FPHLM

Introduction

The Probabilistic Insured Loss Model estimates, for a given exposure, the average annual loss costs due to hurricane winds. The process involved is as follows: input data is supplied by the Provider; the data is checked, cleaned and uploaded into the model by the Developers; the model programs are run to process the data; the model generates a set of output files; the output files are verified by the Verifier; the files are made available to the End user. The old process required that input data be shipped in on CDs or through less secure e-mails, and output files be shipped out through e-mails. Shipping of data can also be time consuming. To both expedite the process and protect it, a website has been created through which data can be directly uploaded by the Provider and output files directly downloaded by the End user in a relatively timely and secure manner. The rest of the process remains the same.

This manual is written for the users who access the website to upload the data, view and/or download the output files. The main purpose of the manual is to introduce the users to some of the basic functions of the website and provide instructions on how to use them. The manual also contains information on the required format for the input files, and the structure and content of the output files. It also discusses what type of information need to be provided by the companies during registration etc.

The website is designed to enable the users to access data efficiently and securely. There are three main users of the website: The Developers, the Company (uploads the data, and can be OIR or the insurance company), and the End user (presumably the Office of Insurance Regulation). Each has separate functionalities and there is role based access. Since the information exchanged through the web site is sensitive, special care is taken to secure all information exchange by using the SSL service. The manual consists of explanation of the step wise execution of the different functionalities accompanied by illustrated screen shots.

The PILM is one of two actuarial components of the Public Hurricane Loss Projection Model. The other is the Scenario Insured Loss Model which estimates losses for a hypothetical or real hurricane event. The PILM can predict annual expected insured residential losses in Florida for an insurance company, zip code, county, or for the entire state. Separate loss estimates are produced for structure, content, appurtenant structure (AP), and additional living expenses.  These losses are reported for portfolios classified by construction type (e.g. masonry, frame, manufactured homes), by county or zip code, by policy form (e.g., HO-3, HO-4 etc.), by rating territory, and combinations thereof.

About The Website

This website link is https://www.cs.fiu.edu:8443/fphlm/. It allows the following functions:

• The companies (or OIR) upload data.
• The Developers are automatically notified that a new data set has arrived.
• The Developers process the data offline and notify the verifier once the output files are uploaded
• The Verifier checks the output files.
• Upon verification the End user (Company or OIR) is automatically notified through e-mail that results are available for retrieval.
• The End-user downloads the output files

User Groups:

There are three main user groups:

1. The Developers comprising of the Administrator and the Verifier.

The Administrator is responsible for maintaining the databases, running the model, and generating the output files. The Verifier is responsible for verifying the results once they are made available by the Administrator.

2. The Companies (or OIR) capable of uploading the data to be executed by the Model.
3. The OIR capable of downloading the verified results.

Please Log in To Access and Download the User Specific User Manual.