captive and ART resources



Captive Pulse Surveys are sponsored by Towers Watson.
Visit their website today!

Ask the Expert Forum
Ask the Expert FAQ
Captive Basics
Captive Daily Wire
Learning Center
Legislative Updates
Pulse Surveys
Shop Captive.com Store
Templates and Tools
Captive.com Members
Captive Yellow Pages
Group & Rental Captives
Insurance Associations
Program Business

Captive Industry Stats
Career Opportunities
Conference Calendar
Domicile Information
News/Library
Website FAQ
Visit the A.M. Best home page
A.M. Best Ratings
Captive.com, llc
Credit Card Authorization
Membership Info
Register for Site Updates
E-mail captive.com

May 2006 Pulse Survey Analysis:
Catastrophe Modeling

 In partnership with captive.com, Towers Watson conducted this pulse survey to take a look at the issues involved in reinsuring your captive.

This analysis was written by analysts at Towers Watson.

The May Pulse survey asked questions about catastrophe modeling, developed from issues raised in the lead Captive Insurance Company Reports article in May. The survey elicited 63 responses.

Respondents included captive owners/risk managers (34.9%), captive managers (27.0%) and a quite a smattering of others (38.1%) such as from insurance underwriters or service providers (six responses), consultants (seven responses), brokers (two responses), and students and risk analysts (two responses), several risk managers or related others who did not own captives (five), among a few others.

When asked whether they have recently done within the past two years or expect to do catastrophe modeling this year, 66.7% said no, while 33.3% said yes.

Of those respondents that answered yes, 17 actually completed the next part of the survey. Of these, 76.5% were doing it for windstorm and flood (hurricane-related), 76.5% for terrorism, 58.8% for earthquake, 47.1% for flood (non-hurricane) and 47.1% for tornado/hail. One each was doing it for bird flu and fire.

Fully 64.7% were using RMS models; 17.6% were using EQECAT, and 11.8% were using AIR. Others said they did not know what models were used, or the model was under development, or it was a customized/risk calculation.

Of the 17 responses, 64.7% said they would do the modeling in-house; 17.6% said they would use a broker or captive manager, and 11.8% said they would use a modeling firm. One said they would use an actuarial firm and another hadn’t decided yet. <b>Comment:</b> We were surprised by this answer. Does this mean captive owners/risk managers have actually purchased these sophisticated models themselves? This seems quite surprising. Then again, maybe this question was answered by those already in the industry (consultants, underwriters, etc.), which explains why they would buy the models. Unfortunately, our survey doesn’t allow us to drill down on this.

What was the purpose of the modeling? Respondents could answer any or all of the following. Again, of the 17 respondents, the modeling was done mostly to determine the appropriate retention levels (76.5%), followed by 52.9% to determine appropriate insurance limits, 47.1% to gauge the reasonableness of premium to be charged, another 35.3% to determine or satisfy capital demands on the captive, and 35.3% to determine loss-control needs. Just 23.5% said they needed to satisfy underwriters’ or marketing needs. A couple of other responses included a need to gauge potential to self-insure, determine potential stop loss needs, and to track accumulations and exposures.

How comfortable were the respondents in understanding catastrophe modeling? Of those 16 that are doing modeling, just one person said he was admittedly weak; 25% said they understand some basic concepts; 50% are comfortable with understanding the results when presented, and 18.8% were capable of challenging the analytics.

Of those 41 respondents that were not doing any catastrophe modeling, these respondents were weaker in understanding the technical aspects of modeling: 22.0% said they were quite weak; 41.5% understood some basic concepts; 22% were comfortable with understanding the results when presented; just 9.8% could actually challenge the results, and two people (4.9%) said they could practically do the analytics themselves.

captive and ART resources