Florida Commission on Hurricane Loss Projection Methodology--July 23,24 Meeting Report
Date Published: 07-28-2009
The Florida Commission on Hurricane Loss Projection Methodology ("FCHLPM") met on Thursday, July 23 and Friday, July 24 in Tallahassee. Below is a summary of items discussed during the two-day meeting. To view the complete meeting agenda, click here.
Discussion of 2009 Legislative Changes Affecting the FCHLPM
As result of House Bill 1495 becoming law on June 1, 2009, the FCHLPM is required to review insurance premium mitigation credits and present a report on its findings to Florida Executive and Legislative leadership by February 1, 2010. To view a summary of legislative changes affecting the FCHLPM, click here.
To generate its report, the FCHLPM will hold a series of public meetings to gain a better understanding of the current process for granting wind mitigation credits. Testimony is expected to facilitate solutions to problems identified by insurers and the public. The first meeting is tentatively scheduled for August 12, 2009.
State Representative Bryan Nelson, who sponsored HB 1495 and attended the July 23 FCHLPM meeting, provided anecdotal evidence of problems with the current wind mitigation system. As an example, he said that some South Florida policyholders of Citizens Property Insurance Corporation could have negative homeowners' insurance rates after implementation of the credits.
According to Representative Nelson, some insurance carriers have related that certain companies providing wind mitigation services are promising Florida homeowners discounted homeowners' insurance rates as a result of their services. However, in some cases, these companies are either facilitating the application of those discounts through illegitimate measures or, upon insurer inspection of the property, it becomes apparent that the mitigation measures allegedly performed by these wind mitigation companies have not been done.
FCHLPM members expressed concern with the method used to calculate wind mitigation credits. Under the current system, discounts are calculated using one model (the HURLOSS model from Applied Risk Associates), and insurance rates are calculated using another. Commission members stated that this presents a problem because, while the HURLOSS model provides the greatest wind mitigation discounts, it also results in the highest homeowners' insurance rates.
Senate Banking and Insurance Committee Staff Director Steve Burgess, explained to the Commission that, in 2005, the FCHLPM was granted a public records exemption for trade secret information. The exemption will sunset in 2010. Because the FCHLPM examines hurricane computer models in closed-door sessions to discuss certain trade secret information, there has been criticism that the public cannot examine it. The Senate Banking and Insurance Committee staff is required to analyze this issue and generate a report to the Committee
Mr. Burgess requested that the FCHLPM examine whether a change should be effected in its public records exemption.
FCHLPM members expressed concern about the amount of information discussed during the public meetings versus closed meetings. It was suggested that the FCHLPM review what types of information are considered trade secrets.
Presentations and Discussion on Storm Surge Issues
Florida Public Hurricane Loss Model
Mark Powell from the National Oceanic and Atmospheric Administrations' Hurricane Research Division discussed storm surge damage, particularly related to Florida's Gulf Coast. According to Mr. Powell, storm surge can occur before, during or after a hurricane or tropical storm makes landfall. To view Mr. Powell's presentation, entitled "Wind vs. Water," click here.
Most homeowners carry windstorm insurance, but not all have flood insurance. For homes covered by both types of insurance, wind damage is covered by a homeowners' or special windstorm policy, but water damage is generally covered by the National Flood Insurance Program. When a claim for damage is filed, there is often uncertainty about whether the damage was caused by the wind or by flooding. This issue is especially pertinent to insureds located in flood plains or areas susceptible to storm surge.
Case studies on storm surge issues are being conducted to gather information on whether storm surge modeling should be added to the Florida Public Hurricane Loss Model.
FCHLPM members suggested that, to eliminate ambiguity and simplify hurricane modeling, a legal standard should be established for water damage caused by wind, versus water damage caused by flooding.
Fair Insurance Rates in Monroe ("FIRM")
In her presentation on wind versus water damage in Monroe County, FIRM Executive Director Colleen Repetto asserted that water causes a greater portion of Monroe-based damage claims than wind, as a direct result of Monroe County having the strongest building code standards in Florida. According to Ms. Repetto, insurers have an unfair negative perception of Monroe County because of its location. Private insurance companies are not writing policies in Monroe County because they do not believe they can charge enough premium for the risk.
After Ms. Repetto's presentation, several modeling companies disputed FIRM's claims.
Florida Division of Emergency Management
Ben Nelson from the Florida Division of Emergency Management reported on the Statewide Regional Evacuation Study Program ("SRESP").
The 2006 Florida Legislature provided funding to update statewide regional evacuation plans and transportation networks, as well as to conduct evacuee behavioral and vulnerability studies. Computer-modeling analyses on the effects of storm surge also was included.
The Florida Division of Emergency Management also is updating Florida's 11 regional evacuation studies, which incorporate computer models used for coastal storm surge/flood prediction. To view more information on these studies, click here.
The first phase of the SRESP analyzed regional hurricane evacuation planning centers throughout the State, as well as transportation and other infrastructure systems to determine whether Floridians are more likely to evacuate to shelters or elsewhere in the event of a hurricane or other natural disaster.
The second SRESP phase involves the update of a Florida topographical map by utilizing the Light Detection and Ranging ("LiDAR") system. Data is being analyzed to account for factors such as tree canopies and houses, and will be utilized to assist in calculating storm surge zones. To view more information on the LiDAR system, click here.
To view Mr. Nelson's presentation, click here.
Presentations and Discussion on Near/Short-Term Hurricane Models
Dr. Pete Dailey from AIR Worldwide gave a presentation on how his company utilizes sea surface temperature data, wind shear conditions and steering ocean currents to predict hurricane landfall rates and storm intensity.
According to Dr. Dailey, studies have shown that, as the ocean's water temperature increases, the number of hurricanes making landfall decreases, but the strength of those storms is greater. Some scientists attribute the frequency of hurricane landfall to natural cyclical temperature changes such as El Niño, while some attribute it to global warming issues such as deforestation and greenhouse gasses.
AIR Worldwide utilizes current climate data and future climate predictions in formulating its hurricane loss projection models.
To view Dr. Dailey's presentation, entitled "Hurricane Risk: Importance of Climate Time Scale and Uncertainty," click here.
Justin Brolley from EQECAT reported on his company's near-term hurricane computer models, which utilize historical, rather than climate change data.
EQECAT relies on the HURDAT data set, which is maintained by the United States National Hurricane Center. The Center provides historical hurricane tracking information for the Atlantic and Gulf of Mexico from 1891 to the present.
FCHLPM members expressed concern that using historic data alone to compile hurricane loss projection models makes the data too easy to manipulate, because, by using historic data only, a large portion of the available information (i.e. climate data) is ignored.
To view Mr. Brolley's presentation, entitled "Marginal Impact of Climate Change on Atlantic Hurricane Risk," click here.
Risk Management Solutions ("RMS")
Katie Coughlin from RMS gave a presentation on the 38 short-term hurricane models developed by her company in an effort to predict hurricane landfall on a five-year scale. To formulate its five-year predictions, RMS has created hurricane landfall prediction models based on different statistical paradigms, such as historical averages and climate change.
FCHLPM members questioned whether predicting hurricane landfall five years into the future is possible. Since available data sources, expert opinions and data solicitation methods are highly volatile, the data used to predict hurricane landfall so far in the future could be "tweaked" to provide different outcomes, based on who is using the information.
To view Ms. Coughlin's presentation, entitled "Risk Management Solutions, Medium-Term View in Hurricane Activity," click here.
RMS' Request for Reconsideration of the FCHLPM Decision to Deny Certification of RiskLink Version 8.0.1.a
During its June 3 meeting, the FCHLPM denied certification of RMS' RiskLink Version 8.0.1.a hurricane model because it did not meet two acceptability standards: M-2 (Hurricane Parameters and Characteristics) and M-4 (Hurricane Windfield Structure).
RMS subsequently sent a letter to the FCHLPM requesting reconsideration of the model. To view RMS' June 10 letter, click here.
At Commissioner Jack Nicholson's suggestion, the FCHLPM approved the addition of a provision to its operating procedures that would allow the RMS issue to be reconsidered, since appeals were not expressly authorized. To view the addendum to the FCHLPM's 2008 Report of Activities, click here.
After an RMS presentation on the contested standards and lengthy discussion by FCHLPM members, RiskLink Version 8.0.1.a was approved for use in establishing Florida residential property insurance rates.
Presentation and Discussion on Commercial Residential Issues
AIR Worldwide Vice President John Rollins gave a presentation on hurricane modeling for commercial residential structures and identified the following factors that create uncertainty in commercial residential models:
- There are fewer hurricane models relating to commercial residential structures than personal residential structures.
- Loss mitigation presents a problem, because even the most effective mitigation measures are not the same for commercial and personal residential properties.
- Claims data for commercial residential buildings is limited and does not always distinguish among condominium association, condominium unit and renters' insurance policies.
- Commercial residential structures present different challenges than personal residential structures because different areas of a commercial residential structure may have different wind loads.
- Wind intensity can vary significantly from ground level to the top of a commercial residential building.
- Engineering issues like rooftop structure and anchorage, as well as rooftop equipment are different for commercial and personal residential structures.
To view Mr. Rollins' presentation, entitled "Development of Commercial Residential Standards for Hurricane Model Acceptability," click here.
Mr. Rollins suggested that an engineer be appointed as a member of the FCHLPM. Commissioner Howard Eaglefeld suggested that the FCHLPM communicate that request to the Florida Legislature.
Discussion and Framework of the Acceptability Process
Dr. Nicholson gave an update on recent changes to the laws governing the FCHLPM. The FCHLPM now is required to review hurricane loss projection computer models every odd year, rather than every year.
In addition to reviewing hurricane loss projections for ratemaking, the FCHLPM's role also has been expanded to provide information relating to probable maximum loss ("PML"). In order to incorporate PML review, commercial residential buildings now will have to be examined, as well as personal residential buildings. FCHLPM members suggested that reasonable, simple modeling standards should first be developed.
The possibility of approving short-term hurricane loss projection models was discussed. Dr. Nicholson stated that he believed there were certain drawbacks to short-term models, and that modelers should not be permitted to submit both a long and short-term model for consideration. He added that there is not enough information on short-term models available to be able to recommend their fair and unbiased use.
The meeting was then adjourned.
Should you have any questions or comments, please contact Colodny, Fass, Talenfeld, Karlinsky & Abate.
To unsubscribe from this newsletter, please send an e-mail to email@example.com