Quantifying Uncertainty and Variable Sensitivity, Billion-dollar Weather and Climate Disaster Cost Estimates

Summary

This is an unformatted version of an article accepted for the journal Natural Hazards, titled Quantifying Uncertainty and Variable Sensitivity within the U.S. Billion-dollar Weather and Climate Disaster Cost Estimates, by Adam B. Smith of NOAA's National Climatic Data Center and Jessica L. Matthews. The abstract states that the paper uses Monte Carlo simulations to quantify the 95%, 90% and 75% confidence intervals around three U.S. natural disaster cost estimate case studies. It states that the simulations provide additional confidence in the U.S. Billion-dollar weather and climate disaster loss estimate report (NCDC 2014). It also states that a sensitivity analysis of Hurricane Ike damage costs determined that commercial losses in Texas are the most sensitive to assumption variability.

Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used

Issuer: Source documents
Document type: Quantifying Uncertainty and Variable Sensitivity (Natural Hazards)
Case: Quantifying Uncertainty and Variable Sensitivity (Natural Hazards)

Cite this document

“Quantifying Uncertainty and Variable Sensitivity, Billion-dollar Weather and Climate Disaster Cost Estimates.” Original at www.ncdc.noaa.gov. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/credit-guarantees-and-new-bank-relationships/, SHA-256 3cf1af1265839ff764b8c2bbbee59ec10f5650eb71e39c94dd1a8c031e5d6590.

Full text

This is an unformatted version of an article that has been accepted for publication in the journal Natural Hazards.
The pre-publication version is available at: http://link.springer.com/article/10.1007/s11069-015-1678-x

Quantifying Uncertainty and Variable Sensitivity within the U.S.
Billion-dollar Weather and Climate Disaster Cost Estimates

1
Adam B. Smith
NOAA’s National Climatic Data Center, Asheville, North Carolina

Jessica L. Matthews
Cooperative Institute for Climate and Satellites -- NC, North Carolina State University, and NOAA’s
National Climatic Data Center, Asheville, North Carolina

---------------------------------------------
1 Corresponding Author Address: Adam Smith, NOAA National Climatic Data Center, 151 Patton Avenue,
Asheville, NC 28801. Telephone: (828) 271-4183 Email: Adam.Smith@noaa.gov

1
Abstract

Research examining natural disaster costs on social and economic systems is substantial. However, there
are few empirical studies that seek to quantify the uncertainty and establish confidence intervals
surrounding natural disaster cost estimates (ex-post). To better frame the data limitations associated
with natural disaster loss estimates, a range of losses can be evaluated by conducting multiple analyses
and varying certain input parameters to which the losses are most sensitive. This paper contributes to the
literature by examining new approaches for better understanding the uncertainty surrounding three U.S.
natural disaster cost estimate case studies, via Monte Carlo simulations to quantify the 95%, 90% and 75%
confidence intervals. This research also performs a sensitivity analysis for one of the case studies
examining which input data variables and assumptions are the most sensitive and contribute most to the
overall uncertainty of the estimate.

The Monte Carlo simulations for all three of the natural disaster events examined provide additional
confidence in the U.S. Billion-dollar weather and climate disaster loss estimate report (NCDC 2014), since
these estimates are within the confidence limits and near the mean and median of the example
simulations. The normalized sensitivity analysis of Hurricane Ike damage costs determined that
commercial losses in Texas are the most sensitive to assumption variability. Therefore, improvements in
quantifying the commercial insurance participation rate for Texas will result in the largest reduction of
uncertainty in the total loss estimate for Hurricane Ike. Further minimization of uncertainty
would continue with improved measurement of subsequent cost parameters in order of descending
sensitivity.

Keywords: natural disasters; costs; losses; uncertainty; statistics of extreme events; sensitivity

2
1. Introduction

The United States and its economy are challenged by weather and climate-related disasters that impart
large social and economic costs (Gall et al. 2011, Field et al. 2012, NCA 2014). Consequently, natural
disaster cost estimates are referenced by a wide variety of users for varying purposes. However, there are
notable differences in the uncertainty surrounding different natural disaster types reflecting the quality of
the data available, methodology and assumptions (Kron et al. 2012). For example, in the United States
drought and flooding events have higher potential uncertainty values around their loss estimates due to
less coverage of insured assets (Smith and Katz 2013). Conversely, severe local storm events have lower
potential uncertainty around their loss estimates due to more complete insurance coverage of wind and
hail damage.

Research examining natural disaster costs on social and economic systems is substantial. Example
studies include: normalizing disaster loss trends over space and time using population and wealth
variables (Downton et al. 2005, Pielke et al. 2008, Barthel and Neumayer 2012, Simmons et al. 2013),
examining how developing countries and smaller economies often suffer more greatly due to natural
disaster impacts (World Bank 2005, Hallegatte and Dumas 2009, IPCC 2014), and exploring how
developed countries have more capacity to rebound from natural disasters impacts due to their wealth
and financial systems (Rasmussen 2004, Toya and Skidmore 2007, Cavallo and Noy 2009). Other research
seeks to quantify total, direct losses (i.e., both insured and uninsured) resulting from specific natural
hazard events using independent estimation methodologies (ECLAC 2003, Munich Re 2014, Swiss Re
2014). However, there are few empirical studies that seek to quantify the uncertainty and confidence
intervals surrounding natural disaster cost estimates (ex-post). To better frame the data limitations
associated with natural disaster loss estimates, a range of losses can be evaluated by conducting multiple
analyses and varying certain input parameters to which the losses are most sensitive (FEMA 2015). This
paper contributes to the literature by examining new approaches for better understanding the
uncertainty surrounding three U.S. natural disaster cost estimate case studies, by running Monte Carlo
simulations to quantify the 95%, 90% and 75% confidence intervals. This research also performs a
sensitivity analysis for one of the case studies examining which input data variables and assumptions are
the most sensitive and contribute most to the overall cost uncertainty of the estimate.

The foundation for this research is the U.S. Billion-dollar Weather and Climate Disaster report
developed by the National Oceanic and Atmospheric Administration’s National Climatic Data Center. This
analysis quantifies the loss from numerous weather and climate disasters including: tropical cyclones,
floods, drought & heat waves, severe local storms (i.e., tornado, hail, straight-line wind damage), wildfires,
crop freeze events and winter storms (NCDC 2014). These loss estimates reflect direct effects of weather
and climate events (i.e., not including indirect effects) and constitute total losses (i.e., both insured and
uninsured). The insured and uninsured direct loss components include: physical damage to residential,
commercial and government/municipal buildings, material assets within a building, time element losses
(i.e., businesses interruption), vehicles, boats, offshore energy platforms, public infrastructure (i.e., roads,
bridges, buildings) and agricultural assets (i.e., crops, livestock, timber). These loss assessments do not
take into account losses to natural capital/assets, healthcare related losses, or values associated with loss
of life. Only weather and climate disasters which cause losses of ≥ 1 billion-dollars in calculated damage
including Consumer Price Index (CPI) inflation adjustment are included in this dataset (Table 1). While this
threshold is arbitrary, these billion-dollar events account for roughly 80% of the total U.S. losses for all
combined severe weather and climate events (Munich Re 2012, NCDC 2014).

These natural disaster cost assessments require input from a variety of public and private data sources
including: the Insurance Services Office (ISO) Property Claim Services (PCS), Federal Emergency
Management Agency (FEMA) National Flood Insurance Program (NFIP) and Presidential Disaster
Declaration (PDD) assistance, and the United States Department of Agriculture (USDA) National
Agricultural Statistics Service (NASS) & Risk Management Agency (RMA), among others. Each of these data
sources provides unique information as part of the overall disaster loss assessment. Previous research

3
analyzed the suitability of these data sources including trends, data accuracy and potential biases (Smith
and Katz 2013) and found the U.S. Billions-dollar disaster estimates had a consistent underestimation bias
of roughly 10 to 15%. This bias was corrected during a reanalysis of the disaster event loss data to reflect
new loss totals (NCDC 2014). However, there are still uncertainty envelopes encompassing these
reanalyzed disaster loss estimates, which this research will offer new approaches for quantifying.

An outline of the paper is as follows. Primary insurance loss data and assumptions for natural disaster
loss estimates are described in Section 2. Next, a Monte Carlo simulation method for estimating
uncertainty surrounding disaster loss estimates, focusing on specific disaster examples, is presented in
Section 3. A sensitivity analysis examining which data and assumptions are most important within one of
the three disaster case study is examined in Section 4. Finally, Section 5 contains a discussion and
conclusions on the uncertainty and sensitivity analysis results, to improve the U.S. billion-dollar disaster
cost analysis.

Disaster Type Number of Percent CPI-adjusted Percent of Average Event
Events Frequency Losses Total Loss Cost
($ billions) ($ billions)
Drought 21 12.4 199 19.1 9.5
Flooding 19 11.2 86 8.3 4.5
Freeze 7 4.1 25 2.4 3.6
Severe Storm 65 38.2 143 13.7 2.2
Tropical Cyclone 34 20.0 530 50.9 15.6
Wildfire 12 7.1 26 2.5 2.2
Winter Storm 12 7.1 35 3.4 2.9

Table 1 Damage cost statistics from U.S. Billion-dollar disaster events (1980-2013) reflecting number of
events, event frequency, CPI-adjusted loss (present year), percent of total losses and average event cost

2. Primary Insurance Loss Data and Assumptions

A number of U.S. insurance participation surveys have been performed over the last several decades
for residential, automotive and commercial lines of insurance. The following paragraphs discuss the data
from these surveys and how it informs our methodology to estimate the total, direct loss for a natural
disaster event. One central adjustment is that insured loss payment data ($) are inflated by a factor
representing the reciprocal of the insurance market participation rate for a specific type of insurance for
each impacted state. Also, we test our disaster cost methodology calculations using a Monte Carlo
analysis, which perturbs each data input value, to estimate 95%, 90% and 75% confidence intervals
surrounding the overall loss estimate.

a. Residential Insurance:

Annual surveys on residential property insurance participation such as the Census American Housing
Survey (1980-present), the Insurance Information Institute survey (2011-present), and All-Industry
Research Advisory Council (1981) have indicated on average that > 90% of U.S. homeowners have multi-
peril property insurance for their residence and contents (Table 2). This coverage includes cost
reimbursement from damage due to wind, hail, lightning, snow and ice, among others, but does not
include coverage for inland or coastal flood damage (discussed in section 2d).

4
Research also shows that residences are often underinsured. In 2013, 60% of homes were
underinsured by an average of 17%; in 2012, 61% of homes were underinsured by an average of 18%
(Marshall & Swift/Boeckh 2013). Scaling these statistics to represent the full housing stock implies that
homeowners are underinsured by an average of ~10%. Therefore, in addition to the 10% who are also
uninsured, we assume that 80%-90% of home losses are covered by insurance policies. The remaining
10%-20% represent uninsured and underinsured property assets.

It is also important to note that wind and water insurance deductibles average about 10% of all paid
insurance claims (ISO Property Claims Service 2014, NFIP 2014) while insurance fraud payments represent
about 10% of total property insurance payments (National Insurance Crime Bureau 2014, Insurance
Information Institute 2014). Therefore, we assume that these two effects are offsetting (i.e., +10% for
deductibles; -10% for fraud; with respect to total insurance payments) and viewed as tangential to our
core analysis since fraud and deductible loss data are not consistently available.

National Insured National Insured
Year Housing Stock (%) Year Housing Stock (%)

1985 93.9 1999 93.6
1987 93.8 2001 93.9
1989 94.3 2003 93.2
1991 93.8 2005 93.7
1993 94.2 2007 94.4
1995 94.0 2009 94.6
1997 93.0 2011 94.1

Table 2 American Housing Survey percentage of U.S. households carrying residential property insurance

b. Automotive Insurance:

Annual U.S. uninsured motorist surveys (1989-2012) by the Insurance Research Council (IRC 2014) have
also found a relatively stable percentage ranging from 80%-90% of automobiles insured (Table 3). Across
the United States, the estimated percentage of insured motorists has increased in recent years (IRC 2014).
In 2001, the average of insured motorists across all states was 85.8%, which increased to 87.8% in 2011
(IRC 2014). Given the variability of automobile insurance over time and space we assume that 80%-90% of
automobiles are insured. The remaining 10%-20% represent uninsured and underinsured property assets.

National Insured National Insured
Year Year
Motorists (%) Motorists (%)

1989 83.7 2001 85.8
1991 84.9 2003 85.1
1993 84.0 2005 85.4
1995 85.8 2007 86.2
1997 86.8 2009 86.2
1999 87.2 2011 87.8

5
Table 3 Percent of Uninsured Motorists by year averaged across the U.S.

c. Commercial Insurance:

Understanding damage to U.S. businesses is more complex. For example, a survey by the National
Association of Insurance Commissioners (NAIC 2007) found that over 90% of the small businesses
surveyed have property insurance. The percentage of companies with property insurance increases with
the size of a company, which may reflect better risk management. However, many businesses lack
business interruption insurance which is a large cost driver in the weeks and months following natural
disaster events. For example, 32% of firms with annual revenue < $1 million had such coverage versus
48% of higher-revenue firms (Table 4). This is further validated by reports indicating that 20-40% of small
businesses that close after a major natural disaster never reopen their business (IBHS 2007, Travelers
2014, III 2014). This can be due to a variety of reasons such as interruption of critical supplies and product
distribution, power outages or other utility failures, loss of customer base and critical data, restricted or
blocked access, employees unable to report to work, etc. Also, like homeowners, it was found that 75% of
commercial buildings (i.e., physical property) are underinsured by an average of 40% (Marshall &
Swift/Boeckh 2011, Travelers 2014). Marshall and Swift/Boekh (MSB) has data for 2,600 locations across
the country and compares the information it has collected with actual reconstruction costs derived from
its insurer clients’ claims experience and adjusts as appropriate. Accounting for this variability and
uncertainty surrounding different forms of commercial insurance, we estimate that 40-60% of natural
disaster-induced business losses are covered by insurance.

Total (%) of
Commercial
companies # of Employees Annual Revenue
Insurance Type
interviewed
$1 M or
1-19 20-99 < $1 M
more
Property* /
91 90 97 91 96
Liability

Commercial Auto 48 47 73 44 67

Business
35 33 58 32 48
Interruption

Table 4 A study by NAIC (2007) found that commercial property, auto and business interruption
insurance coverage varies by business size. *Research by Marshall & Swift/Boeckh (2011) found 75% of
commercial buildings (i.e., building / contents) are underinsured by an average of 40% after examining
data on businesses across 2,600 U.S. locations

In summary, we apply the 80-90% range to both residential and automotive PCS loss data and a 40-60% to
commercial losses. We now take this research a step further to investigate the range of possible error by
using a Monte Carlo analysis perturbing each of the data input values by +/-3% and +/-5% (i.e., reflecting
the insurance survey data uncertainty on which assumptions are based). These values were chosen based
on regulatory audits that routinely confirm the reliability and accuracy of ISO/PCS estimates, finding that

6
final adjusted PCS estimates are within 5% accuracy (Kerney, 2010). Exploring different error levels allows
us to better understand how the assumed insurance coverages affect the 95%, 90% and 75% confidence
intervals surrounding our total loss estimate.

d. NFIP Flood Insurance:

We also have data on flood insurance participation, which factors into our disaster loss analysis.
Residential and commercial flood insurance is most widely provided by FEMA’s National Flood Insurance
Program. Mortgage lenders require any residence within FEMA’s Special Flood Hazard Areas (SFHAs) to
purchase flood insurance. The SFHAs are commonly referenced as those within the 100-year flood plain
boundaries. However, the enforcement and participation is not uniform and many at-risk properties do
not have proper insurance coverage for flood or storm surge-related damage (Table 5). Annual polls on
NFIP participation percentage by region show some consistency on a large scale across the Northeast,
Midwest, and Western regions. Other studies have found the NFIP policy participation across the U.S. is
higher (26%) for eligible parcels (PricewaterhouseCoopers 1999). However, these are still inadequately
low participation rates leading to higher flood cost uncertainty.

all U.S. Northeast Midwest South West
2008 17 20 17 17 15
2009 13 9 14 19 6
2010 10 9 6 14 9
2011 14 5 13 19 12
2012 13 14 6 21 6
2013 13 10 12 15 11
2014 13 11 7 20 8
2008-2014 average 13 11 11 18 10

Table 5 Insurance Information Institute annual survey on NFIP flood insurance participation percentage
by region for all households

There is also a spatial bias in flood insurance policy coverage as NFIP participation is 16% in
communities with 500 or fewer homes in the SFHA, 56% in communities with 501 to 5,000 homes in the
SFHA, and 66% in communities with > 5,000 homes in the SFHA zone (Dixon et al. 2006). Also, the same
research found that the chances of purchasing flood insurance are higher for SFHA communities subject
to coastal flooding/storm surge (63%) compared to communities more at risk to riverine flooding (35%).
One additional factor is that flood insurance coverage outside the high-risk flood areas (SFHAs) is very low
(< than 10%). Yet, NFIP data show that 25% of all flood insurance claims come from the low-to-moderate-
risk areas beyond the 100-year floodplain, which are largely uninsured losses (FEMA 2014). There are also
NFIP coverage limits for residential ($250k structure, $100k contents) and commercial ($500k structure,
$500k contents) properties (FEMA 2014). For these varied reasons, we have defined NFIP policy
participation for the Monte Carlo uncertainty analysis as 10%-25% for inland states and 25%-50% for
coastal states.

7
e. USDA Crop Insurance:

The USDA and associated private crop insurance programs represent over 2 million crop insurance
policies across all states (USDA 2014). USDA data shows that on average across all states, 70% of eligible
acres are insured and most producers select 70% of crop yield to be covered (USDA 2012). Therefore, we
approximate the total crop loss by applying a factor to the crop insurance claims data; that is, 100% all
possible crops/[( x% insured)*(50-75% yield coverage)] = multiplier value(s) for estimating total crop
damage costs. The yield coverage is important as crop insurance is paid only after the crop loss has
surpassed the selected yield coverage. For example, if a producer selected 70% coverage, the crop
producer must first cover the first 30% of crop loss. This is effectively like a deductible, but paid to no one
- just an absorbed cost.

In addition to crop losses, we also incorporate the total livestock feeding cost (i.e., corn and hay for
cattle) when it exceeds the 5-year national average for feedstock (i.e., dollars/per ton). Drought can limit
the availability of corn and hay feed stocks, which increases the cost forcing ranchers to sell off more
cattle than they would during a non-drought year (e.g., increasing long-term meat production costs).
Comparing the 5-year national vs. state feedstock costs against those during a severe drought-year offers
a useful comparison.

f. Framing Uncertainty

In summary, we have set up our disaster cost uncertainty analysis using the following insurance
participation ranges for the each line of insurance (Table 6). The next section discusses the output from
Monte Carlo simulations to determine the 95%, 90% and 75% confidence intervals surrounding our total
loss estimate for the following selected U.S. disaster events:

• the historic U.S. drought (2012),
• the Southeast tornado super-outbreak (late-April 2011),
• and Hurricane Ike (2008).

Minimum Maximum
Wind insurance 80% 90%
(PCS – Residential)
Wind insurance 80% 90%
(PCS – Automotive)
Wind insurance 40% 60%
(PCS - Commercial)
Flood insurance (FEMA) 25% 50%
For coastal states
Flood insurance (FEMA) 10% 25%
For inland states
Crop insurance (USDA) 50% 75%
Multi-peril (drought, flood, etc.)

Table 6 Insurance participation rate ranges used in the Monte Carlo simulation.

8
3. Method for Estimating Uncertainty Surrounding Natural Disaster Loss Estimates

A Monte Carlo approach is employed to assess the uncertainty of disaster loss estimates. For each of
the three events detailed below, the parameters of the equations as outlined in Smith & Katz (2013) were
perturbed to produce a distribution of loss estimates. There are two general categories of input
parameters: 1) insurance participation rates, and 2) loss values. The defined possible ranges for insurance
participation rates were justified in Section 2. We assumed either a +/- 3 or 5% error for the loss
values. For final ISO/PCS insurance loss results, a 3% bound of uncertainty should be adequate.
Regulatory audits routinely confirm the reliability and accuracy of PCS estimates. Historically, after such
regulatory data audits, the final adjusted estimate has differed by at most 5% (Booz Allen Hamilton
personal communication 2013). We also assumed that the parameters were either uniformly or normally
distributed within the ranges. Therefore, there are four cases examined for each of the following events:

+/- 3% error on loss values with uniformly distributed parameters,
+/- 5% error on loss values with uniformly distributed parameters,
+/- 3% error on loss value with normally distributed parameters, and
+/- 5% error on loss value with normally distributed parameters.

10000 simulations were run for each case by random draws within the defined ranges with the defined
distributions. Confidence intervals were calculated by sorting the 10000 loss estimates in ascending order,
then using a percentile method to define the confidence region. For example, to construct the 95%
th th
confidence region, the 250 and 9750 predictions were selected as the lower and upper bounds,
respectively.

a. 2012 U.S. drought

Our first example is the 2012 drought, which resulted in the most extensive drought impacts to affect
the U.S. in decades (NCDC 2014). Moderate to extreme drought conditions affected more than half the
country for the majority of 2012. The most costly drought impacts occurred across the central agriculture
states resulting in widespread harvest failure for corn, sorghum and soybean crops, among others. Using
the USDA crop insurance and feed cost data with the disaster cost methodology described in (Smith and
Katz 2013), the 2012 U.S. drought cost to agriculture was estimated to be $30.0 billion (NCDC 2014).

To verify the uncertainty surrounding this estimate, this research examines 4 different cases of Monte
Carlo simulation for this event. This was performed using a USDA ‘percent of eligible acres insured,’ as a
separate crop value metric from the USDA ‘percentage of insurable crop value’ used to calculate the cost
of the 2012 U.S. drought, as described in section 2. In all cases, the following loss values used are
summarized in Table 7 and the range of crop insurance participation rates (pcrop_yield_cov) are referenced in
Table 6.

The final loss estimate was calculated by:

𝑇𝑇𝑇𝑇𝑇 = [𝑚𝑖 𝑣𝑖 ] + 𝑣𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
𝑖= 𝑎𝑎𝑎 𝑠𝑠𝑠𝑠𝑠𝑠

where

100
𝑚𝑖 =
𝑝𝑎𝑎𝑎𝑎𝑎_𝑖𝑖𝑖𝑖𝑖𝑖𝑖 ∗ 𝑝𝑐𝑐𝑐𝑝𝑦𝑦𝑦𝑦𝑑𝑐𝑐𝑐

9
State Insurance payout ($ million) Percent eligible acres insured
Alabama 19.272608 61
Arkansas 17.157674 67
California 11.133870 34
Colorado 121.977411 70
Delaware 12.641902 81
Iowa 1856.441799 91
Idaho 4.180436 53
Illinois 2861.989110 79
Indiana 1027.963734 72
Kansas 1011.764735 89
Kentucky 359.356293 59
Maryland 121.572572 72
Michigan 131.354503 61
Minnesota 163.322190 90
Missouri 1024.195287 65
Mississippi 17.484280 83
Montana 51.007988 80
North Dakota 111.085759 96
Nebraska 1204.855900 85
New Mexico 23.019718 53
New York 9.711831 27
Ohio 315.336961 68
Oklahoma 94.791212 70
Pennsylvania 11.653171 30
South Dakota 963.773831 94
Tennessee 105.039149 58
Texas 780.123224 76
Virginia 31.021328 48
Wisconsin 363.021359 57
Wyoming 6.106170 29

OTHER
Increased Feed Cost 3614.419032

Table 7 USDA crop indemnity (loss) payout information for each state due to the combined effects of
drought and heat in 2012. The percent of acres in each state that have USDA crop insurance are also
provided. For the Monte Carlo simulation, values in the above table were considered to be within the
associated ranges of +/-3% or +/-5%. State of particular interest have been highlighted in boldface: Illinois
(largest state payout), North Dakota (highest percent of eligible acres insured), and New York (lowest
percent of eligible acres insured).

The results of the Monte Carlo analysis are provided (Table 8). We see that the estimated $30 billion
total, direct cost of the 2012 U.S. drought (NCDC 2014) is reasonably close to the mean and median of the
Monte Carlo simulation. Figures 1a, 1b, 2a and 2b present the histograms of the 10000 estimates for each
case, along with confidence interval bounds.

A histograms of multiplier (mi) values from a selection of states (i.e., Illinois, New York and North
Dakota) affected by the 2012 U.S. drought is shown in Figure 3. Here we see the range of possible
multiplier values applied to the state insurance payout amount. Given the inverse relationship of the

10
multiplier to the percent of acres insured, we see that states with smaller percentages of acres insured (i.e.
New York) will have larger associated multiplier values.

3% 3% 5% 5%
Normal Uniform Normal Uniform
Minimum 27326 27114 27623 27216
Maximum 34661 34230 35402 34903
Mean 30416 30378 30507 30620
Median 30384 30349 30466 30577
75% CI 29398, 31467 29174, 31611 29455, 31591 29383, 31883
90% CI 28986, 31968 28712, 32130 29028, 32102 28922, 32464
95% CI 28728, 32303 28461, 32467 28760, 32459 28648, 32799

Table 8. Results (in $ million) of the four cases analyzing the total cost of the 2012 U.S. drought on crops
and livestock (in original year dollars).

11
Fig. 1a Case 1 results: 3% perturbation on values, parameters normally distributed

Fig. 1b Case 2 results: 3% perturbation on values, parameters uniformly distributed

12
Fig. 2a Case 3 results: 5% perturbation on values, parameters normally distributed

Fig. 2b Case 4 results: 5% perturbation on values, parameters uniformly distributed

13
Fig. 3 Histograms of multiplier values (mi) for several states affected by the 2012 U.S. drought disaster. All
of these are from Case 3 (+/- 5% error, normally distributed parameters). Illinois had the largest insurance
payout (top left). New York had the smallest percent of eligible acres insured (top right). North Dakota
had the largest percent of eligible acres insured (bottom).

14
b. April 25-28, 2011 Tornado Outbreak

Our second example examines a historic tornado outbreak across numerous central and southern
states in late-April, 2011. Several major metropolitan areas were directly impacted by strong tornadoes
including Tuscaloosa, Birmingham, and Huntsville in Alabama and Chattanooga, Tennessee, causing the
estimated damage costs to soar. The total, direct cost for this event was estimated to be approximately
$9.8 billion (NCDC 2014). To verify the uncertainty surrounding this estimate, this research examines 4
different cases of Monte Carlo simulation for this event. In all cases, the following loss values used are
summarized in Table 9 indicating the data values for each state and insurance claim type used in the total,
direct loss calculation.

State PCS PCS PCS FEMA
Commercial Residential Automotive PDD ($ million)
($ million) ($ million) ($ million)
Alabama 1000.0 1500.0 150.0 396.6
Arkansas 45.0 175.0 53.0 -
Georgia 135.0 240.0 20.0 34.9
Illinois 7.0 45.0 12.0 -
Kentucky 31.0 73.5 7.5 48.9
Louisiana 16.0 43.0 11.0 -
Missouri 125.0 125.0 30.0 -
Mississippi 73.0 85.0 11.0 44.5
Ohio 30.0 66.0 2.0 -
Oklahoma 10.0 39.0 16.0 -
Tennessee 410.0 980.0 395.0 76.8
Texas 135.0 320.0 110.0 -
Virginia 19.0 38.0 17.0 -

Table 9 Loss values from the April 25-28, 2011 Tornado Outbreak. For the Monte Carlo simulation, values
in the above table were considered to be within the associated ranges of +/-3% or +/-5%.

The final loss estimate was calculated by:

𝑇𝑇𝑇𝑇𝑇 𝐿𝐿𝐿𝐿
= 𝑚𝑖,𝑃𝑃𝑆𝑐𝑐𝑐𝑐 𝑣𝑗,𝑃𝑃𝑆𝑐𝑐𝑐𝑐 + 𝑚𝑖,𝑃𝑃𝑆𝑟𝑟𝑟 𝑣𝑗,𝑃𝑃𝑆𝑟𝑟𝑟 + 𝑚𝑖,𝑃𝑃𝑆𝑎𝑎𝑎𝑎 𝑣𝑗,𝑃𝑃𝑆𝑎𝑎𝑎𝑎 + 𝑣𝑖,𝐹𝐹𝐹𝐴𝑃𝑃𝑃
𝑖= 𝑎𝑎𝑎 𝑠𝑠𝑠𝑠𝑠𝑠
∗ 𝑣𝑖,𝐹𝐹𝐹𝐴𝑃𝑃𝑃 > 𝑚𝑖,𝑃𝑃𝑃_𝑡𝑡𝑡𝑡𝑡 𝑣𝑖,𝑃𝑃𝑆𝑡𝑡𝑡𝑡𝑡

where

𝑚𝑖,𝑃𝑃𝑆𝑐𝑐𝑐𝑐 is the multiplier for commercial PCS

𝑚𝑖,𝑃𝑃𝑆𝑟𝑟𝑟 is the multiplier for residential PCS

𝑚𝑖,𝑃𝑃𝑆𝑎𝑎𝑎𝑎 is the multiplier for automotive PCS

and these multipliers are defined as:

15
100
Multiplier = .
𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 𝑟𝑟𝑟𝑟

For the comparison part of the equation we define:

𝑣𝑖,𝑃𝑃𝑃_𝑡𝑡𝑡𝑡𝑡 = 𝑣𝑖,𝑃𝑃𝑃_𝑐𝑐𝑐𝑐 + 𝑣𝑖,𝑃𝑃𝑃_𝑟𝑟𝑟 + 𝑣𝑖,𝑃𝑃𝑃_𝑎𝑎𝑎𝑎 (NOTE: no multipliers involved)
𝑚𝑖,𝑃𝑃𝑆𝑐𝑐𝑐𝑐 𝑣𝑗,𝑃𝑃𝑆𝑐𝑐𝑐𝑐 + 𝑚𝑖,𝑃𝑃𝑆𝑟𝑟𝑟 𝑣𝑗,𝑃𝑃𝑆𝑟𝑟𝑟 + 𝑚𝑖,𝑃𝑃𝑆𝑎𝑎𝑎𝑎 𝑣𝑗,𝑃𝑃𝑆𝑎𝑎𝑎𝑎
𝑚𝑖,𝑃𝑃𝑆𝑡𝑡𝑡𝑡𝑡= - 1
𝑣𝑖,𝑃𝑃𝑃_𝑡𝑡𝑡𝑡𝑡

The results of the Monte Carlo analysis are provided (Table 10). Table 6 contains the insurance
participation rates. We see that the estimated $9.8 billion total, direct cost of the April 25-28 tornado
outbreak (NCDC 2014) is reasonably close to the mean and median of the Monte Carlo simulation. Figures
4a, 4b, 5a and 5b present the histograms of the 10000 estimates for each case, along with confidence
interval bounds.

3% 3% 5% 5%
Normal Uniform Normal Uniform
Minimum 8756.4 8759.0 8742.5 8713.5
Maximum 10833 10429 10685 10426
Mean 9519.2 9532.1 9523.7 9529.6
Median 9502.4 9513.2 9505.1 9511.9
75% CI 9246.1, 9801.1 9200.3, 9884.7 9245.1, 9817.8 9192.1, 9891.6
90% CI 9148.9, 9945.3 9097.7, 10016 9130.3, 9964.6 9079.4, 10026
95% CI 9090.5, 10043 9039.5, 10085 9066.5, 10071 9023.2, 10107

Table 10 Results (in $ million) of the four cases analyzing the April 2011 Tornado Outbreak total cost.

16
Fig. 4a Case 1 results: 3% perturbation on values, parameters normally distributed

Fig. 4b Case 2 results: 3% perturbation on values, parameters uniformly distributed

17
Fig. 5a Case 3 results: 5% perturbation on values, parameters normally distributed

Fig. 5b Case 4 results: 5% perturbation on values, parameters uniformly distributed

18
c. September 2008 - Hurricane Ike

In September 2008, Hurricane Ike caused extensive storm-surge and wind damage in Texas. There was
also caused considerable wind and flood damage across many other coastal and inland states (i.e.,
Louisiana, Arkansas, Tennessee, Illinois, Indiana, Kentucky, Missouri, Ohio and Pennsylvania). Severe
gasoline shortages occurred in the southeast U.S. due to damaged oil platforms, storage tanks, pipelines
and off-line refineries. After examining the loss data sources using our tropical cyclone cost methodology
it was determined that the total, direct cost of Ike was approximately $30.0 billion (NCDC 2014). To verify
the uncertainty surrounding this estimate, this research examines 4 different cases of Monte Carlo
simulation for this event. In all cases, the following loss values used are summarized in Table 11. Table 6
contains the insurance participation rates. Note that we consider coastal and inland states separately
when including a multiplier to the FEMA NFIP loss values. Coastal states for this example are: Alabama,
Louisiana and Texas. The inland states are: Illinois, Indiana, and Missouri.

PCS PCS PCS FEMA FEMA
Comm. Residential Auto. PDD ($ million) NFIP ($ million)
($ million) ($ million) ($ million)
Alabama - - - 13.1 1.7
Arkansas 12.5 35.0 8.5 2.5 -
Illinois 50.0 150.0 40.0 108.0 54.3
Indiana 80.0 230.0 20.0 93.0 32.3
Kentucky 110.0 405.0 18.0 18.9 -
Louisiana 50.0 50.0 35.0 263.0 321.0
Missouri 16.0 50.0 10.0 - 42.5
Ohio 255.0 960.0 40.0 39.6 -
Penn. 8.0 63.0 4.0 - -
Texas 4000.0 5500.0 300.0 2464.0 2185.9

OTHER
Oil platform 3000.0
damage
Agriculture & 825.0
Forestry

Table 11 Loss values from the Hurricane Ike disaster. For the Monte Carlo simulation, values in the above
table were considered to be within the associated ranges of +/-3% or +/-5%.

The final loss estimate was calculated by:

𝑇𝑇𝑇𝑇𝑇 𝐿𝐿𝐿𝐿
= 𝑚𝑖,𝑃𝑃𝑆𝑐𝑐𝑐𝑐 𝑣𝑗,𝑃𝑃𝑆𝑐𝑐𝑐𝑐 + 𝑚𝑖,𝑃𝑃𝑆𝑟𝑟𝑟 𝑣𝑗,𝑃𝑃𝑆𝑟𝑟𝑟 + 𝑚𝑖,𝑃𝑃𝑆𝑎𝑎𝑎𝑎 𝑣𝑗,𝑃𝑃𝑆𝑎𝑎𝑎𝑎 + 𝑚𝑖,𝐹𝐹𝐹𝐴𝑁𝑁𝑁𝑁 𝑣𝑖,𝐹𝐹𝐹𝐴𝑁𝑁𝑁𝑁
𝑖= 𝑎𝑎𝑎 𝑠𝑠𝑠𝑠𝑠𝑠
+ 𝑣𝑖,𝐹𝐹𝐹𝐴𝑃𝑃𝑃 ∗ 𝑣𝑖,𝐹𝐹𝐹𝐴𝑃𝑃𝑃 > 𝑣𝑖,𝑃𝑃𝑆𝑡𝑡𝑡𝑡𝑡 + 𝑣𝑠𝑠𝑠𝑠𝑠 + 𝑣𝑈𝑈𝑈𝑈

where

𝑣𝑖,𝑃𝑃𝑃_𝑡𝑡𝑡𝑡𝑡 = 𝑣𝑖,𝑃𝑃𝑃_𝑐𝑐𝑐𝑐 + 𝑣𝑖,𝑃𝑃𝑃_𝑟𝑟𝑟 + 𝑣𝑖,𝑃𝑃𝑃_𝑎𝑎𝑎𝑎 (NOTE: no multipliers involved)

𝑚𝑖,𝑃𝑃𝑆𝑐𝑐𝑐𝑐 is the multiplier for commercial PCS

𝑚𝑖,𝑃𝑃𝑆𝑟𝑟𝑟 is the multiplier for residential PCS

19
𝑚𝑖,𝑃𝑃𝑆𝑎𝑎𝑎𝑎 is the multiplier for automotive PCS

𝑚𝑖,𝐹𝐹𝐹𝐴𝑁𝑁𝑁𝑁 is the multiplier for FEMA NFIP

and
100
Multiplier =
𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 𝑟𝑟𝑟𝑟

The results of the Monte Carlo analysis are provided (Table 12). We see that the estimated $30 billion
total, direct cost of Hurricane Ike is remarkably close to the mean and median of the Monte Carlo
simulation. Figures 6a, 6b, 7a and 7b present the histograms of the 10000 estimates for each case, along
with confidence interval bounds.

3% 3% 5% 5%
Normal Uniform Normal Uniform
Minimum 26539 26669 26308 26518
Maximum 40922 35468 39065 35571
Mean 30356 30451 30334 30445
Median 30233 30382 30191 30355
75% CI 28808, 31974 28592, 32368 28772, 31995 28518, 32435
90% CI 28311, 32903 28003, 33205 28232, 32883 27963, 33327
95% CI 27998, 33661 27720, 33668 27886, 33508 27634, 33802

Table 12 Results (in $ million) of the four cases analyzing the total cost of Hurricane Ike.

20
Fig. 6a Case 1 results: 3% perturbation on values, parameters normally distributed

Fig. 6b Case 2 results: 3% perturbation on values, parameters uniformly distributed

21
Fig. 7a Case 3 results: 5% perturbation on values, parameters normally distributed

Fig. 7b Case 4 results: 5% perturbation on values, parameters uniformly distributed

22
4. Effects of Sensitivity Analysis on Overall Cost Estimate

For each of the total, direct loss estimates for natural disaster events there are a number of input
parameters, each with an associated degree of uncertainty. We wanted to explore how the estimates
respond to perturbations in the input parameters, so we performed a sensitivity analysis for a single case
study. The following analysis is one-dimensional, that is, perturbing only one parameter at a time. We
began with a base set of parameters θ, taken to be the mean of the ranges defined in the examples of
Section 3. The associated loss estimate f(θ) was calculated based on this set of parameters. Then we
applied a 1% perturbation to each parameter forming a new set of parameters θi, for i=1, …n where n is
the number of input parameters. Then the associated loss estimates f(θi) were calculated. Finally, in order
to compare sensitivities fairly, they were normalized by the magnitude of the values. The normalized
sensitivity of f to the ith parameter, si, can be written

𝑑𝑑(𝜃) 𝑓(𝜃𝑖 ) − 𝑓(𝜃) 𝜃
𝑑𝑑 𝜃𝑖 − 𝜃 𝑓(𝜃)

Taking the Hurricane Ike disaster as an example, we explored the sensitivity of the total loss estimate
to each of the input parameters. In all there are 76 inputs for this calculation. The results are presented in
Figure 8. We see that the most sensitive parameters for the Hurricane Ike estimate are the PCS
commercial insured loss value and associated multiplier for Texas. We may interpret this to mean that a
change in the PCS commercial value or multiplier for Texas would impact the total loss estimate of
Hurricane Ike more than changing any other parameter. Therefore, if the goal is to reduce the uncertainty
of the total loss estimate for this Hurricane Ike example, the best way forward is to reduce the uncertainty
in the PCS commercial value and the PCS commercial insurance participation rate for Texas, and then
continue with subsequent parameters in order of descending sensitivity.

Fig. 8 Normalized sensitivity analysis of damage totals caused by Hurricane Ike by state and loss data
category.

23
5. Discussion and Conclusion

The Monte Carlo simulations for all three of the natural disaster case studies provide additional
confidence in our total, direct loss estimates (NCDC 2014) for these events since the estimates are within
the confidence limits and close to the mean and median of the Monte Carlo simulations for each disaster
example. The reason that some differences are present is because the total, direct loss estimates for each
event, as reported in NCDC (2014), use a slightly different approach to loss calculation. Whereas in NCDC
(2014) the PCS loss component values are grouped together, in this analysis we treat the Residential,
Commercial, and Automotive category losses separately. Here we are leveraging the information gathered
through the survey review, as described in Section 2, to incorporate more granular insurance multiplier
values than those used previously as well. In this analysis we defined ranges for insurance participation
and subsequently used these to calculate the associated multipliers.

The confidence intervals are of value because they provide greater clarity to the quality of the
assessment. For each of the three case studies, we examined the impact of two main assumptions, the
error present in the loss values and the distribution of the possible values. For the April 25-28, 2011
tornado outbreak and Hurricane Ike, the impact of assuming 3% or 5% error on loss values was negligible
according to hypothesis testing to determine if the samples were from the same distribution (using
Kolmogorov-Smirnov, Wilcoxon rank-sum, and Wilcoxon signed-rank tests). However, these same
hypothesis tests indicate that the choice of a normal or a uniform distribution of loss values was more
important. This is especially noticeable when comparing histograms, where the simulations with the
normal assumption tend to have longer upper tails than the simulations with the uniform assumption.
Therefore, in the interest of being more conservative in the choice of confidence bounds, we recommend
the confidence intervals as derived from assuming 5% error on uniformly distributed loss values.

The 2012 U.S. drought example had different results from the hypothesis testing. Here, the impact of
both assumptions was important. The cases assuming uniformly distributed values resulted in wider, and
hence more conservative, confidence regions. Increasing the assumed error on loss values from 3% to 5%
also caused the confidence regions to enlarge, and additionally caused a positive shift. Although the
confidence regions from the four cases are not perfectly nested, all things considered assuming 5% error
on uniformly distributed loss values provides the most conservative confidence region estimate.

We also performed a sensitivity analysis examining how the separate loss variables contribute to the
overall loss total for a specific case study. It is not surprising that when examining the Hurricane Ike
disaster we found that the Texas insurance cost variables had the most sensitivity, as Texas experienced
the largest loss values for all data categories; residential, automotive, commercial, and NFIP flood-related
losses. The offshore energy losses were also quite sensitive to the overall cost estimate. The total loss
estimate for Hurricane Ike used 76 input parameters, each of which has some uncertainty which
contributes to the uncertainty in the total loss estimate. The intent of the sensitivity analysis described in
Section 4 was to isolate the most important parameters to focus on reducing the uncertainty, which will in
turn have the greatest impact in reducing the uncertainty in the total loss estimate.

Although we explore Monte Carlo simulations for just several disaster examples, we seek to apply this
methodology to all of the U.S. Billion-dollar weather and climate event loss estimates (NCDC 2014) to
provide better context regarding disaster cost uncertainty. This research is a next step to enhance the
value and usability of estimated disaster costs given data limitations and inherent complexities.

24
Acknowledgments

The work of Jessica Matthews was supported by NOAA through the Cooperative Institute for Climate
and Satellites - North Carolina under Cooperative Agreement NA14NES432003. We thank ISO/Property
Claims Service, FEMA’s National Flood Insurance Program, USDA’s Risk Management Agency and the
Insurance Information Institute for providing key sources and insights to better understand these data
and important caveats. We also thank a number of reviewers including Neal Lott, Mike Squires, Scott
Chapal, John Bates, Tom Peterson, Anthony Arguez, Tom Karl and anonymous reviewers for useful
comments to enhance this research article.

References

All-Industry Research Advisory Council (1981) The availability of homeowners insurance in six major cities
consumer experience and attitudes. Bethesda, Maryland

Barthel F, Neumayer E (2012) A trend analysis of normalized insured damage from natural disasters.
Climatic Change 113:215–237

Cavallo EA, Noy I (2009) The economics of natural disasters: a survey. (IDB working paper series IDB-WP-
124) p 55

Census American Housing Survey (2014) http://www.census.gov/programs-surveys/ahs.html

Dixon L, Clancy N, Seabury SA, Overton A (2006) The National Flood Insurance Program’s market
penetration rate: estimates and policy implications. RAND Corporation, Santa Monica, California

Downton M, Pielke RA Jr (2005) How accurate are disaster loss data? The case of U.S. flood damage. Nat.
Hazards 35:211–228

ECLAC (2003) Handbook for estimating socio-economic and environmental effects of disasters. United
Nations, Economic Commission for Latin America and the Caribbean and International Bank for
Reconstruction and Development

FEMA (2014) Flood Insurance Manual http://www.fema.gov/flood-insurance-manual

FEMA (2015) HAZUS multi-hazard loss estimation methodology https://www.fema.gov/hazus-mh-training

Field CB, Barros V, Stocker TF, et al. (2012) Managing the risks of extreme events and disasters to advance
climate change adaptation. Cambridge University Press

Gall M, Borden KA, Emrich CT, Cutter SL (2011) The unsustainable trend of natural hazard losses in the
United States. Sustainability, 3, 2157-2181

Hallegatte S, Dumas P (2009) Can natural disasters have positive consequences? Investigating the role of
embodied technical change. Ecol Econ 68:777–786

Institute for Business & Home Safety (2007) Open for business. http://www.disastersafety.org/wp-
content/uploads/open-for-business-english.pdf

Insurance Information Institute (2014) Insurance fraud, February 2015 http://www.iii.org/issue-
update/insurance-fraud

25
Insurance Research Council (2014) Uninsured motorists annual report. http://www.insurance-
research.org

IPCC (2014) Climate Change 2014: Impacts, adaptation and vulnerability. Cambridge University Press

ISO Property Claims Services (2014) Available online: http://www.iso.com/Products/Property-Claim-
Services/Property-Claim-Services-PCS-info-on-losses-fromcatastrophes.html

Kerney (2010) The PCS process for identifying catastrophes and estimating the related losses. Reinsurance
Association of America Catastrophe Modeling Conference

Kron W, Steurer M, Low P, Wirtz A (2012) How to deal properly with a natural catastrophe database –
analysis of flood losses. Nat. Hazards Earth Syst. Sci. 12: 535–550

Marshall & Swift/Boeckh (2011) MSB 4Q 2011 RTC Product Bulletin.

Marshall & Swift/Boeckh (2013) MSB ITV index for residential properties. Press release.
http://www.msbinfo.com/Company/News/55

Munich Re (2012) Severe Weather in North America, Perils, Risks, Insurance. Munich Re Group, pp 274

Munich Re (2014) Topics Geo – Annual Review: Natural catastrophes 2014. Munich, Germany.
http://www.munichre.com/en/reinsurance/business/non-life/natcatservice/annual-
statistics/index.html

National Association of Insurance Commissioners (2007) Business & employee insurance issues among U.S.
small businesses http://www.naic.org/documents/newsroom_small_business_summary.pdf

National Insurance Crime Bureau (2014) Insurance Fraud: Understanding the basics.
https://www.nicb.org/ theft_and_fraud_awareness/fact_sheets

NCA (2014) U.S. National Climate Assessment Report. U.S. Global Change Research Program Washington
DC http://nca2014.globalchange.gov/downloads

NCDC (2014) Billion-dollar weather and climate disasters. http://www.ncdc.noaa.gov/billions

NFIP (2014) Flood insurance reform - reports and studies. https://www.fema.gov/flood-insurance-
reformreports-and-studies

Pielke RA Jr, Gratz J, Landsea CW, Collins D, Saunders MA, Musulin R (2008) Normalized hurricane damage
in the United States: 1900–2005. Nat. Hazards Rev. 9:29–42

PricewaterhouseCoopers (1999) Study of the economic effects of charging actuarially base premium rates
for pre-FIRM structures

Rasmussen TN (2004) Macroeconomic implications of natural disasters in the Caribbean. IMF working
paper WP/04/224

Simmons K, Sutter D, Pielke R (2013) Normalized Tornado Damage in the United States, 1950-2010,
Environmental Hazards, 12, 132-147

Smith AB, Katz RW (2013) U.S. billion-dollar weather and climate disasters: Data sources, trends, accuracy,
and biases. Nat. Hazards 67:387–410. doi: 0.1007/s11069-013-0566-5

Swiss Re (2014) Sigma - Natural catastrophes and man-made disasters in 2014. Zurich, Switzerland.
http://www.swissre.com/sigma/

26
Toya H and M Skidmore (2007) Economic development and the impacts of natural disasters. Econ Lett
94:20–25

Travelers (2014) Business risk index. Available online: https://www.travelers.com/prepare-prevent/iw-
images/risk-index/2014-Travelers-Business-Risk-Index-infographic.pdf

USDA/RMA (2012/2014) Summary of business and NASS principal crop acreage
http://www.rma.usda.gov/data/sob.html

World Bank (2005) Natural disaster hotspots: A global risk analysis. World Bank, Washington DC

File and source

File: credit-guarantees-and-new-bank-relationships.pdf
Size: 624,550 bytes
SHA-256: 3cf1af1265839ff764b8c2bbbee59ec10f5650eb71e39c94dd1a8c031e5d6590
Our copy: credit-guarantees-and-new-bank-relationships.pdf
Original: www.ncdc.noaa.gov