Printer Friendly

An actuarial viewpoint.

RISK MANAGERS EXPEND a great deal of effort in attempts to quantify the effectiveness of loss control programs and to compare loss experiences for various companies. Computed statistics such as average claim costs, pure loss costs and claim frequencies are often the basis for these analyses. These statistics are selected because they are readily computed and the information can be used effectively in colorful graphic charts for presentations to senior management.

However, when using compiled statistics, risk managers can encounter many pitfalls, resulting in less meaningful program comparisons. Although the items identified here are not all-inclusive, the ideas presented provide insight as to what is needed to effectively utilize published company statistics. It is important to note that the issues addressed are not unique to a specific line of insurance, but are appropriate for most coverages.

One of the statistics risk managers frequently employ is the average reported claim cost. This is computed by dividing a company's total incurred losses (paid losses plus outstanding reserves) as of a point in time by the number of claims reported as of that date. In using the computed averages, an understanding of the information that is used to calculate the statistics is important.

Among the factors that should be considered in average claim cost comparisons is allocated loss adjustment expense (ALAE), which is defined as an expense that can be clearly assigned to an individual claim. ALAE includes legal expenses and expert witness fees, but typically excludes claim adjusters fees. Depending on the line of coverage, the location and a company's philosophy with respect to claims litigation, ALAE can significantly impact average claim costs. As a result, if ALAE is included in the data for one company while excluded from the information for another company, the comparison of average losses can be misleading. Therefore, it is best to be certain that ALAE is consistently included in or excluded from the loss information for each company analyzed. In addition, some companies do not reserve for ALAE, while other companies establish outstanding liabilities for the estimated value of unpaid ALAE. It is important to be aware of these procedures for each company prior to formulating conclusions about relative average claim costs.


Another factor to be considered in average claim cost comparisons is the evaluation date. The maturity of loss data is measured by the number of months that have elapsed from the beginning of an accident year to the evaluation date of the loss information. For instance, reported losses as of December 31, 1991, for accident year 1/1/91-92 are valued as of 12 months of development. This age of maturity is assigned because December 31, 1991, is 12 months after the January 1 inception date of the accident period. Likewise, as of December 31, 1991, losses for accident year 1/1/90-91 are valued as of 24 months of development. Therefore, if data valued as of December 31,1991, for both accident years are used, losses for the 1/1/90-91 accident year would be considered 12 months more mature than losses for the 1/1/91-92 accident year.

During the 12-to-24-month period, losses for an accident year are likely to increase as a result of changes in the values of reported claims, as well as the addition of losses for claims that become known. Therefore, a comparison of average losses for two accident years that are valued at different ages of development can often be misleading. Comparisons of statistics between accident years are considered to be more effective when losses are valued as of the same age of maturity. Since reserving philosophies and payment schedules can differ from one company to another due to claims administration practices, comparisons of ultimate average costs are often the most meaningful.

Ultimate average losses for two accident years are frequently compared in an attempt to quantify the effectiveness of a cost control program. For this type of analysis, the average loss before the loss control program was implemented is compared to the average cost after the program was implemented. Lower average losses subsequent to the implementation of the program can be used to support the effectiveness of the cost control program.

However, as part of this analysis, it is important to consider the impact that an inflationary trend can have on the average losses. For instance, assuming there are no changes in experience or exposure, it is expected that ultimate average losses for accident year 1/1/91-92 will be higher than ultimate average losses for accident year 1/1/90-91 because the inflationary trend from 1990 to 1991 is expected to increase claim costs.

Therefore, it is necessary to adjust the ultimate losses for both accident years to the same cost level prior to an average loss comparison. Having adjusted the ultimate losses to the same cost level, a risk manager can effectively compare the trended averages.


While the number of no-cost claims can be small for some companies, the number of claims without cost may be significant for other companies. Since the average claim cost is computed by dividing the amount of loss by the number of reported claims, the definition of reported claim used in average loss calculations can significantly affect the computed statistic. If the number of reported claims is defined to include claims without loss cost, a lower average claim cost will be computed than if the number of reported claims used for the computation is defined to exclude no-cost claims. For instance, assuming $1,000,000 of 1oss, 300 claims without loss cost and 500 claims with cost, average loss costs can be computed to equal $1,250 if no-cost claims are included ($1,000,000/800=$1,250) and $2,000 if no-cost claims are excluded ($1,000,000/500=$2,000).

Therefore, including no-cost claims in the computation of average losses will result in lower values than if the averages are computed using only claims that have associated incurred loss values. As a result, a company that includes no-cost claims in reported claim counts may appear to have lower average losses compared to another company that excludes no-cost claims from the average loss computations.

Differences in claim identification, the implementation of incident reporting and the procedures for fast-track reserves can impact the number of reported claims, the number of no-cost claims and the resulting average claim costs. Since these procedures can vary significantly by company and by claims administrator, it is best to exclude no-cost claims in comparisons of average claim costs between two companies.


Reported claims can be administered internally by a company or externally by an insurer or an independent third-party administrator. Because each claims administrator has distinct handling practices, differences in reserving and the payment of claims will impact the average claim size, the number of claims, the ultimate settlement cost and the length of time needed to settle reported claims. Thus, a higher average reported incurred loss for one company does not necessarily mean the company's loss experience is worse. The higher average may instead be due to the company's conservative reserving practices or to a higher level of reserve adequacy.

In addition, if a company changes claims administration practices, comparisons of average incurred losses can be distorted. Any such changes should therefore be recognized and addressed when explaining increases or decreases in average claim costs. Since certain actuarial methods can adjust for these changes when estimating ultimate losses, the use of estimated ultimate losses in average claim cost comparisons becomes even more important when daim reserving practices or claim settlement practices change.


It is reasonable to assume that similar exposures are more likely to produce similar losses. For instance, average size of loss for a manufacturing company is likely to be more comparable to the average loss size for other manufacturing companies. Therefore, it is advantageous to prepare average loss comparisons for similar industries.

However, it is important to note that average costs within a related industry can vary significantly. For instance, consider workers' compensation experience for health care facilities. While acute care hospitals, nursing homes and psychiatric hospitals are all health care entities, each of these groups has a unique exposure that can be significantly different from the exposure of the other two groups. As a result, there can be meaningful differences between the average workers' compensation losses incurred in each of these types of facilities. Therefore, it is advantageous to compare data for similarly related facilities within a given industry when preparing average loss comparisons.

Likewise, cost differentials between rural and urban experiences can be significant, causing large differences in statistics for these areas. Therefore, it is important to be aware of the locations of the companies analyzed. In addition to the rural and urban differences, the effect of legislation may play a role. Comparisons should be avoided between companies that have operations in states with significantly different legal climates.


Statutory coverages, such as workers' compensation, are subject to individual state laws regarding scheduled benefits, claim minimums and claim maximums. State regulations have a direct impact on the average indemnity cost of claims and should be considered. Because these factors can cause differences in the settlement costs for one company compared to another, the effect of location, legislation and statutory regulations on average loss amounts should be evaluated in average claim cost comparisons.

A loss limitation can further impact average claim size. For instance, a company that has compiled data based on losses limited to $100,000 per occurrence may appear to have lower average losses compared to companies for which loss amounts are unlimited. Therefore, it is important to be aware of any loss limitation that may impact the compiled loss statistics so that an effective comparison can be made.

An analysis of statistics regarding the number of losses excess of a specified dollar amount can be helpful. This type of review will provide insights about a company's large claim incidence and the impact that large claims have on average claim costs.

It is also necessary to quantify the effect that ALAE has on the limitation. For instance, some limitations apply only to the indemnity portion of each loss, and unlimited ALAE is included in the claim amount. Other limitations apply to the total claim value including ALAE. Differences in the consideration of ALAE can significantly impact the loss limitation.

For liability coverages, risk managers should be aware of whether data are compiled on a claims-made or occurrence basis. If data are compiled on a claims-made basis, it is necessary to know the retro-date of the claims-made programs for the companies being compared. While most analyses of claims-made data are prepared using mature claims-made information, analysis of immature claims-made experience (i.e., first-year, second-year, etc.) can be effective. However, only data for claims-made years that are valued at the same level of program maturity (i.e., the number of years after the retro-date of the claims-made program) should be used for comparisons.


A pure loss rate is a ratio of the expected dollars of loss cost per exposure unit. The pure loss rate is typically computed by dividing trended ultimate losses by a measure of exposure to loss, commonly called an exposure base. The pure loss rate is not affected by the fixed or variable expenses typically incurred by a company and can be used effectively to compare the loss experience for two companies. Loss limitation and the selected exposure base need to be carefully considered in pure loss rate comparisons.

Similar to the average loss comparison, a loss limitation can significantly affect the ratios computed in a pure loss rate analysis. Therefore, risk managers need to be cognizant of any per-occurrence loss limitation, including a review of the consideration of ALAE, that is used for the pure loss rate computation. An examination of the incidence of large losses may also be advantageous in comparing pure loss rates so that the effect of fortuitous large claims can be determined.

For many coverages, more than one type of unit can be considered an appropriate measure of exposure to loss. For instance, amount of revenue, the number of admissions and the number of transactions could be considered appropriate exposure bases for certain general liability experience. Since the pure loss rates computed using each of these exposure bases could differ significantly, care should be taken that the exposure base used to compute the pure loss rate is consistent for each company analyzed. If consistent exposure data are not available for each company, adjustments should be made to normalize the exposure bases to reflect a consistent foundation for comparison.

Risk managers must also consider the cost level to which losses and inflation-sensitive exposures are trended prior to a comparison of pure loss rates. For instance, the pure loss rate for accident year 1/1/90-91 that is trended to 7/1/92 is expected to be higher than the pure loss rate for this year that is trended to 7/1/91. If pure loss rates are provided for several companies, necessary adjustments to ensure valid comparisons should be made so that all pure loss rates are trended to the same point in time.


Claim frequency is defined as the ratio of the number of claims to a measure of exposure to loss (i.e., the exposure base). Factors that are important to consider in comparisons of claim frequency are similar to the items noted previously in the average claim cost section. As stated previously, some companies include the number of no-cost claims, while other companies exclude no-cost claims from reported claim counts. Therefore, in comparisons of claim frequency, it is important to be certain that the claim counts included for each company are summarized on a consistent basis. It is best to exclude the number of no-cost claims in frequency comparisons since exclusion of these claims will mitigate the impact that differences in reporting procedures and claims administration practices can have on claim frequency.

The comments regarding the need for consistency in the exposure base for each company that were discussed in the pure loss rate section also apply to the comparison of claim frequencies. For a frequency comparison, it is most important that the exposure base used for all companies being compared is the same and, if inflation-sensitive, all exposure bases should be trended to the same cost level.

Prior to formulating conclusions based on comparisons of statistics, it is important to be certain that underlying loss, claim and exposure data are compiled on consistent bases. Loss limitations, trend, maturity of loss data, the inclusion or exclusion of no-cost claim counts and the consideration of allocated loss adjustment expenses are some of the factors that can significantly impact the comparisons of compiled statistics.

If the data underlying the statistics being compared are not consistent, the results of the comparisons may prove to be misleading to management. However, carefully constructed comparisons can be most beneficial to companies in an attempt to measure the effectiveness of cost control programs and as an incentive to continue those already implemented.

Cecilia M. LePere is vice president and consulting actual, for Willis Corroon's advanced risk management services in Nashville, TN.
COPYRIGHT 1993 Risk Management Society Publishing, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:loss comparisons
Author:LePere, Cecilia M.
Publication:Risk Management
Date:Mar 1, 1993
Previous Article:Don't let the ADA disable you.
Next Article:Special events pose Olympic-size risks.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters