Disaster Risk Management and Financing System, and Corresponding
Method Thereof
Field of the Invention
The present invention relates to a disaster management and financing
system for forecasting the impact of disaster mitigation a s well a s and automated
signaled and applied disaster financing and mitigation means based on locationdependent
natural disaster impacts. It especially relates to automated, computerbased
disaster risk management and financing systems.
Background of the Invention
Within the last decade, natural disasters have had devastating impacts o n
the socio-economic and environmental landscape, mainly of developing countries
and emerging market countries, including the so called BRIC countries (Brazil, Russia,
India and China) and MIKT countries (Mexico, Indonesia, South Korea and Turkey) a s
the largest representatives of the latter. For example in the Caribbean, on average, six
natural disasters occurred in the region annually between 1970 and 2006, with higher
incidences in Haiti and the Dominican Republic. The active hurricane season of 2004
resulted in damages in the Caribbean amounting to USD 3.1 billion, with catastrophic
impacts o n the gross domestic product (GDP) of member countries, particularly
Grenada (200 percent of GDP). Similarly, Hurricane Dean in 2007 had a major
destructive impact on the economies of Belize, Jamaica, and Saint Lucia.
Approximately 14 percent of the Saint Lucian population was affected, including 47
percent of the vulnerable community, with costs to the Jamaican and Belizean
economies amounting to USD 329.34 million and USD 89.1 million, respectively.
According to the Centre for Research on the Epidemiology of Disasters, damages from
natural disasters in 2010 showed a different distribution than that seen for previous
events (Source FAO (Food and Agriculture Organization of the United Nations),
February 2013, Status of Disaster Risk Management). The major share of global
damages (45.9 percent) was attributed to the 12 January 2010 earthquake in Haiti. It is
well known that these shocks may cause spillovers at the macroeconomic level, since
fiscal and external pressures can lead to imbalances that spark economic crisis and a n
increased incidence of poverty (Source IMF (International Monetary Fund) Working
Paper WP/04/224).
Although such catastrophic events d o often cause more grievous harm to
the economy and social life of poorer countries, even industrially developed countries
are not immune to the destructive impact of these events on the country's economy.
Overall in 201 3, there were 308 major disaster events, of which 150 were natural
catastrophes and 158 man-made (Source sigma 1/2014, SwissRe). Almost 26,000 people
lost their lives or went missing in the disasters. Typhoon Haiyan struck the Philippines in
November 201 3, one of the strongest typhoons ever recorded worldwide. It killed
around 7,500 people and left more than 4 million homeless. Haiyan was the largest
humanitarian catastrophe of 201 3. The next most extreme in terms of human cost was
the June flooding in the Himalayan state of Uttarakhand in India, in which around 6,000
died. Figure 2 shows the number of victims from the years 1970 to 201 3. The reference
number 1 denotes the Bangladesh storm of 1970, 2 denotes the Tangshan earthquake
in China of 1976, 3 denotes Cyclone Gorky of Bangladesh in 1991 , 4 denotes the Indian
Ocean earthquake and tsunami of 2004, 5 denotes Cyclone Nargis in Myanmar of
2008, 6 denotes the Haiti earthquake of 2010, and 7 denotes Typhoon Haiyan in the
Philippines of 201 3. In Figure 2, the scale is logarithmic, i.e. the number of victims
increases tenfold per band (Source: Swiss Re Economic Research & Consulting). The
total economic losses from natural catastrophes and man-made disasters were around
USD 140 billion last year. That was down from USD 196 billion in 201 2 and well below the
inflation-adjusted 10-year average of USD 190 billion. Asia was hardest hit, with the
cyclones in the Pacific generating most economic losses. Most of the remainder was
caused by weather events in North America and Europe. Figure 1 shows the number of
major catastrophic events from 1970 to 2013 (Source SwissRe Economic Research &
Consulting). Catastrophe losses in 201 3 were equivalent to 0.19% of GDP, also below the
10-year average of 0.30%. Natural catastrophe-related losses were around USD 131
billion in 2013, stemming mostly from floods and other extreme weather events in Asia,
North America, and Europe. Man-made disasters are estimated to have caused more
than USD 9 billion of the total USD 140 billion damages in 2013, up from USD 8 billion in
2012. Table 1 shows the economic losses of 2013 in percentage of the GDP.
Country In USD bn In % of GDP
North America 32 0.1 7%
Latin America & Caribbean 9 0.1 6%
Europe 33 0.1 5%
Africa 1 0.05%
Asia 62 0.26%
Oceania/Australia 3 0.1 6%
Seas / Space 1
Total 140* 0.1 9%
10-year average 190** 0.30%
(Table 1: The economic losses of 20 13 in percentage of the GDP. * denotes rounded numbers
and ** inflation adjusted values (Source sigma 1/20 4 SwissRej)
However, the continental numbers do not show the individual burden of the
concerned countries, which can have massive and severe impacts on a country and its
government due to multiple overruns of the country's GDP (Gross Domestic Product).
Table 2 shows the major disasters of the last 40 years in percentage of the GDP of the
concerned country in the year of the event.
Year Event Country Economic In % of Victims
Losses in GDP
USD millions
2005 Hurricane Katrina US, Gulf of Mexico, 140,000 1. 1% 1,836
Bahamas, North America
2008 Earthquake China 124,578 2.8% 87,449
1995 Great Hanshin Japan 82,399 1.6% 6,425
earthquake
201 0 Floods, mudslides China 53, 113 0.9% 2,480
2008 Hurricane Ike US, Gulf of Mexico, Turks 40,000 0.3% 136
and Caicos Islands, Haiti,
Cuba, Bahamas,
Dominica Republic
201 0 Earthquake Chile 30,000 15. 1% 521
1998 Flooding along China 30,000 3.0% 3,656
Yangtze River
1994 Northridge United States 30,000 0.4% 6 1
earthquake
2004 Chuetsu Japan 29,276 0.6% 39
earthquake
1992 Hurricane Andrew United States 26,500 0.4% 43
2004 Hurricane Ivan United States 22,000 0.2% 124
1999 Earthquake Turkey 20,000 8.0% 19, 118
2008 Snow storms China 20,000 0.4% 130
2005 Hurricane Wilma United States 20,000 0.2% 35
1995 Drought in China 19,669 2.7% 0
Northeastern
China
2008 Hurricane Gustav United States 17,500 0.1 % 135
2004 Hurricane Charley United States 16,000 0.1 % 24
201 0 Wild fires Russia 15,000 1.0% 50
2005 Hurricane Rita United States 15,000 0.1 % 34
201 0 Earthquake Haiti 8,000 114% 220,000
1988 Hurricane Gilbert St Lucia 1,000 386% 341
2004 Hurricane Ivan Grenada 889 203% 124
1991 Cyclones Val and Samoa 278 248% 14
Wasa
1990 Cyclone Ofa Samoa 200 178% 8
1985 Cyclones Eric and Vanuatu 173 143% 25
Nigel
2009 Tsunami Samoa 120 22% 1 9
(Table 2: The major disasters of the last 40 years in percentage of the GDP o f the concerned
country in the year of the event (Source: Swiss Re, Closing the financial gap)
In the above-mentioned example of the Caribbean, the problems related
to disaster risk management and financing can be easily illustrated by the agriculture
sector. The agriculture sector can be subject to different types of hazards, including
cyclones, floods, and droughts. Looking back a t the last 40 years to determine the top
10 natural disasters in terms of loss of life, total number of people affected, and
economic losses, one can see that cyclones often pose the largest threat to human life
and cause the highest economic losses. This applies also for the agriculture sector.
However depending on the country, other hazards might be more significant (droughts
for Africa etc.). In the example of the Caribbean, the regional agriculture sector
continues to be severely undermined a s a result of natural disasters. Hurricane Ivan in
2004 decimated Grenada's agriculture sector and accrued losses in excess of USD 37
million. Ivan destroyed the entire banana industry and approximately 40 percent of
mature cocoa trees of the country. Almost all of the nutmeg trees toppled (90 percent),
with significant negative implications for the local rural economy (Source OECS 2004,
Grenada - Macro-socio-economic assessment of the damages caused by Hurricane
Ivan). Total annual average revenue available to farmers decreased by 89.9 percent,
from USD 18.7 million during 2002-2004 to USD 1.9 million after the disaster (2005-2009)
(Source ITC, July 2010, European Union All ACP Commodities Program, WTO (World
Trade Organization). Similarly, in 2007, Hurricane Dean ravaged Caribbean agricultural
productivity. Jamaica reported damages of approximately USD 43 million. Overall,
56,537 crop farmers and 7,1 70 livestock farmers were seriously affected, with the
greatest impact being among small farms. Belize's agriculture sector documented
damage and loss of USD 54 million, with the majority of costs recorded in the cropping
subsector (90.6 percent). Saint Lucia's agriculture sector reported losses of roughly USD
10 million, with the banana industry accounting for 67 percent of the overall burden of
the sector (USD 6.7 million). The Economic Commission for Latin America and the
Caribbean (ECLAC) posits that Hurricane Dean will have serious implications for future
banana production in Saint Lucia and predicted a reduction in banana exports of USD
5.7 million up to February 2008. Moreover, a Crop and Food Security Assessment Mission
conducted by the Food and Agriculture Organization of the United Nations (FAO) in
Haiti in September 2010 highlighted a decrease in the production of cereals (by 9
percent), legumes (by 20 percent), root crops (by 12 percent), and plantain (by 14
percent) when compared to previous years. Although the earthquake was largely a n
urban event, its effects resounded throughout the rural agricultural areas (Source FAO
(Food and Agriculture Organization of the United Nations), February 2013, Status of
Disaster Risk Management).
A case study of the 2009-2010 El Nino-induced Caribbean drought reported
startling impacts o n the region's agriculture sector (Source FAO (Food and Agriculture
Organization of the United Nations), February 2013, Status of Disaster Risk
Management). Vast amounts of finances were spent by the governments to mitigate
the impacts of the drought. In Guyana, the Government allocated USD 1.3 million to
bring relief to farmers in a first region in February 2010 and spent USD 16,000 a day in
another region to operate pumps and perform other work essential to water delivery.
The banana industry in Dominica reported a 43 percent reduction in production in 2010
compared to previous years. Similarly, the 2010 onion and tomato crops in Antigua and
Barbuda decreased by 25 percent and 30 percent, respectively, due to water-stressed
conditions. Saint Vincent and the Grenadines documented a 20 percent overall
decrease in agricultural productivity during the period. Impacts of the drought were
also reflected to some extent in commodity prices. Tomato prices in Saint Vincent and
the Grenadines rose by 155 percent during the peak of the drought (February-March
2010). The Central Bank of Trinidad and Tobago reported a n increase in the price of
fruits in March 2010 by 20.1 percent when compared to February of the same year.
According to the report, the drought-induced bush fires destroyed many acres of citrus
farms in the two-island republic, resulting in a n increase in the cost of citrus importation
from USD 6.3 million in 2008 to USD 8.3 million by the end of 2010. The study emphasized
that it is imperative that the concerned countries mainstream their forecasting and
alerting systems for drought and for them to develop and implement cost-effective
policies for adapting to and mitigating drought-related impacts.
There is a n urgent need to integrate, automate, and synchronize disaster
risk management (DRM) in governmental risk management through appropriate
systems, allowing a controllable, reproducible, and easily applicable monitoring and
risk transfer/balancing. As mentioned, natural hazards in many countries have a
substantial potential to cause large losses to crops and infrastructure, a great potential
to negatively affect economic and macroeconomic performance, and even have the
potential to destabilize economies on a global scale. For the agriculture sector, the
ό
effects are even more critical in light of the projected impacts of climate change and
variability o n smaller developing states, the peculiar vulnerabilities of these states, and
the moderate to high poverty levels of most of such states (cf. Baas, S. et al., Disaster risk
management systems analysis, 2008). In fact, many disaster-related losses can be
avoided or reduced if appropriate policies and mitigation instruments are implemented
to address the root causes of vulnerability, while also integrating mitigation,
preparedness, and response mechanisms into overall development planning. The
development of sectoral DRM plans for agricultural and other sectors a t the national
level therefore represents a powerful strategy for increasing resilience to natural hazards
and forging a sustainable path to development.
However, natural catastrophes are rare events, which are typically not
subject to the statistics of big numbers. Their occurrence is subject to high fluctuations
that are impossible to forecast in the long term. Hurricanes, cyclones, and typhoons
often show the highest annual rate of recurrence. In many affected countries, this
exceeds 0.7 events per year over a 20-year time span. Note that hurricanes, cyclones,
and typhoons are all the same weather phenomenon. The different names for these
storms are specific to their location. In the Atlantic and Northeast Pacific, the term
"hurricane" is used. The same type of disturbance in the Northwest Pacific is called a
"typhoon," while "cyclones" occur in the South Pacific and Indian Ocean. For this
application, the terms are used a s synonyms describing the same natural hazard
phenomenon. Droughts and floods are typically less recurrent. It is important to note
that due to the limited time horizon, such numbers are often only indicative and cannot
be used for probabilistic risk assessment approaches, i.e. the prior art system and
methods for risk assessment cannot be applied or can only be applied with great
reservations. Further, the interactions among possible instrumental steps for mitigating
the consequences of a catastrophic risk event and for providing more resilient
governance are difficult to understand, judging from the results they achieve when a
disaster event occurs. Moreover, it is almost impossible to acquire indepth experience
a s a person responsible for applying possible mitigation means. Therefore, in order to
provide a better understanding of the possible instrumental means and the
effectiveness thereof, it is important to provide a n automated system for disaster risk
management (DRM) and disaster risk financing (DRF) taking over the role of the Country
Risk Officer (CRO) of a country. It is also important to provide a system for testing a
developed disaster strategy for different perils such a s earthquakes, hurricanes,
typhoons, droughts, and/or floods. The system should allow for improving a country risk
profile of a specific country, expanding and improving test sets based on the present
basic data of a country, self-analyzing trial runs on simulation effectiveness, and
developing a n appropriate electronic automated system.
Summary of the Invention
One object of the present invention is to provide a n automated, selfadjustable
system and method for enabling a better understanding of the effect of
operational adjustments to possible instrumental means and the effectiveness thereof.
Another object of the present invention is to provide a n automated system for disaster
risk management taking over the role of the Country Risk Officer of a country. Another
important task is to provide a system for testing a developed disaster strategy for
different perils such a s earthquakes, hurricanes, typhoons, droughts, and/or floods in a
country-specific setting. The system should allow for improving a country risk profile of a
specific country, expanding and improving test sets based o n the present basic data of
a country, self-analyzing trial runs on simulation effectiveness, and developing a n
appropriate electronic automated system. Finally, the system should provide the
possibility of generating a disaster risk management and strategy thereof for natural
perils to which a specific country is actually exposed, linking temporal, topological,
geographical, social, and population structures of a country.
According to the present invention, these objects are attained in
particularly by the features of the independent claims. In addition, further
advantageous embodiments can be derived from the dependent claims and related
descriptions.
The above-mentioned objects related to the disaster risk management and
disaster risk financing systems for forecasting the impact of disaster mitigation and
financing means based o n location-dependent natural disaster impacts are attained
according to the present invention particularly in that measuring parameters of
historical disaster events are captured in order to determine the impact of natural
disaster events and critical values of parameters of natural disaster events are used a s
triggers in order to generate forecasts of the impacts of disaster events within a
geographic area; in that country-specific parameters of a risk-exposed country are
captured, relating to stored predefined criteria, wherein the country-specific
parameters comprise at least national economic and national budgetary parameters;
in that one or more disaster event types are assigned to a disaster history table, wherein
each disaster event type comprises a plurality of type-specific measuring parameters of
historical natural disaster events and associated type-specific loss frequency function
parameters that provide a corresponding loss frequency function for each natural
disaster event type, and wherein the magnitude of a loss to its expected exceedance
frequency is parameterized by means of the loss frequency function, where the
exceedance frequency is a measure of the annual probability that a n event or loss will
meet or exceed a given magnitude in any given timeframe; in that the system
comprises mapping parameters for capturing and storing a geographic risk map,
wherein for each of the natural disaster event types, corresponding mapping
parameters are captured and stored, which define danger zones for the specific
natural disaster event type; in that the system comprises a plurality of selectable disaster
financing means, wherein each of the selectable disaster financing means is assigned
to a definable cost factor capturing the capital cost of the financing means in relation
to its application for disaster mitigation, and wherein for each of the selectable disaster
financing means, a variable budgetary share factor can be allocated and adapted by
means of an allocation module defining the coverage structure in case of a
catastrophic disaster event; and in that expected catastrophe losses are determined
by means of the loss frequency function and the geographic risk map for various
scenarios of occurring natural disaster event types and a forecast of the effect of the
disaster financing means to cover these losses is prepared based on the coverage
structure, the assigned cost factors, and the determined expected catastrophe losses.
A first disaster financing means can for example be related to a contingency reserves
unit comprising an assigned cost factor set to 1, a second selectable disaster financing
means is related to a contingent debt facility unit comprising a n assigned cost factor
depending on definable credit condition parameters, and a third selectable disaster
financing means is related to a n insurance facility unit comprising an assigned cost
factor set e.g. to 1.7, a factor e.g. based on current market benchmarks. Further, based
on the disaster history table comprising the stored natural disaster event types, at least
four loss frequency curves capturing the perils of hurricanes, floods, earthquakes and
droughts can e.g. be generated together with the corresponding mapping parameters
of the geographic risk map. The system can e.g. comprise a t least country-specific
parameters related to population and/or demographics and/or gross domestic
product and/or sovereign budget and/or inflation rate and/or economic structure
and/or export/import values. Finally, the expected catastrophe losses can e.g. be
determined through numerical integration of the loss frequency curves. The present
invention advantageously provides a system for forecasting expected catastrophe
losses and the effect of the corresponding financing tools the user chooses to cover
these losses. The invention also advantageously provides a system for setting up a
coverage structure which satisfies country-specific needs and for testing the
performance of a specific, defined scheme of disaster mitigation and financing means
for various scenarios in real time.
In one embodied variant, the system comprises a second MonteCarlo
module for generating a probabilistic Monte Carlo loss simulation for a probabilistic
multi-year simulation a s a final test of the effectiveness of a chosen coverage structure
for a specific pre-financing scheme. The MonteCarlo module can e.g. generate the
probabilistic Monte Carlo loss simulation for a probabilistic 30-year simulation. This
embodied variant has, among other things, the advantage that the system allows a
complete monitoring and assessment of the impact of a chosen coverage structure
under different scenarios, in real time.
In a n other embodied variant, the system comprises three selectable input
channels, wherein in a first channel selectable by means of a user interface, a first
budgetary share factor is determined and assigned to the corresponding first disaster
financing means, in a second channel selectable by means of a user interface, a
second budgetary share factor is determined and assigned to the corresponding
second disaster financing means, and in a third channel selectable by means of a user
interface, a third budgetary share factor is determined and assigned to the
corresponding third disaster financing means. The budgetary share factors of the
coverage structure are varied by means of the user interface in order to optimize the
effect of the disaster financing means to cover possible losses. Furthermore, the
allocation module can e.g. comprise a n activating device, by means of which, based
on the generated coverage structure with the allocated budgetary share factors, it is
possible to transmit a corresponding control signal to the monitoring device. This
embodied variant has, among other things, the advantage that the system can
provide a user with the experience of a Country Risk Officer (CRO) by looking a t a
nation's risk profile and creating a n appropriate risk management plan to be tested
through realistic scenarios.
In a further embodied variant, the allocation module comprises a second
Monte Carlo module, wherein by means of the second Monte Carlo module and
based o n the allocated variable budgetary share factors of the coverage structure, a
plurality of data records comprising coverage structures with varied budgetary share
factors are generated, wherein the coverage structure with the allocated budgetary
share factors is optimized by means of the system based on the effect of the disaster
financing means for various scenarios of occurring natural disaster event types. This
embodied variant has, among other things, the advantage that the system automated
generates a n optimized coverage structure with optimized allocated budgetary share
factors.
In still another embodied variant, the allocation module comprises a
signaling device, wherein the selectable disaster financing means are activated based
on the allocated budgetary share factors by means of signal transmission. Furthermore,
the allocation module can e.g. comprise a signaling device, wherein upon triggering
a n optimized coverage structure, the selectable disaster financing means are
activated based on the allocated budgetary share factors by means of signal
transmission. This embodied variant has, among other things, the advantage that the
system can be fully automated, i.e. that the system automatically generates and
optimizes a coverage structure with optimized allocated budgetary share factors, and
also automatically activates the appropriate disaster financing means to provide the
disaster risk management. Furthermore, the system allows efficient and fully automated
monitoring and control of the work of the Country Risk Officer in the testing of possibly
proposed risk management plans through the use of realistic scenarios.
Finally, in addition to the system described above and the corresponding
method, the present invention also relates to a computer program product that
includes computer program code means for controlling one or more processors of the
control system in such a manner that the control system performs the proposed
method; the invention also relates, in particular, to a computer program product that
includes a computer-readable medium containing therein the computer program
code means for the processors.
Brief Description of the Drawings
The present invention will be explained in more detail by way of example in
reference to the drawings in which:
Figure 1 shows diagrams schematically illustrating the number of major
catastrophic events from 1970 to 201 3.
Figure 2 shows a diagram schematically illustrating the number of victims
from the years 1970 to 2013. The reference number 1 denotes the Bangladesh storm of
1970, 2 denotes the Tangshan earthquake in China of 1976, 3 denotes Cyclone Gorky
of Bangladesh in 1991 , 4 denotes the Indian Ocean earthquake and tsunami of 2004, 5
denotes Cyclone Nargis in Myanmar of 2008, 6 denotes the Haiti earthquake of 2010,
and 7 denotes Typhoon Haiyan in the Philippines of 2013.
Figure 3 shows a block diagram schematically depicting a n architecture for
a possible implementation of a n embodiment of the automated disaster management
and financing system 1 for forecasting the impact of disaster mitigation and financing
means 30 based o n location-dependent natural disaster impacts.
Figure 4 shows a n example of loss frequency functions 103 for a fictitious
country for four disaster event types 101 : hurricane 1031 , flood 1032, earthquake 1033,
and droughts 1034 generated based o n the disaster history of a fictitious country and
respectively based on the disaster history table 10.
Figure 5 shows four exemplary risk maps 20 provided by the abovementioned
exemplary numbers of the fictitious country for the four disaster event types
101 : hurricane risk map 2001 , flood risk map 2002, earthquake risk map 2003, and
drought risk map 2004.
Detailed Description of the Preferred Embodiments
Figure 3 schematically depicts a n architecture for a possible
implementation of a n embodiment of the automated disaster management and
management forecast system 1 for forecasting the impact of disaster mitigation and
financing means based o n location-dependent natural disaster impacts. For the
computer-based disaster management and management forecast system 1,
measuring parameters of historical disaster events are captured in order to determine
the impact of natural disaster events and then critical values of parameters of natural
disaster events are used a s triggers in order to generate forecasts of the impacts of
disaster events within a geographic area 501 , 5 11, 521 , 531 ....
Country-specific parameters 121 1, 1212, 12 13 of a risk-exposed country
501 531 are captured, relating to stored predefined criteria 1221 , 1222, 1223. The
country-specific parameters 121 1, 1212, 1213 can comprise at least national economic
and national budgetary parameters. The country-specific parameters 12 11, 1212, 1213
of a risk-exposed country 501 531 provide a country risk profile 12 1 giving a countryspecific
risk and structure scheme. The system 1 can e.g. comprise a t least countryspecific,
predefined criteria 12 11, 1212, 1213 for country-specific parameters 122
related to population 1221 and/or demographic 1222 and/or gross domestic product
1223 and/or sovereign budget 1224 and/or inflation rate 1225 and/or economic
structure 1226 and/or export/import values 1227. The country-specific parameters 121 1,
1212, 12 13 of a risk-exposed country 501 531 can e.g. be captured by means of a
user interface 90. The user interface 90 can e.g. comprise a first channel 901 selectable
by means of the user interface 90, in which a first budgetary share factor 4 11 can be
determined and assigned to the corresponding first disaster financing means 301 by a
user or by a connected input device. In a second channel 902 selectable by means of
the user interface 90, a second budgetary share factor 4 12 can e.g. be determined
and assigned to the corresponding second disaster financing means 302. Finally, in a
third channel 903 selectable by means of the user interface 90, a third budgetary share
factor 413 is determined and assigned to the corresponding third disaster financing
means 303. Table 3 below shows a n example of a country profile 12 1 with the countryspecific
parameter 121 1, 1212, 1213 values, which are masked or defined based on the
parameter criteria (references) 1221 , 1222, 1223.
(Table 3: Example of a country risk profile 12 1with exemplary values of country-specific
parameters 2 , 2 2 , 2 3 defined based on the parameter criteria 122}
One or more disaster event types 101 are assigned to a disaster history table
10. Each disaster event type 101 comprises a plurality of type-specific measuring
parameters of historical natural disaster events and associated type-specific loss
frequency function parameters 102, providing for each natural disaster event type 101
a corresponding loss frequency function 103. The magnitude of a loss to its expected
exceedance frequency is parameterized by means of the loss frequency function 103,
where the exceedance frequency is a measure of the annual probability that a n event
or loss will meet or exceed a given magnitude in any given timeframe. For example, a t
least four loss frequency functions 103 capturing the perils of hurricanes 1031 , floods
1032, earthquakes 1033, and droughts 1034 can be generated together with the
corresponding mapping parameters 201 1, 201 2, 201 3, 201 4 of the geographic risk map
20, based on the disaster history table 10 comprising the stored natural disaster event
types 101 . The expected catastrophe losses can e.g. be determined through numerical
integration of the loss frequency function 103. Each country typically has a countryspecific
disaster history. For example a fictitious country can face several natural
disasters each year, with the most extreme events in terms of economic damage and
affected population for example being earthquakes, hurricanes, floods, and droughts.
A disaster history of such a fictitious country from 1950 - 201 1 for the aforementioned
four major perils could e.g. look a s follows in Table 4 below.
(Table 4: Example of a disaster history o f a fictitious country from 1950 - 20 for the four major
perils o f (i) earthquakes, (iij hurricanes, {Hi} floods, and (ivj droughts as entered in the disaster
history table 0.
In order to quantify how much a country is threatened by each of the
various disaster event types 101 , i.e. perils (for the fictitious country above earthquakes,
hurricanes, floods, and droughts), the implemented assessment of the risks of a country
provides the appropriate loss frequency curves 103 based o n the disaster history. The
loss frequency curve 103 relates the magnitude of a loss relative to its expected
exceedance frequency, where the exceedance frequency is the annual probability
that a n event or loss will meet or exceed a given magnitude in any given year. Figure 4
shows a n example of loss frequency functions 103 for a fictitious country for the four
disaster event types 101 : hurricane loss frequency function 1031 , flood loss frequency
function 1032, earthquake loss frequency function 1033, and drought loss frequency
function 1034 generated based on the disaster history of a fictitious country and
respectively based o n the disaster history table 10.
The system 1 comprises mapping parameters 201 for capturing and storing
a geographic risk map 20. For each of the natural disaster event types 101 ,
corresponding mapping parameters 201 are captured and stored, which define
danger zones for the specific natural disaster event type 101 . The mapping parameters
201 of the geographic risk map 20 can be displayed a s graphical risk maps defining
danger zones for the different disaster event types 101 . Figure 5 shows four exemplary
geographic risk maps 20 provided by the above-mentioned exemplary numbers of the
fictitious country for the four disaster event types 101 : geographic hurricane risk maps
2001 , geographic flood risk maps 2002, earthquke risk maps 2003, and drought risk maps
2004. The geographic risk map(s) 20 can comprise or be build up by the risk maps 20 for
the different perils, a s for example the mentioned geographic hurricane risk maps 2001 ,
geographic flood risk maps 2002, eathquke risk maps 2003, and drought risk maps 2004.
The system 1 comprises a plurality of selectable disaster financing means 30.
Each of the selectable disaster financing means 30 is assigned to a definable cost
factor 301 , 302, 303 capturing the capital cost of the disaster financing means 30 in
relation to its application for disaster mitigation. For each of the selectable disaster
financing means 30, a variable budgetary share factor 4 10, 4 11, 4 12, 4 13 can be
allocated and adapted by means of a n allocation module 40 defining a coverage
structure 401 in case of a catastrophic disaster event. For example, a first disaster
financing means 30 can be related to a contingency reserves unit 3 1 1 comprising a n
assigned cost factor set to 1, a second selectable disaster financing means 30 can be
related to a contingent debt facility unit 3 12 comprising a n assigned cost factor
depending on definable credit condition parameters, and a third selectable disaster
financing means 30 can be related to a n insurance facility unit 313 comprising a n
assigned cost factor set to e.g. 1.7, which factor e.g. can be based on current market
benchmarks. For example, the contingency reserves unit 3 11 can be realized also a s
external functional and accessible unit, e.g. a s a regional development bank system
allowing the system 1 or a user of the system 1 to access to a contingent credit facility.
In addition, the development bank system can offer means to establish a sovereign
catastrophe insurance solution for the country and/or system 1. Each of the financing
instruments has a capital cost, which is measured by the cost factor. The cost factor is
the ratio of costs to loss. Further, the system 1 can comprise means for accessing a
catastrophe reserve fund. Thus, the system 1 autonomously or the user by means of the
system 1 can select in this example the following disaster financing means 30: (i) the
contingency reserves unit 3 11 lets assign the system 1 or a user a portion of the yearly
country budget a s a reserve fund, wherein by means of the contingency reserves unit
3 11 the system 1 is enabled to immediately pay for catastrophe losses/reconstruction
efforts in cash. No direct costs apply to this disaster financing means 30. However, by
setting up a reserve fund parameter, the system 1 foregoes the option to use said
budgetary part for other units or budgetary items or to invest it (opportunity cost). In
addition, appropriately set boundary parameters, a s e.g. budgetary restrictions may
prevent the system 1 from allocating too big shares of the budget a s reserve fund
parameter. As another boundary parameter, time factor may be important in
optimization. If a large event happens and the reserves have not been built up to the
required amount, the difference will have to be post-financed. The assigned cost factor
can be set to 1.00 for the contingency reserves unit 3 11; (ii) the contingent debt facility
unit 3 12 lets assign the system 1 or a user a portion of the yearly country budget a s p re
defined credit lines where drawdown depends on the occurrence of a natural disaster.
Provided by international financial systems or development bank systems, they offer
immediate liquidity to affected countries until other sources of funding can be
accessed. The credit drawn has to be paid back eventually with interest. The cost
factor can be variable set by or to the system 1 depending on boundary conditions a s
credit conditions (interest rate and/or payback period); (iii) the insurance facility unit
3 13 lets assign the system 1 or a user a portion of the yearly country budget a s
insurance premium factor. Once the insurance facility unit 313 is triggered (by the p re
determined event characteristics), it will cover all losses until a n upper threshold value
of cover is reached. In contrast to the contingency reserves unit 3 11, the price of
insurance is not 100% of covered lasses. The yearly premium can e.g. be defined by the
expected loss the insurance facility unit 3 13 has to cover, plus a loading factor in order
to cover expenses of the insurance facility unit 313 and is a fraction of the total limit
provided. Transferred payouts of the insurance facility unit 313 d o not need to be paid
back. As mentioned above, the assigned cost factor can e.g. be set to 1. 70 of
insurance facility unit's 313 expected loss, which factor e.g. can be based o n current
market benchmarks. 1.7 is a n average, which e.g. can be based o n current market
benchmarks or can be assumed or determined otherwise, for example. However, other
values are possible and actual values may vary based on the circumstances of the
insurance parameters' definition. The realized cost factor (i.e. the actual ratio of
transferred premium over losses by the system 1 and/or the country) will depend o n the
losses effectively occurring.
Expected catastrophe losses are determined by means of the loss
frequency function 103 and the geographic risk map 20 for various scenarios of
occurring natural disaster event types 101 and a forecast of the effect of the disaster
financing means 30 to cover these losses is prepared based on the coverage structure
401 , the assigned cost factors 301 , and the determined expected catastrophe losses.
The budgetary share factors 4 10, 4 11, 4 12, 4 13 of the coverage structure 401
are varied by means of the user interface 90 by a user or a n automated input device in
order to optimize the effect of the disaster financing means 30 to cover possible losses.
The system 1 can comprise predefined or otherwise fixed threshold
parameters for each budgetary share factor 410, 4 11, 412, 413, limiting the possibility of
variation of the corresponding budgetary share factor 4 10, 4 11, 4 12, 4 13. Setting the
corresponding threshold parameter, a certain budgetary share factor 410, 4 11, 412, 413
can only by varied up to the assigned threshold value, thus preventing a possible user
or the system 1 from allocating more budgetary value to this budgetary share factor
410, 4 11, 412, 413. As a n embodiment variant, the system 1 can e.g. comprise a
MonteCarlo module 60 for generating a probabilistic Monte Carlo loss simulation for a
probabilistic multi-year simulation a s a final test of the effectiveness of a chosen
coverage structure 401 for a specific pre-financing scheme. The MonteCarlo module 60
can e.g. generate the probabilistic Monte Carlo loss simulation for a probabilistic 30-
year simulation. In this way, the present invention is able to provide a user with
experience and insight into the experience of a Country Risk Officer (CRO) by looking
a t the nation's risk profile 121 and creating a n appropriate risk management plan to be
tested through realistic scenarios. In that sense, the invention also can serve a s
automated training device based o n fictitious countries profiles 121 and respective
country-specific parameters 121 1, 1212, 1213. For example, a user can be appointed to
the role of fictitious Country Risk Officer of a fictitious country, say Costa Azul. By means
of the system 1, the user can be placed in charge of completing the disaster risk
management strategy for the government of the fictitious state Costa Azul. The profile
121 of the fictitious state can be modeled in any appropriate way by means of the
country-specific parameters 121 1, 1212, 1213. For example, Costa Azul can be modeled
a s a n emerging market country located in a tropical climate. For Costa Azul, the
economic expansion can, for example, be combined with the expectation of more
intense natural catastrophes related to global climate change, wherein the total
human and economic costs of natural disasters for Costa Azul are likely to rise in the
future. For a training example, the user can suppose that comprehensive measures
have already been taken for risk mitigation. However, pre-disaster risk financing is still a
component to be elaborated by the CRO, i.e. the user. The budget available depends
on the country profile 12 1, a s defined. The user a s trainee has to set up a n efficient
financing scheme. In this example, the user a s CRO can e.g. suppose that his
responsibilities include: (i) identifying emerging risks, (ii) establishing a frequency/severity
risk landscape; (iii) steering mitigation efforts towards the largest risks (either frequency
or severity); and (vi) developing a risk financing plan for risks that cannot be fully
prevented or mitigated. To be trained by the system 1, the user can suppose that
substantial mitigation efforts have already been made, for example in order to ensure
strict building codes for all new construction, or that barriers are built around
infrastructure to help mitigate the exposure to earthquake and flooding, or bridges are
built to resist strong winds, or a public alarm system for early notification has been put in
place etc. etc. Therefore, the user only needs to set a up a n efficient scheme for
applying the available disaster financing means 30. In other words, to use the system for
training, the user can suppose that the above-mentioned responsibilities have all been
met to date, except for the development of a risk financing scheme. The user now
needs to focus on providing for the most efficient way to transfer a portion of the
country's risks off the government's balance sheet.
To provide the most optimized apportionment by scaling the corresponding
budgetary share factors 4 10, 4 11, 4 12, 4 13 in the risk financing scheme, the user who is
to be trained by the system 1 applies his knowledge based on the country risk profile
121 , the country-specific parameters 121 1, 1212, 1213, the country-specific occurrence
of natural hazards, the risk map 20 and/or risk maps 2001 , 2002, 2003, 2004, their
parameters of occurrence 201 1, 2012, 2013, 2014, and the loss frequency function 103
and/or loss frequency functions 1031 , 1032, 1033, 1034. Thus, the system 1 lets the user
systematically address how the government can assess and reduce the losses from
cyclones and how i† can best prepare for providing relief and reconstruction in the
event of a disaster, i.e. lets the user systematically build up a n appropriate strategy
pattern. It is to be noted, that in the embodiment variant of the fully automated system
1, the appropriate factors and parameter values, a s mentioned above, are captured
by measuring and/or capturing and/or filtering means of the system 1. By means of the
geographic risk map 20, the system allows the user to interactively view their country's
or region's exposure to direct asset risks and (indirect) financial, fiscal, and economic
impacts of disaster scenarios. The outcome for reducing disaster risk can be assessed by
the system 1 and expressed with indicators of interest to the user, such a s the budget
stance, debt, and economic growth. Based on a n assessment of their country's or
region's vulnerability and risk, one of the purposes of the system 1 is to provide a
systematic and automated system 1 for assessing policy options related to financial risk
management, including balancing and allocating the risk-transfer instruments (i.e. the
disaster financing means 30 such a s reserve funds, insurance, and catastrophe bonds)
and their parameters such asthe budgetary share factors 4 10, 4 11, 4 12, 4 13,
respectively. The system 1 can comprise a graphical user interface and is interactive
(including a stand-alone application), that is, users can and should change the model
parameters, given different preferences and parameter uncertainty. For example, the
user can adjust the amount of risk and debt that the country is willing to take on, and
the system 1 will show how this changes the country's vulnerability to disasters and how
it affects different policy paths. One of the purposes of the system 1 is to provide
automated means for reducing a country's risks of experiencing a "resource/financing
gap" or the inability to meet its post-disaster obligations in terms of repairing public
infrastructure and providing needed relief to the private sector. For this purpose, the
user will need to use the above-mentioned information about assessing financial and
macroeconomic risks and vulnerability. The user must also consider how to reduce or
mitigate human and economic losses and finally, he must create a n appropriate risk
management framework for the country (See fig.). Apart from the embodiment variant
of a fully automated system 1 (i.e. operating completely without any human
interaction), the above-mentioned application of the system 1 allows for a different
use. For example, the user can use the above-mentioned interface module to propose
starting parameters, e.g. for the variable budgetary share factor 4 10, 4 11, 4 12, 423,
wherein the system 1 optimizes the starting parameters in order to achieve a global or
local maximum for the parameters. In a n other variant, the system 1 proposes starting
parameters, i.e. budgetary share factor 4 10, 4 11, 4 12, 423, wherein these parameters
are varied in the following by the user, allowing him to understand the effect of
different variations o n the outcome of the risk transfer. The above embodiments also
allow a user to vary the parameters during optimization by means of the system 1 in
order to overcome local maxima or minima, where the optimization operation may
stick.
If a user performs the development of a n appropriate risk management
strategy by adapting the above-mentioned factors, first, the user considers the
interaction of natural hazard-caused losses and the government, i.e. the financial risks
of asset losses and relief expenditures to assist households and business, and the
proportion of financial losses that will be absorbed by the government. Therefore, the
country risk officer must first determine or assess the risk to the country's public sector
assets. For example, the above-mentioned risk depends on the frequency and intensity
of natural hazards, the assets exposed to natural hazards, and their physical
vulnerability to a specific type of natural hazard, a s captured by the loss frequency
functions 1031 , 1032, 1033. Second, based on the limited country resources for reducing
human and economic losses, which are represented, for example, by parameters such
a s the gross domestic product (GDP) of the country, the parameter should be varied in
such a way that the country becomes a s financially resilient a s possible, or provides
sufficient funds for financing reconstruction of public infrastructure and for providing
relief to households and the private sector. On the other hand, financial resilience
depends on how much a natural hazard risk can be reduced so that it has less effect
o n the general economic conditions of the country. Thus, the country risk officer has to
balance the resilience of the country's public sector, based on the risk or on the
achieved risk reduction. Next, it is important that the country risk officer carefully tries to
determine or estimate the so called "resource gap," which is the difference between
the contingent post-disaster liabilities of the country or its government for repairing
infrastructure and providing relief to the private sector and the sources of funding
available to the government. The system 1 can automatically assess this by simulating
the risks to public assets and estimating the government's ability to cover these risks and
provide private sector assistance. The assessment is among other things based o n the
country-specific parameters 121 1, 1212, 1213. When adapting the parameter, the
country risk officer should also try to synchronize the disaster risk with national
development planning, e.g. by incorporating financial disaster risk and potential
financing gaps for funding possible losses into macroeconomic projections of the
country. For the system 1, the consequences can e.g. be related to variables such a s
economic growth or the country's external debt situation.
These indicators represent impacts on economic flows a s compared to
impacts o n stocks addressed by the financial asset risk estimation. Typically, the country
risk officer should be primarily concerned about loss of life from natural hazards and
also about loss of livelihood and productive assets directly or indirectly affecting mostly
the country's public sector assets. It is therefore often important to a risk management
scheme or coverage structure 401 to consider the cost-effectiveness of a n applied
parameter scheme 401 in reducing human and economic losses. Finally, for a n
effective disaster risk plan, the country risk officer uses the allocation module 40 and the
applied coverage structure 401 with the variable budgetary share factors 4 10, 4 11, 4 12,
4 13, to allocate the budget among the available options for reducing the risk of a
resource gap, including insurance, catastrophe bonds, and a reserve fund or
contingent credit arrangements. It is important to balance risk optimization against the
cost-effectiveness of the available disaster financing means 30, i.e. the cost factors 301 ,
301 1, 301 2, 301 3 of disaster financing means, in reducing the resource gap risk. The
system 1 can be used to develop strategies, while the system 1 assesses if a proposed
scheme 401 in fact reduces the risks of disasters and enhances the financial resilience
of the country, or not. The development of a n efficient risk-financing scheme, i.e. a
coverage structure 401 , by means of the system 1 has to be understood a s a n adaptive
process, where measures are continuously revised after their impact o n reducing
financial vulnerability and risk has been assessed.
In the above example of using the system 1 for training a country risk officer,
a user of the system 1 may use the information provided by the country risk profile 12 1,
the disaster history table 10, and/or the geographic risk map 20 to identify the disaster
risks with the most urgent pre-financing need (in the above example: earthquakes,
hurricanes, floods, and droughts). However, by a combination of both frequency and
severity, the most dangerous and costly natural disasters affecting a country can also
be determined by the system 1 or the core engine 2 of the system 1. When using the
system 1 a s training system, a user can once again suppose, for example, that other
mitigation means different from disaster financing means 30 have already been taken
into account. So the user who defines the variable budgetary share factor 4 10, 4 11,
4 12, 4 13 of the coverage structure 40 for the disaster financing means 30 does not have
†o care about the preparedness of a country for natural disaster events. For example
the user can simply suppose that the established risk maps 20 are used for the
introduction of stricter building codes for potentially affected houses and infrastructure;
early warning systems for both hurricanes and earthquakes are in place, schools and
public institutions are expected to conduct regular evacuation exercises etc. etc.
Therefore, the user defining the variable budgetary share factor 4 10, 4 11, 4 12, 4 13 may
use all of the information provided to get a holistic picture of a country's risk landscape
and, using the variable budgetary share factor 410, 4 11, 412, 413 of the coverage
structure 40 for the disaster financing means 30, decide how to create a n appropriate
coverage and mitigation scheme for each disaster event type 101 .
In another embodiment variant, the allocation module 40 can e.g.
comprise a second Monte Carlo module 80. By means of the second Monte Carlo
module 80 and based o n the allocated variable budgetary share factors 4 10, 4 11, 4 12,
413 of the coverage structure 40, a plurality of data records comprising coverage
structures 40 with varied budgetary share factors 410, 4 11, 412, 413 can e.g. be
generated, wherein the coverage structure 40 with the allocated budgetary share
factors 4 10, 4 11, 4 12, 4 13 is optimized by means of a core engine 2 of the system 1
based o n the effect of the disaster financing means 30 for various scenarios of
occurring natural disaster event types 101 .
The allocation module 40 can also comprise, for example, a n activating
device 93 by means of which, based on the generated coverage structure 401 with the
allocated budgetary share factors 4 10, 4 11, 4 12, 4 13, it is possible to transmit a
corresponding control signal to the monitoring device 9 1. The allocation module 40 can
also comprise a signaling device 92, wherein the selectable disaster financing means 30
are activated based on the allocated budgetary share factors 410, 4 11, 412, 413 by
means of signal transmission. As a n embodiment variant, the signaling device 92 can,
upon triggering a n optimized coverage structure 401 , activate the selectable disaster
financing means 30 based on the allocated budgetary share factors 410, 4 11, 412, 413
by means of signal transmission.
To assess the risks of a country, i.e. the risk with respect to the country's
assets and economic operability in case of a natural hazard, the system 1 can e.g.
comprise means to perform a hazard assessment dedicated to the various possible
hazards, e.g. hurricanes, floods, droughts, or earthquakes. The assessment can e.g. be
performed based on historical data of historical events and corresponding losses for a
specific natural hazard. To determine the damage potential of natural hazards,
different techniques can be applied, e.g. stochastic or engineering approaches for
determining physical vulnerability of the assets exposed. However, historical losses can
also be used for direct risk assessment. The catastrophe risk assessment can e.g. be
captured by means of different dedicated modules, e.g. a catastrophe module, a n
exposure module, a vulnerability module, and a loss module, wherein the latter
integrates the results from the first three modules by means of risk metrics or loss
distributions. Loss distributions are cumulative distribution functions where the x-axis
represents the losses, e.g. represented by monetary loss parameters, annual loss
parameters in terms of GDP, or capital stock loss parameters. The y-axis, represents the
probability that losses d o not exceed a predetermined threshold value. For example, a
value of 0.6 o n the y-axis means that with a probability of 60%, the losses will not exceed
the predetermined threshold value of damage. In other words, with a probability of
40%, the losses will exceed this level of damage. However, a 40% probability also means
that a n event happens on average once every 2.5 years (1/0.4= 2.5). This means that
the longer the return period, the lower the probability of the event, but the higher the
losses. The loss distribution function comprises various risk measures that can be derived
from it. For example, (i) the average annual loss (integrated area above the loss
distribution), (ii) the Value a t Risk (VaR) defined a s VaR(p)=F-' (1-p), where F-' is the
quantile function defined a s the inverse of the loss distribution function or (iii) Probable
Maximum Loss (PML), which is associated with a given probability of exceedance.
For the system 1, it is possible, for example, to choose between two possible
approaches in order to carry out the determination of the risk transfer function: (i) via
catastrophe models or (ii) using historical data, i.e. historical events. However, other
methods can be used to derive the risk transfer function. As mentioned, dedicated
modules can be used for performing the catastrophe assessment, e.g. a catastrophe
module, exposure module, vulnerability module, and loss module, all performing
different functions. The catastrophe module comprises e.g. a t least three variables
regarding the source parameters of the hazard: the location of future events, their
frequency of occurrence, and their severity. These parameters can be based o n
filtering historical and/or engineering data, e.g. by simulating potential hurricane tracks
to increase the number observations. The probability of a given event has to be
determined either by time-series analysis or by assuming suitable stochastic models, e.g.
non-homogeneous Poisson distribution of the probability of a hurricane event. In
addition, the intensity is determined. The exposure module captures the spatial
distribution of the assets exposed. An appropriate hash table comprising variable
parameters reflecting regional differences in construction practices and building codes
can be created and comprised in the module. For risk assessment, the spatial resolution
of the exposure data can be used following any order such a s storms, earthquakes,
floods, droughts, and man-made hazards. The process of inventory development can
be a difficult and time-consuming task. However, it is a n important part of the risk
assessment process. For the process, the system 1 can use satellite images and tier
classification with asymmetric mapping. In addition, the vulnerability module quantifies
the physical impact of the natural hazard on the exposed elements. For example, it
expresses the relationship between the intensity of the natural hazard and the
percentage of houses damaged, e.g. a damage ratio parameter. Since the intensity
measure and the level of damage can typically not be captured by means of one
precise value, the damage cannot be expressed a s a precise quantity either, but only
within a range of error or uncertainty. Underlying each damage function is a frequency
component and a severity component. The first determines the probability that a n
exposed element will be damaged and the second determines the percentage of
property damaged, assuming damage has occurred. For example, the relationship
between damage and wind speed depends on the construction of the building, the
age of the building, the topographical and environmental exposure of the building,
etc. Finally, the loss module integrates and transforms the damages into a needed
measure, such a s monetary loss parameters. Various risk metric schemes can be
applied, e.g. value at risk, exceedance probabilities, hazard maps, or loss distribution
functions. Again, the loss module therefore captures and technically implements a
possible function for the total damage e.g. in monetary terms. By means of the above
modularized structure, the system 1 can automatically provide the appropriate loss
distribution functions and loss frequency function 103 a s well a s appropriate
geographic risk maps 20. It is also possible to capture future changes e.g. by
incorporating a dynamic setting into the four modules, which can e.g. be dynamically
adapted by a n appropriate expert unit or system 1. Such future changings can e.g.
comprise a change of the natural hazard intensity and/or frequency, changes in
vulnerability due to economic and social development, or changes in risk exposure.
If the system 1 is implemented based o n historical data module, the system
can use historical data of natural hazards in combination with the extreme value
theory. Based on the total annual natural hazard losses, a n optimization algorithm for
selecting the best fit under the assumption of a n extreme value distribution a s well a s
generalized Pareto distribution can be used. For example, a sequence of parameter fits
can be obtained based on a weighted average function of those data points between
projected return periods, which subsequently can be used a s the next starting point, in
iterative fashion throughout the process. The system 1 can thus provide final results e.g.
for both the GEV and the GP fits (loss distribution based on Generalized Extreme value
distribution and loss distribution based on a Pareto distribution.) Also by means of the
described historical data module, the system 1 can be implemented to automatically
provide the appropriate loss distribution functions and loss frequency function 103 a s
well a s appropriate geographic risk maps 20. In order to determine a robustness
parameter of the operation of the system 1, the two embodiments variants can e.g. be
operated in parallel. If both of the above system's embodiment variants and methods,
i.e. the implementation by means of the dedicated modules a s the above-described
catastrophe module, exposure module, vulnerability module, and loss module, and the
implementation by means of the described historical data module, show comparable
results, then some robustness of the results can be expected.
List of reference numerals
Disaster management system
Core engine
Disaster history table
101 Disaster event types
102 Loss frequency function parameters
103 Loss frequency function
1031 Hurricane loss frequency function
1032 Flood loss frequency function
1033 Earthquake loss frequency function
1034 Drought loss frequency function
1 Country risk profile
121 1, 1212, 1213 Country-specific parameters
2 Predefined criteria for country-specific parameters
1221 Population criteria
1222 Demographic criteria
1223 Gross domestic product (GDP) criteria
1224 Sovereign budget
1225 Inflation rate
1226 Economic structure
1227 Export/import values
Geographic risk map
2001 Geographic hurricane risk map
2002 Geographic flood risk map
2003 Geographic earthquake risk map
2004 Geographic drought risk map
201 Risk mapping parameters
201 1 Hurricane risk mapping parameters
2012 Flood risk mapping parameters
2013 Earthquake risk mapping parameters
2014 Drought risk mapping parameters
Disaster financing mean
301 Cost factor of disaster financing mean
301 1 Cost factor of contingency reserves unit
301 2 Cost factor of contingent debt facility unit
301 3 Cost factor of insurance facility unit
3 11 Contingency reserves unit
3 12 Contingent debt facility unit
313 Insurance facility unit
Allocation module
401 Coverage structure
410 Variable budgetary share factors
4 11First variable budgetary share factor
4 12 Second variable budgetary share factor
413 Third variable budgetary share factor
501 ,51 1,521 ,531 Risk exposed geographic area
MonteCarlo module for probabilistic Monte Carlo loss simulation
Network
Second Monte Carlo module
User interface
901 First selectable input channel for a first budgetary share factor
902 Second selectable input channel for a second budgetary share
factor
903 Third selectable input channel for a third budgetary share factor
Monitoring device
Signaling device
Activation device
Claims
1. A computer-based disaster management and financing method for
forecasting the impact of disaster mitigation and financing means based o n locationdependent
natural disaster impacts by means of a system ( 1 ) , wherein measuring
parameters of historical disaster events are captured in order to determine the impact
of natural disaster events and then critical values of parameters of natural disaster
events are used a s triggers in order to generate forecasts of the impacts of disaster
events within a geographic area (501 531 ) ,
in that country-specific parameters (121 1, 1212, 1213) of a risk-exposed
country (501 531 ) are captured, relating to stored predefined criteria ( 1221 , 1222,
1223), wherein the country-specific parameters (121 1, 1212, 1213) comprise a t least
national economic and national budgetary parameters,
in that one or more disaster event types (101 ) are assigned to a disaster
history table (10), wherein each disaster event type (101 ) comprises a plurality of typespecific
measuring parameters of historical natural disaster events and associated typespecific
loss frequency function parameters (102) that provide a corresponding loss
frequency function (103) for each natural disaster event type (101 ) , and wherein the
magnitude of a loss to its expected exceedance frequency is parameterized by means
of the loss frequency function (103), where the exceedance frequency is a measure of
the annual probability that a n event or loss will meet or exceed a given magnitude in
any given timeframe,
in that the system ( 1 ) comprises mapping parameters (201 ) for capturing
and storing a geographic risk map (20), wherein for each of the natural disaster event
types (101 ) , corresponding mapping parameters (21 ) are captured and stored, which
define danger zones for the specific natural disaster event type (101 ) ,
in that the system ( 1 ) comprises a plurality of selectable disaster financing
means (30), wherein each of the selectable disaster financing means (30) is assigned to
a definable cost factor (301 , 302, 303) capturing the capital cost of the disaster
financing means (30) in relation to its application for disaster mitigation, and wherein for
each of the selectable disaster financing means (30) a variable budgetary share factor
(41 1, 412, 4 13) can be allocated and adapted by means of a n allocation module (40)
defining a coverage structure (401 ) in case of a catastrophic disaster event, and
in that expected catastrophe losses are determined by means of the loss
frequency function (103) and the geographic risk map (20) for various scenarios of
occurring natural disaster event types (101 ) and a forecast of the effect of the disaster
financing means (30) to cover these losses is prepared based o n the coverage
structure (401 ) , the assigned cost factors (301 ) , and the determined expected
catastrophe losses.
2. The method according to claim 1, characterized in that a first disaster
financing means (30) is related to a contingency reserves unit (31 1) comprising a n
assigned cost factor set to 1, a second selectable disaster financing means (30) is
related to a contingent debt facility unit (312) comprising a n assigned cost factor
depending on definable credit condition parameters, and a third selectable disaster
financing means (30) is related to a n insurance facility unit (313) comprising a n assigned
cost factor set to a factor based on current market benchmarks.
3. The method according to one of claims 1 or 2, characterized in that
based on the disaster history table (10) comprising the stored natural disaster event
types (101 ) , a t least four loss frequency functions (103) capturing the perils of hurricanes
(1031 ) , floods (1032), earthquakes (1033), and droughts (1034) are generated together
with the corresponding mapping parameters (201 1, 2012, 2013, 2014) of the
geographic risk map (20, 2001 , 2002, 2003, 2004).
4. The method according to one of claims 1 to 3, characterized in that the
system ( 1 ) comprises a t least country-specific, predefined criteria (121 1, 1212, 1213) for
country-specific parameters (122) related to population (1221 ) and/or demographic
(1222) and/or gross domestic product (1223) and/or sovereign budget (1224) and/or
inflation rate (1225) and/or economic structure (1226) and/or export/import values
(1227).
5. The method according to one of claims 1 to 4, characterized in that the
expected catastrophe losses are determined by means of numerical integration of the
loss frequency function (103).
6. The method according to one of claims 1 to 6, characterized in that the
system ( 1 ) comprises a MonteCarlo module (60) for generating a probabilistic Monte
Carlo loss simulation for a probabilistic multi-year simulation a s a final test of the
effectiveness of a chosen coverage structure (401 ) for a specific pre-financing scheme.
7. The method according to claim 6, characterized in that the MonteCarlo
module (60) generates the probabilistic Monte Carlo loss simulation for a probabilistic
30-year simulation.
8. The method according to one of claims 1 to 7, characterized in that in a
first channel (901 ) selectable by means of a user interface (90), a first budgetary share
factor (41 1) is determined and assigned to the corresponding first disaster financing
means (301 ) , in a second channel (902) selectable by means of a user interface (90), a
second budgetary share factor (412) is determined and assigned to the corresponding
second disaster financing means (302), and in a third channel (903) selectable by
means of a user interface (90), a third budgetary share factor (413) is determined and
assigned to the corresponding third disaster financing means (303).
9. The method according to the claim 8, characterized in that the
budgetary share factors (41 1, 412, 4 13) of the coverage structure (401 ) are varied by
means of the user interface (90) in order to optimize the effect of the disaster financing
means (30) to cover possible losses.
10. The method according to one of claims 1 to 7, characterized in that the
allocation module (40) comprises a second Monte Carlo module (80), wherein by
means of the second Monte Carlo module (80) and based on the allocated variable
budgetary share factors (41 1, 412, 4 13) of the coverage structure (40), a plurality of
data records comprising coverage structures (40) with varied budgetary share factors
(41 1, 412, 4 13) are generated, wherein the coverage structure (40) with the allocated
budgetary share factors (41 1, 412, 4 13) is optimized by means of a core engine (2) of
the system ( 1 ) based o n the effect of the disaster financing means (30) for various
scenarios of occurring natural disaster event types (101 ) .
11. The method according to one of claims 1 to 10, characterized in that
the allocation module (40) comprises a n activating device ( 1 1) , by means of which,
based o n the generated coverage structure (401 ) with the allocated budgetary share
factors (41 1, 412, 413), it is possible to transmit a corresponding control signal to the
monitoring device (91 ) .
12. The method according to one of claims 1 to 11, characterized in that
the allocation module (40) comprises a signaling device (92), wherein the selectable
disaster financing means (30) are activated based o n the allocated budgetary share
factors (41 1, 412, 4 13) by means of signal transmission.
13. The method according to claim 12, characterized in that upon
triggering a n optimized coverage structure (401 ) , the selectable disaster financing
means (30) are activated by the signaling device (92) based o n the allocated
budgetary share factors (41 1, 412, 4 13) by means of signal transmission.
14. A computer-based disaster management and financing system ( 1 ) for
forecasting the impact of disaster mitigation and financing means based o n locationdependent
natural disaster impacts, wherein measuring parameters of historical
disaster events are captured in order to determine the impact of natural disaster events
and then critical values of parameters of natural disaster events are used a s triggers to
generate forecasts of the impacts of disaster events within a geographic area (501
531 ) ,
in that country-specific parameters (121 1, 1212, 1213) of a risk-exposed
country (501 531 ) are captured, relating to stored predefined criteria (1221 , 1222,
1223), wherein the country-specific parameters (121 1, 1212, 121 3) comprise at least
national economic and national budgetary parameters,
in that one or more disaster event types (101 ) are assigned to a disaster
history table (10), wherein each disaster event type (101 ) comprises a plurality of typespecific
measuring parameters of historical natural disaster events and associated typespecific
loss frequency function parameters (102) that provide a corresponding loss
frequency function (103) for each natural disaster event type (101 ) , and wherein the
magnitude of a loss to its expected exceedance frequency is parameterized by means
of the loss frequency function (103), where the exceedance frequency is a measure of
the annual probability that a n event or loss will meet or exceed a given magnitude in
any given timeframe,
in that the system ( 1 ) comprises mapping parameters (201 ) for capturing
and storing a geographic risk map (20), wherein for each of the natural disaster event
types (101 ) , corresponding mapping parameters (21 ) are captured and stored, which
define danger zones for the specific natural disaster event type (101 ) ,
in that the system ( 1 ) comprises a plurality of selectable disaster financing
means (30), wherein each of the selectable disaster financing means (30) is assigned to
a definable cost factor (301 , 302, 303) capturing the capital cost of the disaster
financing means (30) in relation to its application for disaster mitigation, and wherein for
each of the selectable disaster financing means (30) a variable budgetary share factor
(41 1, 412, 413) is allocatable and adaptable by means of a n allocation module (40)
defining a coverage structure (401 ) in case of a catastrophic disaster event, and
in that expected catastrophe losses are determined by means of the loss
frequency function (103) and the geographic risk map (20) for various scenarios of
occurring natural disaster event types (101 ) and a forecast of the effect of the disaster
financing means (30) to cover these losses is prepared based o n the coverage
structure (401 ) , the assigned cost factors (301 ) , and the determined expected
catastrophe losses.
15. The system ( 1 ) according to claim 14, characterized in that a first disaster
financing means (30) is related to a contingency reserves unit (31 1) comprising a n
assigned cost factor set to 1, a second selectable disaster financing means (30) is
related to a contingent debt facility unit (312) comprising a n assigned cost factor
depending on definable credit condition parameters, and a third selectable disaster
financing means (30) is related to a n insurance facility unit (313) comprising a n assigned
cost factor set to a factor based on current market benchmarks.
1 . The system ( 1 ) according to one of claims 14 or 15, characterized in that
based o n the disaster history table (10) comprising the stored natural disaster event
types (101 ) , a t least four loss frequency functions (103) capturing the perils of hurricanes
(1031 ) , floods (1032), earthquakes (1033), and droughts (1034) are generated together
with the corresponding mapping parameters (201 1, 2012, 2013, 2014) of the
geographic risk map (20, 2001 , 2002, 2003, 2004).
17. The system ( 1 ) according to one of claims 14 to 1 , characterized in that
the system ( 1 ) comprises a t least country-specific, predefined criteria (121 1, 1212, 121 3)
for country-specific parameters (122) related to population (1221 ) and/or demographic
( 1 222) and/or gross domestic product (1223) and/or sovereign budget (1224) and/or
inflation rate (1225) and/or economic structure (1226) and/or export/import values
(1227).
18. The system ( 1 ) according to one of claims 14 to 17 characterized in that
the expected catastrophe losses are determined by means of numerical integration of
the loss frequency function (103).
19. The system ( 1 ) according to one of claims 14 to 18, characterized in that
the system ( 1 ) comprises a MonteCarlo module (60) for generating a probabilistic
Monte Carlo loss simulation for a probabilistic multi-year simulation a s a final test of the
effectiveness of a chosen coverage structure (401 ) for a specific pre-financing scheme.
20. The system ( 1 ) according to claim 19, characterized in that the
MonteCarlo module (60) generates the probabilistic Monte Carlo loss simulation for a
probabilistic 30-year simulation.
2 1. The system ( 1 ) according to one of claims 14 to 20, characterized in that
in a first channel (901 ) selectable by means of a user interface (90), a first budgetary
share factor (41 1) is determined and assigned to the corresponding first disaster
financing means (301 ) , in a second channel (902) selectable by means of a user
interface (90), a second budgetary share factor (412) is determined and assigned to
the corresponding second disaster financing means (302), and in a third channel (903)
selectable by means of a user interface (90), a third budgetary share factor (413) is
determined and assigned to the corresponding third disaster financing means (303).
22. The system ( 1 ) according to the claim 2 1, characterized in that the
budgetary share factors (41 1, 412, 4 13) of the coverage structure (401 ) are varied by
means of the user interface (90) in order to optimize the effect of the disaster financing
means (30) to cover possible losses.
23. The system ( 1 ) according to one of claims 1 to 22, characterized in that
the allocation module (40) comprises a second Monte Carlo module (80), wherein by
means of the second Monte Carlo module (80) and based on the allocated variable
budgetary share factors (41 1, 412, 413) of the coverage structure (40), a plurality of
data records comprising coverage structures (40) with varied budgetary share factors
(41 1, 412, 413) are generated, wherein the coverage structure (40) with the allocated
budgetary share factors (41 1, 4 12, 4 13) is optimized by means of a core engine (2) of
the system ( 1 ) based o n the effect of the disaster financing means (30) for various
scenarios of occurring natural disaster event types (101 ) .
24. The system ( 1 ) according to one of claims 14 to 23, characterized in that
that the allocation module (40) comprises a n activating device ( 1 1) , by means of
which, based o n the generated coverage structure (401 ) with the allocated budgetary
share factors (41 1, 412, 413), it is possible to transmit a corresponding control signal to
the monitoring device (91 ) .
25. The system ( 1 ) according to one of claims 14 to 24, characterized in that
the allocation module (40) comprises a signaling device (92), wherein the selectable
disaster financing means (30) are activated based on the allocated budgetary share
factors (41 1, 4 12, 4 13) by means of signal transmission.
26. The system ( 1 ) according to claim 25, characterized in that upon
triggering a n optimized coverage structure (401 ) , the selectable disaster financing
means (30) are activated by the signaling device (92) based o n the allocated
budgetary share factors (41 1, 412, 4 13) by means of signal transmission.