Firms Must Improve Data To Pass Rating Firm Checks, Says RMS

Newark, Calif.-based modeling firm Risk Management Solutions (RMS) said today it has developed strategies to help clients prepare for the increased rating firm scrutiny.

The firm's statement follows an announcement that Oldwick, N.J.-based A.M. Best rating firm will be probing deeper into companies' data quality to assess their risk management processes and the credibility of modeled losses.

RMS said that based on its work with portfolios totaling some $25 trillion in insured value for natural catastrophes, it has shown that it is vital to address both the completeness and accuracy of exposure data.

The firm said that by making data quality improvements to portfolios, RMS has seen model losses change by as much as 30 percent, with far more dramatic changes for individual accounts, which has important implications for risk selection, pricing and overall risk management.

The analysis of factors significantly impacting modeled losses is based on the RMS ExposureRefine service, which the firm said provides a consistent way to measure, monitor and benchmark data quality.

Twenty-one earthquake portfolios totaling $612 billion in exposed limit (or over $8 trillion in total insured value) and 18 hurricane portfolios amounting to over $1 trillion in exposed limit (or nearly $17 trillion in insured value) were assessed, said RMS.

The firm said its results reveal that specific strategies can be employed to both improve the quality of data and prepare companies for the additional questions they will have to answer in the A.M. Best 2008 Supplemental Rating Questionnaire from April this year.

The approaches involved include using metrics for assessing data completeness that uncover data quality issues where they matter most. For example, a location that is geocoded at the ZIP Code level in a coastal area exposed to hurricanes is of greater data quality concern than one that is inland. Use of the more basic approaches can create a false sense of security, RMS explained.

By rectifying systematic issues related to missing, inaccurate and bulk coded data that lead to significant inaccuracies in model results a company can often obtain the best cost-benefit for data improvement.

Information used as part of the underwriting process is available but often not included in the catastrophe modeling. Missing data can potentially create a large range in loss results, said RMS.

Where multiple systems are used for capturing data, underwriting and/or modeling, there is greater potential for process errors, which can be easily fixed once identified, the firm explained.

While using coding assumptions or bulk coding can be valid, the results need to be fully audited, said RMS, because errors can lead to significant inaccuracies and bias model outputs.

Ajay Lavakare, RMS senior vice president and managing director of data solutions, said, "Companies are facing pressure from multiple angles, with balance sheets being hit and increased scrutiny over risk management processes from analysts. This means that robust data quality can no longer be viewed as a 'nice to have'; it's a basic necessity to remain competitive."

Mr. Lavakare said, "With the right analytics and databases, organizations can tighten up their processes and ensure their data is both complete and accurate and, crucially, focus on improving their data quality where it matters most."

The A.M. Best 2008 P&C Supplemental Rating Questionnaire, due April 1, is online at http://www.ambest.com/ratings/methodology/srq2008.PDF.