There can be tremendous value gained to the underwriting processwhen underwriters "know" what is going on in the book of claims.Yet claim operations and technologies have traditionally existed intheir own separate silos, away from the processes technologies thatsupport the underwriting process.

Despite the traditional separation, claim data that developed asa "normal part of good claim handling" can be combined with commonunderwriting data to provide powerful rating information andinsight. When that insight is paired with the appropriate rulesengine and workflow technologies, it can drive actions to improveunderwriting results.

Premium Rating Errors Can Be Very Expensive

For many auto insurers, annual miles driven, commute use, andcommute mileage are key rating factors. According to a 2010 reportby Quality Planning Corporation, these common rating elements arealso the leaders in rating errors that cost insurers an estimated$3.1 billion in lost premiums each year. The report also points todiscrepancies in vehicle usage, vehicle characteristics, ratedterritory, rated operators, and vehicle/driver assignments that addanother $7.7 billion to the estimated $15.9 billion in annualrating error premium loss.

Certainly a series of regular and consistent audits, reviews,and inspections can help reduce these costs as well as ensure thatadequate rates are charged for the risk represented. Each of theseprocesses, however, comes with a cost to the insurer and potentialinconvenience to the insured. The inconvenience could createresistance that may drive a portion of an insurer's book to thearms of competitors with less stringent requirements.

A claim investigation can provide pieces of data thatunderwriters may need in order to validate the type and quality ofthe risk in comparison to what information is contained in theunderwriting file. Smart carriers know this, and have processes inplace by which the claim handler can alert the underwriter to itemsthat may be of rated risk concern.

Most carriers still follow a traditional paper-based method ofcommunicating this type of information. This process is slow,cumbersome, and takes the focus of the adjuster away from handlingthe claim to a non-system supported manual exception process. Theconsistency, timeliness, quality, and value of this exchange can befurther diminished by adjusters if they are:

• Not being fully cognizant of what type of information isimportant to the underwriting community (which is a potentialburden to the underwriting community as well...);
• Not recognizing the importance of reporting potentialunderwriting inconsistencies to an insurer's ability to price riskappropriately and ensure competitive pricing across their policybook or;
• Working under the assumption that underwriting will identifythe issue and resolve them through their standard risk reviewprocesses.

Overlaying these issues are the priorities of the claimadjuster, who place the provision of top quality claim handlingservice over filling out paper forms for the underwritingorganization.

Systems Can Talk to Each Other

A simple auto collision or comprehensive loss necessitates thecollection of the mileage of the insured vehicle. Including the"annual miles driven" data element in the coverage download sent tothe claim system at the initiation of the claim allows the claimsystem to quickly compare the rated mileage to the actual mileage.The system can then provide an automated alert back to theunderwriter as appropriate.

For instance, if the insured vehicle is a 2008 model rated for"less than 10,000 miles per year" and a loss in December 2010 showsvehicle mileage at 40,000 miles, something is amiss and theappropriate premium may not be being collected for the underwrittenrisk. If the opposite is true, the insurer has an opportunity torecalculate the lower annual mileage rate and the potential tooffer the insured a reduction in premium.

Let's explore some other examples based upon the leading reasonsfor auto rating errors. In these cases, automated flags might becreated through a straight-forward comparison of claim andunderwriting data, both within and across lines of business.

Rating Error Claims/Underwriting Flag

Commute
Selected repair location or medical provider more than set radiusfrom garage location or reported work address

Vehicle Usage
Number and ages of passengers in personal use rated vehicle,weekend use of commercial use rated vehicle

Vehicle Characteristics
Repair estimateincludes large number of non-OEM or non-industry standard partssuppliers such as performance part vendors. "Temporary substitutevehicle" listed on loss involving a non-owned high performancevehicle

Unrated Operators
Permissive use indicated as "relation to insured" on vehicle lossnotice

Vehicle/Driver Assignment
Claim incident location in relation to garage location, comparisonof demographics of assigned vehicle driver with items stolen fromvehicle (in the case of an auto break-in with both glass loss andcompanion homeowner's claim).

What About Other Lines of Business? CLICKNEXT

Workers' Compensation

Insurance Services Office (ISO) reports that insurance industrylosses due to workers' compensation fraud are as high as $5 billioneach year. They also estimate that, for every $1 lost throughclaimant fraud, at least $4 to $5 is lost through premiumfraud.

Though it is certain that not all rating errors are due todeliberate malfeasance on the part of the insured, similar types ofcomparisons between data elements contained in the claim file tothose in the underwriting file can highlight areas for thepossibility of further rate adequacy investigation.

If the commercial insured has an automobile accident, whatnumber is the damaged vehicle listed on the policy? Is this numberabove the number of fleet vehicles indicated on the workers'compensation policy? Does the accident or repair location fitwithin the radius of operation for a commercial automobile policyfrom the fleet garaging location?

Comparing the North American Industry Classification System(NAICS) code contained in the underwriting file to the injury typereported in the claim file can also be a key indicator of potentialrating discrepancies. For instance, if the insured's NAICS code islisted as 453110 (florists; fresh flower shops) and the workers'compensation claim is for a burn (ICD-9 range 940-949) this islikely important information for the underwriter to be awareof.

If the insured is pursuing a business interruption claim undertheir commercial property policy, does the payroll amount listed inthe property claim file match the payroll amount contained in theworkers' compensation underwriting file?

General Liability

If a workers' compensation claim codes the injured worker'soccupation as a bartender or cocktail server and the insuredlocation is listed as a liquor store in the general liabilityunderwriting file, an inappropriate rate may be being charged.Likewise, a covered contents claim under the commercial propertypolicy at this same insured premises indicating contents damage toa pool table, juke box, or pinball machines could be a potentialflag for the general liability underwriter.

Claim information that indicates insured contracts containingembedded subrogation waivers or liability limitation limits mayalso be of interest to the underwriter. The insurer may wish to addthese discrete data elements as "yes/no" choices in their claimhandling database.

Homeowners'

When repair estimates are prepared for structure and contentsclaims, do their details as captured in the claim and/or estimatingsystems match the "what should be" information in the underwritingfile? Do structural repair materials make sense for theconstruction type? Does the contents claim match to the reporteduse? For instance, high volumes of computer equipment or a largeamount children's furniture and toys reported in the claim file maypoint to an un-rated home business exposure.

Similarly, contents claims for damaged pet supplies or equipmentcan point to the presence of an unrated animal on thepremises.Liability claims for dog bites are an ever increasingconcern. An August, 2010 report by the Insurance InformationInstitute indicates that dog bites were the basis for more thanone-third of all homeowners insurance liability claims paid in 2009and accounted for $412 million in claim payments. This same reportstated that these payments have grown 30 percent in the past sixyears.

The Importance of Information

Information developed during the claim handling process can addsubstantial benefits to the underwriting rating and pricingdecision process. The straight-forward collection and comparison ofdiscrete data elements commonly found in both claim andunderwriting systems, combined with system enabled business rulesand execution logic can provide significant value to ensuring thatappropriate knowledge of the risk is made available to theunderwriter. This allows the provider with the ability to makeadequate rating and pricing decisions that are consistentlyexecuted across the policy book.

There are a few simple to moderately complex scenarios by whichthis process can bring useful information to the underwriter's deskwithout disrupting the prompt, efficient, cost effective, andservice sensitive claim handling process that every insurer isdedicated to providing.

Mike Mahoney is the Senior Director of Product Marketing,Auto Casualty Solutions at Mitchell International, Inc. He may bereached at [email protected].