AIR Worldwide: Significant Quality Issues with Insurers’ Exposure Data

November 14, 2005

  • November 15, 2005 at 7:29 am
    Rate Maker says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    1st off AIR does not model flood losses so modeling a cat 2 vs. a cat 3 for purposes of breaks in the dikes is a waste of time.
    2nd, the standarad ISO homeowners policy DOES NOT cover flood but there are many, including agents, who want the flood damage covered from the HO policy which specifically excludes flood b/c they do not have a waiver from the insured which now presents them an e&o situation. But if the companies pay for the damage the lazy agents are off the hook.
    3rd, the politicians believe the homeowners policy should pay for these poor people who did not buy flood protection (sorry, a contract is a contract).
    4th and final, the models are only estimates and any company who relies soley on aggregations based off of a model is in big trouble. Strong underwriting for accurate replacement values upfront only increases the accuracy of any model. Junk in = junk out!
    Even MSB has published reports that the majority of homes across America are underinsured by as much as 25%!

  • November 15, 2005 at 1:43 am
    Long Time Underwriter says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Apparently catastrophe modelers aren\’t allowed to interact with the real world. Sampling of data for accuracy, and adjustments for data deficiencies, are a vital and incredibly obvious part of any modeling exercise. From insured, to agent, to underwriter, to data entry to Cat model, property values and risk characteristics are interpreted and revised for many purposes. If cat modelers can\’t handle the real world they shouldn\’t be offering their \”expert\” opinion to the public or the industry.

  • November 15, 2005 at 1:53 am
    André Fredette says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    I believe Karen\’s response while probability accurate about underinsurance values misses the point. It was well known through engineering reports done by Engineers corps that any Hurricane above 3 would breach the dikes. Therefore, armed with this engineering knowledge AIR should have run two scenarios. One with a hurricane at 2 or below and one doomsday version with 3 or above which would assume the catastrophic failures of the dikes. Nature does not follow the Pareto curve. any underwriter worth his salt should know
    that cat models are just the starting
    point not the ending point of accumulation
    evaluation.

  • November 16, 2005 at 10:44 am
    Rick says:
    Like or Dislike:
    Thumb up 0
    Thumb down 0

    Let\’s see……ISO collects statistical data from a significant portion of the industry. Would they like more data and at a greater level of detail (to drive various data products they sell)?? I think so.

    ISO also sells property valuation services….that would correct that undervalue problem that was pointed out. Would they like to sell more? I think so.

    Also, is AIR Worldwide an operating unit of ISO? Oh, yes, I know so.



Add a Comment

Your email address will not be published. Required fields are marked *

*