Content uploaded by Paula Jarzabkowski
Author content
All content in this area was uploaded by Paula Jarzabkowski on Jun 05, 2019
Content may be subject to copyright.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
Professor Paula Jarzabkowski, Dr Konstantinos Chalkias,
Dr Eugenia Cacciatori and Dr Rebecca Bednarek
© Paula Jarzabkowski, th June .
No part of this report to be reproduced or referenced without appropriate citation to: Jarzabkowski,
P., K. Chalkias, E. Cacciatori, R. Bednarek, (). Between State and Market: Protection Gap Entities
and Catastrophic Risk. Cass Business School, City, University of London, th June .
We gratefully acknowledge the nancial support of our sponsors, below, that enabled us to conduct
the research for this report. Their sponsorship does not imply any endorsement of the results
reported and any analysis, ndings or errors are totally our own.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
Executive summary
The protection gap. The economic and social impact of disasters is increasing all around the
world. In , Hurricanes Harvey, Irma and Maria were so destructive as they swept through the
Caribbean that they decimated many of these small island economies. Even in a wealthy country
like the USA, the economic impact of these natural disasters was enormous. In recent years,
earthquakes dealt a blow to Mexico, France saw the worst rains in years with floods peaking in
Paris, typhoons and storms shook the Philippines and Hong Kong, and wildres ravaged California
and Australia.
Yet, the economic losses from such disasters are underinsured. In what is known as the protection
gap, some % of global losses from natural catastrophes are not insured, equating to .
trillion over the past years. In alone, uninsured losses for weather-related disasters
were estimated to be around billion. At the same time, other forms of large-scale risk, such
as terrorism, cyberattacks and pandemics are also increasing, with little nancial protection to
address the aermath.
The social and economic resilience of a country, and its political stability, are dependent on the
ability to recover from disasters. In the short-term, immediate post-crisis nancial response is
critical. Failure to provide a rapid injection of capital in response to disaster puts lives at risk. In
Puerto Rico, the death toll of Hurricane Maria rose above ,, with one-third of the deaths being
caused by delayed or interrupted medical care. In the longer-term, reconstruction of housing,
infrastructure and businesses aer a disaster is essential for recovery. Bridging the protection gap
provides one way to underpin such nancial recovery.
Protection Gap Entities (PGEs): Marrying market solutions to social objectives. In their
quest to address some of their social objectives in protecting their citizens from disaster,
governments are increasingly turning to market solutions, such as innovative means of insuring
for potential loss. They do so through the establishment of Protection Gap Entities (PGEs) that
operate between state and market in developing novel solutions/schemes that mobilize global
(re)insurance capital in addressing the aermath of disaster.
This report draws on a large-scale research study of dierent PGEs around the world, in both
developed and developing economies, to explain their role, their eects and their limitations in
managing risk and alleviating the nancial consequences of disaster. While such government
interventions are growing, lessons need to be learnt about how to maximize their positive eects
and guard against potential unintended consequences that can exacerbate the protection gap.
This report shows the strategic implications of dierent types of PGEs, what they may be best used
for, and how they can evolve to better help society and government to protect against the growing
threats of natural and manmade disasters.
Who should read this report? This report is useful for the dierent stakeholders involved in
the work of PGEs: from policy makers and governments, to NGOs, to those running PGEs or
insurance market organizations. These stakeholders oen have very dierent interests and goals,
which are reflected in the way they understand the purpose of PGEs and insurance. This report
provides valuable insights into the common features of PGEs, and also examines some of their
key dierences, providing an opportunity for learning across stakeholders, PGEs, sectors and
countries.
Contents
Contents
SECTION . INTRODUCTION 3
1.1 Protection Gap Entities 3
1.2 The protection gap 3
1.3 Stakeholders overview 4
1.4 Social and market imperatives 5
1.5 Bridging and reducing the gap 5
1.6 Report map 5
SECTION . ORIGINATION AND OBJECTIVES OF PROTECTION GAP ENTITIES PGES 6
2.1 PGE origination: An ‘uneasy’ truce amongst stakeholders 6
2.2 Objective 1: Resolve disruption in (re)insurance supply in mature markets 7
2.3 Objective 2: Mitigate the threat of unaordable insurance in mature markets 8
2.4 Objective 3: Increase the nancial viabilit y of sovereign states with fragile economies 9
2.5 Beyond markets: PGEs’ roles in framing the protection gap debate and in building exper tise 10
SECTION . IN THE PRESENCE OF A MATURE INSURANCE MARKET: MARKET DYNAMICS 12
3.1 Protection Gap Strategic Response Framework: Types of market inter ventions 12
3.2 Market eect: Strategic positions in the value chain 15
3.3 Summary 19
SECTION . PROTECTION FOR SOVEREIGNS: INSURANCE AS A DISASTER LIQUIDITY PRODUCT 20
4.1 Calculating insurance as disaster liquidity 20
4.2 PGEs in relation to social and market stakeholders 21
4.3 Benets of disaster liquidity products 22
4.4 Enabling conditions for success 24
4.5 Summary 25
SECTION . BRIDGING AND REDUCING THE PROTECTION GAP: PGES’ ROLE IN RESILIENCE 25
5.1 PGEs’ role in risk identication 25
5.2 PGEs role in risk reduction and resilient reconstruction 26
5.3 PGEs role in disaster preparedness 27
5.4 Summary 27
SECTION . PROTECTION GAP ENTITIES OR STOPGAP ENTITIES? 29
6.1 Mismatch 1: PGE remit does not match stakeholders’ expectations 30
6.2 Mismatch 2: When addressing one problem creates another 30
6.3 Mismatch 3: PGE (ir)relevance in the face of an evolving gap 30
6.4 Summary 31
SECTION . CONCLUSION. LEARNING FROM PGE EXPERIENCES: A CALLTOARMS 32
APPENDIX 34
Methodology 34
Glossary 35
REFERENCES 38
TABLE AND FIGURES
Table . The protection gap in developing economies, 4
Figure . Protection Gap Strategic Response Framework 12
Figure . Value chain for risk transfer 16
Figure . Strategic position : Insurer PGE 16
Figure . Strategic position : Reinsurer PGE 17
Figure . Strategic position : Market Capture PGE 18
Figure . PGE Resilience Framework 25
Figure . PGE Evolution Framework 29
Between State and Market: Protection
Gap Entities and Catastrophic Risk
SECTION . Introduction
. PROTECTION GAP ENTITIES
The challenges posed by the growing catastrophe insurance
protection gap, particularly those of rebuilding in the aermath
of disaster, have prompted the generation of entities, which we
label Protection Gap Entities (PGEs). These PGEs bring together
market and non-market stakeholders in an eort to address
the protection gap. They dier considerably in governance,
political economies, points of origin, perils, and means of
funding loss,,(see Exhibit ). Yet PGEs have the same
broad goal:
To transform uninsured risk into insurance-based products
that can be transferred into global nancial markets to provide
capital for recovery following a disaster.
This report examines the nature, characteristics and activities
of PGEs. It provides information for the diverse stakeholders
involved in some aspect of the work of PGEs, from policy
makers to those running PGEs or depending upon them for
nancial products. These stakeholders oen have very dierent
understandings about the insurance market and the remit of
their own and other PGEs. The report will therefore provide
insights into the common features of these entities and some
of their key dierences, providing an opportunity for learning
across stakeholders, PGEs, sectors and countries.
In this section we rst dene the protection gap, and then
provide an overview of some key common elements across the
PGEs examined in our research study: the multiple and diverse
stakeholders with whom they engage, the tension between
the pursuit of market and social objectives that characterises
these stakeholders, and the tension PGEs face between
pursuing their mission strictly by developing nancial products
to ‘bridge’ the protection gap or expanding their remit to
also ‘reduce’ the gap through improved physical resilience
to disaster. This information provides the background for the
analysis presented in this report.
Examples of PGEs
Caribbean Catastrophe Risk Insurance Facility (CCRIF), a
multi-sovereign risk pool set up to provide its member
States with access to rapid capital for responding to the
aermath of natural disasters as diverse as tropical cyclone,
earthquake, and excess rainfall.
California Earthquake Authority (CEA), a privately funded,
publicly managed PGE set up to support the primary
insurance market in providing earthquake insurance to
Californian homeowners.
Australian Reinsurance Pool Corporation (ARPC), a terrorism
reinsurance pool set up to provide reinsurance to insurance
companies oering terrorism cover on commercial
businesses in Australia.
. THE PROTECTION GAP
The term ‘insurance protection gap’ refers to the gap between
the insured and actual economic losses caused by large-scale
catastrophic events. Some % of global losses from natural
catastrophes are uninsured, equating to . trillion over
the past years. Indeed, uninsured natural disaster losses
for weather-related risks are estimated to be around
billion in . Signicant gaps in protection exist not only for
natural disasters but also for other large-scale threats such as
terrorism, cybercrime and epidemics. This gap is a problem in
both developed and developing economies.
Impact. Economic resources are crucial in allowing societies to
recover from devastating disasters. In the absence of adequate
insurance, the burden of paying for losses falls largely on
citizens, governments or aid organizations, with signicant
impact upon already straining government budgets, and
economic and social hardship for those aected.
The growing exposure to disaster was shown in ; a very
active year for natural catastrophes, with earthquakes, floods
and hurricanes inflicting devastation on communities around
the world. As Table . shows, the impact was particularly
severe in developing economies. In these countries, where
insurance penetration is typically low and governments
and citizens have few nancial reserves, losses from
catastrophic disasters can devastate the economy, rolling back
development gains for the country and exacerbating global
inequality. Protecting the GDP is critical for such countries,
which are more vulnerable, proportionally, to the losses from
large-scale events than developed economies.
Section 1
Region Economic losses Insured losses Protection Gap
in USD bn in USD bn % of economic
loses
in USD bn % of economic
loses
Latin America & Caribbean . . .% . .%
Africa . . .% . .%
Asia . . .% . .%
North America . . .% . .%
Europe . . .% . .%
Table 1.1 The Protection Gap in developing and developed economies,
Under-insurance is also a concern for developed economies.,
While the insurance market is developed in these economies,
the protection gap is also increasing. Such countries typically
have high-value assets and infrastructure that, when under-
insured, contribute to a wide protection gap. For example,
in the aermath of Hurricane Harvey only some % of
eligible domestic properties in Texas and Louisiana had flood
insurance. Hence, despite their relative economic strength,
citizens, businesses and governments in such counties
also suer devastating losses that contribute to increased
inequality in the aermath of a disaster, as shown in Table ..
Specifying the protection gap. While the protection gap is
a global problem, aecting all countries, and referring to the
whole uninsured population, the uniform use of the term is
problematic. Typically, states and markets direct their eorts
to address the protection gap at specic and local protection
gaps, rather than aiming to solve the overall problem of under-
insurance. We will see examples in this report of such dierent
protection gaps as: a lack of reinsurance capital available to
insurers who write terrorism cover for city-centre business
districts; insucient emergency capital reserves in developing
economies to maintain essential services aer natural disaster;
or unaordable premiums for homeowners in highly exposed
flood plains, or in earthquake-prone regions (even in mature
insurance markets).
The protection gap as a generic term refers to all uninsured
risk, but the initiatives attempting to address the protection
gap are specic to particular social, political or economic
problems caused by under-insurance in a region. In this report,
therefore, we either specify “the global protection gap” or we
use the term “protection gap” in relation to specic social,
political or economic problems in a region, which may be
addressed by insurance.
. STAKEHOLDERS OVERVIEW
The existence of a protection gap shows, by denition, that
market mechanisms alone are failing to protect against
the specic peril involved. This prompts governments and
inter-governmental organizations to intervene in the market,
and generate some form of PGE. These PGEs draw together
government, market and non-market stakeholders with
dierent expertise and interests relevant to the specic
protection gap problem. They may include:
Insurance market stakeholders: In markets for risk, such
as flood and terrorism, insurers accept the responsibility
for paying claims for post-disaster reconstruction in return
for a premium. Some part of these insurer’s risk of payout is
transferred to reinsurers and other capital providers through
a range of nancial products. Insurers pay a premium to these
capital providers, who in exchange pay a share of the large-
scale claims incurred by the insurers following a disaster.
Reinsurance capital therefore allows insurance companies to
remain solvent aer major claims events. This risk transfer is
facilitated by reinsurance brokers. As procuring agents, they
support the transfer of both capital and information amongst
these parties, helping to make these high-value, highly-
complex deals work.
Modellers: Catastrophe models are used to estimate the
likelihood and severity of nancial loss from catastrophes
before they occur. Modellers in academia, government and
industry provide analyses that inform the sophisticated
mathematical and computing models used to understand and
price the risks being transferred. Their work on the impact of
disasters can also inform improved resilience to disaster.
Government: Central government has political, social and
economic reasons to protect its citizens. Its engagement as
a stakeholder usually takes place through various entities,
including its Financial Ministries, which aim to protect the
government balance sheet; and its Environmental and
Agricultural departments, which promote disaster mitigation
and resilience. Government is a key stakeholder in the
protection gap problem and usually initiates the development
of a PGE. Oen, government departments are gatekeepers of
fragmented, but crucial data for understanding disasters.
Inter-governmental organizations: Given the global and
interconnected nature of the protection gap problem, inter-
governmental organizations can promote mutual interests
across countries in dealing with disasters. Developmental
Banks (e.g. World Bank) and International Development
Organizations (e.g. the UK’s Department for International
Development – DFID) are key actors that bring together
countries, particularly in the developing world, through
PGEs that develop market-based mechanisms to narrow the
protection gap.
Adapted from Swiss Re Institute ,
Between State and Market: Protection
Gap Entities and Catastrophic Risk
Many of these various stakeholders are to some extent already
interdependent, in ways which will shape the activities of the
PGE in which they participate – but the PGE’s activities will in
turn modify those interdependencies.
. SOCIAL AND MARKET IMPERATIVES
These various stakeholders are characterised by a mix of
diering market and social objectives.
Governments tend to have a socially-focused understanding
of the problem, with an objective to protect their citizens and
communities from disasters: “Government’s role is to protect
its citizens and its communities. We have a social responsibility
with the local population. At the end this is our main mandate;
to support them.” (Government Stakeholder)
The (re)insurance industry stakeholders have a market-focused
understanding of the problem, with market objectives that can
oen clash with the social objectives. For example, insurers
charge premiums that reflect the risk they take and thus for
high-risk areas they charge a market price that is oen high
enough to make insurance unaordable:
“Financially it doesn’t make sense [to provide low price
insurance cover for high-risk areas]. I do get the social thing
and there’s a responsibility and a concern for the industry.
Aordability is always a concern for the industry but at the
same time, we’re publicly listed companies, we are not charities.
We have shareholders, so we have to charge an appropriate
premium.” (Market stakeholder)
PGEs are oen formed through joint action between the
government and/or intergovernmental organizations on
the one side; and various market organizations on the
other. Their mandate oen requires them to pursue ‘social’
objectives through market means. PGEs sit at the nexus of a
range of stakeholders, oen coordinating or combining these
market and social objectives. Therefore, the creation of PGEs
introduces a new type of actor, operating on a market basis but
with a clear social mission.
. BRIDGING AND REDUCING THE GAP
When formed, a PGE’s primary mandate is to provide the
capital to support recovery following a disaster. Such initiatives
help bridge the protection gap by providing nancial solutions.
However, if nancial solutions lead purely to reconstruction
of what was destroyed, they leave the underlying vulnerability
unchanged. Therefore, PGEs social objectives oen push them
to attempt to reduce the gap by reducing vulnerability through
more resilient forms of building and other initiatives. PGEs thus
oen face a tension between these two ways of addressing
a protection gap - by bridging it (nding insurance solutions
with greater reach than the existing ones) or by reducing it (risk
reduction through either removing risk or bringing more risk
within tradable range).
. REPORT MAP
The report is structured in sections and an appendix.
Section 1 is the introduction of this report (current section).
Section 2 outlines PGE origination, in which multiple
stakeholders must come together with their dierent interests.
Three objectives of PGEs are introduced, alongside examples
of how PGEs pursue them. How PGEs evolve in relation to
these objectives, and the key challenges which commonly
arise in pursuit of them is then discussed.
Section 3 is about PGEs in developed economies with mature
insurance markets. It presents a Protection Gap Strategic
Response Framework that can be used to dierentiate PGEs
according to the way they intervene in the market for risk.
We identify the pros and cons of three dierent types of PGE
strategic positions in the value chain for risk transfer, providing
a practical model for PGE design.
Section 4 is about PGEs triggered by a desire to protect
sovereign states, with fragile economies that are highly
exposed to natural catastrophe, and where there is low
insurance penetration. We explain how the particular
demands of such economies for immediate capital to cope
with crisis post-disaster has led to innovations in protection
gap products.
Section 5 discusses PGE contributions to disaster resilience.
The PGE Resilience Framework is introduced to examine
how PGEs can support key aspects of resilience. Areas and
challenges that are beyond the direct control or influence of
PGEs are discussed.
Section 6 introduces the PGE Evolution Framework to identify
key mismatches between dierent stakeholder interests, the
evolving protection gap, and the remit of the PGE. Ways of
resolving these mismatches are suggested.
Section 7 issues a call to arms to learn from and make
better use of these already established PGEs to better
understand and address the growing threat of natural and
man-made disasters.
The Appendix presents the methodology of the research study
and a glossary with key terms used in this report.
Section 2
SECTION . Origination and Objectives
of Protection Gap Entities
(PGEs)
Despite the great variety of local circumstances,, and the
highly variable design and remit of the resulting PGEs, our
research study has found that PGEs typically originate at the
instigation of government in consultation with market players.
They come about at their outset to full objectives that fall into
three categories.
Objective 1: Resolve disruption in (re)insurance supply in
mature markets;
Objective 2: Mitigate the threat of unaordable insurance
in mature markets;
Objective 3: Increase the nancial viability of sovereign
states with fragile economies.
We nd that these objectives not only shape the remit and
strategies of the PGEs at their point of origin. They also act
as anchors for the evolution of PGEs. While PGEs are typically
established with a relatively narrow mandate to act on a
specic local gap, some PGEs have evolved over time to tackle
changes in the nature of their protection gap.
. PGE ORIGINATION: AN ‘UNEASY’ TRUCE AMONGST
STAKEHOLDERS
Government has a socio-political and oen an economic
interest in protecting its citizens and communities. When
part of the society is not protected by the existing market
mechanisms, it has the legislative power to bring about
change. One course of action may be to set up a PGE to full
these public interests through the implementation of a market
solution – nding a way to provide insurance.
Establishing a PGE is inevitably a complex and sometimes
protracted process, as can be seen in our case study of the
formation of Flood Re in the UK. It requires negotiations
between multiple stakeholders that have dierent world views,
dierent technical understanding of risk, and crucially, dierent
objectives – social and/or commercial.
It can take many iterations to engineer what can be an ‘uneasy’
truce between these stakeholders into the founding remit of a
PGE. However, only once it begins to operate will the dierent
stakeholders really see how their interests are served by the
PGE and its remit. As a result, the remit of the PGE may have to
evolve; a recurrent theme in this report.
The origin of Flood Re: A ‘truce’ between state and
market stakeholders
The autumn of was one of the wettest ever recorded
in the United Kingdom and prompted a dialogue between
the Government and the insurance industry about how
to manage growing flood losses. Working through their
professional body, the Association of British Insurers (ABI)
the industry drew up a Statement of Principles (SoP) with
Government to stabilize flood insurance provision. Under
this agreement, existing insurer members of the ABI
promised to continue insuring properties at high risk of
flooding and the Government on its side would continue to
invest in flood defences.
The SoP was renewed in June to last until July .
Aer a Flood Summit in September , three working
groups comprising representatives from Government, the
Environment Agency, the insurance industry and related
organizations met regularly to try to agree key principles
for a shared approach to managing flood risk, despite
considerable scepticism about each other’s motivations.
In the move to austerity, government spending on flood
defences had been reduced; a move that was not well
received by insurers. At the same time, a two-tier insurance
market had been created by the SoP, as new insurance
companies entering the market did not have to insure
properties in high risk flood areas, and so could adopt
dierent strategies about which consumers to target. The
SoP conditions were aecting the competitive state of the
market.
Over the following two years there were heated negotiations
about the dierent government and industry objectives.
Insurers preferred a free market where they could decide
which risks made sense to trade commercially, and at what
price. Government did not want high prices to be a deterrent,
as they wanted to make sure that high-risk areas were
covered by insurance. Yet they did not want to take any extra
liability on the public purse.
For insurers, there was the real threat that Government could
legislate an obligation to provide flood insurance, while for
Government, the threat of a free market could leave high-risk
properties uninsured. Leaving a swathe of such properties
uninsured was not good for the reputation of the insurance
industry either, as an insurance stakeholder argued: “the
incentive for us was reputational; we knew there was a
Between State and Market: Protection
Gap Entities and Catastrophic Risk
problem, the market wasn’t servicing these very high flood
risk homes, so how could we deal with that, rather than just
saying it’s too high a risk, we’re not going to insure you”. A
solution had to be found.
The insurance industry drew on modelling and data from
insurers to understand the extent of the problem: how
many homes were at high risk of flood? They also examined
models for the setting up of a PGE, such as the California
Earthquake Authority (CEA), National Flood Insurance
Program (NFIP), and the UK terrorism pool. The Government
proposed dierent options, not wanting to leave that task
completely to the industry. Aer much push back from both
sides, the compromise Flood Re model was agreed in .
Flood Re would be run by the insurance industry, supported
by government legislation to levy all policyholders to fund
the pool, and the Government would increase spending on
flood defences.
More working groups brought together the ABI, representing
the industry, the Department for Environment, Food and
Rural Aairs (DEFRA) for the legislation, and other insurance
market parties such as modellers, brokers and reinsurers.
The oen messy elements of the emerging solution were
presented to a committee of MPs, progressively establishing
the legislation that would be passed as a Bill in UK
parliament on – a full een years aer the start of
discussions. Flood Re was born, and now had to execute
its remit.
. OBJECTIVE : RESOLVE DISRUPTION IN REINSURANCE
SUPPLY IN MATURE MARKETS
Extreme events which cause unexpectedly high losses can
result in sudden large-scale disruption, or even failure, in the
supply of (re)insurance for a specic peril. This is because:
a) unexpected large losses can jeopardize (re)insurers’
capital reserves and thus the ability to pay claims; and
b) industry participants may lose condence in their ability
to quantify and manage exposure to such events.
In such situations, insurers may stop oering policies so that
citizens and/or business cannot get insurance cover. This
disruption may be compounded when reinsurers stop oering
capital to insurers to cover catastrophic losses; the result is
insurers not being able to remain solvent to cover the risk of
a large-scale loss themselves. For example, Pool Re in the
UK was formed following the terrorist bombing of the Baltic
Exchange in London, which disrupted the supply of insurance
and reinsurance capital to cover terrorist attack on high-value
commercial properties.
Supply disruption in either the primary or secondary market
translates into short-term disruption in the ability to transfer
risk, and longer-term loss of trust in the market as a means to
respond to disasters. Such disruption creates a socioeconomic
crisis that may spur the establishment of a PGE whose
objective is to restore supply as quickly as possible, as shown
in the CEA case study (below).
The case of CEA: Solving the problem of supply
The Northridge . magnitude earthquake in southern
California caused insured losses of . billion USD. This
reportedly equated to more than years of premiums for
earthquake in California: a clear message to the market that
existing earthquake pricing had not reflected the actual risk.
Residential insurers in California were concerned about their
ability to price correctly for earthquake risk, for two reasons:
• They could not simply stop oering it because, since
, it was (and remains) mandatory by California law
to oer earthquake cover with all residential-property-
insurance policies. However, oerees are free to decline
the oer, and lenders do not require earthquake cover as a
condition of mortgage issuance.
• California in enacted a new, strict, insurance-rate-
regulation law that dramatically restricted insurers’ ability
to implement rates for personal lines insurance, such as
home insurance.
The eect of Northridge in on the primary home-
insurance market was rapid and severe: insurers simply
withdrew, or severely restricted the availability of, new home-
insurance policies for any peril, not just earthquake. This
market constriction eventually extended to almost % of
the California home-insurance market. Californians requiring
a new home-insurance policy were threatened with inability
to insure new homes.
A local protection gap for home insurance in California was
created. While supply was disrupted, there was still strong
demand, as homeowners could not get a mortgage without
a home insurance policy in place. Although mortgagees
did not, and do not, require mortgagors to buy earthquake
insurance, they still required insurance on a range of perils
beyond earthquake.
Faced with a severe crisis, the State of California, led
by the insurance commissioner, worked with insurance
markets and the state Legislature to nd a solution. This
led to the creation of the California Earthquake Authority
(CEA), a privately funded, publicly managed, not-for-prot
PGE to provide the earthquake cover written when a CEA
participating insurer’s home-insurance customer accepts
that insurer’s mandatory earthquake-insurance oer.
Insurers pay a charge to participate in the CEA and then
remain liable for a signicant part of CEA earthquake
losses. Participation in the CEA is not mandatory for
insurers. Neither is it a mandatory purchase for renters or
homeowners, who can buy a home-insurance policy without
taking up earthquake cover at all. However, where an oer
of CEA cover is taken up, CEA receives premiums which, by
law, must be based on rates that reflect all actual costs of
providing the selected insurance: actuarially sound rates.
Currently, % of the insurance market participates in the
CEA, while about % of residential earthquake insurance in
California is still provided through the private market.
The CEA solved the Northridge-related supply problem
of obtaining residential insurance in California, since it
Section 2
directly enabled CEA participating insurers to come back to
the market. CEA participating insurers could now provide
the mandatory oer of earthquake cover, without having
to be exposed to the full expense of providing the cover
themselves. Residents could now buy home-insurance
policies that enabled them to meet insurance requirements
for new home-purchase mortgages. In this way, the very
specic protection gap of home insurance was bridged. Yet
the long-standing issue of an earthquake protection gap
in California was unresolved. Even before the Northridge
earthquake State-wide residential earthquake take-up was
only some %. Despite the CEA’s best eorts, only about
% of households with home insurance in California today
actually take up earthquake insurance.
Despite concerted eorts to communicate to California
residents the risks that are associated with earthquakes, the
CEA has remained largely dened by its initial remit. It lacks
eective powers to solve the low take-up of earthquake
insurance itself and has not encouraged or proposed
legislative changes such as mandatory take-up.
Evolving gaps and evolving PGEs. PGEs such as the CEA or
Pool Re were set up as an emergency response to a specic
local protection gap, sometimes, like in CEA’s case, connected
to but dierent from the catastrophe gap; sudden contraction
in the supply of (re)insurance for a specic peril in a specic
region, threatening key aspects of economic life.
While PGEs such as CEA and Pool Re are oen successful in
solving the specic problem they were set up to address, they
typically have to evolve, as the related protection gaps emerge
and change over time. For example, as shown in the CEA case
study, the CEA solved the supply problem of homeowner
insurance in California but did not solve problem of earthquake
under-insurance – a related, but ongoing protection gap.
Terrorism PGEs around the world have also faced a change
in the nature of their protection gap, as terrorism itself has
evolved from ‘conventional’ terrorism (such as bombs), to
what is termed CBRN (chemical, biological, radiological and
nuclear attack), to the increasing threat of cyberterrorism and
the rise of small-scale attacks on citizens. These unmodeled
and potentially catastrophic new risks have also been under-
insured by the market, so terrorism PGEs have had to evolve in
order maintain (re)insurance supply.
Solving a temporary supply disruption or an evolving
protection gap? PGEs established to meet Objective aim to
provide capital for a sudden gap in cover, in which existing
demand for insurance is not being met. The PGE’s remit may
(or may not, as in the case study of the CEA) be additionally
intended to increase demand for insurance, or to grow
protection in under-insured areas or for under-insured perils.
Such a remit would imply a longer-term project rather than
simply xing the supply disruption.
Furthermore, supply disruption may be temporary: for instance,
as modelling improves, or the risk is reduced, it is possible that
supply returns to the market. Alternatively, the gap in cover
may evolve, to require new cover which is also in short supply
(as in the case of terrorism). Either eventuality poses questions
in relation to the long-term future of the PGE. Should the remit
be limited to xing the initial supply disruption, aer which the
PGE should wind down? Or should the remit have the flexibility
to evolve alongside the protection gap itself?
. OBJECTIVE : MITIGATE THE THREAT OF
UNAFFORDABLE INSURANCE IN MATURE MARKETS
The combination of more frequent extreme weather events,
high levels of urbanization in weather-exposed areas, and
increasingly sophisticated risk modelling able to pinpoint risks
at an ever-higher level of detail has resulted in a growing
number of insureds in mature insurance markets falling into
the “high-risk” category. Insurance products which are priced
to reflect this high risk may ‘price out’ potential policyholders.
The protection gap created by unaordable insurance is a
social problem. In the aermath of disaster, for instance, social
inequality is widened between those who were covered, and
those who were not. Furthermore, the under-insured parts of
society in high-risk areas may not be able to recover nancially,
unless recovery is met from the public purse. The public purse
is, of course, not a neutral pot of money, but rather represents
subsidization of the under-insured indirectly by taxpayers.
Thus, while pricing constitutes a critical market mechanism, it
also carries considerable social implications.
Yet market pricing should also ensure that insurance
companies remain solvent and continue to play their role
in absorbing society’s risk. As the Northridge earthquake
in California indicates, if insurers are not adequately
compensated for the policies they provide, they risk nancial
collapse when losses need to be paid. Clearly, robust insurers
that can pay claims are critical to the protection of their
policyholders – another social implication of pricing.
It is therefore important to address this problem eectively
from both a market and social perspective. This explains why
governments sometimes attempt to solve the problem by
establishing PGEs that make insurance products aordable to
those in high-risk areas. “Repeated exposure to risk means
that the insurance products become unaordable to citizens
and consumers in developed economies, and so governments
step in with dierent kinds of mechanisms to suppress pricing
or spread pricing in such a way that disadvantaged citizens in
high risk areas can continue to get insurance.”
(Inter-governmental stakeholder)
Lower prices for those at high risk may be counteracted
by spreading the true cost of the highest risk cover across
the wider pool of insureds. This may involve a range of
mechanisms from mandatory insurance for all citizens to levies
on lower-risk policies. Examples of such PGEs are Flood Re in
the UK, the Earthquake Commission in New Zealand (EQC) and
CCR in France, which latter, as a multi-peril PGE, can underpin
insurance provision in highly-exposed areas across a range
of risks.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
PGEs that operate with this remit face the substantial challenge
of not unwittingly creating larger problems.
. Suppressing market signals about high-risk areas:
Increasing prices reflects high risk areas and gives an
important market signal about the growing exposure to a
particular peril in these areas. Suppressing such signals,
by smoothing the price to make insurance aordable
and available, does not discourage people from living in
such areas where the frequency and severity of weather
events is growing. Nor does it incentivize citizens and
governments to increase resilience of the natural and built
environment.
To counteract this problem, PGEs need to be part of an
integrated solution to the problem of highly-exposed risk.
This should include access to policy tools that enhance
resilience, and discourage unsustainable behaviours.
Examples are policies requiring properties that suer
repeated losses to be constructed in more resilient ways –
such as building houses on stilts in flood areas – in order
not to be excluded from the scheme. Flood Re in the UK
goes further, excluding properties built aer from the
scheme to discourage subsidization of new developments
in flood-prone areas.
. Aordable insurance unconditionally? High-risk
policyholders that can benet from a PGE may range
across a wide socio-economic spectrum. At one end of the
spectrum are people of low socio-economic means, who
do not have the capacity to move to safer places or aord
high premiums. At the other end are those who have the
socio-economic means, but who by choice live in desirable
but highly risk-exposed areas, such as sea- and river-
fronts. Cross-subsidization of both types of policyholders
may not be an equally desirable social objective. The
potential for social and nancial inequities and the
related social and political discontent should be strongly
considered when establishing PGEs schemes.
Evolving gaps and evolving PGEs. In general, PGEs
established as a response to the threat of unaordable
premiums for high-risk policy holders do address this
protection gap, as there are specic mechanisms available
to make insurance more aordable (see Section ..).
In addition, where accompanied by mandatory insurance
legislation, they can go a long way towards ensuring that
all citizens receive at least some cover from catastrophes to
which a country is particularly prone. This has been a particular
success of the EQC.
These PGEs also face a challenge to keep up with evolving
protection gaps. They need to be watchful about unintended
social/political consequences of their actions that may
impact upon the protection gap; for example avoid distorting
incentives so that the resilience of the natural and built
environment in risk-prone areas is not compromised. Such
actions might lead the protection gap to widen in new and
initially unexpected directions.
. OBJECTIVE : INCREASE THE FINANCIAL VIABILITY OF
SOVEREIGN STATES WITH FRAGILE ECONOMIES
PGEs are established to protect sovereigns in fragile
economies that have high exposure to natural catastrophe,
and where there is little insurance penetration. These PGEs,
as shown by the following example of the Africa Risk Capacity
(ARC), typically introduce insurance products designed to
provide a rapid injection of capital in the immediate aermath
of a disaster, rather than to support reconstruction. These
products are innovative in addressing a particular protection
gap; the gap between a need for a rapid humanitarian
response, such as providing food, clean water, or shelter aer
a disaster, and the capability of a sovereign in a low-income
economy to meet those costs.
Mauritania is impacted by drought on a frequent basis. It was
among the rst countries to purchase an insurance policy from
ARC for an estimated premium of ,, for a total cover
of ,, for the agricultural season from July through
November . The product worked. Aer a very poor
rainy season, Mauritania received a payout of approximately
,, in January . They were able to use the rapid
payment to alleviate a humanitarian crisis, providing ,
households with kilograms of rice and litres of oil each
over months (April – July ). The timeliness of the
payout had a positive impact; protecting livelihoods, and also
prevented migration and the distressed sale of livestock.
Countries like Mauritania are at risk of severe economic
damage because of the combination of extreme and frequent
weather events, low levels of resilience and increasingly high
levels of public debt. This has resulted in two protection gaps
in low-penetration insurance markets with fragile economies.
First, there is the disaster liquidity gap: the lack of capital in the
immediate aermath of disaster to provide essential services
whilst awaiting the arrival of humanitarian aid. Second there is
the reconstruction gap: the lack of capital to pay for properties
and infrastructure to be rebuilt. It is inevitable in such countries
that the funds for reconstruction aer a disaster will need
to come from international aid. However, at the point of
catastrophe, immediate disaster relief is a priority requiring
relatively small funds - and that cover can be a viable and
aordable product for a PGE to support.
Governments, inter-governmental organizations such as the
World Bank, and, increasingly, aid organizations are therefore
turning to various forms of disaster-liquidity insurance for
low-income countries. Some of these eorts have resulted in
the establishment of PGEs, including both multi-country risk
pools such as the African Risk Capacity (ARC), the Caribbean
Catastrophe Risk Insurance Facility (CCRIF) and the Pacic
Catastrophe Risk Insurance Company (PCRIC); and single-
country entities, such as FONDEN in Mexico. In all of these
PGEs, sovereign states pay a premium to purchase insurance
products that are backed up with market capital, and which
will provide an injection of capital immediately aer a
catastrophic event.
Section 2
The insured is the state, not private citizens or businesses;
and the state may use the payout for any post-disaster
requirements, such as petrol for generators, payment of the
civil service, or the relocation of citizens. The knowledge that
their budget is protected in this way enables sovereigns to plan
for disaster in advance, while the immediate timing of the
payment at the point of the crisis is considered to reduce the
longer-term humanitarian impact of the disaster.
Evolving gaps and evolving PGEs. The PGEs operating in
this area were set up with a very specic remit, focused on a
particular protection gap alicting sovereign states rather than
private citizens or businesses, and providing capital liquidity
rather than funding post-disaster rebuilding.i The countries
involved oen have little technical expertise in assessing
vulnerability and exposure, and poor data availability – factors
which relate to low insurance penetration (for more details see
Section ). These factors point PGEs towards a development
trajectory, with a direct emphasis on developing knowledge
and ability to model the peril; developing the ability to prepare
for it nancially; and improving national resilience. As risk-
modelling capabilities and nancial literacy evolve in such
contexts, a sovereign generally improves its ability to identify
other insurable types of risk. PGEs play a key role in helping
their country members understand both the protection gap
itself, and how insurance-based products can incentivize the
growth of nancial and physical resilience to disasters.
Objectives for PGEs:
. Resolve disruption in (re)insurance supply in mature
markets;
. Mitigate the threat of unaordable insurance in mature
markets;
. Increase the nancial viability of sovereign states with
fragile economies.
Evolution of PGEs and protection gaps:
• PGEs should evolve their remit and operating mechanisms
in parallel with the evolving nature of the protection gap
that they are addressing
. BEYOND MARKETS: PGES’ ROLES IN FRAMING THE
PROTECTION GAP DEBATE AND IN BUILDING EXPERTISE
The previous sections have provided an analysis of PGEs’ roles
in marrying socio-political interests with market mechanisms to
bridge particular localised protection gaps. Their role includes
enabling the market to continue to trade in particularly volatile
or uncertain risks; suppressing price for high-risk insureds; and
developing new markets and products, as well as establishing
the conditions for the market to grow in economies with low
insurance penetration.
PGEs’ roles in addressing the protection gap, however, goes
beyond their direct market eects, to encompass the following
eects on risk framing and developing expertise.
Framing eect. PGEs come about through considerable
work from a range of stakeholders, all of which have
dierent objectives and interests in addressing a protection
gap. Indeed, as we showed at the start of this Section, the
establishment of any PGE will be an ‘uneasy truce’ at a moment
in time between parties. Because of the process through which
they are established, PGEs have a central position with direct
ties to these multiple stakeholders. They thus become a critical
point of interdependence between stakeholders.
As the eects of shiing population distributions, climate
change, and rising inequality change the nature of exposure,
the interests and objectives of stakeholders’ shi. Being at the
nexus of the stakeholders, PGEs become a centre for debate
or informed dialogue amongst them. PGEs, therefore, become
critical actors in framing an evolving understanding of the
protection gap, such as what risks remain under-insured, how
they might be addressed, and who should be responsible.
As a result, even where they have been established as a
temporary solution, PGEs tend to persist. For example, Pool
Re, initially established to cover the risk of bomb blasts in
the city of London, has, as explained above, been part of the
evolving cover for new types of terrorist risks. As they align with
the interests of dierent stakeholders over time they continue
to evolve and to be part of the way that the protection gap is
identied and dened within society.
Expertise eect. PGEs need to trade risk that is either not
insured, or would be priced out of aordability, according
to the existing technical expertise in the market. A new
understanding is needed in order to nd a rationale on which
to trade - and PGEs must therefore develop new technical
expertise, or nd and co-ordinate existing expertise which
may not have been available to the market. For instance,
as part of its initial remit, Flood Re needed to quantify the
potential number of properties at severe risk of flood in the
UK. This entailed combining existing technical expertise about
UK flood peril from dierent parties including insurers, public
databases, modelling companies and the Environment Agency.
In another example, part of the mission of the Pacic
Catastrophe Risk Assessment and Financing Initiative (PCRAFI)
is to develop new technical expertise in understanding the
earthquake and cyclone peril in the Pacic Ocean and quantify
the potential loss from these perils for its sovereign-state
members.
“We had hazard data for small local areas. What we didn’t
have was nationwide hazard information and exposure data.
The data that countries and inter-governmental organizations
had wasn’t sucient to do the work. Initial modelling showed
that the data [even when combined] wasn’t robust enough to
explain the actual losses. That really made the push to look at
how could we improve the data collection and modelling for
the Pacic.” (PGE stakeholder)
PGEs have an incentive and indeed, a unique position in
lling a knowledge gap that oen accompanies a protection
gap. However, these activities have consequences for the
i One exception is FONDEN, which also purchases a reinsurance product from the global market to cover the costs of reconstructing infrastructure.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
competitive dynamics of insurance markets. Technical
expertise can be a source of competitive advantage for market
players, since the ability to quantify and price risk is central to
the ability to trade it. For example, private sector (re)insurers
regard their modelling capability, and technical expertise in
trading risk as one way in which they are better able to price
risk and so, outperform their competitors. This has led to the
development of a market in which data and models are usually
owned and licensed within the private sector. In contrast, a
PGE builds or expands a body of technical expertise with a
view to trade risk that is under-insured or uninsured and hence
not, initially at least, within the market. Its remit and interests
are not necessarily prot-oriented, but aimed at increasing the
spread of insurance for nancial protection as a social good.
Such interests might be served by promoting an open-sharing
approach for its modelling capability in order to promote
wider risk cover and enable more parties to bridge the
protection gap.
In these ways, PGEs can address barriers to knowledge
sharing, overcome knowledge decits and develop a ‘pre-
competitive’ space for technical expertise. However, they can
also generate tensions between the open and proprietary
nature of that technical expertise.
PGEs eects go beyond their direct eect on markets:
. Framing eect: by being at the nexus of multiple
stakeholders, PGEs have a generative role in framing
the debate and evolving understandings about how to
address the protection gap.
. Expertise eect: PGEs generate new knowledge, and
oen make it publicly available as a social good. This
might alter the distribution of expertise in the market, as
well as the incentives for the development of expertise
across the various stakeholders.
Section 3
SECTION . In the presence of a
mature insurance market:
Market dynamics
PGEs typically show considerable variation in governance
structures (e.g. public, private, partnership), risks covered (e.g.
single peril or multi-peril), type of risk solution (e.g. product
used) and their funding model (e.g. policy holders’ premiums,
public or private levy).,, Indeed, these distinctions are oen
so particular to the point of origin and the national political
economy that we may ask: what can we learn from the study of
PGEs, and what can they learn from each other?
We have found that PGEs have important common underlying
principles on how they respond, strategically, to the protection
gap. When PGEs operate in developed economies (Objectives
& in Section ), they must sit alongside the established (re)
insurance market. This raises critical strategic issues:
• Types of market intervention: how they share risk with the
existing market players
• Positions in the value chain: where they sit within the value
chain for risk transfer
Classification of PGEs by their strategic choices
Types of market intervention:
• Removing risk
• Redistributing risk
• Combining risk removal and risk redistribution
Positions in the value chain:
• Insurer PGE
• Reinsurer PGE
• Market Capture PGE
. PROTECTION GAP STRATEGIC RESPONSE
FRAMEWORK: TYPES OF MARKET INTERVENTIONS
Our research study shows that PGEs emphasize primarily either
) removing risk, or ) redistributing risk as their key means of
market intervention (see Figure .).
.. Removing risk
This risk refers to a market intervention in which risk is
removed from the market onto the balance sheet of the PGE
or the Government (vertical axis, Figure .). This is particularly
likely for risk that is seen as too volatile or extreme for the
market to take, such as the threat of what is termed non-
conventional or chemical, biological, radiological or nuclear
(CBRN) terrorism, where the potential losses are beyond
Figure 3.1. Protection Gap Strategic Response
Framework
the capacity or risk appetite of the market. A high position
on this dimension would represent a PGE fully removing
the risk of CBRN terrorism from the market. In this scenario,
insurance companies may accept premiums from insureds
for CBRN terrorism risk, so ensuring policies can still be
issued. However, they then pass the entire percentage of that
premium associated with CBRN terrorism risk to the PGE. The
PGE can then provide the cover because it has access to some
government guarantee (limited or unlimited) to pay for losses.ii
While the extreme position on this dimension is removing
the risk fully from the market, responses may also vary along
the continuum, by removing only some of the most extreme
risk. For example, a PGE might remove a ‘top layer’ of risk as
dened by market signals such as high price, or withdrawal
of insurance supply, while risk below a certain threshold is
retained by primary insurers in the usual way.
ii While the specic policy by which the risk is removed may indicate some future levying of higher premiums on the insurance market and policyholders to recover
some future government costs, the actual mechanism is to remove the risk altogether in the rst instance.
High
HighLow Redistributing risk
across all policyholders to smooth price of high-risk ones
Removing risk
from the market onto the PGE/government balance-shee
t
Redistribution
No responce to
Protection Gap
RemovalCombination
Pool Re
for CBRN
terrorism risk
in the UK
CEA for
earthquake
risk in
California
ARPC for
commercial
property
terrorism in
Australia
Flood Re
for flood peril
in the UK
KGV for
multiple perils
in Switzerland
5
1
2
34
1 Remove all risk from the market to the PGE government
2 Remove risk to the PGE and return only some to the market (e.g. through reinsurance or
insurers retention)
3 Redistribute all of the risk across al the policy holders
4 Redistribute some of the risk actross all policy holders
5 Remove risk from the market to the PGE/government and redistribute across all policy
holders
Between State and Market: Protection
Gap Entities and Catastrophic Risk
Risk removal to a PGE is an eective way to maintain supply,
particularly in the face of sudden market withdrawal from
extreme and highly volatile risk (see Objective in Section ).
When there is no market appetite to provide cover, and yet
there remains commercial demand for an insurance product
such as nuclear terrorism cover, removing that most volatile
risk allows market operations on less volatile risk to be
restored, bringing back stability and security in a market.
Challenges. The issue of what and how much risk to remove
from the market is fraught with challenges. We draw attention
to two unintended consequences that emerge when removing
risk from the market:
. Weakening risk appetite of the market: PGEs that
operate through risk removal are oen set up to address
an urgent crisis of undersupply. Under time pressure to
nd a solution, stakeholders tend to transfer all of the risk
related to the specic peril to the PGE – even when some
of the risk could still be traded by the market, at least when
market conditions stabilize. However, without a deliberate
policy to progressively return at least some of the risk to
the market, the market might lose its appetite for that risk.
Not trading the risk for a long time, can result in the market
no longer having the necessary expertise and being
reluctant to make the investment to develop it.
However, insurance markets exist to trade in risk. Exposure
to risk helps them sharpen their knowledge, technical
expertise and innovation for such risk. Removal of some
risks (e.g. terrorism) or parts of such risks (e.g. particular
flood-prone areas) may dull this competitive edge of
the market for this specic peril. “I think the insurance
industry has grown a little bit fat and lazy when it comes
to [specied risk] because they’ve always seen it as [PGE’s]
remit … They just basically said well I can sleep at night, I’m
not taking that risk.” (PGE stakeholder).
. Skewing PGEs/ Governments to higher expected
losses: If the PGE removes only the most volatile risk,
allowing the market to continue trading by ‘cherry-picking’
the ‘easier’ or more protable forms of risk, the PGE
is essentially set up to absorb the problem of adverse
selection in the insurance market. That is, the PGE takes
the small body of highly exposed risk that does not yield
sucient volume or price to be protable for the industry
to trade.
However, if the PGE takes only the highly volatile or
extreme risk, but does not also have access to the wider
body of risk, with which to diversify its own exposure, then
it has a portfolio skewed to higher expected losses that
will be removed to the PGE and ultimately government
balance sheet. This means that the taxpayer, through
the government’s balance-sheet, will cover these higher
expected losses for a) the purposes of allowing a market
to continue to trade; and b) the protection of only a small,
most exposed sector of society. Risk removal models thus
raise critical questions about market subsidization and
social fairness.
These two challenges are a trade-o, since keeping all the risk
within the PGE can dampen market appetite, whilst leaving
some risk with the market can skew government exposure.
Nonetheless, risk removal remains a key strategic response by
which PGEs can address a protection gap, whilst still allowing
the market to function. It can be an immediate response to a
sudden supply failure (see Objective ), aer which PGEs can
seek to scale down their position on the vertical axis in Figure
., gradually returning more risk to the market as the market
problem rebalances.
.. Redistributing risk
This refers to taking the risk of loss by a relatively small group
of highly-exposed policyholders and redistribute it across the
wider pool of variably-exposed policyholders (horizontal axis,
Figure .). It is typically used in situations where risk-reflexive
pricing makes insurance unaordable for policyholders in
highly-exposed areas (see Objective in Section ).
In this model, low-risk policyholders pay a slightly higher
premium than what would truly reflect their risk, in order to
subsidize an aordable premium for those who are highly
exposed to risk. The PGE, typically formed as a pool, takes
the premiums from all policyholders and uses the resultant
revenue to redistribute and smooth pricing across all
policyholders.
Redistribution as a strategic response essentially uses the
PGE to restore traditional models of insurance. Historically,
insurance losses were less predictable, modelled in coarser
detail, and so pricing could not accurately reflect the risk on a
case-by-case basis. Rather, the premiums of the many, widely
distributed across possible exposures, covered the losses
of the few. Today, however, the plethora and granularity of
available data on risk exposure, combined with improvements
in risk modelling in mature insurance markets, has eroded
these traditional ‘risk-sharing’ models of insurance. Properties
that are most exposed to loss can be pinpointed and their
policyholders charged individual premiums that more
accurately reflect their specic risks.
While such models better specify risk, they do not necessarily
address the protection gap. Rather, as noted in Objective
, Section , they may widen it as some highly-exposed
policyholders fall out of the pool of insureds. “But improved
models for whom? We can identify an individual property and
apply rating factors. Are we doing that to provide more cover?
Why are we doing that?” (Industry stakeholder).
PGEs that adopt the strategic response of redistributing risk
attempt to ameliorate these ‘improvements’ in risk-reflexive
pricing by articially creating a wide pool of insureds, in which
the premiums of the many can continue to cover the extreme
losses of the few. However, they can only do so with some
government legislation, typically supported by an underlying
notion of collectivism in the underpinning national culture, that
enables a levy on lower-risk properties to subsidize higher-
risk properties. Eectively the Government must legislate
for a social objective to take some precedence over market
objectives.
Section 3
Challenges. Risk redistribution via a PGE is eective at
counteracting the problems of highly-exposed actors falling
out of the insurance nest. It can support high collective
insurance coverage in a particular country or region, essentially
bridging the underlying protection gap. However, it also comes
with the following challenges.
. Growth of high-exposure areas. Risk redistribution is
predicated on a model of a large number of insureds, in
which only a few are highly exposed. Take the example of
flooding. We know that climate change and/or growing
urbanization increases exposure to natural disasters such
as flooding. This can result in very large or expensive
areas being repeatedly flooded. In such a situation, the
premiums of the many can no longer necessarily outweigh
the losses of the few.
. Redistribution shis focus away from risk mitigation.
In addition, redistributing risk breaks the market chain
of risk, reward and responsibility. Insurance risk models
send a strong price signal that specic risk-prone areas
are only viable to trade if rewarded with a high premium.
However, the redistribution strategy can induce moral
hazard in which those at the highest risk of repeated loss
are not incentivised to reduce their risk, or change risky
behaviours (for example, through structural changes to
their property to mitigate the eects of flood), since they
do not bear the full costs of their exposure.
More broadly, the market signal of high prices can be
seen as the prompt to improve risk mitigation in risk-
prone areas, for instance through changes in planning
permission to require improved methods of building in
highly-exposed areas (see Section ). Thus, bridging the
protection gap through risk redistribution can not only
suppress the price signal, but also direct attention away
from risk mitigation as another means of reducing the
protection gap. In these ways, risk redistribution can, in the
long-term, exacerbate the very problem it was established
to solve.
.. Combination of risk removal and redistribution
Removing and redistributing risk are not necessarily either/
or responses. As indicated by Figure ., PGEs can combine
risk removal and risk redistribution, albeit not necessarily in
equal measures, or on the same elements of risk. Rather they
may take an approach where they remove some elements of
risk and redistribute others. Oen such changes occur in an
evolutionary way. A PGE may initially be established to solve,
for example, the problem of lack of supply for a very volatile
risk, through a strategic removal response. Once supply begins
to return, it might also employ some redistribution of risk.
To better understand this strategic response, we present the
case study of the Australian Reinsurance Pool Corporation
(ARPC) – a PGE that uses this combination approach.
The case of ARPC: Evolving strategic responses to
address a supply failure
The ARPC was formed in to address the limited capital
supply for terrorism risk globally following the terrorist
attacks on the World Trade Centre. Coming relatively late in
the formation of such pools, the Australian government and
the insurance industry were able to examine and learn from
many existing PGEs. Initially, the Terrorism Insurance Act
was established, which addressed the problem of insurance
supply by over-riding any terrorism exclusions in contracts of
insurance covering ‘eligible’ property (as dened in the Act),
requiring the insurance industry to pay insured policyholders
for losses on eligible property if the Australian government
declared a particular incident as a terrorist act. To this
extent, terrorist insurance was thus mandatory for all insured
policyholders because an insurance company would nd
themselves paying for terrorist losses by default, following a
declared terrorist act, even if they did not deliberately oer
such cover for eligible property.
At the same time, the ARPC was set up with a remit to
respond strategically by removing the risk of paying for
terrorism losses in eligible policies from the insurance
industry. Insurers pay ARPC a premium based on three broad
tiers of risk, from highest premium in central city locations,
to suburban, to lowest premium in rural areas. In return,
they are given full removal of terrorism risk, since the ARPC,
operating as a reinsurance company, would pay all the
claims beyond the insurers’ retentions to a limited liability of
billion AUD. ARPC would meet these payments through
its own capital reserve, the additional reinsurance it bought
in global reinsurance markets as a retrocession product,
and thereaer, from the government balance sheet, which
guarantee the Australian government provides ARPC for
a fee.
While it is voluntary for insurance companies to purchase
terrorism treaty reinsurance from ARPC, if they do choose
to do so, they must transfer the risk for all eligible policies
within their portfolio. Given the legislative situation in which
they are liable for commercial property losses arising from
terrorist attack, almost all of the insurance companies in
Australia have taken up the option to transfer their terrorism
risk to the APRC. This approach thus not only uses risk
removal as a strategic response but, because insurance
companies are required to transfer the risk across all of their
eligible policies in Australia, was also a risk redistribution
strategic response. Specically, those at low risk of terrorism
loss cross-subsidized the highest threats, resulting in a
diversied pool of insured commercial risk across higher and
lower threats around the country.
The ARPC was set up with a remit to remain relevant to
problems of market supply through a three-year legislative
review process. At these reviews, the ARPC has consistently
evolved to both widen the denition of which properties
might be dened as eligible under the Act. For example,
at the review the denition of commercial use of a
Between State and Market: Protection
Gap Entities and Catastrophic Risk
property was expanded to include mixed-use residential and
commercial buildings in which at least % of the property
has a commercial use and all buildings with a sum insured
value of million or more.
At the same time, ARPC has scaled down the level of risk
removal, by pushing insurers to retain more of the risk of
terrorism losses on their own balance sheet to a specied
threshold that is aggregated across the industry. This amount
of retention has grown progressively from an initial threshold
of million AUD to million, as an aggregated
industry loss, from July . The maximum retention for
individual insurance companies has increased from
million AUD to . million AUD.
In scaling down risk removal, the ARPC is thus working
to address the problem of industry supply, by working
with the primary insurance industry to help them supply
some terrorism insurance from their own balance sheets.
At the same time, because ARPC buys reinsurance cover
itself, it has brought reinsurance capital to the market,
further addressing the problem of capital supply. Yet the
government balance sheet remains available for losses
above the industry appetite to supply capital.
Hence, a combination of risk removal and risk redistribution,
evolving within a legislative framework of three-year
reviews have ensured that the ARPC continues to evolve
in consultation with its stakeholders in government and
industry, and the wider needs of Australian society.
PGEs typically start with a removal or a redistribution approach
in response to any particular type of protection gap. For
instance, a PGE might fully remove a risk aer a market shock
when supply fails, but then gradually return risk to the market
as the market’s understanding and appetite return. Or a PGE
might introduce risk redistribution in highly risk prone areas,
but then gradually reduce risk redistribution as risk mitigation
eorts in such areas progressively make insurance premiums
aordable again.
Ideally, PGEs would have flexibility along both axes of Figure
.. They should be able to scale up removal aer a major
shock in which insurance industry supply in a particular region
or on a particular peril is disrupted. Conversely, they should be
able to scale up redistribution where overall capacity within a
region is robust, despite some highly-exposed pockets of risk.
A combination approach is potentially very ecient for PGEs
that have a portfolio of dierent types of risks. Multi-peril
PGEs, such as CCR, can take dierent approaches to perils,
emphasising risk removal on some risks and risk redistribution
on others, according to the available modelling capability and
market appetite for each risk.
Challenges. The main challenge for the combination approach
is how to move between the axes of Figure . – usually from a
starting position wholly on either the removal or redistribution
axis. Critically, there needs to be a highly-flexible approach
to the role and remit of the PGE. A combination approach,
particularly one in which removal and redistribution on
any particular risk may be scaled up or down, necessitates
an adaptable and even nimble approach to legislation
and governance, in which multi-party dialogue between
stakeholders remains open.
.. Summary: evolving remit or mission creep?
As Figure . suggests, risk removal and risk redistribution
are dierent strategic responses that a PGE may take to the
management of a specically identied local protection gap.
However, they should not be seen as static positions. There
is no ‘ideal type’ of strategic response. Rather, as the nature
of the protection gap, and the interdependencies between
the key stakeholders evolves, the strategic response of a PGE
needs to evolve to meet the multiple demands upon it.
While PGEs are well placed to develop new responses as risk
evolves, they also need to be at the heart of continuously-
evolving legislation and multi-party dialogue to manage the
multiple dierent interests involved in bridging the protection
gap. Hence, strategic responses are not simply technical issues
about approaches to risk transfer. Rather, a key feature of PGEs
is the need for high competencies in managing the social and
political economy within which approaches to risk transfer
need to evolve. In particular, PGEs need to balance tensions
between demands for an evolving remit and fears that this may
lead to “mission creep” that distorts the competitive market.
. MARKET EFFECT: STRATEGIC POSITIONS IN THE
VALUE CHAIN
To bring market solutions to socio-politically dened
problems, PGEs intervene in the traditional value chain for
risk transfer (see Figure .). We now explain three possible
strategic positions that they may occupy. These positions are
archetypes, meaning that they illustrate general characteristics
of occupying that space in the value chain. The reality of any
particular PGE, as explained in our case examples, may dier
slightly from the archetype. We address the implications of
each archetype in terms of competitive eects, challenges, and
what that type is ‘Best For’.
Figure 3.2. Value chain for risk transfer
Insureds
Primary
market
Insurers
Secondary
market
Reinsurers
Government
Insurer of
last resort
.. Strategic position : The ‘Insurer’ PGE
PGEs can act as an insurer operating in the primary market to
provide policies on a particular risk that is no longer covered
by the market (Objective ); or one for which cover has become
unaordable for those highly exposed to the risk (Objective ).
As shown in Figure ., this means that they provide insurance
policies directly to insureds in return for a premium. They
also buy reinsurance from the private market to cover their
risk exposure.
Section 3
Figure 3.3. Strategic position : Insurer PGE
Insureds
PGE
Primary
market
Insurers
Secondary
market
Reinsurers
Government
Insurer of
last resort
Examples of the Insurer PGE archetype are EQC, CEA, and NFIP.
This position is typically taken by PGEs oering homeowner
policies and has two main strategic implications:
Co-opetition in distribution. Typically, Insurer PGEs develop
their own insurance policies on the risk they deal with and oer
them to policyholders via traditional insurers. Homeowners
oen buy their insurance as a ‘bundled’ product that covers
them from the risk of re, flood, or earthquake, alongside
other perils. Such policies provide a catch-all of potential
risks and, usually, provide the necessary protection of assets
to underpin a homeowner’s mortgage. When an Insurer PGE,
like CEA in California, intervenes in the value chain to bridge
a local protection gap, then a proportion of the homeowners’
insurance cover, namely earthquake cover in California, can be
provided through the PGE.
There are situations, as in California with earthquake, when the
particular cover for this peril can still be provided by traditional
insurers or they can choose to attach CEA’s cover to their
homeowners’ insurance product. Policyholders then select
whether they buy the PGE cover, the private market cover, or,
if purchasing cover is not mandatory, do not buy any policy
for that particular peril. However, in other situations the cover
for a peril is provided exclusively by the PGE. This means that
the Insurer PGE essentially replaces traditional insurers for this
particular cover. For example, the New Zealand Earthquake
Commission (EQC) provides the earthquake component of all
homeowner insurance policies to a particular threshold and
receives a percentage of the premium charged on
those policies.
This position in the value chain involves collaboration between
insurers and the Insurer PGE. The insurer can continue to
provide comprehensive cover through a bundled product
to policyholders, with the help of the PGE that takes some
risk, usually the most volatile, or most exposed. At the same
time, there is also an element of indirect competition as they
both occupy the same position in the value chain and the
PGE provides a product for which the private market might
otherwise develop risk appetite to supply. This mixing of
collaboration and competition is usually known as
co-opetition.
Risk-reward-responsibility eects. While distribution via
insurance companies is ecient, a potential problem arises.
Since insurers don’t take the risk, don’t get the premium
(reward) and also don’t pay the claims (responsibility) for the
proportion of the policy that is provided by the Insurer PGE,
they are not incentivised to compete to sell more of the cover,
and thus ensure take-up of the PGE’s percentage of the policy.
In addition, aer an event, a key feature of the insurance value
chain is that capital flows to the insurers to pay claims. One
of the key functions of primary insurers is to assess damage
in order to manage the eective payment of claims for which
they are responsible. However, again, unless there is a specic
mandated requirement for primary insurers to assess loss,
in order for the PGE to pay claims, or the PGEs develop their
own apparatus for loss adjustment and payment of claims, the
direct chain of responsibility for accurate risk assessment and
distribution of payments is also disrupted.
While not insurmountable, both elements indicate how
Insurer PGEs aect the relationship between risk, reward
and responsibility.
Best For. The Insurer PGE archetype is best for ensuring
personal lines cover to citizens where a region is heavily
exposed to a key peril, such as earthquake or windstorm.
It can ensure that all members of a society have access to
homeowner cover that could otherwise be unaordable for
some. It may be particularly eective for the risk redistribution
strategic response explained in Section ., especially where
cover is either mandatory, or required by lending institutions in
order for policyholders to obtain a mortgage.
.. Strategic Position : The ‘Reinsurer’ PGE
PGEs can act as a reinsurer operating in the secondary market.
In this scenario, insurers operate as normal in the market,
providing insurance products. However, they then transfer that
proportion of risk specically identied as a protection gap,
such as the risk of flood or terrorism, to the PGE. In the event of
a loss, the PGE provides the capital to pay the claims against
the insurance policies for that particular proportion of the risk.
Some examples of this archetype are Flood Re in the UK, ARPC
in Australia, and CCR in France.
Figure 3.4. Strategic position : Reinsurer PGE
Insureds
PGE
Primary
market
Insurers
Secondary
market
Reinsurers
Government
Insurer of
last resort
This position in the value chain has the following strategic
implications:
Preserving the primary market. The Reinsurer PGE
intervenes in the value chain to preserve the ability of the
primary market to trade in two ways:
Between State and Market: Protection
Gap Entities and Catastrophic Risk
. It can address a supply failure in the secondary market
(see Objective ) by removing risk from the market
(Section ..) to the PGE. The Reinsurer PGE provides
the necessary reinsurance capital that the primary market
needs to continue trading if the traditional secondary
market withdraws capital aer a shock, or has insucient
capacity for the scale of the risk. The Reinsurer PGE may
itself buy a retrocession product from the reinsurance
market, as indicated in Figure . by the presence of
the secondary market in the chain; and/or it may have
access to a government balance sheet to underpin its own
exposure (e.g. Pool Re).
. When insurance premiums become unaordable
(see Objective ) the Reinsurer PGE can act as a
price-smoothing mechanism by redistributing risk
(Section ..).
a. The Reinsurer PGE may act as the ‘transformer’ of price
dierences, where it takes all of the personal lines
risk from the primary market, acting as the dominant
reinsurer. It essentially smooths the reinsurance price
by pooling all risks, of high and low exposure, for a
nation, endeavouring to ensure the overall reinsurance
premium received is sucient to cover the overall
national exposure. In doing so, it is able to oer lower
reinsurance premiums to the primary market, since
it needs only to cover its own diversied exposure,
without concern for prot, or higher premiums to oset
the potential volatility of any one primary insurer. It
thus enables primary insurers to also smooth their
own pricing of personal lines in highly-exposed areas
because they can transfer such risk at a lower price to
the Reinsurer PGE (e.g. CCR).
b. Alternatively, only the most highly-exposed risk may be
transferred from insurers to the PGE. In this scenario,
insurers usually charge a below-market price that is
aordable for policyholders. These premiums, for this
highly-exposed risk, are passed on to the PGE, which
is then liable for the claims against that risk. To aord
the payout for those claims, the PGE has to receive a
subsidy from insurers, levied on all policyholders, that
becomes its capital reserve to cover the high risk of
loss in the PGE portfolio (e.g. Flood Re in the UK). Here,
the Reinsurer PGE operates a risk-sharing mechanism
between its own balance sheet and the primary market
through the levy, but can also use other forms of risk
transfer mechanism, such as buying a retrocession
product and/or transferring risk to the Government, to
manage its own exposure.
In all cases, the primary market is enabled to continue trading
in risk, whilst being able to transfer that proportion of risk,
which comprises the basis of an identied protection gap, to
a Reinsurer PGE as the secondary market. The introduction of
a Reinsurer PGE, providing ample reinsurance cover for less
attractive (or new) risks, can also enable primary insurers to
obtain new business that will expand the market. This means
that insurers can underwrite new policies, which as a result can
narrow the local protection gap.
Competitive eects in the secondary market. The Reinsurer
PGE has access to a pool of primary risk that might otherwise
be transferred to reinsurers, particularly as their appetite for
that risk evolves. If reinsurers have appetite but are restricted
from accessing primary market risk, because it all goes to the
PGE, it may be unclear whether the PGE necessarily oers the
best price, since the eects of competition in driving down
price will be restricted. Thus, regulatory oversight is needed to
ensure that the PGE is not acting as a monopoly for primary
market risk that is anti-competitive to the secondary market, or
that drives up prices.
One way this can be managed is if the Reinsurer PGE acts as an
aggregator for individual highly-exposed primary market risk,
and then transfers part of this risk to the secondary market.
As with Flood Re in the UK, the secondary market can both
directly access flood risk from primary insurers and also access
flood risk from the PGE.
Best For. As discussed in Section ., PGEs can scale
up or down their strategic responses of risk removal and
risk redistribution, in line with the changing nature of the
protection gap. While such strategic flexibility necessitates
a skilful coordination of the interdependencies among
stakeholders, who may have conflicting interests, the
Reinsurer PGE can be an eective archetype for responding
to fluctuations in (re)insurance capital, because of its position
between the primary and secondary market.
For instance, aer a market shock caused by a big catastrophe,
there may be a shortage of reinsurance supply in the market.
At this point, the Reinsurer PGE can facilitate risk removal by
scaling up to take more risk from the primary market, which
it can hold, or pass directly to the government balance sheet
following shortage of supply. At other times, when primary
and/or secondary markets recover their risk appetites, the
Reinsurance PGE can scale down its own share, eectively
increasing primary market retention and passing more risk to
the secondary market (e.g. ARPC or Pool Re). In this way it can
enhance risk redistribution across the entire value chain.
.. Strategic Position : The ‘Market Capture’ PGE
This type of PGE ‘captures’ risk throughout the value chain,
operating in both the primary insurer and the secondary
reinsurer spaces, and also accessing the government balance
sheet. It can act as a primary insurer for all, or at least most,
risks in a country, either alongside traditional insurers as an
additional product, or by ‘co-insuring’ on existing products with
the primary market insurers, to support the market’s ability to
provide cover to and pay claims.
Because the PGE holds all premium transferred over many
risks on a national level, it has a highly diversied portfolio
that enables it to optimize its capital reserves. Essentially
premium generated across the portfolio can pay for losses in
one part of the portfolio. Hence, the Market Capture PGE does
not necessarily need to purchase reinsurance but can choose
instead to act as its own reinsurer, due to the benet of a
diverse portfolio of risks that lowers the overall exposure to
losses. This type of PGE is typically a public-sector organization
with access to the government backstop to secure its balance
Section 3
sheet above the capital reserves it can amass through its
privileged access to the primary market.
Figure 3.5. Strategic position : Market Capture PGE
Insureds
PGE
Primary
market
Insurers
Secondary
market
Reinsurers
Government
Insurer of
last resort
This archetype has the following strategic implications:
Comprehensive national protection. This type of strategic
response can be very eective when a consistent national
approach exists to provide the most comprehensive cover to
all citizens at an aordable price. This presupposes that most
citizens are insured, typically through some mandatory form
of cover, and that any private sector insurance companies are
co-opted into sharing cover with the public-sector PGE. This
way, the Market Capture PGE enables the risk redistribution
element of protection at the primary level, whilst enabling risk
removal from the private-sector elements of the market at the
secondary level.
In eect, this is a means of mutualizing risk cover on a national
level to address the protection gap, albeit not necessarily
by transferring it to a market. In such situations, the
mutualization acts as a price-smoothing mechanism, which
may lower overall cost of cover to individual policyholders.
Crowding out private sector provision of catastrophe
insurance. The strength of this strategic response in
providing comprehensive national cover may also bring
about unintended eects, primarily in deterring private sector
insurers from oering products in the catastrophic-risk market.
This occurs in two ways.
. The primary market may lose its competitive or market
interests. We refer to this eect of this type of PGE strategy
as ‘market capture’ with a nod to the concept of ‘regulatory
capture’. When market parties operate in a context
where the Market Capture PGE is dominant, they can
end up relying on the PGE to take the majority of volatile
risk, lose or never develop expertise in trading that risk,
and therefore lose appetite for such risks. Indeed, the
market players may see this as in their own best interests
to protect their balance sheets. This might happen even
where there are governance structures in place to ensure
that the PGE responds to changes in risk appetite in the
industry, and even where the two parties may feel they
have independently negotiated their positions. While,
as noted above, this may have benets, the flipside is
that it dampens the risk appetite and competitive edge
of market players in ways that may stifle market-based
innovation. For example, primary level players do not have
to worry about calculating the risk-reward-responsibility
relationship in measuring their exposure, pricing policies,
and purchasing risk transfer. The risk and responsibility
of paying for that risk is carried anyway by the national
provider. They are therefore not incentivized to improve
their risk modelling and pricing mechanisms in order to
gain a competitive edge.
. A national barrier to competition at the secondary
market level exists. The Market Capture PGE does not
need to access reinsurance, due to its comprehensive
access to wide diversication, underpinned by a
government guarantee. At the same time, the international
reinsurance market cannot oer cover directly to the
primary market, because the primary market trades solely
with the nationally-owned PGE. While this is benecial
from the perspective that the cost of buying reinsurance
is avoided, it also hinders the inflow of capital from global
markets to pay for national losses.
Best For. The Market Capture PGE eectively nationalizes
the insurance market. It is the best for a country wishing to
use insurance-based mechanisms to provide comprehensive
cover for its citizens, and to control the pricing of that cover. It
is thus eective where cover is largely compulsory and where
the primary aim of the PGE is to bridge the protection gap by
ensuring that citizens have the maximum access to widespread
cover. It is particularly suitable for political economies and
historical contexts in which there is a nationalised approach to
public goods and facilities.
An example of the Market Capture PGE is the Spanish
Consorcio de Compensación de Seguros (CCS). CCS is a public
business entity with assets separate from those of the State. It
has provided insurance for a comprehensive set of catastrophic
risks, including earthquake, flooding and terrorism, to both
private citizens and businesses for several decades. CCS
insurance is provided automatically on the back of regular
insurance policies, and nanced with a geographically-flat
“surcharge” calculated on the basis of the capital insured
and type of policy (residential homes, oces, shops, etc.).
The premium is collected via the insurance companies that
provide the policy. While it is not compulsory to have these
“extraordinary” risks covered with CCS, the surcharge is still
due once a regular insurance policy is set up. This system
has enabled Spain to reach a high level of penetration of
catastrophe insurance. Apart from the fact that all policies
are sold by private insurance companies, collaboration with
primary industry players is ensured, as CCS’ -member
Board includes six high-level executives of private insurance
companies. This board structure aims to ensure that CCS only
takes up the risks that the market is not willing to cover, and
indeed there appears to be little interest in expanding into
catastrophic risk in the Spanish insurance market.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
BEST FOR strategic positions in the value chain
Insurer PGE
Best for ensuring cover to citizens in a region or country
heavily exposed to a peril. It can eectively use risk
redistribution as a strategic response especially where cover
is mandatory, or insurance penetration is high.
Reinsurer PGE
Best for responding to fluctuations in (re)insurance capital
supply for a peril. It can eectively move between risk
removal and risk redistribution or combine both responses.
Market Capture PGE
Best for a country that aims to provide comprehensive
insurance cover for its citizens at a controlled price. Suitable
for contexts with a nationalised approach to public goods
and facilities.
. SUMMARY
In section ., we have shown the strategic responses of risk
removal and risk redistribution through which PGEs use market
mechanisms of risk transfer to address the socio-political
objectives for which they have been established. At the same
time, as we show in Section ., such market-based solutions
intervene in the natural functioning of a market in ways that
may have unintended consequences for the market. Hence,
there will oen be tensions between market and socio-political
interests, particularly as PGEs endeavour to evolve alongside
the evolving protection gap.
If the co-opting of robust and healthy (re)insurance markets
is seen as critical to the ability to solve the protection gap, in
collaboration with government initiatives, it is important that
these markets are maintained. At the same time, as we suggest
in our examples of risk redistribution, the intervention of PGEs
can have unintended social consequences that might even
exacerbate inequality. Hence, none of the strategic responses
we discuss, or the positions in the value chain are necessarily
right or best for PGEs. Rather, each response and each
archetype represent specic ways of managing a localised
protection gap that is itself evolving. Critically, therefore,
ongoing dialogue and cooperation between state and market
parties will be important to allow social and market interests to
evolve in tandem.
Section 4
SECTION . Protection for sovereigns:
Insurance as a disaster
liquidity product
One of the objectives that PGEs are set up to address is the
desire to protect sovereign states with fragile economies that
are highly exposed to natural catastrophe, where there is little
insurance penetration (see Objective in Section ). Such
PGEs address a dierent protection gap from those found in
mature insurance markets. Specically, they are not attempting
to fund reconstruction for private citizens and businesses aer
damage and loss. Here, the gap lies between a need for a rapid
humanitarian response aer a disaster and the capability of a
sovereign to meet those costs. The sovereign is the ‘insured’,
taking responsibility for the nation’s risk, and the level of its
cover through the PGE.
The resulting insurance products are not designed, at least
initially, to supply the funds necessary for governments to fully
rebuild critical infrastructure such as roads, water and sewage
systems, hospitals, and schools. Instead, payouts from these
products enable governments to keep running in the aermath
of a disaster and provide key disaster relief functions,,,
particularly in the critical period between disaster striking and
international aid arriving. Because of this, they are sometimes
referred to as disaster liquidity products, which term we use in
this report.
These PGEs are organizational vehicles aiming to:
a) quantify (through catastrophe modelling and strategic
planning) the level of disaster relief funding which might
be needed aer future catastrophes;
b) transfer some risk for such outlay from the government
budget to a capital market.
In this way, they provide sovereigns with autonomy over at
least some of their own disaster response, so reducing their
reliance on aid. “What we’ve had in the past in terms of
a natural disasters, was a heavy reliance on development
partners coming in with aid to assist us in recovery and
to help us work through rebuilding. We looked at this and
decided that we need to do something ourselves in terms of
preparedness.” (Government stakeholder)
Both the PGEs formed, and the relation they address between
risk and payout, dier from those in mature insurance markets,
as we now discuss.
. CALCULATING INSURANCE AS DISASTER LIQUIDITY
Historically, the idea of fullling a need for disaster liquidity via
insurance was new, and needed an innovative solution. That
solution was found by adapting the nancial instrument known
as a Catastrophe (CAT) bond, which was already well-known
to insurance markets. The key point of this type of product
for the purposes of this report, is the parametric basis for
modelling risk and triggering payment. Products that work on
a parametric basis disburse payment not on the basis of actual
losses, but when a particular parameter (for instance, wind
speed in certain locations, or total rainfall, or modelled losses)
or index of parameters reaches a determined threshold
(see Glossary).
This structure allows insurance products to overcome the
limited data available on exposure, vulnerability, and historic
losses in countries which require disaster liquidity. These
disadvantages make it very dicult to estimate risk and create
insurance products for reconstruction of individual buildings
or facilities. Indeed, traditional insurance penetration in such
countries is low.
However, disaster liquidity does not need models which can
estimate detailed damage and loss to property. Instead, it
needs to relate the probability and the severity of a particular
disaster occurring in a specic region (such as a hurricane at
Category passing through the island of Barbados), with the
overall potential damage to, and human consequences for that
region, and the operations needed to restore civil society. In
such models, the parameters would be the hurricane Category
(), and the direct ‘hit’ on a particular area in Barbados. On the
basis of the models, estimates can be made for the funding
needed by the sovereign government to take eective action.
Triggers can then be established that enable a payout of a
particular size if a hurricane does occur at that wind speed
in a set of agreed geographical locations, as declared by an
independent third party.
A key function of disaster-liquidity PGEs is therefore to develop
the necessary modelling capability to relate the risk of dierent
levels of disaster to the level of funding which might be
necessary to keep Government and society functioning.
All parametric products entail the risk of the parameters or
modelled losses not providing an accurate reflection of the
gravity of the disaster (so called ‘basis risk’). Hence these
products have oen evolved from an initial binary trigger – in
which a single specied level of severity must be reached
to release the entire amount of capital. With such a binary
trigger, if a disaster comes in at just under the severity of the
parameters, there would be no payout despite signicant need
for emergency government expenditure. This situation can
undermine a sovereign’s condence in a product. Increasingly,
therefore, there are staggered threshold triggers that enable
partial payments at dierent levels of agreed severity.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
. PGES IN RELATION TO SOCIAL AND MARKET
STAKEHOLDERS
Disaster-liquidity PGEs are typically an outcome of the
collaboration and technical support between:
• one or more sovereign states;
• one or more development organizations, such as the
World Bank;
• donor organizations, such as DFID in the UK;
• scientic and modelling organizations.
The resultant PGE may be the initiative of a single sovereign,
such as FONDEN in Mexico, or, increasingly, multi-sovereign
risk pools, such as CCRIF (see case study), ARC (with a
potential membership of all African nations), or PCRAFI
(providing disaster risk modelling and transfer to Pacic Island
countries). When multiple sovereigns develop a risk pool,
they benet from greater risk diversication and also enhance
their purchasing power in global markets. As indicated in the
case study of the CCRIF, it is critical that sucient sovereigns
remain members of the pool, to ensure this diversication
and critical mass. Negotiating their ongoing purchase of cover
can thus be a source of tension during the annual renewals
between PGEs and their members.
The history of CCRIF: The Caribbean multi-sovereign
risk pool
In Hurricane Ivan devastated Grenada and caused
heavy damage in other areas such as Jamaica and the
Cayman Islands. Recognising that they were all highly
exposed to similar hazards such as tropical cyclones
and earthquakes, the twenty countries of the Caribbean
Community (CARICOM) decided to approach the World Bank
for help to manage their disaster risk. Under the technical
leadership of the World Bank, a multi-sovereign risk pool, the
Caribbean Catastrophe Risk Insurance Facility (CCRIF), was
proposed to transfer some of the risk of disaster into global
reinsurance markets. Collectively, each of these relatively
small countries would benet from shared technical
expertise, diversication across their member states, and
increased reinsurance purchasing power.
Proprietary insurance models for the region were sparse,
so the pool rst needed to develop data and modelling
capabilities to help members quantify their risk. A Japanese
Government grant helped provide scientic expertise which
came up with a novel proposal: to provide disaster liquidity
insurance products based on parametric triggers, through
which member countries could gain an injection of capital
to manage their cash flow in the immediate aermath
of disaster.
With such a novel solution the process of establishing a
coalition of member countries and donors involved intense
outreach, communication and preparation, including two
international conferences. As a result, operating capital was
provided by contributions to a Multi-Donor Trust Fund (MDTF)
from the Governments of Canada, the UK, France, Ireland and
Bermuda; from the European Union, the World Bank, and
the Caribbean Development Bank, and through membership
fees paid by participating governments.
CCRIF thus formed in as the world’s rst multi-
sovereign risk pool providing parametric insurance. In fact, in
its rst year, the new products demonstrated their value with
Saint Lucia and Dominica receiving payouts totalling almost
US million aer an earthquake in November. The next year
Turks and Caicos Islands received a US. million payment
aer the impacts of Hurricane Ike.
Since then CCRIF has not been static. There have been active
eorts to retain and gain members, in part by developing an
increased range of products, such as excess rainfall, all of
which required additional data and modelling. In CCRIF
was restructured into a segregated portfolio company (SPC)
to facilitate an expansion of the CCRIF membership base
the next year. Through partnership, CCRIF started oering
earthquake, tropical cyclone and excess rainfall policies to
Central American governments.
The expansion required additional funding. In , another
MDTF was established by the World Bank to support the
development of CCRIF SPC’s new products for current and
potential members, and facilitate the entry for Central
American countries and additional Caribbean countries.
The MDTF currently channels funds from various donors,
including: the governments of Canada, the United States,
Germany, and the European Union. In , the Caribbean
Development Bank, with resources provided by Mexico,
provided funding for enhanced insurance coverage to the
Bank’s Borrowing Member Countries.
Today CCRIF has member countries from the Caribbean
and from Central America. As of May , CCRIF has
issued total payouts of US million to member
countries, all within days of disaster. During the annual
policy renewal negotiations with member governments,
CCRIF works actively to help members understand their
risk, provide a business case to support them in nding the
necessary nancial allocation in their budgets to continue
buying risk transfer products, and reassure them that CCRIF
continues to remain relevant to members’ needs through
continuous evolution.
CCRIF soon will be introducing three new products; for
drought, agriculture and sheries. It is also discussing the
feasibility of introducing parametric insurance in other
sectors that are adversely impacted by disasters and where
insurance is not always easily accessible on a timely basis.
Product ideas discussed cover other industries, such as
tourism, and public utilities such as power and telecoms.
Through such eorts CCRIF can both broaden the risk pool
and increase its diversity, and, more importantly, evolve to
meet the needs of a wider range of insureds.
The market players to whom the risk is transferred are typically
global reinsurers, not aid or non-prot organizations. They
price and trade risk for prot and issue payouts on the basis
of commercially agreed market triggers. However, during the
establishment of the PGE, other parties such as development
Section 4
and donor organizations may operate on a dierent basis that
is not strictly commercial.
. Most technical modelling of nancial products is done
on a proprietary or market basis, with the expectation to
recoup the costs of developing the models from either
trading products or licensing those models to others
to trade products. However, in the case of developing
economies, the rate of insurance penetration is too low to
justify development of risk models on a strictly commercial
basis. Rather, donors or development organizations may
pay for the development of the technical infrastructure that
the PGE then draws upon to generate products.
. The ultimate intention is for sovereigns to pay their own
premiums, as part of taking responsibility for their own
disaster response. However, in exceptional cases, such as
the Caribbean Development Bank’s payments of Haiti’s
premiums for membership of the CCRIF, donors may pay
the premiums for some sovereigns, enabling them to
participate in the risk-transfer product. Sometimes these
payments are on a sliding scale, decreasing each year as
the sovereign assumes greater understanding of both the
product, and command of their resources to pay their
own premiums.
As these points suggest, such PGEs bring together an
especially complex set of parties with both market and
social objectives. In particular, donors and development
organizations, with their social mission, play a key role
in establishing these PGEs and enabling the transfer of
sovereign risk to nancial markets. However, an important
feature of these schemes is that they enable a path through
which countries become increasingly able to access nancial
markets, self-fund, and expand their insurance cover.
. BENEFITS OF DISASTER LIQUIDITY PRODUCTS
While these disaster liquidity products are dierent in aim
and scope from traditional catastrophe insurance products,
they have certain advantages for countries with emerging
insurance markets, which allow them to play an important role
in recovery from catastrophe. The main advantages include:
. Financial protection. These products, with their limited
scope, have smaller premiums (and payouts) than
traditional insurance products, and are therefore
more aordable.
. Rapid payout. Payouts from these products generally fall
far below the cost of the devastation actually experienced
by a country. However, their eect is considered amplied
because of the rapidity of payment. Cash flow in the
immediate aermath of disaster can provide services
that prevent the humanitarian crisis from escalating. For
example, in a drought, the provision of food to people
within their communities can prevent migration and the
development of a refugee crisis. Similarly, the provision of
clean water and shelter aer a hurricane can prevent the
development of illness that escalates the crisis. Hence, it
is not simply the size of the payment, but its timing in the
disaster response that is critical. “I remember when we
saw the payment from CCRIF aer the earthquake, it was
very quick… The President declared a state of emergency
and so we put the money in an emergency budget. Then
the Government could use this money to respond.”
(Government stakeholder)
. The introduction of these products generates new
knowledge and expertise. Data access and collection
improves, resulting in stronger modelling capability. This
can have positive eects in two ways:
a. Growth of insurance. Governments and other
public organizations gradually develop a deeper
understanding of how these products work and
are educated on the benets of insurance-based
risk transfer mechanisms. The inflow of data and
modelling capability can work for the mutual benet
of stakeholders’ social and market objectives. It can
expand the appetite of governments to increase their
risk cover and of market players to provide capital for
such cover in these regions. For instance, CCRIF, aer
years in operation, it is introducing additional disaster
products such as for excess rainfall, and additional
insurance products such as for sheries, in response
to the needs of its member countries. “Here people
were not very nancially literate about it … but I think
the work we are doing is very important for all the other
stakeholders that are now interested in this specic
market because we kind of opened the door to it.”
(PGE stakeholder). The establishment of insurance
products for disaster liquidity may also stimulate the
progressive development of insurance products for
reconstruction of national infrastructure, as and when
countries can aord to purchase such cover.
b. Empowerment over disaster response. As data and
modelling expertise improve, governments can better
understand their vulnerability from potential disasters
and so improve disaster mitigation and resilience. For
example, they can gain a better understanding of the
risk they are carrying and its impact on their budgets,
enabling better nancial planning for catastrophic risk.
“Thanks to [the PGE] we now have exact gures.
Knowing that three years ago there was [a disaster] and
it cost you this, but if you had insurance it would have
cost you this … When you have those gures, it is really
clear to explain to a Minister that you can better plan for
your nances … then you realize that OK, yes, maybe we
should consider [insurance] going forward.”
(PGE stakeholder)
These products can also enable governments to take
more responsibility for and have more control over their
response to disasters – for instance, by formulating
costed contingency plans – and reduce their reliance
on international aid organizations. “The fact that it is
the [sovereigns] that initiated the creation of [this PGE]
is also a big signal that [this region of the world] now
wants to take that issue into their own hands and not
just wait for the solution to come from outside.”
(PGE stakeholder)
Between State and Market: Protection
Gap Entities and Catastrophic Risk
In general, PGEs aimed at disaster liquidity have proved
a successful marriage of social and market objectives.
Participants from the insurance markets benet because a new
market is made accessible, and because modelling techniques
are expanded; governments and other social agents learn to
achieve social objectives via market mechanisms, which can
leverage their resources.
. ENABLING CONDITIONS FOR SUCCESS
The disaster liquidity products being developed by PGEs must
also overcome some challenges to realize their many positive
eects. Key enabling conditions include:
. Their aim should be clearly communicated, so that they
do not engender misplaced expectations about the
nature and extent of the cover provided. This is critical
as misunderstanding about what these products can do
(nance disaster response with a specied amount in a
specied time period) and cannot do (nance property
reconstruction) can lead to backlashes that might reduce
trust in the product and reduce future levels of insurance
protection.
“So at rst I was not a fan. But then I spent time getting
more acquainted with the product, how it works, the trigger,
understanding the trigger. And at that point I recognised
that like any other business, they are selling a product, and
really and truly the failure was at the level of governments
not spending enough time to appreciate the product and
what it can or cannot do for them.”
(Government stakeholder)
. Improving the metrics for modelled losses, pricing and
triggers for payments is also vital to increase sovereigns’
condence in their ability to use these products as part
of the ex-ante preparation for disaster. Better modelling
also helps in attracting market players to invest in creating
their own products to transfer these types of risks into the
capital market.
. Adherence to market signals is a vital principle in
promoting market condence. In the long term it will
enable the market to make the fullest contribution to the
social development objective – that of giving sovereigns
autonomy over their own aairs.
In the shorter term, and more particularly in the early
stages of model and product development, market
interests to stick to the modelled triggers may conflict
with the social development interests to provide relief.
For example, in the case of Malawi, triggers were not met
per the contract agreement placed by the government
but conditions on the ground demonstrated a clear case
of a drought event (with this discrepancy driven largely
by some of the modelling customisation assumptions).
Being proactive about investigating such situations and
nding technical solutions grounded in the models and
methodological triggering approach are critical, as well
as the flexibility to accompany such learning process in
these countries.
For developing countries, where insurance is seen as a
luxury in extremely scally constrained circumstances,
simply sticking to contracts that may be incompatible with
the development mandate of PGEs could prove counter
productive. Indeed, embedding market principles and
developing nancial capabilities, gives condence to
market-players and empowers sovereign states to manage
their own risk. Likewise, the flexibility to accommodate
the required learning process among multiple parties with
dierent interests is an ongoing challenge for these PGEs,
necessitating multi-party dialogue and flexibility.
. The success of disaster liquidity products as tools to
address the protection gap depends on the governments’
ability to spend the money rapidly and eectively.
Eectiveness of these products is likely to be enhanced
where governments are accountable for the use of payouts
to their citizens and sponsors.
“One of the biggest problems is the lack of accountability
at national level in terms of how these payouts are used.
Visibly you are unable to link payouts in to the recovery
and early reconstruction phases or in serving as any sort of
mechanism for reducing risk, going back to this argument
about how insurance can drive risk reduction. Because
there’s no accountability around how it does that.”
(Government stakeholder)
Eorts to promote accountability involve developing and
reporting on contingency plans for the use of resources,
as a condition of payment (e.g. ARC), or may involve
more voluntary forms of reporting (e.g. CCRIF, PCRAFI).
For example, ARC members are rst required to provide
contingency plans showing how beneciaries will be
targeted before they can access insurance and second are
subject to process and nancial audits when they receive
a payout in order to ensure ecient and targeted spent of
the money. This ex ante approach is innovative in linking
accountability for the use of public funding for insurance
with the targeted delivery of relief to the most vulnerable.
. Most multi-sovereign PGEs act as a risk-pooling
mechanism. To be eective as a source of capital
protection, they need to have a large enough member
base, and these members need to be retained over a
number of years. This ensures diversication across
geographic regions and perils, improving the pools’ own
capital optimization and ability to pay for member losses,
and also their purchasing power. Hence, as indicated in
the CCRIF case, eorts need to focus on building members’
common purposes for remaining with the PGE.
A PGE which attends to the conditions outlined above, in its
remit or its evolution, should enjoy the condence of sovereign
states, their citizens and wider stakeholders that government
money is well placed in purchasing risk transfer products, and
that the PGE itself meets the various market, humanitarian, and
political objectives of the dierent parties.
Section 4
Road map to success for PGEs’ disaster liquidity products
• Eective communication about the product as immediate
relief post-disaster, rather than as more large-scale
reconstruction
• Continuous improvement of the modelling for product
development and metrics that trigger payment
• Supporting eective and accountable use of funds to
alleviate hardship post disaster
• Ensuring eective pooling through diversication and
retention of members
. SUMMARY
Overall, the establishment of PGEs and disaster liquidity
products in the context of low-income countries or those
with low insurance penetration redenes the protection
gap. Such schemes redene who or what may be insured,
extending the ability of insurance to protect against disaster to
sovereigns for the purposes of providing cash flow to manage
disaster response. Disaster liquidity products were therefore
a signicant innovation in the treatment of under-insurance:
disaster relief and humanitarian crisis had not previously
been considered areas in which ex-ante market mechanisms
might be deployed. But their creation has prompted new ways
of looking at the protection gap debate. For instance, some
donations to poor countries which might previously have gone
directly into government expenditure are now used to buy
disaster liquidity cover, or to provide operating capital for PGEs
to support countries to buy their own disaster liquidity cover,
so leveraging the donation.
Certainly, PGEs of this sort are growing both in their regional
spread with new multi-sovereign PGEs being considered
and existing PGEs growing their membership, and in terms
of the range of cover oered. At the same time, these types
of products are being applied to a wider range of gaps, as
evidenced by the recent development of the World Health
Organization (WHO) catastrophe bond for pandemics.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
SECTION . Bridging and reducing the
protection gap: PGEs’ role
in resilience
Disaster resilience is the ability of individuals, communities,
organizations and states to survive, adapt to, and recover
from catastrophic events; and to adapt their structures and
behaviours to reduce the potential damage from future
perils., There is an important relationship between PGEs
as providers of insurance to those who might otherwise be
un-insured, and the disaster resilience of a society. One aspect
of this relationship can be improved nancial resilience,
where insurance provides the capital to support recovery.
Such initiatives help bridge the protection gap by providing
nancial solutions. However, if nancial solutions lead purely
to reconstruction of what was destroyed, they leave the
underlying vulnerability unchanged. Given the importance
of social as well as market objectives, reducing the gap by
reducing vulnerability through more resilient forms of building
can also help to full the social objectives of PGEs. One way
that PGEs can support both bridging and reducing the gap is
through their eects in generating technical knowledge for
identifying, measuring and better preparing for risk.
Paradoxically, PGEs, with some exceptions, are not in a
position to influence resilience measures directly. Rather,
governments are generally the central agents in improving
resilience. Through their many sub-entities at both national
and local levels, they have the legislative power to drive
resilience through tools such as environmental policies,
land-use planning, building codes and standards, defensive
infrastructure, and disaster relief. Hence, PGEs must work with
and through their interdependent stakeholders, particularly
government, in supporting the relationship between insurance
and resilience.
We present a Protection Gap Entities (PGEs) Resilience
Framework, adapted from the World Bank pillars
framework, to show how PGEs, through their eects in
increasing insurance cover, can also support key aspects
of resilience.
PGEs are put in place specically to increase the nancial
resilience of both insurance markets to shock (Objective
in Section ) and policyholders and governments to large
and/or frequent disasters (Objectives & in Section ).
This function is indicated in Figure . by the thick arrow to
financial protection. However, risk identification is central
to managing the various elements of disaster resilience,
since society cannot prepare for, mitigate against, or arrange
nancial protection from risks that are not identied. We
position PGEs, via their role in supporting the insurance-
based nancing of disaster, as one critical mechanism for
enhancing risk identication. Of course, there are other ways to
promote risk identication. For instance, public organizations
and technologists may use open data and open soware
tools to identify the risks posed by natural disasters and
humanitarian organizations may work with local communities
to map vulnerable areas. However, because of their role in
supporting insurance against disaster, PGEs are especially
well-positioned to enhance risk identication; and thereaer,
PGEs have some unique characteristics that enable them to
play a potentially pivotal role.
Figure 5.1. PGE Resilience Framework
Resilient
Reconstruction
Risk Reduction
Financial Protection
Risk
Identification
Preparedness
P
G
E
. PGES’ ROLE IN RISK IDENTIFICATION
Risk identication refers to building the capacity to identify,
assess and analyse risk, typically as a technical capability
supporting the quantication of risk. The insurance industry,
in collaboration with natural and environmental scientists,
follows a specied risk-identication process to enable the
transfer of risk within the value chain: from modelling the
hazard; assessing the vulnerability of a particular region to that
hazard; then calculating the exposure of property and citizens
to that hazard, and the potential loss arising from a severe
occurrence of the hazard. As a result, it can price cover based
on the probability and severity of a loss.
Data and model generation. Since the early s, there
have been signicant developments in catastrophe modelling,
largely driven by advances in the availability of computing
power and data, and by the commercial needs of the (re)
insurance industry to price and trade in risk.
Section 5
Where such data and modelling capabilities are lacking, as
in emerging markets (see Section ), or in relation to new
forms of risk such as terrorism, PGEs are critical catalysts
in establishing data for risk identication. PGEs can be
particularly eective in promoting risk identication and
making it relevant to government policy, because they
construct a direct link between risk identication and nancial
solutions that may act as complementary or substitutive forms
of funding for governments. The search for a nancial solution
drives the coordination of existing data and models and the
development of new ones.
Even in mature insurance markets where data exists, PGEs
oen have a key role in identifying the most exposed areas of
risk, because they can collect, rene and recombine existing
modelling and data analytics that were generated for purposes
outside the insurance industry, such as environmental projects.
For example, Flood Re, at its inception drew upon a wide
range of data from the insurance industry but also the UK
Environment Agency to identify those properties most likely
to be at risk. Hence, PGEs promote the collection of data and
identication of what may otherwise have been over-looked or
unmodeled risk.
Information transfer for social and market objectives.
While pricing risk may be specic to its transfer, other
aspects of insurance industry risk identication can have
wide application. In particular they support the estimation
of aggregate losses in particular regions, as a basis for
various government initiatives. As they are not-for-prot
organizations, PGEs do not need to make such modelled
information proprietary but can provide it as a ‘public good’ to
other parties. “It would be a shame to just collect this data
for insurance and not make it accessible for other uses. So it
was good to identify other data formats that [the modelling
company] was willing to share that wouldn’t actually jeopardise
any proprietary uses.” (PGE stakeholder)
PGEs can therefore be a valuable two-way conduit for
information between the insurance market and society.
Disparate data gathered in social, environmental or
population-based projects can be used to improve insurance
modelling. Once available, insurance modelling increases the
ability to identify and understand risk, both in general terms
and in terms of economic consequences. Identifying risk can
provide the motivation and information necessary to increase
risk mitigation and disaster resilience. Indeed, the insurance
industry is involved with eorts to improve and share ‘pre-
competitive’ risk-modelling data in several countries.
Creating an ecosystem. In promoting risk identication,
PGEs stimulate the development of an ecosystem, comprising
professionals from aid, science, development, policy,
commercial (re)insurance, and government ministries, for
producing information that underpins all forms of disaster
resilience. Interaction across this ecosystem creates at least
two major spill-over eects:
. Better understanding of the hazard and the associated
vulnerability allows governments to better assess the
economic costs of disasters and how they impact the
budget of dierent ministries, as this quote on the role of
terrorism PGEs indicates:
“There is nowhere in government where they are looking
at these terrorist events in terms of economic loss. They
will look at those terrorist events as terrorist acts, they will
be looking at them from an operational and a political
perspective … but they won’t be looking at those events as
something that will cost them a lot of money. And there’s
a role for [PGE] I think to conduct that kind of research
on government’s behalf… it would then be able to inform
government; ‘Look, if we’d had a similar attack in [our
country], the consequential loss [from loss of tourism,
disruption to transport etc] would have been X billion and
[PGE] doesn’t cover that’.” (Government stakeholder)
. Improved risk identication also supports the design
of new nancial products (whether provided by PGEs
themselves or not) that may complement and extend the
relief provided by insurance-based products. An example
is forecast-based nancing (FbF) as a technique to ll gaps
in humanitarian aid. It uses the science of weather and
climate to anticipate possible impacts in risk-prone areas
and to mobilize resources automatically before an event.
Such forecasting can also support insurance triggers.
For example, satellite data can now be used to forecast
drought based on levels of green-ness, and so to put in
place humanitarian strategies before catastrophe occurs.
In these ways, PGEs can have a direct eect on the relationship
between risk identication and nancial resilience in the face
of potential disaster.
. PGES ROLE IN RISK REDUCTION AND RESILIENT
RECONSTRUCTION
Insurance can enhance resilience by providing incentives
for behaviours that either reduce risk or improve resilient
reconstruction. Examples might be the implementation of
risk reduction measures such as levee banks and drainage
systems that change the built environment, or improved
building codes for reconstruction aer disasters. In improving
risk identication, PGEs can also indirectly influence these
areas even so that they rarely have direct responsibility for
implementing physical and behavioural resilience measures.
Under pure market conditions, a property highly exposed to
risk would nd it dicult to get aordable cover; this would
dampen the market for high-risk property (since mortgage
conditions typically require insurance cover), and thereby
discourage building in high-risk areas or with insucient
risk mitigation features. Such a situation could motivate risk
reduction behaviours both in policyholders, who may build to
dierent structural codes in order to improve the insurability of
their properties, and in planning authorities, in terms of where
to build and what mitigation features to include. However, as
developed economies already have signicant property that
has been built to varying structural codes in highly-exposed
areas, such pure market conditions do not exist.
Incentivizing risk reduction. PGEs ameliorate some of these
legacy eects by providing nancial support to properties and
regions that do not have inbuilt risk reduction features. PGEs
Between State and Market: Protection
Gap Entities and Catastrophic Risk
can still incentivize better risk reduction behaviours going
forward. For instance, in the UK, properties eligible for the
Flood Re scheme had to be built prior to January . This
was to ensure that the PGE was not incentivizing development
of new buildings in flood prone areas. Flood Re’s mandate
also includes promoting flood management action through
the use of data to support decision-making by Government,
the Devolved Administrations, Local Authorities and others,
including the Environment Agency and the Committee on
Climate Change.
Placed between governments and markets, PGEs are well-
equipped to coordinate resilience action on both fronts.
The insurance industry might have loss data and/or the
technical expertise to identify and calculate high risk areas
and consult on ecient ways of resilient rebuilding; but only
governments can introduce and enforce policy-making such
as land-use planning and building codes. However, our results
show that while PGEs sit at the nexus of various stakeholders
within governments and markets, they typically lack the
formal power to introduce and impose the uptake of risk-
reduction measures. Usually their remit is limited to providing
consultation on resilience and attempting to influence the
various stakeholders based on their position in the risk
transfer value chain. In Figure . therefore, we depict the link
from the PGE’s role in risk identication to risk reduction as
relatively weak.
Nonetheless, PGEs coordinate multiple sectors (e.g. insurance,
resilience, governments) and disciplines (natural, nancial,
social science), enabling knowledge-sharing and capacity-
building. They are increasingly eager to provide information
on risk mitigation to improve risk reduction behaviours. For
example, the ARPC terrorism pool in Australia has identied
some key structural features, such as bollards that limit the
positioning of explosives near key points in buildings. As such
features will reduce the risk of certain terrorism activities, the
PGE is oering premium discounts for companies that install
such measures. Price is thus used as a means to incentivize
behavioural change.
Other PGEs, notably the CEA, have invested heavily in
understanding how reconstruction of legacy properties can be
linked to mitigation of loss from earthquake damage. Their risk
identication activities have enabled them to understand how
to retrot structural adaptations to existing at-risk properties,
prior to an event, so as to limit the damage caused by
earthquake. As part of their statutory requirements, they use
this information to actively support risk reduction by oering
retrot packages that take advantage of tax discounts and that
can even be cost neutral for the homeowner.
Constraints around PGE influence over risk reduction.
Despite this, evidence from post-flood reconstruction grants
in the UK and retrot packages in California, shows
that uptake appears to be low. The reasons are various,
ranging from diculties in communicating and distributing
reconstruction grants, to poor understanding by policyholders
of the potential risks of damage, to concerns over the cost,
timing and inconvenience of reconstruction.
Such ndings indicate the diculty for PGEs in implementing
risk reduction measures, and ex ante retrotting in particular,
without legislative support. However, these same measures
can be used in resilient reconstruction, where the PGE pays
the claim on the proviso that rebuilding follows specic
reconstruction codes. PGEs that provide insurance cover to
highly-exposed properties may have a more direct influence
over resilient reconstruction, particularly where the payment
of claims may be linked to enforceable structural codes for
rebuilding. In Figure . we therefore indicate this influence on
resilient reconstruction by a relatively thick arrow.
. PGES ROLE IN DISASTER PREPAREDNESS
The link to disaster preparedness is the weakest link shown
in Figure .. Financial protection, risk reduction and
resilient reconstruction can all be tied together through risk
identication in a logical virtuous cycle, in which better-
constructed properties that reduce risk are more aordable
to protect nancially, whilst nancial protection enables
more resilient reconstruction post-disaster that subsequently
reduces risk. Nonetheless, as explained in section ., this is
not easy for a PGE to influence directly. Disaster preparedness,
involving early warning systems and contingency planning,
is even more remote from PGE influence. A PGE may have
the data to indicate the likelihood of loss, and so to support
understanding of how to prepare for it, but may not have the
remit to encourage such preparedness.
Some PGEs, however, are set up with a remit to build links from
risk identication to disaster preparedness. ARC is a case in
point. A key feature of the way ARC inducts its member states
is to take them through a process of risk modelling, which is
used to develop early warning systems of impending disaster,
and to undertake contingency planning for how specically,
they will use any payouts to address the disaster, prior to
taking out the risk nancing component of membership.
The establishment of such risk preparedness procedures in
ARC is helped by the fact that most of the scheme deals with
drought, a slow-onset disaster in which the impact can be
observed, and systems put in place as the disaster worsens.
In addition, the payments are intended to support food
security, for which, again, there can be contingency planning in
advance. Preparedness for other types of peril, where onset is
more rapid, may be dierent. Nonetheless, some early warning
systems to support, at the least, evacuation in the path of
a hurricane, and other similar measures, may be spillover
benets from the risk modelling and identication promoted
by PGEs.
. SUMMARY
The primary purpose of a PGE is nancial protection of those
who are potentially under-insured. We might describe this
extension of the reach of insurance as bridging the protection
gap, and the role of the PGE as increasing nancial resilience
However, for a society to solve the protection gap, increasing
nancial resilience is not generally enough. In high-risk
areas particularly, it is essential to reduce risk, and thereby
reduce the protection gap, which means that other aspects
Section 5
of resilience – resilient reconstruction, risk reduction and
preparedness – are essential (see Figure .).
We have shown that PGEs have a key role in risk identication,
and thus they can influence or enable these other aspects
of resilience. Increasing the uptake of resilience measures
has an impact on the social goal of reducing the protection
gap. In addition, it can reduce the frequency and severity of
loss, which means that PGEs will have to pay lower claims,
decreasing the nancial impact on their balance sheet.
To enhance PGEs’ role in resilience, we suggest:
• Build formal power in resilience: Being identied as
independent, and being a centre of risk expertise, places
PGEs in a unique position regarding resilience. However,
they usually lack the formal power to enable the uptake
of resilience measures and are restricted to, at best, an
advisory role. For instance, Flood Re is doing research to
understand how to incentivize the uptake of resilience
measures, but without formal powers, it can only cooperate
with the insurance industry and the government to promote
its position in risk reduction. Potential ways to increase
formal power could be to give PGEs a mandate to enforce
resilient reconstruction aer a disaster or to nancially
incentivize citizens to take risk reduction measures. CCS
for instance can request resilience measures to be taken in
order to continue to provide cover.
• Enhance useful links with(in) government: PGEs can be
public or private organizations or public-private partnership
organizations. They are either part of the government or at
least collaborate directly with governmental departments.
Even the PGEs that are private organizations have a strong
public purpose. PGEs hold information and develop technical
expertise that is critical for central and local governments
in supporting resilience initiatives. For instance, in both
FONDEN and CCR, the data used for risk modelling to
develop (re)insurance products is integrated with the data
used for disaster management. By linking government
departments involved with planning for disaster, planning
for infrastructure and nancial planning, PGEs can help
build a more comprehensive approach to resilience. Formal
channels of communication and collaboration between PGEs
and parts of the government that have powers over these
dierent aspects of resilience should be strengthened.
• Evolve remit but avoid mission creep: A PGEs is developed
to address a specied protection gap by invoking a market
mechanism. PGEs may be a means of coordinating the
dierent interdependencies involved in resilience, but their
specic expertise is in providing insurance-based measures,
rather than initiating resilience measures. Expanding PGE’s
remit in resilience should avoid mission creep, losing focus
and bloating their role. We suggested that PGEs need to
evolve over time alongside their stakeholders and their
local protection gap. However, PGEs should be alert not
to change their mission unintentionally in ways that may
neglect a continuing necessity to provide aordable nancial
protection, when commercial markets cannot supply it.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
SECTION . Protection Gap Entities or
Stopgap Entities?
The previous sections have shown that PGEs are usually set up
with objectives related to a specic and local protection gap.
These objectives fall into three main categories, (i) addressing
a sudden disruption in the supply of (re)insurance; (ii) ensuring
that a portion of the currently insured population that is
classied as ‘high-risk’ is not priced out of the market; and (iii)
ensuring the nancial viability of governments in the aermath
of catastrophic events in countries with low insurance
penetration (see Section ). These objectives emerge in the
aermath of major catastrophic events that promote political
engagement and influence opinion in civil society.
They thus form the basis of a remit around which stakeholders
with diverse and sometimes conflicting interests can agree
to join forces in setting up the PGE. Crucially, those interests
may lean more towards market objectives, or towards social
objectives; they may prioritise bridging the gap or reducing it;
and tensions between such interests will need to be managed.
The need to forge consensus in a short period of time to
address a crisis means that a PGEs’ remit is oen tightly
constructed around the objective leading to its establishment.
In such cases, PGEs, as initially set up, can provide only partial
solutions to the protection gap.
Paradoxically, while the PGE’s activities may contribute to
a wider, deeper and more rened understanding of the
protection gap (see Section ), their institutional set-up can
signicantly constrain any evolution to address either the
underlying reasons for the gap or changing circumstances.
Indeed, stakeholder attention, particularly political attention,
moves on so that there is typically little will to address change
until aer another disaster exposes a new protection gap.
In fact, a ‘stopgap’ character is oen built into PGEs and their
remits from the very beginning. Narrow and rigid remits can
create problematic situations in which stakeholders assume
the threat of a particular disaster has been covered (because
the remit has been ‘ticked’), leading to criticisms of the PGE
when it is unable to respond to losses that were never in its
scope. However, stakeholders might reasonably expect that
the PGE is there to address the protection gap, rather than a
shopping list of objectives related to a narrow part of it. From
this point of view, PGEs should be able to modify their remit to
continue to address the gap.
An opposing point of view holds that complex organizations
have a well-documented tendency to “mission creep” –
the expansion of the organization beyond its expertise or
usefulness. For this reason, many PGEs are explicitly set up
to be temporary: “We’re a government agency and we’re a
temporary organization because we were created as a short-
term stopgap measure for market failure. So we actually have
a social purpose for existing and we’re not a private sector, for
prot or private interest organization.” (PGE stakeholder).
These tensions around flexibility or mission creep, around an
evolving gap or a xed purpose, must in fact be constantly
monitored and managed by all PGEs. The PGE’s remit, the
stakeholders’ expectations, and the evolving protection
gap exist in a dynamic balance, as we depict in Figure ..
Particular problems in managing this balance occur when
there are mismatched understandings between parties to this
balancing act, as we now explore.
Figure 6.1. PGE Evolution Framework
Evolving
Protection gapStakeholder’s
expectations
PGE
PGE's remit
. MISMATCH : PGE REMIT DOES NOT MATCH
STAKEHOLDERS’ EXPECTATIONS
Stakeholders may see a PGE as inadequate even when it
fullled its remit. For instance, since its inception in ,
CCRIF has clearly performed extremely well to its remit, making
payouts totalling US million to member governments;
all within days of the event. Following Hurricane Irma in
, CCRIF issued payouts to member governments within
days of the event. Nonetheless, as one country member
explained to us, while praising the eciency with which they
received their payout:
“It is a drop in a bucket in terms of the actual loss… The
payout for the impact of Hurricane Irma was just over 6 million.
It helps because Irma was major. We just had a damage and
loss assessment putting the bill around 338 million, so for
us that’s just over 100% percent of GDP” (Government
stakeholder)
Section 6
Such cases exemplify what we call the Expectations mismatch.
Disaster-liquidity PGEs (as discussed in Section ) are not
intended to address the protection gap for reconstruction in
such countries – but stakeholders may expect them to do so.
Disaster-liquidity PGEs generally seem to perform as intended,
and with good leverage for the premium paid. Yet they can
have only the most marginal of eects in closing the far larger
protection gaps for reconstruction which are associated
with the same disasters. Furthermore, their existence may
distract from other ways of addressing the immensity of the
reconstruction protection gap, if stakeholders have unrealistic
expectations of them.
“They have seen sovereign risk pools in my view as a silver
bullet, and so they do not invest in other means of reducing
risk and exposure and they do not invest in other areas of
nancing that gap that you speak of aer an event. … You
know, when you look at the catastrophic events and you go to
the donor community to have these donor conferences, when
you look at the losses versus what is raised even in these
donor conferences, it’s less than one tenth of the losses. So
it means there is a signicant nancing gap for recovery and
reconstruction.” (Inter-governmental stakeholder)
The problem is that in highly-exposed countries with fragile
economies, it is dicult to think of any self-funded insurance
solution that could enable a country’s full reconstruction aer
recurrent disasters that cause losses of the order of magnitude
of their entire GDP.
This Expectations mismatch is not only a problem in emerging
insurance markets, but can also occur in mature markets. For
example, a terrorist attack, in which a truck drives into crowds
of people in a tourist area, may cause no property damage. If
the remit of the PGE is to cover only property damage caused
by terrorism attack, no payments would be made. But this truck
attack may cause signicant loss of life and will probably have
a huge economic impact on tourism, due to cancellations and
last-minute changes by travellers concerned about the threat
of terrorism. The non-cover of these nancial losses does not
make the PGE unsuccessful, since they were not relevant to
its remit. Yet, ex-post, some stakeholders, not to mention the
public, may expect that this ‘should’ have been the remit of
the PGE.
Tackling the mismatch: There are two key ways to deal with
an Expectations’ mismatch:
. Communicating remit and success: It is critical for PGEs
to communicate their remit clearly to stakeholders. The
fact that the PGE deals with ‘the protection gap’ can be
confusing as that term is quite abstract. PGEs are formal
organizations that have a set of specic goals to achieve
in relation to a specic protection gap and a specic risk,
rather than a mission to close the global protection gap.
Thus, PGEs need to dene what constitutes success in
terms of their remit, so that stakeholders can evaluate
them based on this predened set of objectives.
. Building other forms of resilience. PGEs, while
performing many valuable functions, are not a silver
bullet. Rather, they should be seen as a critical mechanism
for nancial and physical resilience (see Section ), by
showing the feasibility and eectiveness of insurance
products and helping build the expertise around them.
PGEs support nancial resilience (that is, protection from
nancial loss), which in turn can lead to other resilience
features such as resilient reconstruction, preparedness
and risk reduction (Figure .).
. MISMATCH : WHEN ADDRESSING ONE PROBLEM
CREATES ANOTHER
While PGEs provide an initial solution to an identied
protection gap, these solutions can have unintended
consequences. One such consequence is moral hazard.
This is the situation in which policyholders can engage in
risky behaviour knowing that other parties (insurers and/or
government) will incur the cost. Specically, as explained in
Section .., PGEs, by paying for losses, may unintentionally
reduce the incentive to engage in risk mitigation, both by
policyholders and more widely by planning authorities. Indeed,
aer repeated events in flood prone areas, leading to the
rebuilding of highly exposed properties on more than one
occasion, the National Flood Insurance Program (NFIP) in the
USA has come under attack for precisely this issue, with the
criticism that “it is subsidized floodplain development”.
This problem is exacerbated by a risk redistribution strategy
(Section .), in which the PGE adopts a pooling mechanism,
so that low-risk areas cross-subsidize the higher-risk areas.
When the actual risk of living on a floodplain or an earthquake
fault-line is not reflected in the premium, then policyholders
have little incentive to engage in risk mitigation strategies
that would lower the cost of their premiums, or the cost of
rebuilding. As time goes on, and flood or earthquake strikes
repeatedly, the costs of a risk-redistribution response rise to
pay for repeated high losses, increasing the expense to others
in the pool.
Tackling the mismatch: Co-evolving nancial and physical
resilience measures. As highlighted in Section , while
increasing the aordability of cover for policyholders, PGEs
can aect the risk-reward-responsibility balance. Removing
responsibility for high risks from policyholders may have
unintended consequences for those policyholders’ behaviours.
The issue of subsidizing insurance policies has been analysed
extensively in the debate on equity vs eciency – where the
overall welfare benets resulting from universal protection
in a society are weighed against the overall costs for society
of moral hazard. This literature has generated proposals on
how to combine policy instruments such as tax incentives or
building regulations to address the problem.,,
In Section . we have considered the issue by placing PGEs
and insurance within the broader landscape of resilience. We
discussed how PGEs can play a critical role in coordinating
the dierent interventions necessary for nancial resilience
through market measures, and physical resilience through
public policies for risk reduction and resilient reconstruction.
This is a critical area in which the remit of PGEs might need to
be expanded to match the changing nature of the protection
gap and the expectations of stakeholders.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
. MISMATCH : PGE IRRELEVANCE IN THE FACE OF AN
EVOLVING GAP
PGEs are set up to deal with a local protection gap at a specic
moment in time. However, protection gaps evolve, so that the
original gap that the PGE was set up to address might become
less relevant; or while it remains important, other gaps of equal
signicance may emerge over time.
One example is terrorism risk, which has evolved signicantly
since the origin of most terrorism PGEs. For example, the
threat of sophisticated, highly-coordinated, severe events
like / may be ever-present. However, increasingly terrorist
events are ‘lone wolf’ attacks with less catastrophic impact,
using familiar, accessible objects, such as cars and knives,
as weapons. Many of the nancial impacts of such attacks,
such as the business interruption caused by site closure
aer an attack, the loss of revenue from diminished footfall,
or indeed, the capacity to pay for loss of life, may not be
covered by PGEs established to cover catastrophic damage
to commercial property. Thus, a swathe of uninsured risk
emerges, demonstrating the evolution in the protection gap
and, post-disaster, the emergence of stakeholder interests and
expectations where the risk is not addressed within the remit
of the PGE. Quite simply, even if it goes beyond their initial
remit, PGEs need to remain relevant to the evolving protection
gap, as this will be expected of them in case of future disasters.
On the other hand, PGEs that can influence the closure
of a protection gap may also need to co-evolve – into
obsolescence. For example, while flood risk has increased
signicantly over the last decades it can be signicantly
reduced if risk resilience (Section ) is taken seriously.
Measures that promote resilient reconstruction, risk reduction
and preparedness, alongside nancial protection (PGEs
Resilience Framework; Figure .) can have a real impact on
reducing the protection gap. Essentially, if the protection gap
that a PGE was set up to address is not signicant anymore,
the PGE itself may become redundant.
Tackling the mismatch: Evolving policy dialogue, with regular
review. Our research study shows that the response to this
situation can vary signicantly. A rst strategy is to adapt the
remit of the PGE, as political or societal awareness of evolving
threats emerges. For example, as awareness of terrorism threat
has grown, most terrorism PGEs have evolved from oering
‘conventional’ terrorism cover for re and explosion to also
oering non-conventional cover for chemical, biological,
radiological and nuclear (CBRN) threats. In this way, the
dynamic balance illustrated in Figure . is preserved as
stakeholder interest, protection gap, and PGE remit co-evolve.
Oen this process of adaptation is precipitated by a crisis,
similar to the process leading to the initial establishment of
PGEs. For instance, some PGEs operating in the terrorism area
originated only in the wake of /, such as ARPC in Australia;
elsewhere many existing terrorism PGEs responded by
expanding their remit to include new categories of CBRN risk.
Such changes of remit can better be brought about through
regular policy or legislative review of the PGE. That is, through
foresight and planning, rather than being forced upon the PGE
in the aermath of a new disaster.
Another strategy is to give PGEs a limited lease of life. This is
done in the assumption that the problem as perceived today is
temporary and that the PGE will, by its termination date, have
stimulated sucient market development to close the gap. For
instance, the recently established Flood Re in the UK has been
set up with a specic remit to ‘solve’ the problem for which it
originated and then dissolve over a -year period. While such
deadlines may simply demonstrate that the PGE has not, in
fact, solved the protection gap, or may call into question what
other gaps have arisen, the termination date should generate a
point of dialogue about the role and remit of the PGE.
. SUMMARY
The mismatches detailed in this Section are indicative of the
inherent tensions that exist when the PGEs’ remit, the interests
of stakeholders and the protection gap are evolving in parallel
(Figure .). It would be wrong to understand potential
mismatches as fundamental problems that need to be xed.
Rather, these are naturally-occurring tensions that would
always exist at the nexus of competing demands upon PGEs,
which they would have to mediate.
By actively managing these mismatches, PGEs can powerfully
frame both the way we understand risk in society, and the role
of governments, markets and wider stakeholders in managing
that risk. Although dierent stakeholders will always have
dierent interests in the potential solutions, more use can be
made of PGEs, not solely as temporary stopgap solutions to
initial problems but also as spaces in which to host an ongoing
dialogue about a growing problem.
Section 7
SECTION . Conclusion. Learning
from PGE experiences:
A call-to-arms
This report has provided an impetus for learning across regions
and sectors, about the nature, characteristics and activities
of Protection Gap Entities (PGEs); those organizations and
initiatives set up in collaboration between States and markets
to bring nancial solutions to the socio-economic problems
caused by the increasing prevalence of natural and man-
made catastrophes. PGEs, while in many ways heterogeneous,
are set up with the objective to transform uninsured risk
into insurance-based products that can be transferred onto
government balance sheets or into global nancial markets in
order to provide capital for recovery following a disaster. This
report has developed the following ndings:
PGEs classification. Beyond their widely recognized sources
of variation, PGEs may be usefully grouped according to their
common features:
. PGEs are partial and temporary solutions to a specic and
local protection gap – rather than entities set up to tackle
the overall problem of underinsurance against disasters.
Specically, we have shown that PGEs are oen set up
in response to major crises that push stakeholders with
dierent market and social interests to collaborate in
tackling one of three objectives:
Objective : Resolve disruption in (re)insurance supply in
mature markets;
Objective : Mitigate the threat of unaordable insurance
in mature markets;
Objective : Increase the nancial viability of sovereign
states with fragile economies.
. PGEs tackle these objectives by adopting dierent
positions in our Protection Gap Strategic Response
Framework, in terms of their choices on:
a. how they share risk with existing market players -
removing versus redistributing risk;
b. what position they occupy in the value chain for risk
transfer - Insurer PGE, Reinsurer PGE or Market
Capture PGE.
Each of these choices brings advantages and
disadvantages.
. PGEs in sovereign states with fragile economies that are
highly exposed to natural catastrophe, where there is little
insurance penetration, develop disaster liquidity products
as immediate relief post-disaster, rather than as more
large-scale reconstruction
PGEs functions. Beyond their function in providing nancial
solutions, because of their unique position at the nexus of a
diverse set of stakeholders, PGEs perform a range of other
critical functions that are being increasingly recognized as
important:
. They act as the locus in which a conversation about the
evolving nature of the protection gap can take place,
thereby contributing to framing the debate and policy
initiatives connected to the protection gap.
. They play an important role in fostering the development
of knowledge about perils and how they aect specic
areas that were not previously tackled by markets; and
in changing the dynamics of knowledge distribution
and ownership.
. They facilitate risk identication and contribute to making
it more policy relevant to Government, by providing better
knowledge about perils and by linking this knowledge to
the availability of additional nancial resources in case
of disasters.
In doing so, our PGE Resilience Framework outlines the
important spillover eects that PGE’s have in informing and
improving society’s resilience to disasters.
PGE challenges. PGEs face signicant challenges, as
demonstrated by our PGE Evolution Framework, to:
. Balance the competing demands of dierent stakeholders,
both in terms of balancing social and market objectives.
. Support the more vulnerable portions of society without
undermining long term resilience by distorting incentives
and providing grounds for moral hazard.
. Adapt their remit so that it matches the evolving
characteristics of the protection gap but also avoid
mission creep and loss of eectiveness.
. Manage stakeholders’ expectations, so that they are not
perceived as silver bullets to the problem of the protection
gap, displacing other forms of risk management and
reduction, particularly those related to resilience.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
This report is by no means a nal word on the complex and
challenging issue of the protection gap. Rather, it has provided
some new insights about PGEs from which to continue the
dialogue of how it may be better addressed in dierent
contexts by multiple stakeholders.
This report should be considered as “call to arms” to learn
from, and make better use of those already established PGEs,
in order to better address the increasing threat of natural
and manmade catastrophic disaster, and the growing
protection gap.
Appendix
Appendix
METHODOLOGY
This report presents the full results from a qualitative research
study, looking at Protection Gap Entities (PGEs) around the
world.
Dataset. The data for this global research study of PGEs
includes:
• in-depth interviews of key stakeholders within and around
PGEs ( interviews with participants);
• ethnographic observations within a sample of PGEs;
• participant-observation at key conferences, workshops,
and meetings, as well as social events;
• more than , pages of documentary data such as annual
reports, press releases and media articles.
Methodological approach. This research study incorporates
global breadth and variation in cases. We studied PGEs in
dierent countries, employing dierent governance structures
and covering dierent types of risk. We adopted a multi-
stakeholder approach to all cases, accessing data with various
stakeholders from both the public and private sector.
Global breadth. We have collected primary data for
PGEs and their stakeholders across countries spanning
the globe. In addition, our dataset includes PGEs from both
developing and developed economies. This captures dierent
understandings of the protection gap and variation in the
solutions developed.
Variation. We selected the PGEs purposively to include
variation. First, the PGEs represented dierent governance
structures. In particular, the selected PGEs range from public
to private to partnership entities. Second, the selected
PGEs covered dierent types of risk, ranging from flood to
earthquake, tropical cyclone, drought, excess rainfall, and
terrorism. In addition, some were mono-risk, covering only one
peril, while others covered multiple perils. While we maintain
condentiality over our cases, those participants happy to be
named are listed on the back page of this report.
Multi-stakeholder approach: As well as global breadth this
research study focused on developing a holistic, polyvocal
understanding of PGEs. The study informants consist of:
• the PGEs;
• insurance market players such as insurers, reinsurers,
modellers and brokers;
• government actors such as ministers, governmental
departments and government-based organizations;
• intergovernmental organizations such as the World Bank and
OECD;
• other key protection gap actors such as independent
consultants and resilience teams.
The focus was thus not on any particular organization but
on developing a holistic understanding of the particular
local solutions to the protection gap via a multi-stakeholder
approach.
We thank the PGEs and their stakeholders for the generosity
and transparency that has made the analysis in this report
possible. Our engagement with these participants is ongoing
as we continue to explore the issues outlined in this report.
PGEs
Brokers
Modellers
Insurers
Reinsurers
Other actors
Government
Inter-governmental
8%
8%
13%
10%
2%
13%
10%
36%
Between State and Market: Protection
Gap Entities and Catastrophic Risk
GLOSSARY
ARC (African Risk Capacity) was established in as a
Specialised Agency of the African Union (AU), with Member
States that signed the Establishment Agreement initially, which
has grown to Member States in . ARC aims to provide
insurance products that help protect food security in the face
of extreme weather events, such as drought.
ARPC (Australian Risk Pool Corporation) is a corporate
Commonwealth entity established in and provides
insurance cover against terrorism in Australia.
CBRN Chemical, Biological, Radiological and Nuclear terrorism
perils.
CEA (California Earthquake Authority) is a not-for-prot,
publicly managed, privately funded entity established in
and provides insurance cover against earthquake in California,
US.
CCR (Caisse Centrale de Reassurance) is a public-sector
reinsurer established in and provides insurers operating
in France with multi-peril coverage against natural catastrophes
and other risks.
CCRIF SPC (Caribbean Catastrophe Risk Insurance Facility) is
an entity established in that provides insurance cover for
hurricane, earthquake ad excess rainfall to its, as in ,
Caribbean government-members and one Central American
government member.
CCS (Consorcio de Compensación de Seguros - Insurance
Compensation Consortium) is a state-owned entity established
in that provides insurance cover for natural and terrorism
disasters in Spain.
Crowding out eect is a phenomenon discussed in
economics in which rising public sector spending drives down
or even eliminates private sector spending.
Disaster liquidity is the short-term liquidity necessary in
the aermath of disasters to start recovery eorts while
maintaining essential government service.
EQC (Earthquake Commission) is a public entity established
in that provides insurance to residential property owners
against earthquake and associated perils such as natural
landslip, volcanic eruption, and hydrothermal activity in
New Zealand.
Exposure refers to the inventory of elements such as citizens,
infrastructure, housing, production capacities and other
tangible human assets in an area in which hazardous events
may occur. Measures of exposure can include the number of
citizens or types of assets in an area.
Financial protection is the nancial resilience of
governments, private sector and citizens through insurance-
based mechanisms.
Flood Re is an insurance pool established in to provide
insurance cover against flood in the UK.
FONDEN (Fondo de Desastres Naturales - Natural Disasters
Fund) is a public entity established in that provides
insurance cover to the Mexican States and the Federal
Agencies against natural disasters.
GAREAT (Gestion de l’Assurance et de la Réassurance des
risques Attentats et actes de Terrorisme – Management of the
Insurance and Reinsurance of Risks of Terrorist Attacks and
Acts of Terrorism) is a private-public partnership established
in that provides reinsurance cover against terrorism in
France.
Hazard is the condition that can cause a disaster such as a
hurricane or a tsunami.
Market capture a term that we use to reference the concept
of ‘regulatory capture’, in which a state organization set up to
regulate a market in the interests of citizens, ends up instead
adopting the point of view of the market and so serving its
interests. The market capture PGE can have the opposite issue,
in which market parties operating in a context where the PGE
is highly dominant, can end up assuming that their interests
are best met through the PGE, so dampening their own risk
appetite.
Modelled loss is the loss estimated through an insurance
catastrophe model on the basis of the parameters of the
event (e.g. wind speed). The event parameters are fed into a
catastrophe model to calculate the modelled loss.
Parametric insurance is a type of insurance that uses a
parameter or an index of parameters of the catastrophic event
as triggers for issuing a payout. Such insurance products may
combine a mix of triggers from indemnity to industry loss, to
the occurrence of specic parameters of a peril, such as wind
speeds within a specied zone. Such products can also be
linked to modelled losses (as opposed to actual claims for
losses), triggering a payment when losses exceed a particular
threshold. However, they do not have to be linked specically
to claims for property loss.
Peril is the direct cause of loss such as flood or earthquake.
Pool Re is a reinsurance pool established in to provide
reinsurance cover against terrorism in the UK.
Risk is the possibility of loss.
PCRAFI (Pacic Catastrophe Risk Assessment and Financing
Initiative) is an entity established in to provide the Pacic
Island Countries (PICs) with climate- and disaster-related
insurance.
Protection gap (global) is the gap between the insured and
actual economic losses caused by large-scale catastrophic
events. It refers to a global problem, aecting all countries, and
referring to the whole uninsured population.
Protection gap (local) refers to specic manifestations of
underinsurance in a particular region, such as lack of terrorism
cover for city-centre business districts; insucient emergency
capital reserves in developing economies to maintain essential
services aer natural disaster; or unaordable premiums for
homeowners in highly exposed flood plains, or in earthquake-
prone regions.
Glossary
Protection Gap Entity (PGE) is the entity that brings together
dierent market and non-market stakeholders in an eort to
address the protection gap by transforming uninsured risk
into insurance-based products that can be transferred onto
government balance sheets or into global nancial markets in
order to provide capital for recovery following a disaster.
Preparedness is the development of early warning systems,
support of emergency measures and contingency planning to
prepare for disasters.
Resilient reconstruction is the ex-post reconstruction of
property and the built environment for quicker, more resilient
disaster recovery.
Risk mitigation refers to taking ex-ante action to reduce the
adverse eects of disasters.
Risk identification is building the capacity to identify, assess
and analyse risk, typically as a technical capability supporting
the quantication of risk assessments and risk communication.
Risk reduction is the reduction of risks in society by
implementing structural and non-structural measures in policy
and investment.
Vulnerability refers to the propensity of exposed elements
such as individuals, a community, and assets to suer adverse
eects of hazardous events.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
References
References
CCS () Natural catastrophe Insurance cover. A diversity of systems. Consorcio de Compensacion de Seguros. Madrid.
OECD (). National terrorism risk insurance programmes of OECD countries with government involvement. OECD. Paris.
World Bank (). Sovereign catastrophe risk pools: World Bank technical contribution to the G. World Bank. Washington.
Swiss Re (). Under-insurance of property risks: Closing the gap. Sigma. No.
Patankar, A. & A. Patwardhan (). Estimating the uninsured losses due to extreme weather events and implications for
informal sector vulnerability: A case study of Mumbai, India. Natural Hazards-Journal of the International Society for the
Prevention and Mitigation of Natural Hazards (). pp. -.
World Bank (). Bringing resilience to scale. Global Facility for Disaster and Recovery (GFDRR), Annual Report.
Swiss Re (). Natural catastrophes and man-made disasters in : A year of record-breaking losses. Sigma. No.
Browne, M. & R. Hoyt (). The demand for flood insurance: Empirical evidence. Journal of Risk and Uncertainty. (). pp.
-.
Evans, S. (August ). Harvey economic loss could be up to bn: Moody’s analytics. Artemis. Retrieved from www.artemis.
bm.
Pool Re (). Pool Re to extend cover to include physical damage from cyber terrorism from April . www.poolre.co.uk/
pool-re-extend-cover-include-physical-damage-cyber-terrorism-april-.
World Economic Forum (). The global risks report – th Edition. World Economic Forum. Geneva.
Flavelle, C. (April ). Latest climate threat for coastal cities: More rich people. Bloomberg. www.bloomberg.com/news/
articles/--/the-latest-climate-threat-for-coastal-cities-more-rich-people?utm_source=slipcase.
Oxford Policy Management (). Independent evaluation of the African Risk Capacity. Oxford Policy Management. United
Kingdom.
Cummins, J. D. & O. Mahul (). Catastrophe risk nancing in developing countries: principles for public intervention. World
Bank Publications. Washington.
Clarke, D. J. & S. Dercon (). Dull Disasters? How planning ahead will make a dierence. Oxford University Press. Oxford.
Talbot, T., S. Dercon & O. Barder (). Payouts for perils: how insurance can radically improve emergency aid. Center for
Global Development. Washington.
Golnaraghi, M., I. Branagan, S. Fraser, J. Gascoigne & A. M. Gordon (). Guidelines for risk assessment to support sovereign
risk nancing and risk transfer. A joint publication of The Geneva Association and Insurance Development Forum (IDF).
Oasis Hub (December ) New tools to help understand environmental risk and resilience. Oasis Hub. oasishub.co/oasis-
blog/oasis-hub-prepares-launch-exciting-new-tools-section-help-understand-environmental-risk-and-plan-resilience.
McAneney, J., D. McAneney, R. Musulin, G. Walker & R. Crompton (). Government-sponsored natural disaster insurance
pools: A view from down-under. International Journal of Disaster Risk Reduction. . p. .
RMS (). A guide to catastrophe modelling. The Review – Worldwide Reinsurance.
Jarzabkowski, P., R. Bednarek & P. Spee (). One man’s trash is another’s treasure: Competing in a consensus market. In:
Making a Market for Acts of God. Oxford University Press, Oxford, Chapter . pp. -.
Schwarze, R., N. Holthausen, P. Locher, C. Quinto & G. G. Wagner (). The Swiss model - Economic and legal aspects of
integrated local public insurances. Berne: prevention foundation of the cantonal building insurance.
Stigler, G. J. (). The theory of economic regulation. The Bell Journal of Economics And Management Science. (). pp. -.
Le Den, X., M. Persson, A. Benoist, P. Hudson, M. de Ruiter, L. de Ruig, & O. Kuik (). Insurance of weather and climate-
related disaster risk: Inventory and analysis of mechanisms to support damage prevention in the EU. Publication Oce of the
European Union.
GFDRR (). The making of a riskier future: How our decisions are shaping future disaster risk. Global Facility for Disaster
Reduction and Recovery. Washington.
Between State and Market: Protection
Gap Entities and Catastrophic Risk
World Bank () FONDEN: Mexico’s Natural Disaster Fund - A Review. World Bank. Washington.
World Bank (). Disaster risk nance as a tool for development: A summary of ndings from the disaster risk nance impact
analytics project. World Bank. Washington.
Artemis (November ). CCRIF working on parametric insurance for Caribbean sheries industry. www.artemis.bm/
blog////ccrif-working-on-parametric-insurance-for-caribbean-sheries-industry
African Risk Capacity (November ) Press Release - Malawi to receive USD M insurance payout to support drought-
aected families. www.africanriskcapacity.org////press-release-malawi-to-receive-usd-m-insurance-payout-to-
support-drought-aected-families.
UNISDR (). Hyogo Framework for Action -: Building the Resilience of Nations and Communities to Disasters.
World Conference on Disaster Reduction. - January , Kobe, Hyogo, Japan.
OECD (). What Does “Resilience” Mean for Donors? An OECD Factsheet. OECD.
Ghesquiere, F. & O. Mahul (). Financial Protection of the State against Natural Disasters: A Primer. Policy Research
Working Paper . World Bank. Washington.
Yepes, C., V. Silva, T. Rossetto, D. D’Ayala, I. Ioannou, A. Meslem & H. Crowley (). The Global Earthquake Model Physical
Vulnerability Database. Earthquake Spectra, Online.
GFDRR (). Managing Disaster Risks for a Resilient Future, -. Global Facility for Disaster Risk Reduction. World
Bank. Washington.
World Bank (). The Sendai report: Managing disaster risks for a resilient future. World Bank, Washington.
Surminski, S. & A. H. Thieken (). Promoting flood risk reduction: The role of insurance in Germany and England. Earth’s
Future. (). pp. -.
Red Cross Red Crescent Climate Centre () ‘Forecast-based Financing’ of preparedness: Developing an operational
protocol. Red Cross Red Crescent, Climate Centre. www.climatecentre.org/downloads/les/programs/RCCC_FbF%of%
preparedness%pager%V%web.pdf.
Artemis (February ). World Bank, Swiss Re backed Kenya livestock index insurance pays out. www.artemis.bm/
blog////world-bank-swiss-re-backed-kenya-livestock-index-insurance-pays-out.
Jenkins, K., S. Surminski, J. Hall & F. Crick (). Assessing surface water flood risk and management strategies under future
climate change: Insights from an Agent-Based Model. Science of The Total Environment. (). pp. -.
Flood Re. (). Transitioning to an aordable market for household flood insurance: The rst Flood Re transition plan.
Report. London, UK.
Boneld, P. () The property flood resilience action plan. Department for Environment and Rural Aairs (DEFRA).
Rabinovici, S. (). California Earthquake Authority South Napa home impact study. California Earthquake Authority.
Johannesburg.
African Risk Capacity (). African Risk Capacity Strategic Framework -. African Risk Capacity.
Flood Re (). Incentivising household action on flooding and options for using incentives to increase the take up
of flood resilience and resistance measures. Flood Re’s response to the Social Market Foundation report. www.floodre.
co.uk////flood-res-position-on-incentivising-household-action-on-flooding.
CCRIF Press Release. CCRIF to make st payout to Trinidad & Tobago aer October rains. October .
Phil Bedient, Professor of Engineering, Rice University in: New York Times. Lessons From Hurricane Harvey: Houston’s Struggle
Is America’s Tale. Nov .
Picard, P. (). Natural disaster insurance and the equity-eciency trade-o. Journal of Risk and Insurance . pp. -.
Linnerooth-Bayer, J. & Amendola, A. (). Global change, natural disasters and loss-sharing: issues of eciency and equity.
The Geneva Papers on Risk and Insurance - Issues and Practice . pp. -.
Kleindorfer, P.R. & Kunreuther, H. (). The complementary roles of mitigation and insurance in managing catastrophic risks.
Risk Analysis . pp. -.
Pantucci, R, Ellis, C. & Chaplais, Lorien. () Lone-actor terrorism: Literature review. Royal United Services Institute, London.
Gill, O. (November ). Government criticised for failing to act over flawed terror insurance that le London traders exposed.
City A.M. www.cityam.com//government-criticised-failing-act-over-flawed-terror.
We gratefully acknowledge the generous participation of dierent stakeholders from the PGEs on which we collected primary
data for this report. Those PGEs that have agreed to be named are listed below. Their acknowledgement does not indicate any
endorsement of this report, and all analyses, ndings and errors therein remain our own.
• ARC (African Risk Capacity
• ARPC (Australian Risk Pool Corporation)
• CEA (California Earthquake Authority)
• CCR (Caisse Centrale de Reassurance)
• CCRIF SPC (Caribbean Catastrophe Risk Insurance Facility)
• CCS (Consorcio de Compensación de Seguros)
• Flood Re
• GAREAT (Gestion de l’Assurance et de la Réassurance des risques Attentats et actes de Terrorisme)
• Pool Re
Participating PGEs
facebook.com/cassocial
instagram.com/cassbusinessschool
@cassbusiness
youtube.com/user/cassprogrammes
Cass Business School
Disclaimer: All the information contained within this brochure was correct at the time of going to print. Published June 2018.
Cass Business School
In 2002, the School was renamed
Sir John Cass Business School
following a generous donation
towards the development of its new
Bunhill Row premises.
Sir John Cass’s Foundation
Sir John Cass’s Foundation has
supported education in London since
the 18th century and takes its name
from its founder, Sir John Cass, who
established a school in Aldgate in
1710. Born in the City of London in
1661, Sir John ser ved as an MP for
the City and was knighted in 1713.
City, University of London is an independent member of the University of London which was established by
Royal Char ter in 1836. It consists of 18 independent member institutions of outstanding global reputation
and several prestig ious central academic bodies and activities.
Cass Business School
106 Bunhill Row
London EC1Y 8TZ
T: +44 (0)20 7040 8600
www.cass.city.ac.uk
