Activity Number:
|
670
- Statistical Computing in Data Science
|
Type:
|
Contributed
|
Date/Time:
|
Thursday, August 3, 2017 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Statistical Computing
|
Abstract #324707
|
|
Title:
|
Florida Public Hurricane Model - Estimation of Extreme Losses
|
Author(s):
|
Florence George* and Sneh Gulati and Shahid Hamid
|
Companies:
|
Florida International University and Florida International University and Florida International University
|
Keywords:
|
Picklands Estimator ;
Hills Method ;
Extreme loss
|
Abstract:
|
Natural disasters are increasing worldwide and consequently so are the economic losses caused by these events. Estimation of probable maximum loss is an important component of any loss model. In this paper, we investigate the use of semi-parametric methods to estimate Probable Maximum Loss (PML) for the Florida Public Hurricane Loss Model (FPHLM). Detailed analysis of the data shows that the annual losses from FPHLM do not tend to be very heavy tailed and therefore neither the popular Hill's method nor the Moment's estimator work well. However, Pickand's estimator with threshold around the 84th percentile provides a good fit for the extreme quantiles for the losses.
|
Authors who are presenting talks have a * after their name.