Seismic risk management of insurance losses using extreme value theory and copula - Seismic Risk Analysis and Management of Civil Infrastructure Systems - page 754

billion US dollars, respectively (Swiss Re, 2011), which rank in the top 20
costliest natural catastrophes since 1970 (in terms of insurance loss). More-
over, two most recent devastating events in Christchurch (New Zealand)
and Tohoku (Japan) are expected to reach insurance claims of 12 and 20-30
billion US dollars, respectively. (Note that the insured loss is only a fraction
of the total economic seismic loss.)
To assess potential impact of future destructive earthquakes accurately
and to implement risk mitigation measures effectively, there is a need for
quantitative seismic loss estimation and decision-making tools. The tools
should take into account key uncertainty and dependence in earthquake
occurrence, ground motion intensity, and nonlinear structural behaviour.
For such purposes, many methods and tools have been developed (FEMA/
NIBS, 2003; Goulet et al. , 2007; Ellingwood et al. , 2008; Yucemen et al. , 2008;
Goda et al. , 2011). For instance, the HAZUS-Earthquake, developed by
FEMA and NIBS (2003), is capable of assessing regional seismic loss to a
number of buildings and infrastructure due to a scenario earthquake, and
serves as a critical risk management tool for both pre-disaster planning and
post-disaster relief activities. However, it cannot evaluate the probability
distribution of aggregate seismic loss; such information is valuable for eval-
uating/comparing viable mitigation options and for analysing insurance/
reinsurance portfolio under earthquake risk (Kleindorfer et al. , 2005;
Bazzurro and Park, 2007). In fact, there is an urgent research need to further
develop more rigorous tools/methods that can be used for complex risk
management problems. One such problem is insurance/reinsurance portfo-
lio analysis of properties protected by earthquake insurance (Dong and
Grossi, 2005; Kleindorfer et al. , 2005; Bazzurro and Park, 2007; Goda and
Yoshikawa, 2012).
An important aspect in assessing seismic loss of a portfolio of buildings
and infrastructure is the adequate consideration of uncertainty and depen-
dence of physical effects on multiple structures due to ground shaking. The
probabilistic approach laid out by the Pacifi c Earthquake Engineering
Research (PEER) Center is a rational and comprehensive framework
(Goulet et al. , 2007), and serves as a good initial model for general purposes.
Furthermore, consideration of spatiotemporally correlated seismic excita-
tions, when analysing multiple structures, must be emphasised (Goda et al. ,
2011). Recently, Bazzurro and Park (2007) investigated the effects of port-
folio relocation and aggregation (i.e. modelling many buildings originally
located in a wide area as a single macro-structure at a single site) on the
accuracy of seismic loss estimation for multiple properties, and found that
such artifi cial manipulation, which is routinely done in insurance portfolio
analysis, could bias the assessment. Aiming at avoiding such inaccuracy in
portfolio relocation and aggregation, Goda and Ren (2010) investigated
a copula-based approach to combine seismic losses from two building