Reserving

Question 1

Adequacy of case outstanding is increasing for an insurer. How each of the following would be influenced by this change and briefly explain why:

i. Reported Claim Development Technique

Answer 1

i. The rept dev technique would overestimate ult claims since it would apply historical
development patterns based on lower case outstanding adequacy to a higher rept claims
number, resulting in higher estimates of ult claims.

Question 2

Adequacy of case outstanding is increasing for an insurer. How each of the following would be influenced by this change and briefly explain why:

ii. Expected Claim Technique

Answer 2

ii. The EC technique would not be affected by the change in case outstanding (estimates would
remain accurate) to the extent that the ECR calculation is not influenced by the changing case
adequacy (if the changing adequacy only started with the latest accident period, then it will not
impact the ECR).

Question 3

Adequacy of case outstanding is increasing for an insurer. How each of the following would be influenced by this change and briefly explain why:

iii. Reported Bornhuetter-Ferguson Technique

Answer 3

iii. The rept B-F technique would overestimate ult claims since the % unrept would not reflect the
changing case adequacy. However, it will not overestimate ultimates by as much as the rept dev
technique.

Question 4

Adequacy of case outstanding is increasing for an insurer. How each of the following would be influenced by this change and briefly explain why:

iv. Reported Cape Cod Technique

Answer 4

iv. The rept CC technique would overestimate ult claims since the ECR calculation would be influenced by the changing case adequacy in the most recent year, causing the ECR to be higher.
Also, the % unrept would not reflect the changing case adequacy. As such, CC will overestimate by more than the rept B-F technique.

Question 5

• The company recently began dedicating more legal resources to defend claims at earlier stages
in the claim cycle in an attempt to reduce ultimate claim costs.
• Prior to this claims initiative, the company had a relatively stable claim history.
• Paid legal expenses are tracked separately and are considered allocated claim adjustment expenses for the company.

Describe a potential challenge of using the paid development technique to estimate unpaid allocated claim adjustment expenses for this company.

Answer 5

The speedup in spending legal costs will overstate the ult ALAE when using the paid dev method since the higher paid ALAE will be multiplied by LDFs that assume a slower ALAE payment rate. Also, the LDFs may be highly leveraged (compared to using a paid-to-paid ratio approach).

Question 6

• The company recently began dedicating more legal resources to defend claims at earlier stages in the claim cycle in an attempt to reduce ultimate claim costs.
• Prior to this claims initiative, the company had a relatively stable claim history.
• Paid legal expenses are tracked separately and are considered allocated claim adjustment expenses for the company.

Describe a potential challenge of using a paid ALAE-to-paid claims only ratio technique to estimate unpaid allocated claim adjustment expenses for this company.

Answer 6

The speedup in ALAE spending will result in higher paid-to-paid ratios at earlier maturities,and the ult ratios will likely increase compared to prior years if less claims are paid as a result of the initiative.
This will look like a speedup in the paid-to-paid ratio triangle, and can result in overestimating ALAE since the higher paid-to-paid ratios will be multiplied by LDFs that assume a slower ALAE payment rate relative to claims. However, it is easier to judgmentally adjust for this using the ratio approach than with the development method applied to ALAE directly.

Question 7

• The company recently began dedicating more legal resources to defend claims at earlier stages in the claim cycle in an attempt to reduce ultimate claim costs.
• Prior to this claims initiative, the company had a relatively stable claim history.
• Paid legal expenses are tracked separately and are considered allocated claim adjustment expenses for the company.

Comment on whether using a frequency-severity technique is appropriate to estimate legal expense reserves for this company.

Answer 7

A freq-sev approach would probably not be appropriate since it isn’t clear what the definition of claim counts should be as regards legal expenses. For example, paid claim counts would not be appropriate since there may be claims with significant ALAE paid but no indemnity paid amount. Furthermore, freq-sev approaches assume that severity would develop as it has in the past, which would not be true with the new claims initiative as ALAE would be paid sooner.

Question 8

The Benktander technique can be viewed as a credibility weighting of other common techniques. Identify these techniques.

Answer 8

1. B.F

2. Development technique

Question 9

an insurer uses the reported development technique based on its historical accident year data to set reserves.

i. Discuss the effect on estimated ultimate claims and
ii. Identify either an alternate technique or an adjustment to the reported development technique to improve the estimate, if necessary.

a.
Mid-year the company institutes a new policy for setting case outstanding for open claims, in which case outstanding is set at policy limits.

Answer 9

i. Rept claims will be higher after the increase in case OS, so multiplying by a historical LDF would overstate ult claims.

ii. Any 1 of:
- The expected claims technique would not be impacted by the change in case OS, so it may be more accurate.
- The paid dev technique would not be impacted by the change in case OS, so it may be more accurate.
- The rept Berquist-Sherman technique could be used to bring all years to the current case OS adequacy level, and then LDFs could be determined from the adjusted rept triangle to bring claims to ult more accurately.

Question 10

An insurer uses the reported development technique based on its historical accident year data to set reserves.

i. Discuss the effect on estimated ultimate claims and
ii. Identify either an alternate technique or an adjustment to the reported development technique to improve the estimate, if necessary.

b.
The company had historically stable writings, but undertakes an advertising initiative in the second quarter and increases its premium volume written through the end of the year by 300%.

Answer 10

i. Since the company is growing rapidly during the year, the avg accident date will be later in the year than in historical years. As such, applying a historical LDF to the less mature latest AY would understate ult claims.

ii. The expected claims technique would not be impacted by the change in the avg accident
date, so it may be more accurate.

Question 11

An insurer uses the reported development technique based on its historical accident year data to set reserves.

i. Discuss the effect on estimated ultimate claims and
ii. Identify either an alternate technique or an adjustment to the reported development technique to improve the estimate, if necessary.

c.
At the beginning of the year, the company began offering a general liability product covering losses in excess of its basic limits.

Answer 11

i. Using historical LDFs would not consider the presumably longer-tailed development pattern from this new liability product, so ult claims would be understated.
ii. The expected claims technique could be used to estimate the ult loss ratio for the company’s products, which could be used to derive an ult claims estimate that would not be impacted by the new development pattern.

Question 12

Describe why linear interpolation may not be appropriate for estimating the expected reported claims for an immature accident year.

Answer 12

Linear interpolation assumes losses will develop evenly throughout the year. However, development tends to be higher in earlier maturities and decreases over time, so we would expect more development earlier in the year and less later in the year.

Question 13

State the key assumptions of the classical technique, and briefly comment on the appropriateness of utilizing the classical technique in estimating unpaid ULAE for a company that just started.

Answer 13

The key assumptions are:
• The insurer’s ULAE-to-claim relationship has achieved a steady-state.
• The relative volume and cost of future claims management activity on not-yet-rept claims and rept-but-not-yet-closed claims will be proportional to the dollars of IBNR and case OS, respectively.

Since this is a new growing company, it is unlikely that the ratio of paid ULAE to paid claims has stabilized, so using the classical technique would result in inaccurate estimates of unpaid ULAE.

Question 14

Describe a refinement to the classical technique that can be used to derive a reasonable estimate of unpaid ULAE for a new company.

Answer 14

Any 1 of:
• Using the Kittel approach would rely on both paid and incurred claims, and may result in more stable ratios of ULAE to claims than the classical technique.
• Using the Mango-Allen approach would rely on expected paid claims instead of actual paid claims, and this may result in more stable ratios of ULAE to claims since this is a small company and actual paid claims may be volatile.

Question 15

Catastrophe (Storm)

Identify a technique that will understate the estimate of ultimate claims.

Answer 15

• The paid dev technique will understate ult claims since it does not incorporate the storm losses into the estimate.
• The expected claims technique will understate ult claims since it does not incorporate the storm losses into the estimate.

Question 16

Catastrophe (Storm)

Identify a technique that will overstate the estimate of ultimate claims.

Answer 16

The rept dev technique will overstate ult claims since it will assume IBNR should also be higher due to the storm losses, which is not true.

Question 17

Catastrophe (Storm)

Identify a technique that will result in a reasonable estimate of ultimate claims.

Answer 17

The rept Bornhuetter-Ferguson method will recognize the higher losses from the storm while still calculating IBNR that isn’t impacted by the higher storm losses.

Question 18

The Loss Ratio for a book of business is improving.
There have been no changes in either loss emergence patterns or the company’s claim reserving practices.
IBNR have been calculated with 3 different methods :
1. Development method
2. BF
3. Expected Claims method

Discuss the accuracy of each of the three methods in the situation described above.

Answer 18

The development method will be accurate since the development pattern is not changing.

The Expected Claims method will overestimate the IBNR since it will not recognize the improvement in loss ratio.

The B-F method will also overestimate the IBNR, but not to the same extent as the EC method.

Question 19

Briefly explain whether the Bornhuetter-Ferguson or the Cape Cod technique is more appropriate in the following scenarios:

i. Decrease in underlying claims ratio.
ii. Thin or volatile data.

Answer 19

i. When the underlying claim ratio is changing, CC is more appropriate than the B-F technique since it uses more recent data to calculate the ECR.
ii. B-F is more appropriate with thin or volatile data since the ECR calculation for CC requires a sufficient volume of credible rept claims in order to calculate a reliable estimate.

Question 20

Discuss how to adjust the development paid techniques to make it appropriate for larges claims

Answer 20

Split the data into separate triangles for small and large claims. Use the Berquist-Sherman settlement rate adjustment on both triangles (assuming there would be a corresponding slowdown in large claim handling) to restate paid claims using disposal rates from the latest diagonals.

Question 21

List three considerations when establishing a large claim threshold for the purpose of estimating unpaid claims.

Answer 21

• Number of claims over the threshold each year
• Size of claim relative to policy limits
• Size of claim relative to reinsurance limits
• Credibility of internal data regarding large claims
• Availability of relevant external data
• Whether the threshold is defined qualitatively or quantitatively

Question 22

Contrast the effect that large claims have on the development technique and the Bornhuetter-Ferguson technique for estimation of unpaid claims.

Answer 22

SOLUTION 1: Large paid claim only impacting year for which we want unpaid claims.

Assuming LDFs and BF ECR are not impacted by large claims, a large paid claim in the current year would cause the paid development method to overstate unpaid claims. Since the paid BF method relies on expected data only for unpaid claims, it would be accurate.

Question 23

Discuss the applicability of the Bornhuetter-Ferguson technique when cumulative claim development factors are less than 1.00.

Answer 23

With CDFs less than 1, the credibility interpretation of the BF method is no longer reasonable, but the method can still be used to produce estimates. You can continue to use the method as is, you could limit CDFs to a minimum value of 1, or you could rely on a different technique to select ultimates for years with CDFs below 1.

Question 24

Discuss whether the frequency-severity technique is appropriate for determining an estimate of unpaid claims for general liability.

Answer 24

SOLUTION 1: It is appropriate

The freq-sev technique is appropriate for use for general liability. Frequency is likely to be relatively stable, while severity is volatile, so actual freq can be used in combination with expected sev or a fitted sev distribution to produce estimates.

SOLUTION 2: It is not appropriate

Freq and especially sev for general liability is likely to be very volatile, so trying to develop them to ult will produce volatile estimates of unpaid claims. As such, it would not be appropriate.

Question 25

The insurer recently changed their offering from large deductible policies to small deductible policies. Discuss the impact of this change on the frequency-severity technique, including an assessment of the appropriateness of the technique.

Answer 25

The move to smaller deductibles will increase freq since more claims will be above the lower deductibles, but will decrease sev because the inclusion of smaller claims (net of a lower deductible) will bring down the avg claim size. Overall, losses should be higher with lower deductibles, so the freq increase should outweigh the sev decrease. The change will likely impact development patterns, so using historical freq and sev LDFs will not be accurate. However, using a freq-sev technique with judgmental adjustments to historical frequencies and severities could be appropriate.

Question 26

Suggest an improvement using Freq Sev technique if a company changed the deductibles to smaller deductibles

Answer 26

The analysis could be done on a PY basis, which would isolate the deductible change, and thus development patterns would not be distorted by the change. The actuary could then continue to use the freq-sev technique for older years, and for newer years use judgmental selections of ult freq and sev.

Question 27

Discuss a possible distortion when using the Berquist-Sherman paid claim development adjustment.

Answer 27

BS technique does not recognize that change to settlement could be restated to claim size. This means that if you have smaller claims for the latest year compared to historical this could be a problem. Since the BS adjustment assumes no change in the prioritization between small and large claims, it will underestimate ult claims since it will assume the percent of ult claims paid will be too high.

Question 28

An actuary is using the development technique based on accident year data to calculate ultimate claim estimates at 12 months maturity. Briefly discuss how it may impact the analysis and propose an appropriate response to mitigate the issue. a. The actuary observes a long development pattern.

Answer 28

A long development pattern in itself is not a problem as long as LDFs are stable. However, for a long-tailed line, they are not likely to be stable, so a method that relies to some degree on expected claims will be appropriate here. Since LDFs at early maturities will be highly leveraged and volatile, the estimates of ult claims using the development method will be very volatile as well. As such I would recommend using any 1 of: * the expected claims method, since it doesn’t rely at all on the volatile LDFs. * the Bornhuetter-Ferguson method, since it combines actual data with expected future development to produce more stable estimates. * the Benktander method, since it combines actual data with expected future development to produce more stable estimates, and gives extra weight to the development method compared to the BF method.

Question 29

An actuary is using the development technique based on accident year data to calculate ultimate claim estimates at 12 months maturity. Briefly discuss how it may impact the analysis and propose an appropriate response to mitigate the issue. b. Tort reforms anticipated to decrease severity of all open and future claims were recently enacted.

Answer 29

The reduction in benefits will reduce development for open claims. To deal with this, I would recommend any 1 of: * on-leveling all claims and restating the triangles before performing the development method (especially if ults are being used in ratemaking). * using a freq-sev method, since freq would remain unchanged and sev could be adjusted for the benefit reduction. * using an expected claims method and judgmentally select ECR values that take the change into account. * using a BF method and judgmentally select ECR values that take the change into account.

Question 30

An actuary is using the development technique based on accident year data to calculate ultimate claim estimates at 12 months maturity. Briefly discuss how it may impact the analysis and propose an appropriate response to mitigate the issue. c. In recent years, policies have been written with higher deductibles than in prior years.

Answer 30

The higher deductibles would reduce ult claims, and would change development patterns since smaller claims would be taken out of the data. To account for this, I would recommend any 1 of: * using PY triangles to isolate deductible changes to separate rows. * re-stating historical data at higher deductibles and then performing the development method. * applying a loss elimination ratio to expected claims with low deductibles to get the expected claims with the higher deductibles.

Question 31

An actuary is using the development technique based on accident year data to calculate ultimate claim estimates at 12 months maturity. Briefly discuss how it may impact the analysis and propose an appropriate response to mitigate the issue. d. The insurer has implemented a new claims system that allows faster processing of claims.

Answer 31

I assume “faster processing of claims” impacts reported development, not just settlement rates. The faster processing of claims would cause the development method to overestimate ult claims. As such, using the Berquist-Sherman paid method would account for the speedup in settlement and could accurately estimate ult claims.

Question 32

Briefly describe a situation that would lead to the downward development in high reported claim counts from 12 to 24 months which is increasing every year .

Answer 32

The reported triangle could exclude claims closed without payment, so some claims open at 12 months might close without pay by 24 points.

Question 33

Describe a scenario where the paid Cape Cod technique is preferred to the reported Cape Cod technique.

Answer 33

If there are changes in case adequacy, the reported Cape Cod would overestimate the ECR and thus will overestimate ultimate claims. The paid Cape Cod method would be unaffected and would be accurate.

Question 34

List three components of an unpaid claims estimate.

Answer 34

Any 3 of: • Case outstanding on known claims • Provision for future development on known claims • Estimate for reopened claims • Provision for claims incurred but not reported • Provision for claims in transit OR you could have answered with these 3: • Case outstanding on known claims • Incurred but not enough reported (IBNER) • Pure IBNR

Question 35

Briefly describe a scenario where it would be appropriate to use the case outstanding development technique.

Answer 35

Any 1 of: • It is appropriate for claims-made policies since there is no pure IBNR. • When the only data available is case outstanding. • When doing a RY analysis. • When doing an AY analysis and most claims are reported by the first maturity (e.g., first 12 months).

Question 36

Briefly discuss two advantages of the ratio approach as compared to the development approach for Salvage and Subrogation.

Answer 36

* The development factors for the ratio approach tend to be less leveraged than the development factors based on received S&S dollars. * The ratio approach produces ratios of ultimate S&S to ultimate claims, which can be used as a diagnostic.

Question 37

The company pays a large shock loss early in the life of the accident year. Is it appropriate to use Reported BF Paid BF

Answer 37

The paid development technique would multiply the larger losses (due to shock loss) by CDFs that are too high, which would overestimate future development. Reported BF would be appropriate since the expected future development is based on expected losses and CDFs, which are not impacted by the large loss.

Question 38

The company has difficulty compiling an accurate history of rate changes. Is it appropriate to use Reported BF Paid BF

Answer 38

The paid development method would be fine since it doesn't rely on premium data. I think you could answer this part either way, depending on whether you believe the ECR used is from company data or not. Solution 1: Assume the ECR is based on company data. Reported BF would not be appropriate since premium needs to be brought to a common level in order to determine an appropriate expected claim ratio for the method.

Question 39

The company begins to settle claims earlier. Is it appropriate to use Reported BF Paid BF

Answer 39

The paid development technique would multiply the larger losses (due to the speedup) by CDFs that are too high, which would overestimate future development. Reported BF method would be fine since reported claims aren't impacted by changes in settlement rates.

Question 40

Severity trend is higher than expected. Is it appropriate to use Reported BF Paid BF

Answer 40

The paid development method would be fine as it only depends on development patterns and not trends. I assume this question is referring to the trend in a single new AY, which implies that the ECR would be incorrect as a result. Reported BF would underestimate loss since the ECR would be too low, as it was estimated assuming the severity would be lower.

Question 41

Briefly describe the relationship between gross and net tail factors for each of the following reinsurance arrangements: i. Stop Loss ii. Quota Share

Answer 41

i. The net tail factor will typically be smaller than the gross tail factor since the net losses may be capped by the reinsurance protection. ii. The gross and net tail factors will be the same since the net triangle is a constant multiple of the gross triangle.

Question 42

Identify a scenario that distorts the classical technique and briefly explain how it is addressed by the Kittel refinement.

Answer 42

When the book of business is growing, the ULAE tends to grow faster than the paid claims (especially for longer tailed lines), which distorts the paid to paid ratio used by the classical approach. The Kittel approach will address this by incorporating incurred claims (in addition to paid claims), which will increase faster than paid claims for a growing book.

Question 43

Briefly describe the reported claim development technique to determine IBNR.

Answer 43

CDFs are determined for each accident period based on recent emergence patterns of rept claims. Ult claims are equal to the product of the rept claims to date and the cumulative CDFs. The IBNR is the difference between the estimated ult claims and the rept claims to date.

Question 44

A company’s reserving actuary observes that one segment for a particular line of business has much higher severity and a longer-tailed settlement pattern than the remaining segments. Exposures in the high-severity, longer-tailed segment are growing faster than in the other segments. Briefly discuss how the considerations of homogeneity and credibility apply in this situation.

Answer 44

In this case, it may be beneficial to estimate the unpaid claims separately for the high-severity, longer-tailed segment than for the other segments, so long as the data is still credible at that level of granularity. The increased homogeneity should result in more reliable estimates if each segment is credible.

Question 45

An insurance company decides to book a workers compensation unpaid claim estimate that exceeds the indicated value, reasoning that the additional reserve will give the company an extra layer of protection against future adverse development. Discuss four reasons this action could put the insurer at risk.

Answer 45

i. It may be against state law, as state laws may require accurate reserving. ii. It might lead to the company raising rates, resulting in less competitiveness and/or adverse selection. iii. It might result in detracting investors to the company because the business will appear less profitable. iv. It might cause the company to decide to purchase more reinsurance unnecessarily. v. It might cause underwriters to further restrict underwriting guidelines, resulting in less competitiveness. vi. It might cause the company to stop writing workers compensation business in a particular territory. vii. It might lead to regulatory intervention if the company appears to be in danger of failing to meet financial obligations. viii. It might lead to unnecessary employee layoffs in order to reduce company costs to maintain profitability.

Question 46

Discuss two actuarial considerations for designing a reserve study when you merge 2 companies A and B

Answer 46

i. Homogeneity: The business from Company A and B may have different development patterns, so combining them might lead to less accurate unpaid claim estimates. ii. Credibility: Even if former Company A and B business exhibit different development patterns, they may not each be credible enough on their own to be analyzed separately. iii. Data Availability: We may not be able to distinguish former Company A business from former Company B business after the companies combine, so we’ll be forced to reserve them on a combined basis.

Question 47

Briefly describe how a redundant unpaid claim estimate can impact decisions for internal management, investors, and regulators.

Answer 47

* Internal Management: It could lead to decisions such as raising rates, tightening underwriting guidelines, exiting a line of business or territory, or purchasing additional reinsurance. * Investors: It would lower the insurer’s profit, making the company appear worse to investors. * Regulators: It could result in a regulator restricting the insurer’s ability to write business.

Question 48

Briefly describe how an inadequate unpaid claim estimate (lower than expected) can impact decisions for internal management, investors, and regulators.

Answer 48

• Internal Management: It could lead to decisions such as lowering rates, relaxing underwriting guidelines, pursuing aggressive growth, or purchasing less reinsurance. • Investors: It would raise the insurer’s profit, making the company appear better to investors. • Regulators: It could make the company look better than it does, so regulators don’t get involved until it is too late to save the company.

Question 49

One advantage and one disadvantage of Chain Ladder

Answer 49

Advantage: Any 1 of: • Uses the latest information about paid or reported losses for the period for which ultimates are being estimated. • Produces accurate results when the loss ratio is changing (all else being equal). Disadvantage: Any 1 of: • Estimates are highly sensitive to paid or reported losses to date for immature years, especially for longer-tailed lines of business. • Produces inaccurate results when there are changes in settlement rates (and using a paid loss triangle). • Produces inaccurate results when there are changes in case reserve adequacy (and using a reported loss triangle). • Produces inaccurate results when there are changes in the mix of business. • Produces inaccurate results when there are changes to the average accident date between periods.

Question 50

One advantage and one disadvantage of Bornhuetter-Ferguson

Answer 50

• Uses actual data where it is more credible (mature), and an a priori estimate where there is less credibility. • Random fluctuations at early maturities do not significantly distort the projections. • Less distorted than chain-ladder when there are changes in settlement rates (and using a paid loss triangle). • Less distorted than chain-ladder when there are changes in case reserve adequacy (and using a reported loss triangle). • Less distorted than the expected claims method when the loss ratio is changing. • Provides more stable estimates than the chain-ladder method. • Can be used for a new line of business with no historical data. • Relates the IBNR reserve to the exposure. Disadvantage: Any 1 of: • It is difficult to select an appropriate expected loss ratio for use in the method. • It is less responsive to changes in experience than the chain-ladder method. • Produces inaccurate results when there are changes in settlement rates (and using a paid loss triangle). • Produces inaccurate results when there are changes in case reserve adequacy (and using a reported loss triangle).

Question 51

Identify one situation in which it would be preferable to use the Bornhuetter-Ferguson method rather than the case incurred loss development method to estimate the IBNR.

Answer 51

* For earlier maturities for a long-tailed line of business, so that estimates are more stable. * For a new line of business with no historical data.

Question 52

Identify one situation in which it would be preferable to use the case incurred method rather than the Bornhuetter-Ferguson method to develop the IBNR.

Answer 52

• If the loss ratio is changing, the chain-ladder method would give more accurate results. • When the historical development pattern is consistent, and is expected to be consistent in the future. • When a reasonable expected loss ratio for use in the B-F method cannot be determined.

Question 53

What is the key innovation of the Cape Cod method over the Bornhuetter-Ferguson method?

Answer 53

The key innovation is using actual data to calculate the ECR instead of it being an a priori estimate (perhaps selected arbitrarily).

Question 54

What is one disadvantage of the Cape Cod method?

Answer 54

Any 1 of: • Premiums must be brought to a common (usually the current) rate level to calculate the ECR, and it may be difficult to obtain appropriate on-level premium for older years of data. • Produces inaccurate results when there are changes in case reserve adequacy as the rept CDFs (% unrept) will not account for this change and the ECR will be impacted. • Produces inaccurate results when the claim ratio is changing. While the method is somewhat responsive, it uses multiple years of data to calculate the ECR, so the ECR will not fully reflect the newest claim ratios.

Question 55

Which method between Chain Ladder and Cape Cod should receive more weight in recent accident years? Explain.

Answer 55

The CC method should receive more weight in recent AYs as it will produce a more stable estimate by using multiple AYs worth of data in the estimate, while the C-L would be highly sensitive at early maturities.

Question 56

Which method between Chain Ladder or Cape Cod should receive more weight in older accident years? Explain.

Answer 56

The C-L method should receive more weight in older AYs since the data will be more mature and thus the C-L method will produce a better picture of the true reserve. The CC method would be influenced by less relevant data from other AYs, and it would also be more difficult to obtain on-level premium for the older years.

Question 57

Berquist and Sherman describe an approach to adjust the paid loss triangle for distortion. Describe the technique to make the necessary adjustment.

Answer 57

First, you need to obtain projected ult claim counts for each AY by developing rept counts to ult. Then you calculate disposal rates for the latest diagonal as the cumulative closed counts for each AY divided by the ult counts for that AY. Then you use those disposal rates along with interpolation (perhaps linear) to restate the paid claims triangle so that all years of the triangle have the same disposal rate for each maturity.

Question 58

Explain two reasons why using paid loss data to estimate a severity trend can be inappropriate for medical malpractice losses.

Answer 58

i. The slow payment of claims for medical malpractice substantially reduces the data available by accident year that can be used to estimate the paid severity trend at early maturities. ii. Severity trends can be distorted by irregular settlements and variation in the rate of claims closed without payment.

Question 59

Identify three operational changes that could affect the accuracy of loss reserve estimates for a book of business written by an insurance company.

Answer 59

3 of the most common ones i. A speedup in the settlement rate of claims. ii. An increase in case reserve adequacy. iii. Writing more profitable business with lower loss ratios (but the same development patterns) .

Question 60

Briefly discuss one advantage in using the ratio method to determine salvage and subrogation recoverables.

Answer 60

Any 1 of: • The development factors for the ratio approach tend to be less leveraged than the development factors based on received S&S dollars. • The ratio approach produces ratios of ult S&S to ult claims, which can be used as a diagnostic. If a ratio for a particular year seems unreasonable, a more reasonable S&S ratio can be selected for that year.

Question 61

Briefly explain how the selection of tail factors for both net and gross reported claims should be impacted by the presence of an excess of loss reinsurance contract.

Answer 61

The tail factor for the gross claims should be calculated normally. The tail factor for the net claims should likely be less than the tail factor for gross claims since net claims are capped by the excess of loss coverage.

Question 62

Briefly describe one advantage and one disadvantage of the ALAE Loss ratio method

Answer 62

Advantage: Any 1 of: • The development factors for the ratio approach tend to be less leveraged than the development factors based on ALAE dollars. • The ratio approach produces ratios of ult ALAE to ult claims, which can be used as a diagnostic. If a ratio for a particular year seems unreasonable, a more reasonable ALAE ratio can be selected for that year. • It recognizes the inherent relationship between claims and ALAE. Disadvantage: Any 1 of: • For some lines of business, there may be claims with no claim payment but have substantial ALAE. • An error in the estimation of ult claims will lead to an error in the estimation of ult ALAE.

Question 63

Identify two problems with the ULAE reserving methods that use calendar year paid-to-paid loss ratios as a starting point.

Answer 63

Any 2 of: • The paid-to-paid ratios will result in inaccurate estimates if the book of business is growing or shrinking. • They don’t address the scenario when ULAE and claim costs are trending at different rates. • They assume that 1 $10,000 claim results in the same ULAE as 10 $1,000 claims, which may be not a good assumption. • They don’t consider the difference between paying and closing claims, since some claims may not be paid when closed. • They don’t work as well for long-tailed lines of business.

Question 64

Identify two limitations to the self insurer's Case oustanding technique

Answer 64

Any 2 of: • Industry CDFs may not be appropriate for the particular self-insured entity. • The projections can be distorted by case reserves for large losses. • The CDFs for immature years may be highly leveraged, making estimates highly volatile.

Question 65

Briefly describe a situation when the self insurer's Case oustanding technique is particularly useful.

Answer 65

Useful when only current case outstanding data is available so other techniques cannot be used

Question 66

Briefly describe one situation where the credibility-weighted assumption underlying the Bornhuetter-Ferguson method may not apply.

Answer 66

Would not apply if % paid is greater than 100%(Violate credibility definition)

Question 67

Compare and contrast the Cape Cod method and Bornhuetter-Ferguson method by providing one similarity and one difference.

Answer 67

Similarity: Both methods use a credibility weighted estimate of the C-L ult and EC ult. Difference: The EC ult is calculated differently for the 2 methods. The B-F method uses an a priori estimate, while the Cape Cod method includes the experience for the latest year in determining the ECR.

Question 68

Briefly discuss which approach, the development or ratio approach, to select in recommending an ultimate salvage and subrogation estimate in general

Answer 68

Ratio approach generally provdes more stability, less subject to leveraging at early maturities

Question 69

Best Reserving technique for Property catastrophe coverage in a year with higher-than-expected catastrophe losses reported to the insurer but not yet paid.

Answer 69

With property catastrophe coverage, the actual losses will vary significantly from year to year, so a regular development method could produce erratic results. Given the higher than expected rept losses, an a priori expected claims estimate would be too low. As such. the rept B-F method would be appropriate so that the ult losses consist of the higher actual rept losses plus the unrept losses based on an a priori expected claims estimate.

Question 70

Best Reserving technique for Self-insured workers compensation for a large corporation in a state where the statute of limitations for filing a claim has been recently reduced.

Answer 70

Because the statute of limitations has been reduced, I expect that claims will be reported sooner, which will impact both the paid and rept development patterns. As such, the Expected Claims method that doesn’t rely on past patterns would be appropriate. A Berquist-Sherman adjusted paid method might also be acceptable.

Question 71

Best Reserving technique for Basic limits auto liability for an insurer that has recently implemented a new claims processing system to make faster payments to insureds without changing the company’s reserving methodology.

Answer 71

Because of the speedup in settlements, the unadjusted paid development method will produce estimates that are too high. Therefore, using the Berquist-Sherman adjustment to paid claims for changes in settlement rates would allow you to estimate ult claims based on adjusted paid claims.

Question 72

Best Reserving technique for Excess of loss reinsurance with an average attachment point of $100 million on product liability policies.

Answer 72

At this high of an attachment point, the actual data will be so thin as to not be reliable, so a method needs to be used that relies on expected claims or a mix of actual and expected claims. As such, a Bornhuetter-Ferguson method (or an Expected Claims method) would be appropriate.

Question 73

State 2 assumptions of the reported claim development technique

Answer 73

Any 2 of : i. Development of future claims will be similar to development in our prior periods. ii. Claims observed for an immature period tell you something about claims yet to be observed iii. Consistent with reserve adequacy

Question 74

Can you find an alternative for Reported development method for this assumption does not hold i. Development of future claims will be similar to development in our prior periods.

Answer 74

An expected Claims method can be used, since it does not rely on historical development patterns (Depending on how the ECR is calculated)

Question 75

Can you find an alternative for Reported development method for this assumption does not hold ii. Claims observed for an immature period tell you something about claims yet to be observed

Answer 75

BF since it assumes claims yet to be observed should be based on a priori expectations instead of claims observed to date.

Question 76

Can you find an alternative for Reported development method for this assumption does not hold iii. Consistent with reserve adequacy

Answer 76

Reported B-S can be used to adjustd for changes in case reserve adequacy

Question 77

Identify two situations when the ultimate claims estimate from the reported claim development technique will equal the ultimate claims estimate from the reported Bornhuetter-Ferguson technique.

Answer 77

i. In a steady state environment where the expected loss ratio for the B-F is based on the historically stable loss ratio, both methods will produce equal results. ii. When the % unreported is 0 (i.e., at ultimate).

Question 78

Identify two situations when the Bornhuetter-Ferguson technique is preferable to the claim development technique and briefly explain why.

Answer 78

Any 2 of: i. When there is significant volatility in losses at early maturities, the B-F estimate will be more stable than the dev method. ii. The paid B-F method is less distorted by a change in settlement rates than the paid devtechnique since the B-F method relies partially on an a priori estimate, so the paid B-F method would be preferred. iii. The rept B-F method is less distorted by a change in case reserve adequacy than the rept dev technique since the B-F method relies partially on an a priori estimate, so the rept B-F method would be preferred. iv. The B-F method can be used for a new line of business where there is no historical data to determine dev factors for use in the dev technique.

Question 79

The actuary has observed that the adequacy of case outstanding is increasing for an insurer. How's the Reported Claim Development Technique is influenced ?

Answer 79

i. The rept dev technique would overestimate ult claims since it would apply historical development patterns based on lower case outstanding adequacy to a higher rept claims number, resulting in higher estimates of ult claims.

Question 80

The actuary has observed that the adequacy of case outstanding is increasing for an insurer. How's the Expected Claim Technique is influenced ?

Answer 80

The EC technique would not be affected by the change in case outstanding (estimates would remain accurate) to the extent that the ECR calculation is not influenced by the changing case adequacy (if the changing adequacy only started with the latest accident period, then it will not impact the ECR).

Question 81

The actuary has observed that the adequacy of case outstanding is increasing for an insurer. How's the rept B-F technique is influenced ?

Answer 81

The rept B-F technique would overestimate ult claims since the % unrept would not reflect the changing case adequacy. However, it will not overestimate ultimates by as much as the rept dev technique.

Question 82

# Compare techniques The actuary has observed that the adequacy of case outstanding is increasing for an insurer. How's the Reported Cape Cod Technique is influenced ?

Answer 82

The rept CC technique would overestimate ult claims since the ECR calculation would be influenced by the changing case adequacy in the most recent year, causing the ECR to be higher. Also, the % unrept would not reflect the changing case adequacy. As such, CC will overestimate by more than the rept B-F technique.

Question 83

Briefly explain how the Berquist-Sherman case outstanding adjustment can be considered in the reported Bornhuetter-Ferguson technique.

Answer 83

Use the Berquist-Sherman adjusted cumulative rept losses and adjusted CDFs instead of regular rept losses and regular CDFs in the rept B-F method.

Question 84

Briefly describe one similarity and one difference between the Cape Cod and Bornhuetter-Ferguson techniques.

Answer 84

* Both methods calculate IBNR as an ECR times EP times % unrept. * Both methods calculate ult claims as a credibility weighting of the development ultimate and the EC ultimate.

Question 85

Is Frequency and Severity method appropriate for A very long-tailed line of business.

Answer 85

Freq-Sev techniques are appropriate for long-tailed lines since freq can usually be estimated fairly accurately and sev can be estimated using a combination of trended historical data from more mature years and judgment.

Question 86

Is Frequency and Severity method appropriate for A line of business with a significant proportion of reopened claims.

Answer 86

It would not be preferable to use a freq-sev technique for this line since the reopened claims can create overestimation of claim counts and possibly underestimation of sev.

Question 87

Is Frequency and Severity method appropriate for A line of business with a recent increase in high severity claims during the experience period.

Answer 87

A freq-sev technique would likely estimate claim counts accurately, and sev can be estimated using a combination of trended historical data and judgment to reflect the sev increase going forward.

Question 88

Is Frequency and Severity method appropriate for A line of business that has experienced changes in case reserving philosophy during the experience period.

Answer 88

A freq-sev technique based on paid claim data can be used to estimate ult claims since it won’t be impacted by the case reserving philosophy change.

Question 89

This insurance company expects a significant change in their mix of business that will impact the prensent AY and future years. Briefly describe one advantage and one disadvantage of using the paid ALAE-to-paid claim only technique during this change in mix of business.

Answer 89

Advantage: any 1 of: * The relationship between claims and ALAE might be stable even if claims and ALAE separately are not. * The development factors for the ratio approach tend to be less leveraged than the development factors based on ALAE dollars. * The ratio approach produces ratios of ult ALAE to ult claims, which can be used as a diagnostic. If a ratio for a particular year seems unreasonable, a more reasonable ALAE ratio can be selected for that year. Disadvantage: An error in the estimation of ult claims will lead to an error in the estimation of ult ALAE.

Question 90

Identify two potential limitations of Case outstanding method with industry data .

Answer 90

Any 2 of: * In most lines of business, case outstanding do not provide sufficient information about pure IBNR. * There is a lack of industry benchmark data for accident year applications of this method. * The projections can be distorted by case reserves for large losses. * The approach depends on industry CDFs, which may not be appropriate for the particular self-insurer. * The CDFs for the approach may be highly leveraged for immature years, making estimates highly volatile.