© 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is...

20
© 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is confidential and intended solely for the information and benefit of the immediate recipient hereof. Martin Lewis, FCAS, MAAA

Transcript of © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is...

Page 1: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

© 2005 Towers Perrin

STRIMA 2005 Conference

Premium Development/Cost Allocation

This document is incomplete without the accompanying discussion; it is confidential and intended solely for the information and benefit of the immediate recipient hereof.

Martin Lewis, FCAS, MAAA

Page 2: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

2© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Cost Allocation General Principles

Attributes of Sub-Optimal Systems

Work Steps

Summary

UnderwL
Page 3: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

3© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Cost Allocation General Principles

“Optimal” Definition: the process of equitably allocating property/casualty risk financing costs within an organization in a manner that promotes loss control

Cost Allocation vs. the Actuarial Report

Actuarial report: changes with new information/data

Cost allocation analysis: flexible

Cost Allocation vs. Rating

Cost Allocation Rating

Number of Participants <50 >50

Policies Issued Not always Yes

Fixed Number of Participants

Yes No

Entry/exit Little High

Split/merge Yes No

Page 4: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

4© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Cost Allocation General Principles

No single system is appropriate for all state entities

The system cannot be properly designed until the level of risk sharing has been selected

A properly designed and implemented system must address two distinct and classic tradeoffs: stability vs. responsiveness, and simplicity vs. equity

A successful system requires a reasonable investment in data acquisition and maintenance

Systems that initially focus on distributions, rather than absolute dollars, are easier to implement, explain, change, maintain, and update

Page 5: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

5© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Attributes of Sub-Optimal Systems

Attributes of a sub-optimal system Does not promote loss control Participants frequently challenge results Participants do not understand the system No “buy-in” from participants Too much volatility Too simple Too complex No incentives to reduce losses Data problems Improper level of risk sharing Unwanted subsidies Improper application of loss limitation No capping of year-to-year changes Improper use of exposure data

Page 6: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

6© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Evaluate risk sharing

Meet with participants

Obtain/maintain data

Evaluate distributions

Create universe of solutions

Presentation of results

Page 7: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

7© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Evaluate Risk Sharing/Meet with Participants

Should exposures be used? Exposure weight?

Is appropriate exposure data available?

How many years of loss data should be used?

Which years of loss data should be used?

Should individual losses be limited?

How should a new department be rated?

Should year-to-year changes be capped?

What is a reasonable cap?

Page 8: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

8© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Obtain/Maintain Data: Losses

“Traditional” Approach:

Use 3-5 accident years, excluding the most recent and/or current year

Losses and allocated loss adjustment expenses (ALAE)

Reported (paid plus case reserve) loss and ALAE

Losses not developed

Losses not trended

Page 9: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

9© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Obtain/Maintain Data: Losses

Example of problem with “traditional” approach: report lag

20052004200320022001

AYAYAYAYAY

Report Lag

7/1/00 7/1/02 7/1/03 7/1/047/1/01 7/1/05

Page 10: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

10© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Obtain/Maintain Data: Losses

Example of problem with “traditional” approach: payment lag for under reserved claim.

20022001200019991998

AYAYAYAYAY

Case reserve = $50,000

7/1/97 7/1/99 7/1/00 7/1/017/1/98 7/1/02

Cycle 1

Page 11: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

11© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Obtain/Maintain Data: Losses

Example of problem with “traditional” approach: payment lag for under reserved claim.

20032002200120001999

AYAYAYAYAY

Case reserve = $50,000

7/1/98 7/1/00 7/1/01 7/1/027/1/99 7/1/03

Cycle 2

Page 12: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

12© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Obtain/Maintain Data: Losses

Example of problem with “traditional” approach: payment lag for under reserved claim.

20042003200220012000

AYAYAYAYAY

Case reserve = $50,000

7/1/99 7/1/01 7/1/02 7/1/037/1/00 7/1/04

Cycle 3

Page 13: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

13© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Obtain/Maintain Data: Losses

Example of problem with “traditional” approach: payment lag for under reserved claim.

20052004200320022001

AYAYAYAYAY

Settlement = $1,000,000

7/1/00 7/1/02 7/1/03 7/1/047/1/01 7/1/05

Cycle 4

Page 14: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

14© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Obtain/Maintain Data: Losses

Solution to problems of (1) report lag and (2) payment lag for under reserved claim:

Traditional: AY Losses

Revised: AY LossesRY Losses (report year)FY Paid Losses (cash flow)

Page 15: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

15© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Obtain/Maintain Data: Exposures

Exposure Bases for State Programs

Sub-Optimal or Biased Better Data Available? Relativities Available?

Workers Compensation

Payroll Payroll by NCCI Class Sometimes – periodic audit?

Yes*

Automobile Liability Vehicles Vehicle by type Yes Yes*

General Liability FTE’s or expenditures FTE’s by Type ? No

Hospital Professional Liability

FTE’s, visits, budget Visits by type, surgeries, births, beds

? Yes*

Physician/Surgeon Professional Liability

FTE’s FTE’s by specialty ? Yes*

Bond/Crime FTE’s FTE’s by type ? ?

*Industry data

Page 16: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

16© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Evaluate Distributions Prior Allocation Exposures Accident Year Reported Losses (by Limit) New Allocation

Loss Limit

10%

12%

14%

16%

1,000 50,000 500,000

Prior

Exposures

Losses

New

Dis

trib

utio

n

Page 17: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

17© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Evaluate Distributions

Avoid multiple allocation cycles by focusing on distributions, not dollars

Dollars Distributions

Allocation Steps 1-9

Submission

Review

Revision

Allocation Step 10

Acceptance

Allocation Steps 1-9

Submission

Review

Revision

Allocation Step 10

Acceptance

Page 18: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

18© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Work Steps

Create Universe of Solutions(Sensitivity Testing)

Example:

Cap on Change Years Reviewed

± 15% 3, 4, 5

± 30% 3, 4, 5

± 40% 3, 4, 5

Universe of Solutions

Page 19: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

19© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Department

Change due to

Allocation

Change due to

Funding Total

A 20% 25% 50%

B 10% 25% 37.5%

C -10% 25% 12.5%

D -20% 25% 0%

Total 0% 25% 25%

Work Steps

Presentation of Results Present the distributions to each department Components of change exhibit

Components of Change

Page 20: © 2005 Towers Perrin STRIMA 2005 Conference Premium Development/Cost Allocation This document is incomplete without the accompanying discussion; it is.

20© 2005 Towers Perrin

STRIMA 2005 ConferencePremium Development/Cost Allocation

Summary

Goals defined

Buy-in Required

Data Challenges

Focus on Distributions

Sensitivity Testing

Presentation