X hits on this document





5 / 20

Transforming Underwriting

MARCH 2004


Nothing in insurance is more important than risk selection and risk pricing. Accepting risks indiscriminately, or for inadequate premiums, is a ticket for a fast ride into insolvency. Alternatively, being too selective or charging the highest quoted premium will lead to stagnation and decline. There is an old saying among actuaries and underwriters, “any risk is a good risk, if it is priced properly.” Today, in the current environment of “going bare” and alternative risk mechanisms, one should amend it to, “any risk is a good risk, if it is priced properly and you can get the policyholder to pay the quoted rate.”

Beyond the truth and humor of both sayings, there are some critical questions. What data and information are necessary in order to really understand the nature of a given risk? What is the proper price? How much will an applicant or policyholder actually pay? And in operational terms, how does an insurer obtain and analyze the necessary information, set prices, make underwriting decisions, and determine rates with superior levels of speed and accuracy?

Celent believes that four technologies, integrated and working together, provide answers to these questions for personal lines property/casualty insurers (and possibly in other lines as well). These four technologies are: business rules engines, predictive scores/analytics, data management, and optimization.

Giving insurers usable answers to these perennial questions would be a major accomplishment. But Celent also believes that the same technologies have the potential to transform the underwriting function:

  • From selection and pricing of individual risks,

  • To managing risk portfolios containing sets of territories, producers

and/or applicants and policyholders,

  • And even to managing portfolios of risks and premiums, growth

and underwriting profits, market perceptions and channel positioning.

© 2004, Celent Communications. Authorized reproduction permitted.



Document info
Document views31
Page views31
Page last viewedMon Oct 24 14:57:52 UTC 2016