Evaluating the Impact of theEvaluating the Impact of the Increase in Hurricane Frequency q yUsing and Internal Model. A
Simulation Analysis
Enrique de AlbaINEGIandand
University of Waterloo
Ricardo AndradeINEGI
Hurakan (“One legged")Hurakan ( One legged )
Hurakan, also known as Heart of Heaven, was a Mayan god of wind, storm and fire. Dwelt indarkness and mist he repeated the word "earth" until land rose from the sea He planned thedarkness and mist, he repeated the word earth until land rose from the sea. He planned thecreation of life and participated at creating mankind. When gods became angry with the first“wood” humans, Hurakan sent a resin flood to destroy them.
In appearance he has one leg the other being transformed into a serpent a zoomorphic snout orIn appearance he has one leg, the other being transformed into a serpent, a zoomorphic snout orlong‐nose, and a smoking object such as a cigar, torch holder, or axe head which pierces a mirror onhis forehead.
INTRODUCTION
Tropical Cyclone ClassificationTropical Cyclone Classification
Saffir‐Simpson Scale Category Wind Speed (Km/h)
Tropical Storm (TS) 61 to 118 Hurricane Category 1 (H1) 119 to 153Hurricane Category 1 (H1) 119 to 153 Hurricane Category 2 (H2) 154 to 177 Hurricane Category 3 (H3) 178 to 209
S N ti l O i At h i Ad i i t ti U S
Hurricane Category 4 (H4) 210 to 249 Hurricane Category 5 (H5) More than 249
Source: National Oceanic Atmospheric Administration, U.S.
H1 and H2 hurricanes thathit Mexico between 1970hit Mexico between 1970and 2006
NORTH ATLANTIC OCEANOCEAN
EAST PACIFICEAST PACIFIC OCEAN
Hurricane Category 2 H i C 1Hurricane Category 1
Source: National Meteorological Service, Mex. (2009).
H3, H4 and H5 hurricanesthat hit Mexico betweenthat hit Mexico between1970 and 2006
NORTH ATLANTIC OCEANOCEAN
EAST PACIFICEAST PACIFIC OCEAN
Hurricane Category 5H i C 4Hurricane Category 4 Hurricane Category 3
Source: National Meteorological Service, Mex. (2009).
Most Expensive Catastrophes in MexicoMost Expensive Catastrophes in Mexico
Highest Losses for Mexican Insurance SectorHighest Losses for Mexican Insurance SectorCatastrophe Loss (million USD)
Hurricane Wilma (2005) 1,752Tabasco Flood (2007) 700Hurricane Gilbert (1997) 567Mexico City Earthquake (1985) 473Mexico City Earthquake (1985) 473Hurricane Isidore (2002) 308Hurricane Emily (2005) 302Hurricane Stan (2005) 228Hurricane Kenna (2002) 176Hurricane Julliette (2001) 90
Source: Mexican Association of Insurance Institutions (2007).
Hurricane Julliette (2001) 90Hurricane Pauline (1997) 62
Hurricane Hits in 2005 (North Atlantic)Hurricane Hits in 2005 (North Atlantic)-120 -110 -100 -90 -80 -70
4030
lat
200
210
Emily (H4, H3)Stan (TS, H1)Wilma (H4)
lon
240 250 260 270 280 290
Increase in Hurricane Activity?Increase in Hurricane Activity?
• K t T R d T l R E (2004) I t f CO2 i d d
(In Favour)
• Knutson, T.R., and Tuleya, R.E. (2004), Impact of CO2‐induced warming on simulated hurricane intensity and precipitation: Sensitivity to the choice of climate model and convective parameterization Journal of Climate 17(18) 3477 3495parameterization. Journal of Climate 17(18), 3477‐3495.
• Emanuel, K. (2005), Increasing Destructiveness of Tropical Cyclones O th P t 30 Y N t 436 686 688Over the Past 30 Years. Nature 436, 686‐688.
• Webster, P.J., Holland, G.J., Curry, J.A., Chang, H.‐R. (2005), Changes , , , , y, , g, ( ), gin Tropical Cyclone Number, Duration, and Intensity in a Warming Environment, Science Vol. 309, 1844‐1846.
• Agata, H. (2007), El Cambio Climático: Repercusiones Económicas para el Seguro, XIX Seminario Internacional de Seguros y Fianzas de la CNSF.la CNSF.
Increase in Hurricane Activity?Increase in Hurricane Activity?
L d CW (2005) H i d Gl b l W i
(Against)
• Landsea, C.W. (2005), Hurricanes and Global Warming, Nature 438(22), E11‐E12.
• Pielke, R.A. (2005), Are there trends in hurricane destruction?, Nature 438, E11‐E13.
• Landsea, C.W., Harper, B.A., Hoarau, K., Knaff, J.A. (2006), Can We Detect Trends in Extreme Tropical Cyclones?,Can We Detect Trends in Extreme Tropical Cyclones?, Science Vol. 313, 452‐454.
• Kerr, R.A. (2008), Hurricanes Wont Go Wild, According to Climate Models, Science Vol. 320, 999.
Tropical Cyclone Frequency (1950 2007)Tropical Cyclone Frequency (1950‐2007)Tropical Cyclones per Year in the North Atlantic
quen
cy
15202530 29 Years
All Tropical CyclonesHurricanes
Freq
50 60 70 80 90 00 10
05
10
195
196
197
198
199
200
20
Tropical Cyclones per Year in the East Pacific
y 2530
All Tropical CyclonesHurricanes
29 Years2530
Freq
uenc
y
05
101520
Hurricanes
05
101520
1950
1960
1970
1980
1990
2000
2010
0
1950
1960
1970
1980
1990
2000
2010
0
Intense Hurricane Activity (H4 and H5)Intense‐Hurricane Activity (H4 and H5)
Intense Hurricanes1975‐1989 1990‐2004
N b P t N b P tNumber Percentage Number Percentage
North Atlantic 16 20 25 25
Source: Webster et al. (2005).
East Pacific 36 25 49 35
Intense Hurricanes in North Atlantic and East Pacific Basins
Percentage of Intense Hurricanes (5-Year Moving Average)
70%go
ry
50%
60%
urric
anes
/Cat
eg
40%
50%
erce
nt T
otal
Hu
20%
30%
Pe
10%
1975 1980 1985 1990 1995 2000 2005
0%
THE LOSS MODEL
Tropical Cyclones Considered in ERN Software
D t f N ti l O i d At h i• Data from National Oceanic and Atmospheric Administration (NOAA), US.
North Atlantic Data: since 24/June/1851– North Atlantic Data: since 24/June/1851.– East Pacific Data: since 10/June/1949.
• Only those that reached hurricane category• Only those that reached hurricane category (Saffir‐Simpson scale) at some point of their trajectory.j y
• Only those that get closer than 200 km from Mexican coasts.
• 269 hurricanes considered.
Gaussian Perturbations of the TrajectoriesGaussian Perturbations of the Trajectories
60
-120 -110 -100 -90 -80 -70 -60
4050
lat
304
200
10
WilmaSimulations
lon
240 250 260 270 280 290 300
Wilma Trajectory Yucatan Peninsula-92 -90 -88 -86 -84
Wilma Trajectory‐Yucatan Peninsula(Readings every six hours)
23
Wilma´s Trajectory 23/10/05 18:00 H2
122
22/10/05 12:00 H322/10/05 18:00 H2
23/10/05 0:00 H223/10/05 6:00 H2
23/10/05 12:00 H2
lat
202
21/10/05 6:00 H4
21/10/05 12:00 H421/10/05 18:00 H422/10/05 0:00 H4
22/10/05 6:00 H322/10/05 12:00 H3
YUCATAN PENINSULA
819
20/10/05 0 00 H420/10/05 6:00 H4
20/10/05 12:00 H420/10/05 18:00 H4
21/10/05 0:00 H4
21/10/05 6:00 H4YUCATAN PENINSULA
171
19/10/019/10/05 18:0
20/10/05 0:00 H4
lon
268 270 272 274 276
RESULTADOS DE LA EVALUACIÓN DE RIESGO HIDROMETEOROLÓGICO DE LA CARTERA
Fecha de creación:Octubre 25, 2007
Fecha de corte:Septiembre 30, 2007
Responsable:ERN, Evaluación de Riesgos Naturales
Archivos analizadosTipo 1 D:\HIDRO_PML\2007\0907\Eval00\Cartera Independientes.mdbTipo 2 D:\HIDRO_PML\2007\0907\Eval00\Cartera Colectiva.mdb
AsegurableSuma $ 294,034,316,768.20
Prima pura total $ 165,721,091.73 0.564al millarPrima pura total $ 165,721,091.73 0.564al millar
Prima pura devengada $ 89,908,029.04 0.306al millar
Prima pura no devengada $ 73,499,610.37 0.250al millar
PML $ 5,090,101,229.93 1.731%
RetenciónSuma $ 148,208,815,642.91 50.405% a retención
Prima pura total $ 133,122,681.25 0.453al millar
Prima pura devengada $ 68,415,630.00 0.233al millar
Prima pura no devengada $ 61,960,192.75 0.211al millar
PML $ 4,067,898,539.50 1.383%
PMLPeriodo de Retorno Total Retenidae odo de e o o o e e d
50 $ 1,145,466,254.99 0.39% $ 895,054,421.03 0.30%
100 $ 1,655,063,203.53 0.56% $ 1,306,720,915.40 0.44%
200 $ 2,310,906,381.48 0.79% $ 1,840,514,265.06 0.63%
500 $ 3 421 428 457 69 1 16% $ 2 735 722 879 08 0 93%500 $ 3,421,428,457.69 1.16% $ 2,735,722,879.08 0.93%
1,000 $ 4,436,054,144.60 1.51% $ 3,546,556,914.15 1.21%
1,500 $ 5,090,101,229.93 1.73% $ 4,067,898,539.50 1.38%
2,500 $ 5,966,994,230.56 2.03% $ 4,767,003,433.55 1.62%Análisis realizado con el sistema RH-Mex v1.0 desarrollado por ERN Evaluación de Riesgos Naturales
Additional Output (Fuentes File)Additional Output (Fuentes‐File)
Loss Statistics of Each Tropical Cyclone
Site MagnitudeExpected Loss
Expected Loss (%)
Occurrence Probability
Variation Coefficient
P0 P1 a b Exposure( ) y
1 0 1.08E+06 3.24E‐04 7.04E‐03 0 0E+00 0E+00 0.0541 166.9303 3.32E+09
2 0 2.27E+07 3.48E‐03 7.04E‐03 0 0E+00 0E+00 0.1450 41.4889 6.52E+09
3 0 7.86E+07 7.66E‐04 7.04E‐03 0 0E+00 0E+00 0.1439 187.7655 1.03E+113 0 7.86E 07 7.66E 04 7.04E 03 0 0E 00 0E 00 0.1439 187.7655 1.03E 11
4 0 1.42E+07 2.72E‐03 7.04E‐03 0 0E+00 0E+00 0.2415 88.3976 5.21E+09
5 0 8.40E+06 4.39E‐03 7.04E‐03 0 0E+00 0E+00 0.1277 28.9752 1.92E+09
6 0 4.54E+06 1.27E‐03 7.04E‐03 0 0E+00 0E+00 0.1906 150.1830 3.58E+09
7 0 1.56E+06 2.26E‐04 7.04E‐03 0 0E+00 0E+00 0.0544 240.8914 6.89E+09
Reinsurance SchemeReinsurance Scheme
Non‐Proportional Reinsurance Scheme
i i CReinstatement
l iLayers Priority CoverReinstatement
PremiumRol Reins
1 $ 7,500 $ 7,500 $ 1,586 21.1% 22 $ 15,000 $ 15,000 $ 1,890 12.6% 2$ , $ , $ ,3 $ 30,000 $ 30,000 $ 2,268 7.6% 14 $ 60,000 $ 40,000 $ 1,548 3.9% 15 $ 100,000 $ 130,000 $ 2,574 2.0% 1
(Monetary Quantities in Thousands of Dollars)
$ , $ , $ ,
10% Quota share before $ 7,500.
Simulation 1: Internal Model ResultsSimulation 1: Internal Model Results
Loss Summary: Simulation with Internal Model Results
Min Q1 Median Mean Q3 Max
Gross loss $ 0 $ 42 $ 720 $ 8,087 $ 5,381 $ 1,189,000
Net loss $ 0 $ 42 $ 720 $ 2,927 $ 5,142 $ 952,000
Net loss w.o. reins. $ 0 $ 42 $ 720 $ 2,399 $ 5,000 $ 944,100
R t ti 5% 100% 100% 88% 100% 100%Retention 5% 100% 100% 88% 100% 100%
(Monetary Quantities in Thousands of Dollars)
Our ModelOur ModelSimulation :
NOAA Data (1950‐2007)(217 tropical cyclones)
Tropical Cyclone FrequencyModel 1) Number of
occurrences by year.(217 tropical cyclones)
Tropical Cyclone IntensityModel
2) Intensity of each tropical cyclone.
3) Loss due to each tropical cyclone.
ERN Data(269 tropical cyclones)
4) Annual Insurer losses.
5) 150 000 yearsLoss Distribution
by Intensity(269 tropical cyclones) 5) 150,000 years.by Intensity
What We Did with NOAA DataWhat We Did with NOAA Data
d if hi h i l l hi i• Identify which Tropical Cyclones hit Mexico.– Between 1970 and 2007: Registries from Mexican National Meteorological Service.
– Before 1970: • Hit: If the tropical cyclone got closer than 100 km from Mexican coasts.C t M i t i t d t th i t• Category: Maximum category registered at those points in the trajectory that were closer than 100 km from Mexican coasts.
• 217 hits (1950‐2007).
Additional Information Requested to ERNAdditional Information Requested to ERN
Complementary Information to Fuentes‐File
# By Catalog # By Ocean Ocean Name Date Max Speed Max Category
91 91 Atl NOT NAMED 1.82E+04 212.7903 H492 92 Atl BAKER 1.85E+04 194.2867 H393 93 Atl ITEM 1.85E+04 175.7832 H294 94 Atl CHARLIE 1.89E+04 212.7903 H495 95 Atl HOW 1 89E+04 175 7832 H295 95 Atl HOW 1.89E+04 175.7832 H296 96 Atl FLORENCE 1.96E+04 203.5385 H397 97 Atl ALICE 19899 129.5245 H1
What We Did with ERN DataWhat We Did with ERN Data
Cl if h 269 i l l di h• Classify the 269 tropical cyclones according to the category they had when they hit Mexico.
Between 1970 and 2007: Registries from Mexican National– Between 1970 and 2007: Registries from Mexican National Meteorological Service.
– Before 1970:Before 1970: • Hit: If the tropical cyclone got closer than 100 km from Mexican coasts.
• Category: Maximum category registered at those points in the• Category: Maximum category registered at those points in the trajectory that were closer than 100 km from Mexican coasts.
• Only 184 hits.y• The remaining 85 tropical cyclones were considered as tropical storms (similar range of losses).
Classifications ComparisonClassifications ComparisonClassification of Hurricanes in Fuentes‐FileTS H1 H2 H3 H4 H5
ERN Classification
0 115 68 41 32 130% 43% 25% 15% 12% 5%124 81 35 13 11 5New
Classification124 81 35 13 11 546% 30% 13% 5% 4% 2%
Expected Loss by Category Expected Loss by Categoryp y g y(ERN Classification)
on U
SD
)
50
60
p y g y(New Classification)
on U
SD
)
50
60
d Lo
ss (M
illio
20
30
40
50
d Lo
ss (M
illio
20
30
40
50
Exp
ecte
d
H1
H2
H3
H4
H5
0
10
20E
xpec
ted
TS H1
H2
H3
H4
H5
0
10
20
Category
H H H H H
CategoryT H H H H H
Rank of Losses per BasinRank of Losses per BasinExpected Loss by Ocean and Categoryp y g y
(New Classification)
SD
)
6070
North AtlanticEast Pacific
ss (M
illio
n U
S
040
506 East Pacific
Exp
ecte
d Lo
s
1020
30
E
TS H1 H2 H3 H4 H5
0
Category
SCENARIOS
IPCC Criteria for Assessing Future Climate Scenarios
• Consistency at regional level with global projections.
Ph i l l ibilit d li• Physical plausibility and realism.
• Appropriateness of temporal and spatial scale.pp p p p
• Representativeness of the potential range of future
regional climate change.
• Accessibility of information• Accessibility of information.
Frequency Distributions FittedFrequency Distributions FittedDistribution Fitted to North Atlantic Data
(1950 2007)Distribution Fitted to East Pacific Data
(1950 2007) (1950-2007)
0.4PoissonTD-Gumbel
(1950-2007)
0.4PoissonTD-Gumbel
Mean = 1.15 Mean = 2.58
0.3 0.3
Den
sity
0.2
Den
sity
0.2
0.1 0.1
0.0 0.0
Tropical Cyclone Hits per Year
0 2 4 6 8
Tropical Cyclone Hits per Year
0 2 4 6 8
Intensity ModelIntensity Model
Hits by Category (1950 2007)Hits by Category (1950‐2007)TS H1 H2 H3 H4 H5
North Atlanti34 13 3 6 6 5
North Atlantic51% 19% 4% 9% 9% 7%
East Pacific82 46 12 5 4 155% 31% 8% 3% 3% 1%55% 31% 8% 3% 3% 1%
Simulation 2: Base ScenarioSimulation 2: Base Scenario
Loss Summary: Base Scenario
Min Q1 Median Mean Q3 Max
Gross loss $ 0 $ 61 $ 1,136 $ 1,910 $ 8,236 $ 1,027,000
Net loss $ 0 $ 6 $ 114 $ 1,682 $ 853 $ 780,000
Net loss w.o. reins. $ 0 $ 6 $ 114 $ 905 $ 750 $ 770,100
R t ti 0% 10% 10% 10% 10% 78%Retention 0% 10% 10% 10% 10% 78%
(Monetary Quantities in Thousands of Dollars)
Change in Mean FrequencyTropical Cyclones per Year in the North Atlantic
Change in Mean Frequency
quen
cy20
30
40 TotalHitsMeans
Period 1 Period 2
+19%Fr
eq
950
960
970
980
990
000
0
10+19%
‐14%
Year
19 19 19 19 19 20
Period 1: 1950‐1978Period 2: 1979‐2007
Tropical Cyclones per Year in the East Pacific
cy 30
40 TotalHits
Period 1 Period 2
Freq
uenc
0
10
20
30Means
+38%
‐23%
Year
1950
1960
1970
1980
1990
2000
0
Extreme Frequency ScenarioExtreme‐Frequency ScenarioFrequency Distribution (North Atlantic) Frequency Distribution (East Pacific)
0.4 19% mean increase
Base Scenario (lambda=1.15)Extreme Scenario (lambda=1.37)
0.4 38% mean increase
Base Scenario (lambda=2.58)Extreme Scenario (lambda=3.57)
0.3 0.3
Den
sity
0.2
Den
sity
0.2
0.1 0.1
0.0 0.0
Tropical Cyclone Hits per Year
0 2 4 6 8 10
Tropical Cyclone Hits per Year
0 2 4 6 8 10
Scenario ComparisonScenario ComparisonExtreme‐Frequency Scenario v s Base ScenarioExtreme‐Frequency Scenario v.s. Base Scenario
Min Q1 Median Mean Q3 Max
Gross lossBase $ 0 $ 61 $ 1,136 $ 11,910 $ 8,236 $ 1,027,000
Extreme $ 0 $ 259 $ 2 483 $ 15 330 $ 12 800 $ 1 285 000Gross loss Extreme $ 0 $ 259 $ 2,483 $ 15,330 $ 12,800 $ 1,285,000(+/‐ %) 0% 328% 119% 29% 55% 25%
Net lossBase $ 0 $ 6 $ 114 $ 1,682 $ 853 $ 780,000
Extreme $ 0 $ 26 $ 248 $ 2 153 $ 1 635 $ 1 064 000Net loss Extreme $ 0 $ 26 $ 248 $ 2,153 $ 1,635 $ 1,064,000(+/‐ %) 0% 328% 119% 28% 92% 36%
Net loss Base $ 0 $ 6 $ 114 $ 905 $ 750 $ 770,100
Extreme $ 0 $ 26 $ 248 $ 1 156 $ 770 $ 1 054 000w.o. reins.
Extreme $ 0 $ 26 $ 248 $ 1,156 $ 770 $ 1,054,000(+/‐ %) 0% 328% 119% 28% 3% 37%
RetentionBase 0% 10% 10% 10% 10% 78%
Extreme 0% 10% 10% 11% 10% 83%
(Monetary Quantities in Thousands of Dollars)
Retention Extreme 0% 10% 10% 11% 10% 83%(+/‐ pp) 0.00 0.00 0.00 0.30 0.45 4.75
Observed Intensity in 29 Year PeriodsObserved Intensity in 29‐Year PeriodsPercentage by Category (North Atlantic Total) Percentage by Category (North Atlantic Hits)
20%
30%
40%
50%
Period 1Period 2
30%
40%
50%Period 1Period 2
TS H1
H2
H3
H4
H5
0%
10%
20%
TS H1
H2
H3
H4
H5
0%
10%
20%
Percentage by Category (East Pacific Total) Percentage by Category (East Pacific Hits)
20%
30%
40%
50%
Period 1Period 2
30%
40%
50%Period 1Period 2
TS H1
H2
H3
H4
H5
0%
10%
20%
TS H1
H2
H3
H4
H5
0%
10%
20%
Period 1: 1950‐1978Period 2: 1979‐2007
Observed Intensity in 29 Year PeriodsObserved Intensity in 29‐Year Periods
Change in Category Proportions in the North AtlanticTotal Hits
P1 P2 +/‐ pp P1 P2 +/‐ ppTS 39% 45% 6.1 53% 48% ‐4.4H1 22% 25% 2 1 19% 19% 0 1H1 22% 25% 2.1 19% 19% ‐0.1H2 11% 9% ‐1.4 3% 6% 3.7H3 15% 9% ‐6.5 8% 10% 1.3H4 9% 9% ‐0.2 11% 6% ‐4.7H4 9% 9% 0.2 11% 6% 4.7H5 4% 4% ‐0.1 6% 10% 4.1
Change in Category Proportions in the East PacificTotal Hits
P1 P2 +/‐ pp P1 P2 +/‐ ppTS 52% 44% ‐7.8 61% 46% ‐15.0H1 30% 19% 11 6 32% 29% 2 5H1 30% 19% ‐11.6 32% 29% ‐2.5H2 6% 10% 3.9 1% 17% 15.7H3 5% 10% 5.1 2% 5% 2.3H4 6% 15% 9.2 2% 3% 0.7H5 1% 2% 1.2 1% 0% ‐1.2
Period 1: 1950‐1978Period 2: 1979‐2007
MaximumWind Speed RegisteredMaximum Wind Speed Registered Maximum Speed (North Atlantic Total 1950-2007) Maximum Speed (North Atlantic Hits 1950-2007)
ensi
ty
0.006
0.008
0.010
0.012
ensi
ty
0.006
0.008
0.010
0.012
De
50 100
150
200
250
300
0.000
0.002
0.004 De
50 100
150
200
250
300
0.000
0.002
0.004
Wind Speed (Km/h)
Maximum Speed (East Pacific Total 1950-2007)
Wind Speed (Km/h)
Maximum Speed (East Pacific Hits 1950-2007)
ensi
ty
0.006
0.008
0.010
0.012
ensi
ty0.006
0.008
0.010
0.012
De
50 100
150
200
250
300
0.000
0.002
0.004 De
50 100
150
200
250
300
0.000
0.002
0.004
Wind Speed (Km/h) Wind Speed (Km/h)
Fitted Generalized Pareto Distribution to Wind Speed (North Atlantic Total)
Probability Plot Quantile Plot
0.6
0.8
1.0
el 0025
030
0
cal
0.0
0.2
0.4
0
Mod
e
100
150
20
Em
piri
0.0 0.2 0.4 0.6 0.8 1.0
0
Empirical
100 150 200 250 300
Model
300
Return Level Plot
l
Density Plot
0.00
8
100
200
Ret
urn
leve
f(x)
00.
004
1
Return period (years)
0.1 1 10 100 1000
x
50 100 150 200 250 300
0.00
0
Fitted Generalized Pareto Distribution to Wind Speed (East Pacific Total)
Probability Plot Quantile Plot
0.6
0.8
1.0
el 200
250
300
cal
0.0
0.2
0.4
0
Mod
e
100
150
2
Em
piri
0.0 0.2 0.4 0.6 0.8 1.0
0
Empirical
100 150 200 250 300
Model
250
300
Return Level Plot
l
Density Plot
.008
100
150
200
2
Ret
urn
leve
f(x)
00.
004
0.
Return period (years)
0.1 1 10 100 1000
x
100 150 200 250 300
0.00
0
Extreme Intensity Scenario AExtreme‐Intensity Scenario A
• Proportions according to the Generalized Pareto Distribution fitted.
Extreme‐Intensity Scenario A v.s. Base ScenarioTS H1 H2 H3 H4 H5
Base 51% 19% 4% 9% 9% 7%North Atlantic
Base 51% 19% 4% 9% 9% 7%Extreme A 43% 21% 11% 11% 9% 5%
(+/‐ pp) ‐7.28 1.12 6.61 2.31 ‐0.01 ‐2.75Base 55% 31% 8% 3% 3% 1%
East PacificBase 55% 31% 8% 3% 3% 1%
Extreme A 49% 21% 11% 10% 7% 2%(+/‐ pp) ‐5.49 ‐9.45 2.64 6.56 4.05 1.69
Proportion of Intense Hurricanes is reduced!
Extreme Intensity Scenario BExtreme‐Intensity Scenario B
• East Pacific proportions as in Extreme‐Intensity Scenario A.
• Increase Intense Hurricanes in the North Atlantic in the same ti i E t P ifiproportion as in East Pacific.
Extreme‐Intensity Scenario B v.s. Base ScenarioTS H1 H2 H3 H4 H5
Base 51% 19% 4% 9% 9% 7%North Atlantic
Base 51% 19% 4% 9% 9% 7%Extreme B 49% 18% 3% 8% 13% 9%
(+/‐ pp) ‐1.44 ‐1.44 ‐1.44 ‐1.44 4.05 1.69Base 55% 31% 8% 3% 3% 1%
East PacificBase 55% 31% 8% 3% 3% 1%
Extreme B 49% 21% 11% 10% 7% 2%(+/‐ pp) ‐5.49 ‐9.45 2.64 6.56 4.05 1.69
Extreme Intensity Scenario CExtreme‐Intensity Scenario C
Extreme‐Intensity Scenario C v.s. Base ScenarioTS H1 H2 H3 H4 H5
North AtlanticBase 51% 19% 4% 9% 9% 7%
Extreme C 49% 18% 7% 7% 10% 8%/(+/‐ pp) ‐2.17 ‐0.96 2.88 ‐1.48 1.10 0.64
East PacificBase 55% 31% 8% 3% 3% 1%
Extreme C 49% 21% 11% 10% 7% 2%( / )(+/‐ pp) ‐5.49 ‐9.45 2.64 6.56 4.05 1.69
Scenario ComparisonScenario ComparisonExtreme‐Intensity Scenario A v.s. Base Scenario
Min Q1 Median Mean Q3 Max
Gross lossBase $ 0 $ 61 $ 1,136 $ 11,910 $ 8,236 $ 1,027,000
Extreme A $ 0 $ 92 $ 1,700 $ 15,134 $ 11,525 $ 1,319,968(+/‐ %) 0% 52% 50% 27% 40% 29%
Net lossBase $ 0 $ 6 $ 114 $ 1,682 $ 853 $ 780,000
Extreme A $ 0 $ 9 $ 170 $ 2,162 $ 1,421 $ 1,100,584(+/‐ %) 0% 52% 50% 29% 67% 41%
Net loss w o reins
Base $ 0 $ 6 $ 114 $ 905 $ 750 $ 770,100Extreme A $ 0 $ 9 $ 170 $ 1,171 $ 752 $ 1,090,718
w.o. reins.(+/‐ %) 0% 52% 50% 29% 0% 42%
RetentionBase 0% 10% 10% 10% 10% 78%
Extreme A 0% 10% 10% 11% 10% 83%(+/‐ pp) 0.00 0.00 0.00 0.10 0.00 5.33
(Monetary Quantities in Thousands of Dollars)
Scenario ComparisonScenario ComparisonExtreme‐Intensity Scenario B v.s. Base Scenario
Min Q1 Median Mean Q3 Max
Gross lossBase $ 0 $ 61 $ 1,136 $ 11,910 $ 8,236 $ 1,027,000
Extreme B $ 0 $ 92 $ 1,778 $ 17,036 $ 12,846 $ 1,237,688(+/‐ %) 0% 52% 56% 43% 56% 21%
Net lossBase $ 0 $ 6 $ 114 $ 1,682 $ 853 $ 780,000
Extreme B $ 0 $ 9 $ 178 $ 2,511 $ 1,684 $ 1,018,019(+/‐ %) 0% 52% 56% 49% 98% 31%
Net loss w o reins
Base $ 0 $ 6 $ 114 $ 905 $ 750 $ 770,100Extreme B $ 0 $ 9 $ 178 $ 1,416 $ 753 $ 1,008,153
w.o. reins.(+/‐ %) 0% 52% 56% 56% 0% 31%
RetentionBase 0% 10% 10% 10% 10% 78%
Extreme B 0% 10% 10% 11% 10% 82%(+/‐ pp) 0.00 0.00 0.00 0.13 0.05 4.20
(Monetary Quantities in Thousands of Dollars)
Scenario ComparisonScenario ComparisonExtreme‐Intensity Scenario C v.s. Base Scenario
Min Q1 Median Mean Q3 Max
Gross lossBase $ 0 $ 61 $ 1,136 $ 11,910 $ 8,236 $ 1,027,000
Extreme C $ 0 $ 90 $ 1,670 $ 15,958 $ 11,956 $ 1,349,655(+/‐ %) 0% 49% 47% 34% 45% 31%
Net lossBase $ 0 $ 6 $ 114 $ 1,682 $ 853 $ 780,000
Extreme C $ 0 $ 9 $ 167 $ 2,341 $ 1,506 $ 1,123,610(+/‐ %) 0% 49% 47% 39% 77% 44%
Net loss w o reins
Base $ 0 $ 6 $ 114 $ 905 $ 750 $ 770,100Extreme C $ 0 $ 9 $ 167 $ 1,306 $ 752 $ 1,113,744
w.o. reins.(+/‐ %) 0% 49% 47% 44% 0% 45%
RetentionBase 0% 10% 10% 10% 10% 78%
Extreme C 0% 10% 10% 11% 10% 83%(+/‐ pp) 0.00 0.00 0.00 0.11 0.00 5.20
(Monetary Quantities in Thousands of Dollars)
Scenarios ComparisonScenarios ComparisonExtreme Intensity Scenarios v s Base ScenarioExtreme‐Intensity Scenarios v.s. Base Scenario
Min Q1 Median Mean Q3 MaxBase $ 0 $ 61 $ 1,136 $ 11,910 $ 8,236 $ 1,027,000
Extreme A (+/ %) 0% 52% 50% 27% 40% 29%Gross loss
Extreme A (+/‐ %) 0% 52% 50% 27% 40% 29%Extreme B (+/‐ %) 0% 52% 56% 43% 56% 21%Extreme C (+/‐ %) 0% 49% 47% 34% 45% 31%
Base $ 0 $ 6 114 $ 1 682 $ 853 $ 780 000
Net loss
Base $ 0 $ 6 114 $ 1,682 $ 853 $ 780,000Extreme A (+/‐ %) 0% 52% 50% 29% 67% 41%Extreme B (+/‐ %) 0% 52% 56% 49% 98% 31%Extreme C (+/ %) 0% 49% 47% 39% 77% 44%Extreme C (+/‐ %) 0% 49% 47% 39% 77% 44%
Base $ 0 $ 6 $ 114 $ 905 $ 750 $ 770,100
Net loss Extreme A (+/‐ %) 0% 52% 50% 29% 0% 42%Extreme B (+/‐ %) 0% 52% 56% 56% 0% 31%
w.o. reins.Extreme B (+/‐ %) 0% 52% 56% 56% 0% 31%Extreme C (+/‐ %) 0% 49% 47% 44% 0% 45%
(Monetary Quantities in Thousands of Dollars)
Scenario ComparisonScenario ComparisonExtreme Frequency & Extreme Intensity (B) Scenario v s Base ScenarioExtreme‐Frequency & Extreme‐Intensity (B) Scenario v.s. Base Scenario
Min Q1 Median Mean Q3 Max
Gross lossBase $ 0 $ 61 $ 1,136 $ 11,910 $ 8,236 $ 1,027,000
Extreme $ 0 $ 400 $ 3 888 $ 21 770 $ 19 326 $ 1 197 047Gross loss Extreme $ 0 $ 400 $ 3,888 $ 21,770 $ 19,326 $ 1,197,047(+/‐ %) 0% 561% 242% 83% 135% 17%
Net lossBase $ 0 $ 6 $ 114 $ 1,682 $ 853 $ 780,000
Extreme $ 0 $ 40 $ 389 $ 3 238 $ 2 753 $ 967 474Net loss Extreme $ 0 $ 40 $ 389 $ 3,238 $ 2,753 $ 967,474(+/‐ %) 0% 561% 242% 92% 223% 24%
Net loss Base $ 0 $ 6 $ 114 $ 905 $ 750 $ 770,100
Extreme $ 0 $ 40 $ 389 $ 1 857 $ 833 $ 956 750w.o. reins.
Extreme $ 0 $ 40 $ 389 $ 1,857 $ 833 $ 956,750(+/‐ %) 0% 561% 242% 105% 11% 24%
RetentionBase 0% 10% 10% 10% 10% 78%
Extreme 0% 10% 10% 11% 11% 81%Retention Extreme 0% 10% 10% 11% 11% 81%(+/‐ pp) 0.00 0.00 0.00 0.48 1.48 3.03
(Monetary Quantities in Thousands of Dollars)
ReferencesReferences
( )• Agata, H. (2007), El Cambio Climático: Repercusiones Económicas para el Seguro, XIX Seminario Internacional de Seguros y Fianzas de la CNSF.la CNSF.
• Bernardo, S.M. and Smith, A. F. M. (1994), Bayesian Theory, Wiley and Sons.
• Clark, K. M. (2002), The Use of Computer Modelling in Estimating and Managing Future Catastrophe Losses, Geneva Papers on Risk and Insurance 27: 181 195and Insurance, 27: 181‐195.
• Coles, S. and Casson, E. (1998), Extreme Value Modelling of Hurricane Wind Speeds. Structural Safety, Vol. 20, 283‐296.p f y, ,
• Eeckhoudt, L., Gollier, C. and Lambson, V. (1996), Risk: Evaluation, Management and Sharing, Harvester Wheatsheaf.
ReferencesReferences
( )• ERN (2007), Estudio Metodológico y Sistema de Cómputo Sobre la Estimación del Valor de las Pérdidas Originadas por EventosHidrometeorológicos. Technical Note.Hidrometeorológicos. Technical Note.
• Gilleland, E. and Katz, R.W. (2006), Analyzing seasonal to interannual extreme weather and climate variability with the extremes toolkit (extRemes), 18th Conference on Climate Variability and Change, 86th American Meteorological Society (AMS) Annual Meeting Atlanta Georgia P2 15 (contributed poster)Meeting, Atlanta, Georgia. P2.15 (contributed poster).
• Goldberg, S.B., Landsea, C.W., Mestas‐Nuez, A., Gray, W.M. (2001), The Recent Increase in Atlantic Hurricane Activity: Causes and Implications, Science Vol. 293, 474‐479.
ReferencesReferences
( )• McNeil, A.J., Frey, R. and Embrechts, P. (2005), Quantitative Risk Management. Concepts, Techniques and Tools, Princeton University Press, Princeton.Press, Princeton.
• Verlaak, R. and Beirlant, J. (2003), Optimal reinsurance programs: An optimal combination of several reinsurance protections on a heterogeneous insurance portfolio, Insurance: Mathematics and Economics 33,2, 381‐403.
• Woo G (1999) The Mathematics of Natural Catastrophes Imperial• Woo, G. (1999), The Mathematics of Natural Catastrophes, Imperial College Press.