lloyd’s sao review process...6 ©lloyd’s statements of actuarial opinion - uk actuarial...

Lloyd’s SAO review process

Jerome Kirk

Senior Manager, Lloyd’s Market Reserving & Capital

27 November 2009

Presentation to Society of Actuaries in Ireland

© Lloyd’s2

Agenda

Background

Review Process

Results

Impact over time

Closing thoughts / Questions

© Lloyd’s3

Background

© Lloyd’s4

Lloyd’s is an active Market

80 Syndicates

51 Managing Agents

£16bn Premiums

£28bn Net Reserves

….which need professional oversight if it is to work

© Lloyd’s5

Statements of Actuarial Opinion

Worldwide technical provisions on each open syndicate year of account are subject to annual Statement of Actuarial Opinion (SAO)

“One-way” test. The provisions held must be at least as large as actuary’s best estimate

SAOs cover Gross and Net Solvency Technical Provisions which: Are not discounted for the time value of money Cover Outstanding Claims, IBNR, UPR and URP Do not allow the release of profits on unearned business Allow for reinsurance bad debt and claims handling expenses

Large loss wordings highlight areas of increased uncertainty

© Lloyd’s6

Statements of Actuarial Opinion- UK Actuarial Standards

Specific UK actuarial guidance issued by Board for Actuarial Standards:

GN12: General Insurance: Actuarial Reports

GN20: Actuarial Reporting under Lloyd’s Valuation of Liabilities Rules

GN33: Actuarial Reporting for Lloyd’s syndicates writing US business

Practising certificates required

Provides additional technical advice papers

Bad debt / ULAE

Regular liaison with Lloyd’s

© Lloyd’s7

Review Process

© Lloyd’s8

Formal Reports support SAOs

SAOs are supported by formal actuarial reports

Reports are submitted to Lloyd’s

113 reports for year-end 2008

42 signing actuaries

9 consulting firms and 9 in-house

Deadline is mid-April

After the opinions have been submitted !

Electronic copies help our review

© Lloyd’s9

MRC reviews these reports…..

MRC team at Lloyd’s review the reports

12 people / 1 day per report

Structured review

Fixed questions

Based on GN12

Includes subjectivity / qualitative comments

Consistency is important

Guidance for reviewers

Peer review

© Lloyd’s10

…and feeds back to the producers

Reports are scored per section

Leads to an overall “score” per report

Feedback document

For producers not recipients

Highlights strengths and weakness and includes “scores”

Including examples of good practice

Provide (anonymous) market data for comparison

Introduces some competition

Follow up meetings

Very positive feedback from producers

© Lloyd’s11

Example feedback chart…..

Performance Across All Syndicates

62%67%

80%

70%76%

55%

75% 72%68%

65%

Syndic

ate 1

Syndic

ate 2

Syndic

ate 3

Syndic

ate 4

Syndic

ate 5

Syndic

ate 6

Syndic

ate 7

Syndic

ate 8

Syndic

ate 9

Syndic

ate 10

Average

Unacceptable

Poor

Meets Requirements

Good

Excellent

Note: Based on dummy data

© Lloyd’s12

Lloyd’s role extends beyond SAOs

We actively manage reserving risk

Main corporation tools to understand reserving risk are:

- Independent projections

- Monitor IBNR usage through year / Early warning

- Benchmark syndicate reserves

- Monitor catastrophe claims development

Strong link to capital setting

Can’t cover this in detail today

but will give some ideas on benchmarking syndicate reserves

© Lloyd’s13

Benchmarking syndicate reserves

Calculate the relative reserving levels of syndicates NOT a measure of reserve adequacy

Index results to rank syndicates

Analyse key areas via KPIs

Results are fed back to syndicates

Detailed report for technical staff

Summary for senior staff

Input into ICA process as linked to reserving risk

© Lloyd’s14

Benchmarking syndicate reserves

1 2 3 4Reserve Benchmarking Quartile

Av. Reserve Risk as % of

Reseves

VisualFeedback

Linked to ICA

© Lloyd’s15

Benchmarking syndicate reserves

GROSS QuartileReserve Benchmarking Index Portfolio Benchmark Deviation

1. Reserve Strength 1.1 Reserve as % Ultimate Claims 31.6% 29.1% 2.5%1.2 IBNR as % Ultimate Claims 12.2% 12.8% (0.6%)1.3 Case Reserve as a % of Incurred 22.1% 18.6% 3.4%1.4 Survival Ratio (years) 2.3 2.3 0.0

2. IBNR Utilisation 2.1 IBNR Burn (1 year) During 2006 50.3% 35.8% (14.5%)2.2 IBNR Burn (1 year) During 2005 47.4% 39.2% (8.1%)2.3 IBNR Burn (1 year) During 2004 20.6% 30.4% 9.8%

3. Reserving over time 3.1 Ultimate(@now) as % Ultimate(@YOA) 97.3% 98.8% 1.6%3.2 Ultimate(@now) as % Ultimate(@YOA+2) 93.2% 99.4% 6.1%

4. Quality of business 4.1 Paid Loss Ratio 62.9% 60.3% (2.6%)

5. Speed of Development 5.1 Paid Claims Development Deviation 0.4%

Other KPIs5. Reinsurance N/A 6.1 % Premium Ceded 32.8% 30.5% 2.3%

6.2 RI ULR 97.3% 100.9% (3.6%)6.3 RI Ultimate as % of Gross Ultimate Claims 34.7% 36.2% (1.5%)6.4 RI Reserve as % Gross Reserve 28.1% 28.3% (0.1%)

6. Ultimate Loss Ratio N/A 7.1 ULR 91.9% 85.0% (6.9%)

Source: Dummy Reserve Benchmarking Report based on Lloyd’s SRD database

© Lloyd’s16

Results

© Lloyd’s17

What are we aiming for in the reviews?

Must be realistic

Reports need time to produce (and time is money)

Marginal returns as quality improves beyond good

High level objectives:

No poor reports

Maximise use to Lloyd’s

Understand budgetary constraints

Aim: All Reports Are assessed as “GOOD”

© Lloyd’s18

We see high quality in nearly all areas…

Range of Results for all producers

0%10%20%30%40%50%60%70%80%90%

100%

Con

tent

Nat

ure

&Ac

cura

cy o

fD

ata

Met

hods

Assu

mpt

ions

Em

ergi

ngE

xper

ienc

e

Res

ult

Sel

ectio

n

Res

ults

Pre

sent

atio

n

Unc

erta

inty

Bus

ines

sK

now

ledg

e

Ove

rall

com

mun

icat

ion

Unacceptable Poor Meets Requirements Good Excellent

© Lloyd’s19

Overall Communication

Business Knowledge

Uncertainty

Results Presentation

Result Selection

Emerging Experience

Assumptions

Methods

Nature & Accuracy of Data

Content

Produc

er1Prod

ucer2

Produc

er3Prod

ucer4

Produc

er5Prod

ucer6

Produc

er7Prod

ucer8

Produc

er9Prod

ucer1

0

ExcellentGoodMeets RequirementsPoorUnacceptable

…and across all producers

© Lloyd’s20

Impact over Time

© Lloyd’s21

The burning question…….

Has Lloyd’s review made a difference?

© Lloyd’s22

….can analyses the results over time

Look at:

Overall impact

- via average score per year

Consistency

- via spread of scores between reports

Specific impact

- via areas Lloyd’s has specifically focussed on

This form of review gives 4 years of data to work with

© Lloyd’s23

We have seen a steady increase in average scores….

Getting very close to our goal

Overall performance for all producers by year-end

60% 61%65%

69%

2005 2006 2007 2008

Unacceptable

Poor

Meets Requirements

Good

Excellent

© Lloyd’s24

2005 SAO - Range of Results for all producers

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Con

tent

Nat

ure

&A

ccur

acy

of D

ata

Met

hods

Ass

umpt

ions

Em

ergi

ngE

xper

ienc

e

Res

ult S

elec

tion

Res

ults

Pre

sent

atio

n

Unc

erta

inty

Bus

ines

sK

now

ledg

e

Com

mun

icat

ion

Unacceptable Poor Meets Requirements Good Excellent

…and across every section2006 SAO - Range of Results for all producers

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Con

tent

Nat

ure

&A

ccur

acy

of D

ata

Met

hods

Ass

umpt

ions

Em

ergi

ngE

xper

ienc

e

Res

ult S

elec

tion

Res

ults

Pre

sent

atio

n

Unc

erta

inty

Bus

ines

sK

now

ledg

e

Com

mun

icat

ion

Unacceptable Poor Meets Requirements Good Excellent

2007 SAO - Range of Results for all producers

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Con

tent

Nat

ure

&A

ccur

acy

of D

ata

Met

hods

Ass

umpt

ions

Em

ergi

ngE

xper

ienc

e

Res

ult S

elec

tion

Res

ults

Pre

sent

atio

n

Unc

erta

inty

Bus

ines

sK

now

ledg

e

Com

mun

icat

ion

Unacceptable Poor Meets Requirements Good Excellent

2008 SAO - Range of Results for all producers

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Con

tent

Nat

ure

&A

ccur

acy

of D

ata

Met

hods

Ass

umpt

ions

Em

ergi

ngE

xper

ienc

e

Res

ult S

elec

tion

Res

ults

Pre

sent

atio

n

Unc

erta

inty

Bus

ines

sK

now

ledg

e

Com

mun

icat

ion

Unacceptable Poor Meets Requirements Good Excellent

© Lloyd’s25

…..and an improvement in consistency

Note: 2 highest and lowest scores for each year have been excluded

Range of Results for all producers

0%

20%

40%

60%

80%

100%

2005 2006 2007 2008 2009

Mean Maximum Score Minimum Score

Excellent

Good

Meets Requirements

Poor

Unacceptable

Target?

Estim

ated

© Lloyd’s26

Specific points of focus (1)Peer review

Reports stating work had been peer reviewed (by year-end)

7%

55%

74%

93%

2005 2006 2007 2008

© Lloyd’s27

Specific points of focus (2)Quantifying uncertainty

Reports where uncertainty had been quantified (by year-end)

69%

79%

89%

2006 2007 2008

© Lloyd’s28

How do people quantify uncertainty?

Methods Used to Quantify Uncertainty - 2008 year-end

15%

22%

48%

4%

11%

Range of ReasonableOutcomesStress/Sensitivity test

Bootstrap

Mack

Other

© Lloyd’s29

Specific points of focus (3)Large loss wordings

Reports describing how large loss wordings were selected (by year-end)

47%

83%

95%

2006 2007 2008

© Lloyd’s30

Specific points of focus (4)Emerging experience

Reports which included A vs. E (by year-end)

38%

54%

2007 2008

© Lloyd’s31

Closing thoughts

Reviewing SAO reports is a valuable exercise

to both Lloyd’s and the profession

Requires significant resources

To be effective

consistency is required through structure

- but review can include subjectivity

feedback is essential

Lloyd’s review extends beyond SAOs

Seen a positive impact of the SAO reviews over last 4 years

© Lloyd’s32

Questions

© Lloyd’s33