 |
 |
 |
|
Getting the most from the numbers MarketStance provides agents and insurers with a "scorecard" of underwriting and market data By Len Strazewski Despite years of controversy, most state insurance regulators have accepted the integration of credit scoring into the personal lines underwriting process. But as data analysis evolves, so has the application of credit scoring to personal and commercial lines. Credit scoring experts from industry leaders Equifax and D&B Insurance Solutions told agents, brokers and underwriters attending a recent MarketStance seminar in Chicago that the financial analysis techniques of credit scoring do more than help identify individuals who may default on premiums. They say that the same techniques, applied to personal property and commercial lines of insurance help underwriters estimate claims, predict losses and build competitive advantages. MarketStance is a Middletown, Connecticut-based analytical services company that provides integrated data analysis tools to the insurance industry that support agency management, marketing, product development and underwriting. Reza Barazesh, senior vice president of Equifax in Atlanta, explained the value of new business credit scoring systems for estimating insurance losses from commercial as well as personal lines. “Credit scoring isn’t a new device. The statistical model has been used for more than 30 years and generic scoring is designed solely to predict the likelihood of delinquency of payments,” he said. “But is there a relationship between delinquency and insurance claims?” Credit scoring for personal lines has proved the value of the technique, Barazesh said, but continues to pose problems for insurers. The use of credit scoring for personal lines requires the design of a customized data application model that reflects company policy and can be explained to cautious state regulators. “Unless you build models that are very simple and can show a direct connection to your underwriting policy, you will come under difficult scrutiny,” he pointed out. However, applying commercial credit scoring to commercial lines underwriting is less treacherous, he says, and has drawn less concern from regulators. Combined with other traditional underwriting factors such as business classification, longevity in business and geographic market information, credit scoring contributes to the development of a valuable “scorecard” for ranking a risk portfolio, according to Barazesh. Thomas Gersbach, executive director of D&B Insurance Solutions in Short Hills, New Jersey, confirmed that most commercial insurers are already integrating commercial credit scoring into their underwriting process, along with more traditional factors. “Credit scoring is only one piece of the puzzle,” he said, “but one thing is clear. Money or the lack of it, changes behavior.” Gersbach said that reports from D&B clients that used the company’s commercial credit scoring database as part of a review of their risk portfolios noted a strong relationship between claims and low scores, reflecting a “financial stress factor.” That relationship, Gersbach explained, is reflected in both the moral hazard and the morale hazard. Insureds with the lowest commercial credit scores have a greater propensity for “moral hazard,” or damaging acts that lead to claims, such as arson-for-profit or unjustified claims. Policyholders with low credit scores but not at the bottom of the heap have a greater propensity for “morale hazard,” or unwillingness to spend money on appropriate behavior such as routine maintenance or workplace safety, producing a higher level of miscellaneous property claims and workers compensation claims. However, the relationships are not simple or absolute. Gersbach pointed out that large companies tend to have low credit scores as they use their financial clout to delay payments to vendors and stretch their corporate credit, but are not necessarily poorer commercial risks. Also, geography continues to be a risk factor, as economically disadvantaged areas sport a higher level of workers compensation and other claims related to employees fearing loss of employment, he says. Growing use of sophisticated data analysis, however, provides its own risks and opportunities for agents, brokers, insurers and their reinsurers, said Tom Quinn, MarketStance’s vice president of underwriting services. As insurers rely more on increasingly more narrow and sophisticated risk classifications, mis-classification becomes more likely, he explained. “And mis-classification can mean missed opportunities for agents and brokers seeking the best rates and most comprehensive coverage for their clients. In this competitive environ-ment, charging too much because of an improper rating can mean the difference in keeping or losing a client.” The same classification issues apply to insurers and their reinsurers, he noted. Insurers seeking the best treaty reinsurance rates from reinsurers need to present the most accurate risk classifications for their book of business. Quinn provided a demonstration of the classification confusion that results from incorrect links between GL risk management industry classifications, NAICS/SIC codes and geographic market data. These linking errors yield premium rate variations of 300% to 400% and coverage restrictions on usually included product liability. “If there is a disconnect between a NAICS/SIC classification, a GL code or the geographic distribution of the type of risk, that’s an automatic red flag,” he explained. “For an agent or broker trying to get the most precise rate, it will pay to get and analyze all relevant classifications.” Quinn noted that MarketStance has directly mapped NAICS classification codes to the industry GL classifications, adding to its established NAICS to SIC mapping and business class to workers compensation class mapping systems. Agents and brokers don’t have to find huge discrepancies to get a better deal, he added. “Even an incremental improvement in classification accuracy can improve a producer’s or an insurer’s return on investment and provide broader coverage or administrative benefits.” MarketStance executives also highlighted new data products and services for the insurance industry. Among the new products is an enhancement of MarketStance Version 5.1 to include premium estimates for Business Owners Policies (BOPs) according to geographic territory by county or ZIP codes, business class and account size. According to MarketStance estimates, the BOP market totals about $17 billon in premium or about 57% of the commercial multi-peril insurance market. The estimates include mid-year revised growth rate forecasts that reflect the impact of the 2005 hurricane season. The estimates include eight general account categories: office, habitational, retail, wholesale, restaurants, service, specialty contractors and motels. MarketStance Vice President of Business Development Jim Hearn said the new BOP features were the result both of analysis of industry commercial multi-peril premiums divided into appropriate classes, and also advice from insurers about BOP product design. “Because they are often uniquely defined, carrier input was received to determine how to report the information in the most useful manner,” he said. Hearn added that the data was designed to provide better insight into the gray area between small commercial and middle market coverage, some of which is written with BOP polices. MarketStance has also released its first homeowners data analysis which integrates economic, industry and household demographic data. The product also provides for analysis by geography to the county level, market value of residences, head of household age, gender and marital status, annual household income and age of residence. The “Homeowners Cube” allows agents, brokers and insurers to measure premium for homeowners policies, umbrella liability, and flood insurance and includes two- and four-year growth forecasts for premium volume and number of households. * For more information: |
|
||||||||||||||||
| |||||||||||||||||
