Group Chap 25: Estimating Disability Claim Costs

Question 1

Introduction

Answer 1

1. Own experience is preferred, but usually is unavailable or insufficienctly reliable or credible
   a. Alternative sources of Data for Disability Claims:
   - Intercompany experience studies
   - Rate filings by other insurers
   - Governmental and business publications
   - Reinsurance and consulting firms

Question 2

Long Term Disability

1. Estimating Claims Costs
2. Adjustments for Plan Variations
3. Adjustments for Group Characteristics
4. Miscellaneous

Answer 2

1. Expected Claim Costs for LTD involves:
a. Present value of disabled life annuity as well as claim incidence

b. Annuity depends on interest rate, claim termination rates, maximum duration of benefits, and benefits offsets

c. Pricing and reserving interest assumptions should reflect investment strategy for assets to back the reserves

d. Claims adjudication practices may impact expected duration of claims, and reserve assumptions

2. Sources of Data
a. Insurer studies: loss ratio studies, actual-to-expected claim incidence studies, A/E claim termination rate studies, benefit offset studies

• Loss Ratio Studies
  i. Calendar Year loss Ratio Study
• Computes ratio of incurred claims to earned premium where incurred claims are paid claims plus change in claim reserves during the year
• Closest relation to company’s financial statements, not the clearest picture of historical trends, because affected by claims incurred long ago
• May overstate morbidity unless remove required
  interest on reserves

ii. Incurral Year Loss Ratio Study
* Incurred claims divided by earned premium where incurred claims are PV (claim payments to date) plus current claim reserve, all discounted to year of incurral

• Earned premium is the earned premium for year of incurral
• Do not correspond directly to financial statements, better historical trend
• No further interest adjustment is required
• Failure to include IBNR will result in understated values for recent years
• A/E Claim Incidence and Claim Termination Studies
  i. 100% signifies actual experience equal to expected
  ii. Expected values often based on published table
  iii. Large block of business may rely on its own historical experience, rather than published table
• Studies can be segmented into age group, gender, elimination period, benefit percentage, type of offset, employee contribution, size of group, industry, area
• Useful in identifying experience trends

b. SOA 2008 GLTD Experience Table (and 2012 GLTD Experience Table)
i. Recent SOA Study

ii. Modifications from old studies - longer select period, breaks total claim elimination ratios aprt for deaths and recoveries, add adjustments for own occupation vs any occupation definition and added detail on claim termination rates by diagnosis

iii. 2012 GLTD Valuation Table will be used for determining minimum standards for claims incurred on or after October 1, 2016 (per the NAIC model regulation)
* 2012 table is based on 2008 table with 15% margin on termination rates and a provision for mortality improvement
iv. 2008 and 2012 tables don’t include incidence rates
* No good industry source of LTD incidence rates, so many companies look at own experience and loss ratio studies, in addition to using individual disability tables
with adjustments

3. Claim Costs for Standard Plans
a. Claim Cost = Incidence Rate x Sum of (Benefit x Continuance x Interext Discount) over benefit period

b. Incidence rates and claim termination rates may be separated into diagnosis groupings (e.g. mental disorders, maternity, cancer, and all remaining diagnoses)

c. More often, pricing using aggregate assumption for combination of all diagnosis groups

d. Social Security Offsets
- After 6 months, SSDI offsets can be anticipated
- One approach calcualtes using probabilities of Social Security disability awards, formular for Social Security disability benefits, and features of plan
(1) Features include benefit percentage, minimums and maximums, and type of Social Security offset (direct offset or indirect offset)
(2) Probabilities should represent all levels of appeal
(3) Also consider timing of Social Security approvals
- SSDI Benefit depends on “bend points” adjusted for inflation each year
(1) Primary Insurance Amount (PIA) = 90% x (Bend Point 1) + 32% x (Bend point 2 - Bend point 1) + 15% x (AIME - Bend Point 2); AIME = Average Indexed Monthly Earnings
- Family SSDI is 150% of PIA (cannot exceed 85% of AIME)
- Insurers usually use simplified formuals to estimate expected offsets, based on age and salary

4. Canadian Integration
a. Canadian Pension Plan (CPP) and Quebec Pension Plan (QPP) available after five months of disability

b. Primary awards are less than Social Security, but dependent awards can be large

c. Some benefits on after-tax replacement level
- Tax rate can be approximated, based on salary

5. Adjustments for Plan Variations
a. Benefit Percentage
- Higher income replacement provides greater incentive to claim and less incentive to return to work
- Savings from offsets become proportionately smaller as benefit amounts increase, resulting in higher rating factors for higher benefit percentages

b. Maxium and Minimum Benefit
- Higher claim rates result from the incentives implicit in large benefits

c. Elimination Period
- Affects both claim incidence and claim termination rates

d. Benefit Period
- Variations can be priced by net premium calculations on the basis of 1987 CGDT or 2008 CLTD Table

e. Definition of Disability
- Majority of plans have “own occupation” period of 2 years from start of benefit payment
- After 2 years, benefits continue only if unable to perform “any reasonable occupation”

f. Other offsets
- Including state disability and Paid Family and Medical Leave (PFML), workers’ compensation, pension benefits, sick pay, and part time work

• Impact of sick pay may be priced as explicit offset or requiring that sick pay be exhausted before LTD benefits begin
• Part time work offsets handled through offset formulars and provisions
• State-specifics credits not necessary, because WC provided in all states with considerable uniformity
• Specific rate credits unncessary for emloyers’ pension and profit sharing plans
  i. Credits determined on an ad hoc basis
  ii. Issue is most significant when claimant can retire before 65 with a defined benefit pension plan
• Rate credits desirable for offsets of temporary disability benefits mandated by some states

g. Limits on Mental Illness and Substance Abuse
- Most plans impose lifetime limit of 24 months on mental and nervous when not confined to institution
i. Many impose similar limit on alcoholism and chemical dependency

• Rating adjustments for coverage variations
  i. Loading should be greater for certain industries
• Be aware of recent court decisions affecting two year mental nervous limitations

h. Optional Features
- Pension Supplement - Extra premium for a pension supplement; amount of additional disability benefit repressented by pension contributions

• Survivor Benefits - Involve probabilities claimants dying and having eligible dependents
• COLA riders
  i. Loading depends on terms, CPI-related or flat percentage adjustment; annual limit on changes; cumulative limit on increases; and limit on number of years of adjustment
  ii. Catastrophic disability riders - use assumptions similar to pricing long-term care

i. Underwriting Variations
- Variations in actively-at-work requirements, evidence of insurability (EOI) and preexisting condition limitaions (PEL)
- Preexisting condition limitation “3-3-12”
(1) First 3 - indicates preexisting if during the 3 months prior to effective date
(2) Second 6 - Indicates exclusion expires after 3 months without treatment
(3) Final number, indicates exclusion expires after insured 12 months, regardless of treatment

6. Employee Contribution / Participation
a. For plans paid with pre-tax dollar, benefits are taxable
- Plan paid with after-tax dollars, beneifts are non-taxable
- Non-taxable 60% benefit may replace 70% or more on after-tax basis
- Rating factor for contributory plans should reflect level of after-tax replacement

b. Load increases as participation decreases
- At participation below 25%, the load may be replaced by individual underwriting

7. Adjustmets for Group Characteristics
a. Age and Gender
- If factors appropriately reflected in claim cost tables, then the age and gender mix of the group will be reflected in the resulting premium rate, even though final premium chaarged may not vary directly by age and gender

b. Occupation
- Establish factors for broad categories of occupation, such as
i. Hourly, salaried, and commissioned
ii. Blue collar, grey collar, and white collar

c. Industry
- Change premiums according to industry of group, rather than occupation of the individual
- Avoid double-counting with earnings factor

d. Average Earnings per Employee
- Widespread belief that disability rates better for higher paid workers
- Not all industries experience more favorable claims with higher earnings
i. High income, may have high rates of disability due to physical requirements, economic uncertainty or stress

e. Area
- Area variations in LTD significant
- Credible experience by area not as available for LTD as for medical. Generally distinguish only between entire states (some may be more precise on ZIP code ratings)
- May get appropriate area adjustments by looking at rate filings of leading carriers, from state insurance departments
- Reasons for adverse experience in Califonia include entitlement culture, litigiousness, attitudes of lawyers and judges
- Variations by territory also caused by quality of field underwriting

f. Size of Group
- Claim costs highest for the largest and smallest employers
- Insurers provide volume discounts reflecting commission scales which vary by size

8. Miscellaneous Factors Affecting LTD Claim Cost
a. Economic Cycle
- During economic downturn, lower claim termination rates and increased claim incidence rates
- Overall rate adjustment factor to change all manual rates simultaneously, to reflect changes due to economic cycle
- Many insurers avoid overreacting to cyclical swings
- Tightening of underwriting a more flexible response

b. Distribution System
- Groups covered under association programs experience greater adverse selection than under single empoyer cases

c. Catastrophic Risks
- Since 2001, cost of catastrophe reinsurance has increased significantly; need to consider explicitly in pricing of disability insurance
- COVID-19 - example of catastrophic risk for disability insurers - Early indications - greater cost increase for STD than LTD

Question 3

Short-Term Disability

• Unlike LTD, not necessary to separately determine rates of disablement and disabled life annuities
• Studies of own experience based on loss ratio studies
• Loss ratios studies available separately by plan, size of group and industry, but usually not by age or
  gender

Answer 3

1. Sources of Data
   a. Due to high frequency and low severity of STD, credibility faster than LTD o Best source of data is company’s own experience
   b. Littleusefulinformationpublished
   c. Source of STD Data
   ▪ TSA 1983 Reports
   ▪ 1985 Commissioners Individual Disability Table A (1985 CIDA)
   i. Useful in developing adjustment factors for group STD by age and occupation. not useful in establishing overall level of claim costs
   ii. Note normal maternity claims are excluded from 1985 CIDA
   ▪ Individual Disability Experience Committee Studies
   i. Detailed experience on short benefit period and occupation groupings is useful to STD insurers
   ii. Excludes claims due to normal maternity
2. Adjustments to Experience Studies
   a. Experience of a group insurer will depend on its marketing strategy, distribution, field and home office underwriting, claim administration, and accuracy of pricing
   b. Should smooth or graduate data, so that final rates have reasonable internal relationships o Trends should be recognized
   c. Make appropriate allowance for maternity costs, which aresignificant for STD
3. State Mandated Cash Sickness Benefits
   a. Statistical reports published by government agencies
   b. Insurer can use regular STD manual rates adjusting for unusual features