September 22, 2004 Maine Event 2004 Thinking about Decision Making Rich Clement:Underwriting Decisions Rick Leavitt: Actuarial Decisions - Issues - Simulation.

September 22, 2004

Maine Event 2004Thinking about Decision Making

Rich Clement: Underwriting Decisions

Rick Leavitt: Actuarial Decisions- Issues- Simulation- Game Theory

September 22, 2004

Maine Event 2004

Underwriting Decisions

September 22, 2004

Maine Event 2004

Difficulties In Underwriting

Limited to poor information

Quick turnaround time

Competition – Acquisition/Irrational Pricing

Multiple Products

Tenure of underwriters – size of accounts

Economic conditions

Evaluation periods are too short

Most volatile time period

September 22, 2004

Maine Event 2004


Limited Tools/Consistently Changing Tools

Changing plan designs – move to employee contributions

New sales and underwriting territories

Aging workforce

Obesity – co-morbid conditions

Pressure on underwriting gain from poor investment income

Limitations to Home Office Partners

September 22, 2004

Maine Event 2004


Mind space for LTD

Increasing medical costs

Churning business

Employer’s “attitude” towards employee’s

Regulatory issues

Internal meetings

Profitable Growth Goals

September 22, 2004

Maine Event 2004

Traditional U/W Response

September 22, 2004

Maine Event 2004


Risk Analysis

Read and Understand the RFP

Why is case out to bid?

Who is the broker?

Partner with sales on strategy.

Who is employer? – Financial Underwriting

Manual rates

SIC, OCC, Area factors

September 22, 2004

Maine Event 2004


Risk Analysis (continued)

Plan design features

Experience evaluation

Cause and effect

Most recent period is volatile

Credibility

Other perspectives

Claims, actuarial, similar customers

trust yourself

September 22, 2004

Maine Event 2004

New Approaches

Track Decisions

Competitive advantage

Learn from your mistakes

Evaluate # of claims

Strategic direction

Understand where you make money

Maintain underwriting discipline – mistakes sell

“Hit them were they ain’t” – Ty Cobb

September 22, 2004

Maine Event 2004

Actuarial Decisions

…. Harder than I thought

“All you need in this life is ignorance and confidence; then success is sure”

- Mark Twain

“The greatest obstacle to discovery is not ignorance - it is the illusion of knowledge”

- Daniel Boorstein

September 22, 2004

Maine Event 2004

Actuarial Decision Making

1980’s actuary: Conduct an experience study and recommend rate modifications

21st Century Actuary: ???

“Actuaries are professionals who provide expert advice and relevant solutions for business and societal problems that involve economic risk” ….SOA Promotional Material

Given difficult business environment, it is not enough just to get the pricing right.

September 22, 2004

Maine Event 2004

2004 Disability Product Strategy

1. “Slow and Steady Wins the Race”

- Develop Appropriate Pricing and Stay the Course

- You might need investor and management education

*** Strategy Options ***

2. “Decisive and Aggressive”

- Identify Target Markets and Attack Aggressively

- Fix Mistakes as you Go

September 22, 2004

Maine Event 2004

2004 Disability Product Strategy

*** Strategy Options ***

3. Market Intelligence: “Hit ‘em where they ain’t”

-Exploit Undeveloped Market Niches

4. “Execution, execution, execution”

-Improve Distribution Efficiency/UW Decision Making

-Reduce Administrative Expenses and Cost of Capital

-Risk Evaluation Tools = Smart Decision Making

September 22, 2004

Maine Event 2004

Actuarial Decisions: LTD

Why is it so hard to get the pricing right?

1. Volatility of Loss

September 22, 2004

Maine Event 2004

Loss Volatility

Social Security Age-Gender Adjusted Incident Rates

0.300.350.400.450.500.550.600.650.700.75

1970 1975 1980 1985 1990 1995 2000

September 22, 2004

Maine Event 2004



1. Volatility of Loss – Random versus external impacts

2. Loss takes a long-time to develop

September 22, 2004

Maine Event 2004

LTD Components of Loss

0%

20%

40%

60%

80%

100%

120%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15Duration in Years

DLR

IBNR

Payments

September 22, 2004

Maine Event 2004



1. Volatility of Loss – Random versus external impacts

2. Loss takes a long-time to develop

3. Product Complexity

September 22, 2004

Maine Event 2004

LTD Product Complexity

Risk Drivers are Correlated (Isolation of Risk is Impossible)

Age Elimination Period Work-Place EnvironmentGender Definition of Disability Societal AttitudesOccupation Replacement Percent Health TrendsSalary Maximum Benefit Local and National EconomyIndustry Definition of Earnings Financial Health of EmployerRegion Pre-ExCase-Size M&N Limit

LTD Components of Risk

September 22, 2004

Maine Event 2004Example of Risk Dependence

Distribution of Ages by Occupational Group

8%

9%10%

11%

12%

13%14%

15%

16%17%

18%

30 35 40 45 50 55

Executives

Other

Blue-Collar

September 22, 2004

Maine Event 2004

Actual Manual Rates

A B CMax Rate / Min Rate

Base Rates 0.57 0.64 0.64 113% Age 0.57 0.65 0.61 Gender 1.05 1.04 1.00 Elim 0.77 0.78 0.92 Benefit Dur 0.98 0.97 0.99 SS Credit 0.70 0.67 0.71Rate 1 0.18 0.22 0.25 137%

Companies

September 22, 2004

Maine Event 2004

Actual Manual Rates (Cont)


Rate 1 0.18 0.22 0.25Salary 0.96 1.00 1.00Occupation 1.31 1.25 1.73Industry 0.75 1.16 0.87Area 0.90 1.00 1.00Size Lives 1.00 1.00 1.27Rate 2 0.16 0.32 0.47 306.0%

September 22, 2004

Maine Event 2004

Actual Manual Rates (Cont)


Rate 2 0.16 0.32 0.47Plan Provisions 1.40 1.08 0.96Net Rate 0.22 0.34 0.46 210.2%

Expenses 45.0% 44.0% 45.8%

Final Rate 0.39 0.61 0.84 213.3%

September 22, 2004

Maine Event 2004

Manual Rate Volatility

0

0.5

1

1.5

2

2.5

0 0.5 1 1.5 2 2.5Average Rate

Car

rier

Rat

e

September 22, 2004

Maine Event 2004

Actuarial Decision Making

So Let’s see….

Disability Rates are all over the place

Trends are hard to predict and hard to track

Some Carriers are buying Market share

Some Carriers are behaving, uh, irrationally

Investors want short-term results on a long-term product

What is an actuary to do??

September 22, 2004

Maine Event 2004

One Approach: Simulation

Simulate Reality and Test Various Strategies

But….

Need to simplify may eliminate key dynamics

Be careful not to assume your outcomes

Test outcomes and relationships for “Reasonability”

(Need Reality Check)

September 22, 2004

Maine Event 2004Example: Sales/Profit Model

Assumptions:

All Carriers have identical expertise

Average Rate = Best Estimate of Cost

Sale depends on Rate (5% Threshold)

Incumbent has rate advantage on renewal

Some Cases will take almost any rate increase

Some Cases will leave regardless of renewal action

September 22, 2004

Maine Event 2004

New Sales Price ElasticityMaximum Discount 20%Std Dev of Rates 10%

Manual Rate

MarginClose Ratio

Sold Discount

Sold Rate

Margin

15% 2.7% 11.0% 0.3%13% 3.2% 10.5% -1.4%10% 3.8% 10.0% -3.3%8% 4.4% 9.2% -4.7%5% 4.9% 8.7% -6.3%2% 5.6% 8.0% -8.1%0% 6.2% 7.2% -10.0%-3% 6.9% 6.8% -12.0%-5% 7.8% 5.8% -14.5%-8% 8.6% 5.2% -16.5%-10% 9.6% 4.4% -19.0%-13% 10.8% 3.8% -21.8%-15% 12.0% 3.2% -24.1%

-25.0%

-20.0%

-15.0%

-10.0%

-5.0%

0.0%

5.0%

10.0%

15.0%

-15% -10% -5% 0% 5% 10% 15%

Manual Rate Margin

Close Ratio

Sold Discount

Sold Rate Margin

September 22, 2004

Maine Event 2004

Persistency Model

IncreasePct Not Shopped

Shopped Persist

Total Persist

0% 100% 53% 93%

5% 100% 25% 93%

10% 86% 24% 83%

15% 76% 25% 78%

20% 67% 27% 73%

25% 58% 28% 68%

30% 40% 27% 58%

0%

20%

40%

60%

80%

100%

120%

0% 5% 10% 15% 20% 25% 30%

Increase

Percent not Shopped

Persistency on Shopped Cases

Total Persistency

September 22, 2004

Maine Event 2004

Scenario 1: On Target Pricing No Rate Increase on Renewal

YearClose Ratio Persist Gain

Inforce Premium

InforceGrowth 1-Persistency

1 4.1% 9.7% 120,952K2 3.7% 97.6% 9.7% 122,626K 0.8% 2.4%3 3.6% 95.9% 9.9% 129,439K 1.4% 4.1%4 3.7% 93.2% 10.0% 138,579K 5.6% 6.8%5 3.7% 96.0% 9.9% 151,229K 7.1% 4.0%6 3.5% 94.7% 10.0% 161,662K 9.1% 5.3%7 3.7% 94.2% 10.0% 175,382K 6.9% 5.8%8 3.8% 95.3% 10.0% 189,274K 8.5% 4.7%9 3.8% 95.2% 10.0% 207,111K 7.9% 4.8%10 3.7% 94.2% 9.9% 219,888K 9.4% 5.8%

Average Growth Gain

Year 1-5 4.7% 9.9%Year 6-10 7.8% 10.0%Years 1-10 6.2% 10.0%

0%

2%

4%

6%

8%

10%

12%

1 2 3 4 5 6 7 8 9 10

Year

0M

50M

100M

150M

200M

250M

Close Ratio1-PersistencyGainInforceGrowthInforce Premium

September 22, 2004

Maine Event 2004Scenario 2: Modest Acquisition:Break-even Pricing, 10% Increase on Renewal


Inforce Premium


1 7.9% -0.9% 132,193K2 7.3% 91.7% 1.9% 144,582K 10.2% 8.3%3 7.3% 83.7% 4.3% 156,896K 9.4% 16.3%4 7.5% 86.4% 6.6% 177,806K 8.5% 13.6%5 7.5% 81.4% 7.9% 181,538K 13.3% 18.6%6 7.6% 79.1% 8.0% 188,833K 2.1% 20.9%7 7.4% 77.9% 8.3% 187,089K 4.0% 22.1%8 7.4% 75.4% 7.7% 187,088K -0.9% 24.6%9 7.4% 80.7% 7.9% 197,112K 0.0% 19.3%10 7.6% 81.5% 8.1% 201,981K 5.4% 18.5%

Average Growth Gain

Year 1-5 8.6% 5.3%Year 6-10 2.2% 8.0%Years 1-10 5.3% 7.1%

-5%

0%

5%

10%

15%

20%

25%

30%

1 2 3 4 5 6 7 8 9 10

Year

0M

50M

100M

150M

200M

250M


September 22, 2004

Maine Event 2004Scenario 3: Aggressive Acquisition:Priced at a Loss, 20% Increase on Renewal


Inforce Premium


1 10.4% -5.7% 141,580K2 10.1% 85.6% -2.2% 158,711K 18.0% 14.4%3 10.2% 74.2% 0.4% 169,974K 12.1% 25.8%4 10.2% 73.3% 2.1% 181,380K 7.1% 26.7%5 9.8% 69.3% 2.6% 176,732K 6.7% 30.7%6 10.2% 72.8% 2.9% 178,114K -2.6% 27.2%7 10.0% 71.5% 3.1% 183,033K 0.8% 28.5%8 10.2% 70.1% 2.3% 183,227K 2.8% 29.9%9 10.2% 70.9% 1.9% 184,402K 0.1% 29.1%10 10.0% 72.0% 1.5% 189,517K 0.6% 28.0%

Average Growth Gain

Year 1-5 8.1% 0.5%Year 6-10 1.4% 2.3%Years 1-10 4.7% 1.6%

-10%

-5%

0%

5%

10%

15%

20%

25%

30%

35%

1 2 3 4 5 6 7 8 9 10

Year

0M

50M

100M

150M

200M

250M


September 22, 2004

Maine Event 2004

Scenario 4: Light Acquisition:Two Renewal Increases limited to 5%


Inforce Premium


1 7.0% 2.3% 131,330K2 6.5% 94.9% 3.7% 142,877K 9.4% 5.1%3 6.6% 93.5% 5.0% 162,392K 8.8% 6.5%4 6.3% 88.8% 6.3% 181,092K 13.7% 11.2%5 6.5% 90.9% 7.3% 194,838K 11.5% 9.1%6 5.6% 88.9% 8.5% 205,957K 7.6% 11.1%7 5.0% 86.5% 9.3% 210,020K 5.7% 13.5%8 5.2% 88.4% 9.9% 219,274K 2.0% 11.6%9 5.2% 86.6% 10.4% 226,125K 4.4% 13.4%10 5.0% 89.0% 10.8% 235,147K 3.1% 11.0%

Average Growth Gain

Year 1-5 10.2% 5.7%Year 6-10 3.8% 9.9%Years 1-10 7.0% 8.5%

-1%

1%

3%

5%

7%

9%

11%

13%

15%

1 2 3 4 5 6 7 8 9 10

Year

0M

50M

100M

150M

200M

250M


September 22, 2004

Maine Event 2004

Scenario 5: Classic Underwriting Cycle


Inforce Premium


1 10.4% -5.1% 140,374K2 10.0% 94.1% -5.9% 158,514K 17.0% 5.9%3 9.9% 95.5% -3.5% 187,305K 12.9% 4.5%4 3.3% 87.3% 5.8% 177,076K 18.2% 12.7%5 3.2% 76.1% 12.2% 153,933K -5.5% 23.9%6 3.3% 72.7% 14.9% 132,051K -13.1% 27.3%7 8.7% 75.2% 7.0% 144,912K -14.2% 24.8%8 8.6% 85.3% 4.4% 171,704K 9.7% 14.7%9 8.5% 88.5% 3.8% 195,680K 18.5% 11.5%10 5.1% 84.6% 7.1% 202,541K 14.0% 15.4%

Average Growth Gain

Year 1-5 5.1% 2.0%Year 6-10 5.6% 6.8%Years 1-10 5.4% 4.9%

-15%

-10%

-5%

0%

5%

10%

15%

20%

25%

30%

35%

1 2 3 4 5 6 7 8 9 10

Year

0M

50M

100M

150M

200M

250M


September 22, 2004

Maine Event 2004

Scenario Outcomes

Inforce Premium

100M

120M

140M

160M

180M

200M

220M

240M

260M

280M

1 2 3 4 5 6 7 8 9 10

Year

Target Pricing

Modest Acquisition

Aggressive Acquisition

Light Acquisition

Underwriting Cycle

Annual Gain

-10%

-5%

0%

5%

10%

15%

20%

1 2 3 4 5 6 7 8 9 10

Year

Target Pricing

Modest Acquisition

Aggressive Acquisition

Light Acquisition

Underwriting Cycle

September 22, 2004

Maine Event 2004

Scenario Outcomes (Years 6-10)

1 Priced at Target: No increase on renewal 5 Underwriting Cycle2 Priced at break even: 10% increase on renewal 6 Increasing Rates3 Priced at a loss: 20% increase on renewal 7 Decreasing Rates4 Initially priced below target: 5% increase on renewal 8 Low rates

0%

2%

4%

6%

8%

10%

12%

70M 75M 80M 85M 90M 95M 100M 105M 110M 115M 120M

5Yr Premium

5 Y

r G

ain

4

6

3

2

5

17

8

September 22, 2004

Maine Event 2004

Other Approaches?

Can Game Theory Help??

The theory of games is a theory of decision making in complex situations… when the outcome is uncertain (chance) and there are competing goals (other players)

1944, John Von Neumann – “Theory of Games and Economic Behavior”

1948, John Nash – “Nash’s Equilibrium” – identifies existence of optimal Strategy

September 22, 2004

Maine Event 2004

Game TheoryPure Skill Pure Luck Luck and Skill

Chess War PokerGo Stock Market

Checkers LTD ??Tic-Tac-Toe

Chutes and Ladders

Trivial !

Categories

Zero-Sum versus Non Zero-Sum

2 Person versus Multi-person

Cooperative versus Non-Cooperative

September 22, 2004

Maine Event 2004

Optimal Strategies??

Equilibrium Strategy

Defined as a strategy in which you will have no regrets regardless of what the other

players do

Nash Equilibrium – 1948

An equilibrium strategy exists for all two-person games

zero-sum and non-zero sum

September 22, 2004

Maine Event 2004

Prisoner’s Dilemma

Question: What is the equilibrium strategy?

Option 1 Option 2

Option 1 2,2 0,3

Option 2 3,0 1,1

Pla

yer

1 Player 2

September 22, 2004

Maine Event 2004


You and your partner are caught leaving the scene of a crime with no other evidence. You are placed in separate rooms and pressured to testify. What should you do?

Keep Silent Testify

Keep Silent Both go FreeYou go to jail, but your

partner goes free

Testify You go free, but your partner goes to jail

Both convicted of a lesser charge

Your Partner

You

September 22, 2004

Maine Event 2004


Best Possible Payoff: Keep Silent

If you keep silent and your partner testifies, you will regret decision

If you testify:

Partner keeps silent: You go free (Good decision)

Partner Testifies: Lesser charge (Good decision)

Both Partners testify = Nash Equilibrium

September 22, 2004

Maine Event 2004


Nash Equilibrium minimizes expected payout.

Is this really the optimal strategy?

What strategy would you use over multiple games?

Option 1 Option 2

Option 1 2,2 0,3

Option 2 3,0 1,1

Pla

yer

1 Player 2

September 22, 2004

Maine Event 2004

Multi-player Games Complicated by Possibility of Alliances

Start-up company offers the following bonuses

24K

Actuary Underwriter

20K 16K

Claims

27K if all three join

Who should join? How much should each get?

September 22, 2004

Maine Event 2004

Multi player Games

Option 1: All three join and get $9K each

Option 2: Actuary and Underwriter get $12K each

… but wait,

Claims person offers the underwriter $13 (keeping $3)

Actuary offers the claims person $4 (keep $16)

Underwriter offers the claims person $5 (keeping $11)

… where will the negotiation end up?

September 22, 2004

Maine Event 2004

Multi-player Games

Any solution is unstable. There can always be an alliance with another player that works better for one member of the alliance.

No Optimal Solution: Too often all parties end up with nothing

September 22, 2004

Maine Event 2004

Game Theory

Interesting, but…

N-player, non-zero sum, cooperative games are poorly understood.

Game Theory can characterize optimal solutions, but does not do well at identifying actual solutions

Defining player utility in real life is difficult

Games are more fun to play than they are to study…

September 22, 2004 Maine Event 2004 Thinking about Decision Making Rich Clement:Underwriting Decisions Rick Leavitt: Actuarial Decisions - Issues - Simulation.

Documents

maine event

underwriting decisions

impossible slide

volatility of loss slide

loss volatility slide

components of loss slide

product complexity slide

manual rate volatility