do

• Price for the experience of the account
  – Adjusted for quantifiable changes in the
    account from re-underwriting or market
  – Allowing for events that haven’t happened
     • If the reinsured is getting premium for the risk, we
       should get our share
A risk profile is so versatile….

•   Exposure rating – obviously
•   Number of vessels and TSI
•   Average rate by band
•   Level and utilisation of maximum line size
Adjusting Experience

• Change in • % line, adjust losses and subject
  line size    premium
             • Dollar line, harder to adjust. Need
               risk profiles to help
• Change in • Ideally work from gross losses
  reinsurance and apply current RI

• Territory   • Cat risk and usual perils of the
Adjusting experience (2)

• Usually use on level premium to adjust
  historic frequency / avg severity
• Can use historic risk profiles
  – Index for frequency based on exposed
    vessels count to layer
  – Index average severity based on average
    exposed sum insured to layer
Exposure rating

• Current risk profiles gives indication of
• Assume size of loss curve appropriate
• Underlying loss ratio assumption key.
  – But this may vary through risk profile
• With all adjustments, often significant
  difference between experience and
What is available?

•   Historic risk profiles usually are
•   Commentary on line size
•   Commentary on reinsurance
•   Fleet mix

• But no use asking 2 days before renewal
A few cargo issues

•   Policy limit or shipment
•   How long is exposure on risk
•   Where is most of trade
•   What sort of commodity
•   Accumulation risk
•   Start and end of exposure
Energy – the opportunity

• After 2005 windstorm, full review
• Great time to question
  – Every aspect of coverage under review
• Significant change in data presentation
  – Sub limits in GOM for wind
     • Per policy not per platform
     • Does not effect risk rate
• Increased modelling

•   Do not stick to basic methods
•   What else can we use data for
•   Learn subject
•   Engage underwriter and client
•   Demonstrate use

Shared By: