Florida Commission on Hurricane Loss Projection Methodology Workshop Report: January 29 and 30

Feb 2, 2009

On January 29 and 30, 2009, the Florida Commission on Hurricane Loss Projection Methodology (“Commission”) held a workshop on tropical cyclone activity, near and short-term hurricane models and commercial residential issues. 

Chairman Randy Dumm opened the meeting and stated that the workshop is intended to provide an opportunity for modeling companies and the Commission to engage in dialogue about the use and acceptability of short and medium-term hurricane loss projection models, which many policymakers consider controversial.

Prior to discussing the models, Hugh Willoughby, Chairman of the Commission’s Meteorological Standards Committee, gave a presentation entitled “Episodic Changes in Western Atlantic Tropical Cyclones.”  To view a copy of the presentation, click here.

Mr. Willoughby’s report was based on data collected from the Atlantic Basin Hurricane Database HURDAT, which provides historical hurricane data from 1851 to the present.   Among Mr. Willoughby’s conclusions were that sea temperatures fluctuate, and warmer sea temperatures have given rise to larger and more damaging storms. 

Mr. Willoughby also surmised that global warming trends do not indicate increased hurricane landfall.  Other factors such as wind shear, El Niño, La Niña and an increase in southward movement affect landfall activity.  Following Mr. Willoughby’s presentation, workshop participants questioned his conclusion about the effects of global warming.

Glen Daraskevich, Senior Vice President, Karen Clark and Company, gave a presentation entitled “Near-Term Hurricane Models, How Have They Performed?”  To view a copy of the presentation, click here.

Mr. Daraskevich’s presentation was informational and he did not provide definitive statements regarding the use of near-term models, but instead commented that the credibility of a short-term model should include “evaluation of skill over the forecast period of time.” 

Mr. Daraskevich remarked that the 2004-2005 hurricane seasons triggered development of shorter-term models.  He noted that while North Atlantic hurricane activity increased, landfall activity remained static.  He also stated that there are no trends regarding insured losses. 

Following Mr. Daraskevich’s presentation, representatives from AIR Worldwide suggested that all modeling companies should not be “lumped into one basket.”

 

Modeler Presentations

 

Risk Management Solutions

Shree Khare from Risk Management Solutions, Inc. (“RMS”) provided an overview of RMS’ use of medium-term models.  To view a copy of the presentation, click here.

RMS, which began incorporating five-year timeframe models in 2005, uses a suite of approximately 20 models in its process.  According to Mr. Khare, the medium-term model is useful because it combines multiple forecasts. 

Mr. Khare noted that RMS utilizes expert opinions in formulating its hurricane predictions.  During the presentation, Commission members emphasized the need for RMS to justify its conclusions with more scientific data.  Commission members also stated concern that the “suite” of models used by RMS is too complex to fairly evaluate. 

 

AIR Worldwide

AIR Worldwide (“AIR”) representatives gave a presentation on the company’s use of short-term models.  To view a copy of the presentation, click here.

The AIR presentation focused on the impact sea surface temperatures (“SSTs”) have on U.S. hurricane landfalls.   AIR short-term forecasting methods include the use of SSTs in connection with historical data collected since 1900 and an “ensemble approach” to viewing risk.  AIR representatives contend that SSTs are the most dependable predictor of hurricane risk and stated that the warmer the SST, the greater the risk. 

AIR representatives also explained that climate condition models have more uncertainty than traditional models, which have been in use longer.  AIR representatives stated that their models are peer-reviewed to assure they are reasonable, transparent and reproducible. 

Following AIR’s presentation, Commission members discussed whether an insurance company can use one or multiple models in its Florida rate filings.  Commission member and Florida Hurricane Catastrophe Fund Chief Operating Officer Jack Nicholson noted that the applicable Statutes vary in use of the terms “model” and “models,” which can lead to ambiguity.  The Florida Office of Insurance Regulation generally will not approve rate filings that include the use of multiple models.

 

EQECAT

The final modeling presentation was given by Justin Brolley from EQECAT, Inc.  To view a copy of the presentation, click here.

Mr. Brolley gave a brief report that focused primarily on warm and cool phases of the Atlantic Basin.   He stated that hurricane activity has been greater during “warm phase” years (1900-1925 and 1970-1994) and lesser during “cool phase” years (1926-1969 and 1995-present).

Discussion between the Commission and meeting attendees followed the modeling presentations.  Mr. Willoughby stated that the discussion was healthy, but suggested that a five-year time horizon for modeling was too short.  He also reaffirmed his position that science must dictate every step in the modeling process.   Meeting participants indicated that the five-year model includes the best science with the best view of risk.  Chairman Dumm noted he would like further detail regarding the elicitation process and more scientific evidence to support the models. 

AIR Representatives provided their perspective on the model development process and noted that AIR would not submit its new wind field model for approval as previously indicated but will attempt to submit the new model in 2010.  The AIR representatives also stated that the regulatory process was very expensive.

 

Commercial Property Insurance Presentations

Additional modeler presentations and discussion took place regarding commercial-residential property insurance issues. 

Presentations were given by Dr. Mark Powell from the Florida Public Hurricane Loss Model and Dr. Jayanta Guin from AIR. 

  • To view Mr. Powell’s presentation, entitled “Vertical Wind Profiles in Hurricanes – Florida Public Hurricane Loss Model,” click here.
  • Mr. Guin’s presentation entitled “Development of Commercial-Residential Standards for Hurricane Model Acceptability” may be viewed by clicking here.

The commercial-residential property insurance presentations outlined meteorological issues such as vertical wind profiles, engineering issues such as high-rise buildings, actuarial issues such as deductibles, and acceptability process issues.

The above information is a brief summary of the activities that took place during the two-day workshop.  It is not intended to be a comprehensive analysis of the issues discussed.  The meeting notice, agenda and presentations may be viewed by clicking here.

 

Should you have any questions or comments, please do not hesitate to contact Colodny Fass.

 

To unsubscribe from this newsletter, please send an e-mail to ccochran@cftlaw.com