culture clash: us v them doug lacoss care - london casualty pricing approaches 16 th july 2007
Post on 31-Mar-2015
213 Views
Preview:
TRANSCRIPT
Culture Clash: US v Them
Doug Lacoss
CARe - London
Casualty Pricing Approaches
16th July 2007
16 July 2007 Casualty Pricing Approaches 2
Number of Qualified Actuaries
Source: Tim Aman CARe presentation. 1. Actuaries.org 2. SOA.org 3. CAS.org
18,390
23,000
EuropeNorth America
16 July 2007 Casualty Pricing Approaches 3
Source: T Aman CARe 2007 Actuaries.org, SwissRe.com, CAS.org, SOA.org
Actuarial Saturation
7.20
5.72
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
North America Europe
Number of actuaries per USD 1 billion premium, non-life
16 July 2007 Casualty Pricing Approaches 4
Actuaries in ReinsuranceCAS Members in Reinsurance
1000
1500
2000
2500
3000
3500
4000
4500
5000
To
tal
CA
S M
em
bers
0
100
200
300
400
500
600
Mem
bers
wo
rkin
g i
n R
I
Total CAS Members Number of Members in RI
16 July 2007 Casualty Pricing Approaches 5
Reinsurance Actuaries – US v. Europe Number of Fellows + Associates in the US – 23,000 Percent of Non-life actuaries - 19.0% Percent of Non-life actuaries in Reinsurance – 12.1% Number of Non-life, Reinsurance actuaries in US – 529
Number of Actuaries in Europe – 18,930 Percent of Non-life actuaries – 15% Percent of Non-life actuaries in Reinsurance – 18.62% Number of Non-life, Reinsurance actuaries in Europe - 529
Source: Tim Aman presentation CARe 2007, CAS website
16 July 2007 Casualty Pricing Approaches 6
Reinsurance Results
Source: Benfield 2006 Reinsurance Renewal Report
16 July 2007 Casualty Pricing Approaches 7
Avoid the Valleys
Rate monitor Capture all business Produce regular reports Measure against exposure Adjust for
Terms & conditions Changes in coverage Varying limits / retentions Underlying rate changes
16 July 2007 Casualty Pricing Approaches 8
Adjusted Loss Ratios
YearLoss
Ratios
Primary Rate
ChangeRI Rate Change
Claims Trend
Impact of T&C
Change
Limits Change Impact
1995 80.0% 2% 5%1996 75.0% 5%1997 92.3% -1% -5% 5% 6%1998 107.1% -3% -6% 5%1999 113.3% 2% 5%2000 111.1% 10% 5% -12%2001 92.3% 6% 8% 5%2002 82.1% 8% 18% 5% -3%2003 75.0% 3% 6% 5% 6%2004 72.2% 7% 5%2005 76.9% -5% 3% 5% 3%2006 85.0% -8% 6% 5% 3% 3%2007 88.1% -4% 8% 5%
16 July 2007 Casualty Pricing Approaches 9
Adjusted Loss RatiosLoss Ratios for Line of Business
40.0%
50.0%
60.0%
70.0%
80.0%
90.0%
100.0%
110.0%
120.0%
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Historical Loss Ratios Adjusted For Rates, Trend Adjusted for T&C
16 July 2007 Casualty Pricing Approaches 10
Adjusted Loss Ratios
Loss Ratios for Line of Business
40.0%
50.0%
60.0%
70.0%
80.0%
90.0%
100.0%
110.0%
120.0%
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Historical Loss Ratios Adjusted For Rates, Trend Adjusted for T&C
16 July 2007 Casualty Pricing Approaches 11
Adjusted Loss Ratios
Loss Ratios for Line of Business
40.0%
50.0%
60.0%
70.0%
80.0%
90.0%
100.0%
110.0%
120.0%
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Historical Loss Ratios Adjusted For Rates, Trend Adjusted for T&C
16 July 2007 Casualty Pricing Approaches 12
Avoid the Valleys
Understand the risk Discussion with underwriters Discussion with cedents Incidental classes matter Read non-actuarial reports
Underwriting submission Claims reports Trade press
16 July 2007 Casualty Pricing Approaches 13
Avoid the Valleys
Locate the floor Objective measurement Unbiased analysis Portfolio view
16 July 2007 Casualty Pricing Approaches 14
Avoid the Valleys
Check for reasonable results Applies to both experience and exposure rating Include consideration of unknown risk
16 July 2007 Casualty Pricing Approaches 15
Testing XS Layer Projections Assumptions
Number of vehicles = 10 million Average premium = 1,000 Total subject premium = 10 billion Loss Ratio = 100%
Selected XS Factor, above 1 million => 10% Total Loss Cost above 1 million = 1 billion Estimate an average excess loss cost, say => 2 million Then the implied claim count = 500
Is this reasonable?
16 July 2007 Casualty Pricing Approaches 16
Celebrating Differences
Different data available
Different coverages
Different pricing approaches
16 July 2007 Casualty Pricing Approaches 17
Different Data AvailableEUROPE
In general, almost no market data available for Europe
Exception: Germany
Amount of client specific data varies by country
Even in “good” countries (UK), the data varies by company
Standard Motor market questionnaire
Rate changes - Rate change info provided by cedent, sometimes with supporting calcs
USA ISO collects data from all
member companies Loss costs, exposure curves
Amount of data varies by company Companies more open to
reinsurer visits in the US Rate changes – Due to
regulation, usually filed with state insurance departments, esp. personal lines
16 July 2007 Casualty Pricing Approaches 18
Different Data
Insurance Service Organisations Data from member companies in US
90% of insurance premium in some lines
Excess Loss Development
ILF tools
Very detailed info By state
Rating class
Coverage
16 July 2007 Casualty Pricing Approaches 19
Different Coverages Motor (Auto)
In US, basic limits quite low In Europe, unlimited liability exists
Selby – Private motor claim Public Liability (General Liability)
Limits in US $1m/$2m, then umbrella on top No umbrella in Europe, PL policies purchased to make total limit
Employers Liability In US, Workers’ Comp, 1st party coverage, unlimited EL is 3rd party coverage, though standard of negligence has been
eroded in many countries. In UK, standard limit of £10m
16 July 2007 Casualty Pricing Approaches 20
Different Pricing Approach
With no market data, reinsurers rely more on experience Credibility of exposure rating is lower
Experience Rating Issues Small data sets – analysis of small, regional business is volatile
from year to year; don’t overreact to random fluctuations Note “free cover”, “unused layer”
Make judgmental adjustment Fit distribution to limited losses, then extrapolate
16 July 2007 Casualty Pricing Approaches 21
Different Pricing Approach
Exposure rating issues Reinsurers attempt to build their own exposure curves European “Countrywide” data is much smaller data set than
US countrywide data Data is not homogeneous* High layers still based heavily on judgement* Reliability of curves*
*Some issues impact US curves also
16 July 2007 Casualty Pricing Approaches 22
Different Pricing Approach
General Issues Index Clauses
Standard in Europe, rare in US Cover including/excluding ALAE
Mainly a US issue Trend
Unreliable information in some countries Reinsurance Contract Structuring
16 July 2007 Casualty Pricing Approaches 23
Exposure rating Data set – loss triangles Can use data to estimate claims trend Get exposure data – premiums/ vehicle years Use exposures/losses to estimate frequency – poisson, neg
bin Use loss size to estimate severity – pareto, lognormal
Use distributions for non-standard layers Most loss severity data is not easily fit by curves How to get ultimate shape for casualty? Different claim costs by country make comparisons difficult
top related