Loss Reserve Survey Reveals Diverse Solutions to New LAE Definitions

Effective January 1, 1998, the NAIC adopted a change in how loss adjustment expense (LAE) is split into categories within Schedule P of the property and casualty statutory Annual Statement. Through 1997, LAE was split into Allocated Loss Adjustment Expense (ALAE) and Unallocated Loss Adjustment Expense (ULAE). Starting in 1998, new categories were introduced that are now called Defense & Cost Containment (DCC) and Adjusting & Other (A&O). Generally, DCC expense includes all defense and litigation-related expenses, whether internal or external to a company, while A&O expense includes all claims adjusting expenses, whether internal or external to a company.

Earlier this year, the CAS Committee on Reserves surveyed loss reserve practitioners on how the change in the definition of loss adjustment expense affected loss reserving since 1998 and how it may impact future years and other aspects of actuarial work. This survey will be discussed at the upcoming CAS Annual Meeting at a concurrent session entitled, "NAIC Redefinitions of Loss Adjustment Expense." The survey was a 17-item questionnaire developed by the Committee on Reserves and approved by the CAS Executive Council. Out of 3,239 questionnaires mailed, 74 were completed. The survey also contained several open-ended, write-in questions. The following are the key findings of the survey:

• Of the respondents, 63.5% reported that they were company reserving actuaries, while 25.7% reported that they were consulting reserving actuaries.

• In describing how their company classified ALAE vs. ULAE prior to the change on January 1, 1998, 58.1% reported using claim-specific/nonclaim-specific as their criteria as opposed to external vs. internal classifications (13.5%) or a combination (23%).

• For 56.8% reporting, the major change for their company was to reclassify External Claim Adjusters from ALAE to A&O.

• 82.4% reporting said that they implemented changes with the 1998 Annual Statement.

• 55.4% said that their company selected the Calendar Year (all accident years for calendar year 1998 and beyond) method to implement the new LAE split.

• 54.1% reported using an Expense Tracking System, while nearly one-quarter (23%) used Formula Allocations.

• 55.4% said that their company is currently maintaining internal expense reporting under the former categorization while adopting the new categorization for statutory reporting.

• Over half (58.1%) said that their company was not using the new expense categorization for any purposes other than Annual Statement reporting.

• As to areas they believed needed further research regarding the impact of the new LAE categories, the most popular responses were Reinsurance Contracts (18.9%) and Ratemaking Practices (13.5%).

For the write-in questions, respondents were asked to explain how their companies accomplished a reclassification of expenses from categories where claim detail was not maintained (for example, internal defense attorney costs, formerly categorized as ULAE) to categories such as DCC, where at least some detail (for example, accident year) would be required.

A sample of the answers follows. It illustrates the wide range of solutions among the respondents.

• Claims staff estimate their time between the categories DCC and A&O—not revisited in 1999. Use loss payments to allocate between accident years.

• Estimate Total Paid ULE/DCC as a percent of Total Paid ULE using Salaries plus Overhead. Allocate Paid ULE/DCC to Line of Business using judgment percentages. Allocate Paid ULE/DCC to Accident Year using Calendar Year Closed Claim Costs + Open Counts.

• Obtain as much detail as possible and use interviews of claims personnel and gut feeling to make projections.

• Make a wild guess. Make data look like what it should look like or what they want it to look like.

• Characterize insignificant internal expenses as DCC and call all internal expenses A&O. Continue to use the old ULAE accident year allocation rule of 45/5.

• For treaty reinsurance, use an arbitrary formula reallocation, varying by subject treaty. (The treaties follow the old definition!) Survey MGAs. If they respond, use what they gave. For those who didn't respond, prorate following the pattern of those who did.

• If the detail is unavailable, just use allocation procedures to put the new DCC dollars somewhere. Nothing in the regulation says what type of detail one has to maintain on the new expenses.

• Calculate ULAE reserves under the old definition by coverage and accident year, then estimate the percentage attributable to internal defense attorney costs, based on input/claims data from our law department. These percentages are mainly based on actuarial judgment.

The next write-in question gave the following background and then asked what the reserving challenges of dealing with this are and what solutions they had found. It also asked how they had changed their reserving practices.

On a calendar year basis, the new categorizations apply to the incremental calendar year change across all accident years beginning January 1, 1998. From a Schedule P standpoint, this means for accident years 1997 and prior, the 12/31/98 evaluation of ALAE (that is, the current column) and all future evaluations (or columns) will reflect a mixed definition. Accident year 1998 and future accident years will be under the new DCC definition. On an accident year basis, the new categorizations will apply to only accident year 1998 and future accident years. Prior accident years will continue to runoff under the old definition of ALAE.

The following are a few of the many detailed responses we received on this topic. The feedback illustrates the diversity of reserve challenges that companies are facing.

• CY Basis: Compare total LAE projections—new vs. old definition—to benchmarks that have not changed (premium, loss reserves, paid LAE, etc.). We've found that percent of total LAE which is A&O is greater than that which was ULAE. We apply "paid-to-paid" method to determine A&O/Loss ratio as the basis of projecting A&O reserves. We have tried to establish "paid-to-paid" DCC/Loss factors as well.

• The company has maintained internal expense reports that utilize the old ALAE/ULAE segregation. They will continue to do this until sufficient experience has been gathered using the new categories.

• Significant judgments are required in the selection of projection factors. We calculate total LAE reserve needs using historical triangle of ULE & ALE (DCC + A&O) and make sure that the judgment calls we're making for DCC and A&O individually, yield an overall LAE reserve similar to what we develop in total.

• Internally we have not changed our reserving practices and still review ALAE and ULAE reserves separately. There is just an extra step required to split ALAE into DCC and A&O for statutory reporting.

• Reserving practices not changed. Still analyze ALAE separately from ULAE. The differences between ALAE & DCC and ULAE and A&O are dealt with in the data reconciliation of the actuarial report.

• For older years we did not have the detail to construct expense triangles under the new definition; therefore we had to combine all expenses to form a LAE triangle to determine expense development.

• Easy for those who don't change. We will let you know when they move to new definitions. On CY basis, restating screws up triangles.

• Won't runoff calendar year 98 and subsequent be under the new DCC definition? Our reserving practices have not changed at all. We are still developing our ALAE/external and ULAE/internal reserves the same way we have in the past. What has changed is that we now have to allocate our developed ALAE/external reserve to the new DCC and A&O categories. We are allocating to these categories based on paid DCC and A&O expenses collected for calendar year 98 and subsequent. We consider all ULAE/internal reserves to be A&O. We are also collecting and building historical triangles of external DCC and A&O paid expenses for AY 1998 and subsequent. As soon as sufficient history is available we will use triangular analysis to develop our external DCC and A&O reserves.

• The biggest issue is separating the A&O component of ALAE reserves. We can separate the historical payments of ALAE by component. We looked at historical A&O payments as percent of total ALAE payments by accident year at different evaluation points (12, 24, 36 uses, etc.) From that, we could derive the percent of total bulk ALAE reserves for A&O by accident years at different evaluation points.

• We analyze our ALE and ULE reserves separately using an accident year change in Paid ALE to Incurred L/R estimate for ALE and a calendar year Paid-to-Paid and Paid-to- Paid plus O/S as an estimate of the relationship of the ULE O/S to Loss O/S for ULE. In all but one Reserve Analysis there appeared to be no distortion in the rate of ALE to ULE. In the one we used the latest year diagonal, which effectively eliminated the distortion. We will probably change to a Paid-to-Paid method and analyze the reserve in total.

• Internally, we have maintained the old definitions, so we have not changed our basic reserving practices. Our challenge is in estimating how much "old definition" ULAE to move to DCC for statutory purposes. Our system can capture ALAE (old definition) moving to A&O, so that hasn't been as difficult.

• We continue our reserving practices based on old definitions and allocations.

As to the impact the categorization change will have on industry Schedule P data as individual companies make different choices on how they will handle the change, a sample of the responses follows:

• Industry Schedule P data for all companies combined will be distorted by the change and by different ways of handling the change. Could impact companies that use industry data for benchmarking.

• A big mess.

• Will only be able to analyze total LAE expenses and reserves for accident years prior to 1998.

• Minimal impact.

• Industry Schedule P data will be a mishmash of various company definitions.

• Industry conglomerate data could be rendered useless for 10 years, especially for small companies who employ outside adjusters.

• Schedule P data will be distorted. The reliability of any triangular analysis based on Schedule P data during the 10-year phase-in period must be questioned.

• This change in practice makes absolutely no sense whatsoever. I have been told that the reason for the change is that ALE is not comparable between companies (some companies utilize outside adjusters more than others). Well, companies are different and though old rule measured that difference, the proper place to break out legal and adjusting was in Part 4 (A/S) expense class.

• There will be less consistency in industry data going forward. It was my understanding that the primary reason that this change was adopted was to allow improved direct comparisons between companies. I do not support that reason as being more important than ratemaking, pricing, reserving, underwriting, and reinsurance reasons for continuing allocating as many claim-specific dollars as possible to individual claims. Regulations should benefit policyholders and not simply add to the expense dollars policyholders should pay.

The respondents were also asked how users of Schedule P adjust for possible distortions in the data. A sample of the answers follows:

• The same way you raise teenagers—any way you can.

• Not use Schedule P. Ask for actuarial report.

• Ignore 1998/1999 Schedule Ps for the development of ULAE/ALAE ratios. We've already seen this with a 1998 financial exam—the auditors ignored the 1998 Schedule P and used 1997 Ps to develop ULAE ratios. We were told that this was how they were handling the problem.

• This is a good question. Historically, Parts 2 and 3 have been prepared on a loss plus allocated basis given the fact that, by their very nature, unallocated loss adjustment expenses could only be assigned by accident year on a judgment basis. I am not clear what is expected to show up in Parts 2 and 3 now. To the extent any ULAE-type losses get into Part 2, loss development measures will be distorted. To the extent they get into Part 3, paid loss development patterns will be distorted. Who knows how users will adjust for these distortions, or if they even can?

Twenty-four respondents wrote additional comments. The following comments reflect the consensus of that group.

• In my opinion, this change was poorly thought out and serves no purpose at all. I am unaware of any beneficial purpose that will ultimately be served.

• I think the whole change was pointless.

• This resulted in a greatly increased amount of work on my part and on company personnel parts for no discernible benefit. It also caused great confusion as there is no clear definitions of what is A&O and what is DCC.

• It causes headaches for me regarding statutory reporting but since my company management has decided against making internal reporting changes the burden of the change pretty much falls on me.

• It is possible that companies that have to reclassify internal litigation expenses will find their reserves less adequate and companies reclassifying adjuster expenses will find their reserves more adequate.