Reserving - ULAE Flashcards

1
Q

Approaches to Estimating ULAE

A
  • can be grouped categorically as either dollar based techniques that relates ULAE dollars to claims dollars, count based tech that assign ULAE cost per claim or transaction, or triangle based tech that rely on ULAE payment pattern
  • can also use market value approach which = cost that TPA would require to take over handling book of claims
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Dollar-based Tech: Generalized Approach

A

-common assumption: ULAE cost track with claims cost in both timing and amount

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

selecting ratio W*

A

-after obtaining M and B for several past CYs, can calc W=M/B for each year and use judgment to obtain selected ratio W*

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

3 ways to estimate unpaid ULAE for group of AYs as of given ultimate claims for same group (L)

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

When is approach not appropriate?

A

it does not account for when ULAE inflation is different than claims inflation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Simplified Generalized approach

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Classical Approach: key assumptions

A

ULAE costs are proportional to $s of claims being handled

  • half of ULAE is spent when opening and half spent when closing (U1=50% and U3=50%)
  • no partial payments
  • when estimating B, need additional assumption that insurer’s BOB has reached stead state so R=P
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Classical Approach formulas

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Classical Approach: not appropriate

A

-approach does not work well for long-tailed LOBs, when ULAE inflation rates differ from claims, when insurer is significantly growing/shrinking, or when 50/50 not appropriate

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Kittel Approach & its assumptions

A
  • only difference from Classical is claims basis estimation, B
  • steady state assumption of R=P would lead to inaccurate results for BOB growing/shrinking

2 assumptions:

  1. change in pure IBNR during any given CY is 0 (same assumption made under generalized approach)
  2. prior year total reserves are accurate
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Kittel Approach formulas

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Kittle appropriateness

A
  • improvement over classical when book is growing/shrinking and for long tailed
  • not good when 50/50 not appropriate or inflation differs
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Mango-Allen

A
  • only difference from classical is that expected paid claims are used instead of actual paid when estimating B
  • useful when insurers have limited or volatile CY paid
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Count-Based Techniques: recognizes 2 issues

A

Recognizes 2 issues with dollar-based approaches

  1. ULAE is not directly proportional to claims
  2. Using ULAE-to-claims ratios results in volatile ULAE when claims are volatile
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Count-based Techniques: assumption

A

Key assumption = same kind of transaction costs the same ULAE regardless of claims size

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Count-Based Challenge

A

Challenge = obtaining accurate and consistent claim count data

17
Q

Brian Tech

A
  • ULAE split into 5 types: opening, maintaining open, making payments, closing, and reopening claims
  • ULAE costs for each of these transactions was similar and could be estimated using historical ratios of ULAE to number of transactions
18
Q

Wendy Johnson

A
  • ULAE split into 2: opening and maintaining claims
  • ULAE could be estimated using historical ratios
  • allowed cost per transaction to vary for different types unlike Brian