development and expenses

Question 1

development ensures

Answer 1

rates in future policy period will be adequate to cover ult costs coming from policies written using those rates

Question 2

development represents

Answer 2

-development represents changes over time to costs or premiums within given exposure period

Question 3

why can exposures, prem, loss and ALAE develop

Answer 3

• exposures and prem can develop due to prem audits and retrospective rating adjustments
• losses and ALAE can develop due to claim payment being made or case reserves being changed

Question 4

order of development for different loss aggregation

Answer 4

• CY < RY < AY < PY
• PY takes longer to develop than AY since avg accident date for PY is later than avg accident date for AY

Question 5

4 statements in estimating ultimates

Answer 5

1. exploratory analysis of data
2. apply appropriate tech to estimate ultimates
3. evaluate conflicting results of different techs
4. monitor projections of actual versus expected development

Question 6

diagnostic triangles

Answer 6

• closed claim counts/reported claim counts: speedup in closing claims over time
• closed without pay claim counts/closed claim counts: higher % of claims being closed without pay
• avg paid on closed: reflect severity rends and speedups in closing of small claims relative to large claims
• avg case reserves: reflect severity rends, speedups in closing of small claims relative to large, and changes in case reserve adequacy
• paid losses/reported loss: reflect speedups in closing of claims and changes in case reserve adequacy

Question 7

group claims by LOB based on

Answer 7

Similarity in coverage

Volume of claim counts

Reliability of case reserves

Report lag

Settlement lag

Likelihood of claims reopening

Claim severity

Question 8

ult can be made more accurate by choosing level of granularity that best balances

Answer 8

homogeneity and credibility of data

Question 9

Main development methods

Answer 9

• chain ladder: uses past development patterns to newer data to forecast ultimates
• expected claims: ignores newer data and uses a priori estimate
• BF: mix of Chain Ladder and Expected claims
• Cape Cod: similar to BF but expected portion is calc differently
• Freq and severity techniques: develop claim counts and severity separately
• case outsanding: looks at paid on prior case reserves
• Berquist Sherman: restate data for changes in case reserve adequacy or settlement rates

Question 10

Reasons for negative development

Answer 10

• case reserve decreases
• deductible recovers
• subrogation and salvage

Question 11

selecting tail factor

Answer 11

-when we observe significant development still occurring in last columns of a triangle, need to select tail factor that bring losses to their ultimate level

Common methods for selecting:

• doing special study that contains more years of data
• using industry benchmark tail factor
• fitting curve to LDFS and extrapolating
• use reported-to-paid ratios at latest paid development period
• judgment

Question 12

how do one time and continuous changes impact LDFs

Answer 12

-one-time changes do not typically impact LDFs -> if needed changes can be usually isolated into separate rows of triangle

-continuous changes: trends are observes as you go down columns and development is observes as you go across rows

Question 13

3 common methods used to incorporated UW expenses in RM

Answer 13

1. All variable expense: treat all UW expenses as variable to prem -> assumes future expense ratios will be consistent with hist ERs
2. Prem-based projection: assumed future ERs will be consistent with hist ERs, but separately calculates fixed and variable ERs
3. Exposure/Policy-based projection: divide fixed expenses by exposures and variable expenses by premium
   - all methods divide variable expenses by prem
   - 1&2 divide fixed expenses by prem and 3 uses exposures
   - divide expenses incurred through the policy by EP/EE

Question 14

when does using EP/EE and WP/WE make a difference?

Answer 14

this only makes difference is company size is changing in which written and earned amounts will not be equal

Question 15

all variable expense method: when is it inaccurate and how is expense trend used

Answer 15

• method can result in inaccurate cost estimates if some expenses are truly fixed
• if some are fixed -> undercharge risks with lower than avg prem and overcharge with higher than avg prem as it bases expenses fully on premium
• distortion can be accounted for by using prem discount or expense constant in rating algorithm
• no expense trend needs to be applied to historical expenses since all expenses are fully based on prem

Question 16

problems with prem-based projection method

Answer 16

• one-time changes can causes hist ERs to be different than expected future ratios -> can correct for by using OLP before calc ratios even though it can be difficult to on-level CW prem
• premium and/or expense trends can also impact expense ratios -> trend all prem and expenses to projected future level before calc ERs
• can create inequitable rates across states when CW fixed expenses are allocated to state level because states with higher than avg prem will get higher allocation -> calc fixed ERs by state

Question 17

possible enhancements to exposure based projection method

Answer 17

• finding more scientific way to split expenses into fixed and variable components
• finding more equitable way to allocate CW expenses to states
• some expenses that are considered fixed actually do vary by char. such as RB vs NB -> asset share pricing to differentiate
• existence of economies of scale in changing BOB will lead to increasing or decreasing future avg expenses per exposure -> impacts could be identified more directly and quantified

Question 18

permissible loss ratios

Answer 18

variable PLR=1-var exp%-target profit

• portion of each $ of indicated prem that is permitted to be spent on loss, LAE, and fixed expenses
• denominator in PP and LR methods to determine ind. avg prem or ind. RC

total PLR=1-total exp%-target profit

• portion of each $ of indicated prem that is permitted to be spent on loss and LAE
• will be same as VPLR if all expenses are treated as variable

Question 19

2 methods for calculating a rate level indication

Answer 19

1. Pure premium method: indicated average rate
2. Loss ratio method: indicated average rate change

Question 20

2 main differences between PP and LR methods

Answer 20

1. LRM relies on LR which requires prem -> current prem must be available and it must be on-leveled, developed if necessary, and trended
   - PPM relies on PP which requires exposures in denominator -> exposures must be well defined and future exposures will need to be estimated (if estimate is based on historical exposures, exposures need to be developed if necessary and trended)
2. output is different