Casualty Accumulation Risk
October 2015
Authors:
Brad Fischtrom (AIG)
Luc de Lignières (Axa)
Tim Jandeck (Generali)
Michael Brauner (Munich Re)
Guillaume Ominetti (Scor)
Eric Schuh, Andrea Scascighini and Sabrina Wulf (Swiss Re, workgroup lead)
CRO Forum Secretariat:
Kuba Szczygielski
CRO Forum – October 2015 3
Contents
1 Executive summary 6
2 Understanding casualty accumulation risk 8
2.1 Defining casualty accumulation risk 11
–
Classic clash 11
–
Serial aggregation 12
–
Systemic loss 12
2.2 Examples of historical accumulation losses 13
–
Asbestos 13
–
Deepwater Horizon oil spill 14
–
The medical malpractice crisis in France 15
–
Economic, societal and legal environments 15
2.3 Casualty accumulation matters in an interconnected and fast evolving society 16
–
Cyber 18
–
Pandemic 20
–
Nanotechnology 20
–
E-cigarettes 21
–
Concussions 21
–
Climate change 22
2.4 Embedding casualty risks in the overall accumulation risk 23
–
Crossing the lines of business 23
–
Accumulation with the asset side of the balance sheet 24
3 Assessing the potential effects of uncontrolled casualty accumulation 25
3.1 Why is casualty accumulation risk more challenging to assess and model than
property catastrophe accumulation risk? 25
–
Impact on insurers 26
3.2 Scenario-based modelling 27
–
The unfolding of a liability disaster 27
–
The scenario approach to quantification 28
–
Mapping scenarios to portfolio exposures 28
3.3 A forward-looking approach 28
–
Limitations of current actuarial techniques 28
–
Studying loss generation: from natural catastrophe modelling to liability exposures 29
–
Considerations on modelling for solvency purposes 30
4 Towards an effective management of the casualty accumulation risk 32
4.1 The role of Enterprise Risk Management 32
– Data quality and risk identification 32
–
Risk quantification / prioritization 32
–
Governance 32
–
Incentives 33
–
Mitigating actions 33
CRO Forum – October 2015 4
4.2 Managing the assumed risk 33
–
Monitoring of risk exposures 33
–
Management actions on underwriting 33
–
Risk transfer solutions 33
4.3 Increasing understanding to shape the industry 34
–
Why is increased market transparency the future? 34
–
Internal need for better data 35
–
Industry data standards for measuring casualty accumulation 35
–
Possible external demand for more information and thorough understanding 35
4.4 External stakeholders 35
–
Risk transfer market 35
–
Shareholders 36
–
Rating agencies & regulators 36
5 Conclusions 38
6 References 39
CRO Forum – October 2015 6
1 Executive summary
Casualty accumulation is the concentration of insured risks or insurance coverages that may be
affected by events or circumstances that cause substantial losses under several insurance policies,
and potentially over multiple years and geographies. In the past, casualty accumulation has led to
well-known claims complexes such as asbestos, the Mont Blanc Tunnel accident and the
Deepwater Horizon event.
The increased interconnectivity and interdependency of the world due to globalization, technology
advances, regulatory changes and macro-economic factors heightens the challenges faced by the
re/insurance industry in terms of detecting and managing accumulation potential within the casualty
portfolio. As supply chains span countries and companies, and new technologies develop, the risk of
casualty accumulation increases. Monitoring these developments is a priority for a Chief Risk
Officer.
Casualty catastrophes can be classified into three categories:
■ Sudden and accidental events, which we call classic clash, (eg, Mont Blanc/Deepwater Horizon).
■ Serial aggregation losses, where multiple insurance policies are triggered out of one single
defect, such as losses linked to the hazardous properties of diacetyl.
■ Systemic losses, where a repeatable process/procedure or industry/business practice results in
a series of losses, such as IPO laddering practices in the financial industry.
Exposure analysis of a portfolio should include two aspects. Firstly, past/known loss complexes
need to be managed and can serve as guidance for what could happen in the future. Secondly, the
Chief Risk Officer has to keep in mind that past events have less predictive power, compared to
property, towards assessing future casualty catastrophes. Asbestos is not going to be the "next
asbestos"!
In property re/insurance, accumulation risk is geographically defined and easier to model. Casualty
accumulation
1
risk modelling and assessment is more challenging. Current methodologies are based
on statistical analysis of triangles, expert judgement and top-down calibrations. Modelling casualty
accumulation is difficult in view of time dimensions of the exposures and also because societal
trends can significantly alter the risk landscape. However new technologies like big data and new
forward-looking modelling techniques have matured to a point where it may become possible to
significantly improve the accuracy, prediction power and quality of casualty accumulation models.
This trend should enable additional risk-taking activities, without loss of control over systemic and
accumulating exposures. In parallel, in a deterministic approach, the industry has to build
sophisticated and standardized scenarios for monitoring accumulation. These scenarios will help the
industry better understand casualty accumulation risk, and allow it to test sensitivities aimed at
setting limits for the risk.
Most re/insurers manage casualty accumulation risk by limiting coverage in their policy forms and
keeping track and limiting aggregate exposure. The Chief Risk Officer can enable a controlled risk
taking by sponsoring improvements in modelling techniques, and also by fostering better
understanding of casualty risks by capturing essential exposure information and key coverage
aspects on a standardized basis for the whole portfolio. Improved understanding of assumed
1
Insurance ERM (2015), “A new breed of casualty cat model” Retrieved from https://www.insuranceerm.com/analysis/a-
new-breed-of-casualty-cat-models.html
CRO Forum – October 2015 7
exposure through more complete data and better risk monitoring by building standard scenarios will
support the development of new products. In turn, that will enhance the ability of the re/insurance
industry to better-address earlier the casualty insurance needs of a fast evolving society.
CRO Forum – October 2015 8
2 Understanding casualty accumulation risk
Large, devastating catastrophes do not often generate big headlines about casualty insurance. But
there is a real risk associated with the possible build up – or accumulation – of known and unknown
casualty exposures. It is essential that a Chief Risk Officer fully understands this risk, both from an
historical perspective and in terms of the exposures and potential losses that could materialise in
the future. By nature, casualty risks are correlated with the general economy and societal
developments and, potentially, also with the investment portfolio of a re/insurance company.
The World Economic Forum Risk Report 2014 indicates that "over the past decade, risk
management has assumed a much more important role in many firms across different industry
sectors. In general, there is a trend away from technical planning for individual risks and towards
holistic planning for a range of unspecified risks. A spate of crises and extreme events in recent
years has convinced many companies that the benefits of globalization have been accompanied by a
much greater degree of interdependency and interconnectedness, bringing new vulnerabilities from
unexpected directions." In other words, casualty catastrophes (ie, large loss events due to
accumulating risks) are set to increase, and this will require the attention of Chief Risk Officers.
In this chapter the authors define a framework for classifying casualty accumulation risk and provide
examples of large – and potentially catastrophic – casualty losses emanating from undesired
concentration of risks. We review well known casualty-loss events from an accumulation
perspective to show the broad spectrum of potential scenarios. In order to help Chief Risk Officers
assess the future in a holistic way, we also discuss the implications of some emerging exposure
scenarios.
CRO Forum – October 2015 9
The global casualty market
The size of the global casualty insurance market per 2014 is estimated at a premium income for
primary insurance of USD 640bn. About USD 330bn, or 52%, is motor third party liability
insurance, general liability is USD 160bn, whereas accident, including workers' compensation, is
USD 150bn.
Source: Swiss Re Economic Research & Consulting
Global liability direct premiums totalled around USD 160bn both in 2013 and 2014, with 90%
coming from developed and 10% from emerging markets. Liability premiums are expected to
grow at a compound annual growth rate (CAGR) of 5% in developed markets and by 12% in
emerging markets over the next decade, to a total of around USD 280bn by 2025.
Commercial Liability, 2013
Premiums & GDP (USD billions) Percentage shares
Rank
Liability Total non-life GDP Liability/total non-life Liability/GDP
1
US
84.0
531.2
16,802
15.8%
0.50%
2
UK
9.1
99.2
2,521
9.2%
0.36%
3
Germany
7.8
90.4
3,713
8.7%
0.21%
4
France
6.8
83.1
2,750
8.2%
0.25%
5
Japan
6.0
81.0
4,964
7.3%
0.12%
6
Canada
5.2
50.5
1,823
10.3%
0.29%
7
Italy
5.0
47.6
2,073
10.6%
0.24%
8
Australia 4.8 32.7 1,506 14.8% 0.32%
9
China
3.5
105.5
9,345
3.3%
0.04%
10
Spain
2.2
31.0
1,361
7.0%
0.16%
Top 10
135
1,152
46,857
11.7%
0.29%
World
160
1,550
61,709
10.3%
0.26%
Note: Non-life excludes health insurance.
Source: Swiss Re Economic Research & Consulting, sigma 4/2014.
Within the casualty sector, accident (including personal accident and workers' compensation,
health business is removed where reported separately) is an important line of business, with
global premiums of USD 145bn in 2014. More than 80% came from developed markets,
although growth in the developed markets is projected to slow and even turn negative in the
next 10 years. In the emerging markets, premiums could grow by a CAGR of more than 16%.
Global accident premiums are forecast to reach USD 190bn by 2025, with over half coming from
Motor
TPL
52%
Liability
25%
Accident
23%
Global Casualty Market
Premiums, 2014: USD 640bn
CRO Forum – October 2015 10
The US casualty market
The US is by far the largest casualty market in the world, due to the size of the US economy and
the high penetration of liability insurance (0.5% of GDP). In 2013, US businesses spent USD 84bn
on commercial liability covers, of which USD 50bn was on general liability, including USD 12bn on
errors and omissions (E&O) and USD 5.4bn on directors and officers (D&O) policies. US
businesses spent another USD 13bn on the liability portion of commercial multi-peril policies, USD
9.5bn on medical malpractice, and USD 3bn on product liability covers.
By line of business, auto liability direct premiums written in the US were USD 138bn in 2014, 51%
was casualty sector premiums and about 25% was US P&C market premiums. The majority (84%)
of auto liability premiums came from personal lines
The "other liability" line is the second largest category, with USD 58bn in premiums in 2014. It
refers to the definition in the "schedule P" US reporting standards and includes umbrella business
as the biggest segment not allocated to a specific line of business. Other liability is closely
followed in size by workers' compensation with USD 56bn, each comprising approximately 21% of
US casualty premiums. The remainder were from medical professional liability (USD 10bn),
accident & health (USD 6bn) and product liability (USD 3.5bn).
Source: AM Best, Swiss Re Economic Research & Consulting
Medical professional
liability
4%
Product Liability
1%
Workers'
Compensation
21%
Accident & Health
2%
Other Liability
21%
Auto Liability
51%
US Casualty Direct Premiums Written, by Line of Business (2014)
CRO Forum – October 2015 11
2.1 Defining casualty accumulation risk
Casualty accumulation risk originates from the concentration of insured risks or coverages that may
be affected by events or circumstances that cause substantial losses under several insurance
policies,
2
and potentially over multiple years and geographies. Such events are called "casualty
catastrophes." They can be a single event, a complex of losses, or systemic events that generate
large losses. In the following, we define these different types of scenarios and provide examples.
Classic clash
This is when multiple claims are generated by (a) sudden accident(s), occurrence(s) or event(s) such
as the unexpected collapse of a building. The collapse can spark general liability, employers' liability
and professional indemnity claims from a single insured across multiple classes, and/or multiple
insureds across single or multiple classes.
2
There could be a liability catastrophe market loss which is not insured, but for our purposes, we're concerned with insured
losses.
Mont Blanc tunnel accident
The Mont Blanc tunnel connects Italy and France. On 24 March 1999, a truck transporting flour and
margarine caught fire in the middle of the tunnel, creating intense heat and toxic fumes. The fire
lasted 53 hours, 39 lives were lost and the economic damages were significant. Several parties
involved in the catastrophe were sued and many motor third party, general liability and product
liability claims were triggered. This example shows how easy it is for multiple parties including
municipalities, security authorities, operators of the tunnel and manufacturers to be involved in the
same casualty catastrophe.
Deepwater Horizon
On 20 April 2010, while drilling at the Macondo Prospect in the Gulf of Mexico, an explosion on the
Deepwater Horizon rig killed 11 crew members. The resulting fire could not be extinguished and,
on 22 April 2010, Deepwater Horizon sank leaving the well gushing oil onto the seabed and
causing the largest spill ever in US waters. Multiple general, employer, product and environmental
liability and also D&O insurance policies for multinational corporations were triggered to cover the
resulting economic and environmental losses.
CRO Forum – October 2015 12
Serial aggregation
This is where a series of losses can be linked back to one problem. For example a defect in the
design or manufacture of a product causes multiple losses, which can all be clearly linked back to
the faulty product or a single corporate failure, out of which multiple professional indemnity and
D&O losses could arise. Cases involving a single insured and single class are more likely to be serial
aggregations than clash losses.
Systemic loss
Here, a repeatable process/procedure or industry/business practice, rather than a faulty product,
results in a series of losses.
Diacetyl and the "popcorn workers'" lung disease
Diacetyl is a volatile organic chemical compound added to food to give a butter flavour. Workers
in popcorn factories who have inhaled diacetyl for long periods have been found to suffer from a
rare lung disease, bronchiolitis obliterans. The large number of companies producing and using
diacetyl, and the relative young age of the work force exposed to it gives rise to a broad
aggregation issue scenario. This is called the serial aggregation loss complex, triggered by one
defect, in this case the production and use of diacetyl.
The Parmalat default
The bankruptcy of Parmalat is one of the most significant financial events in Europe. Driven by a
mis- and non-transparent representation of the company's financial conditions, leading to a
USD 20bn hole in its accounts, several banks, auditors and Parmalat itself have been involved in
litigations. The example shows that even outside the US (with its high propensity to sue
environment), a bankruptcy can be a casualty catastrophe event given the resulting accumulation
of D&O and professional lines claims.
IPO laddering
An example of a systemic loss scenario is Initial Public Offering (IPO) laddering, which arose in
the late 1990s as the tech bubble and stock market speculation fuelled a boom in IPOs.
Laddering describes one particular type of claim against the practice of investment banks in
which, as a condition for receiving an allocation of shares in an IPO, they get investors involved in
another offering. With time, laddering has become a more generic label for a number of other
claims arising from IPOs and the payment of undisclosed commissions to investors in exchange
for preferential share allocation, or the obligation for investors to buy further shares in an issuing
company post IPO. The common factor in laddering claims is the creation of a false market for
shares. This has given rise to repeated claims by investors and regulatory actions against issuing
companies and investment banks.
CRO Forum – October 2015 13
The above examples for casualty accumulation risk and casualty catastrophes align well with
accumulation insurance products and risk management scenarios. A classic clash is similar in spirit
to the standard per risk and per event casualty reinsurance treaties common in Europe and Asia.
Serial aggregation is usually dealt with by batch clauses in insurance policies and also by product
failure risk management scenarios, the best example being any asbestos-related claims scenario. A
business disaster is sometimes used to model the capital requirements for D&O and professional
liability business in view of correlation with market risk.
For the Chief Risk Officer, the key observation is that losses from casualty accumulations can arise
from multiple sources and cover a wide spectrum of insurance situations. A proactive, open minded
and visionary approach is needed to anticipate and avoid the next big surprise.
2.2 Examples of historical accumulation losses
Multi-billion casualty catastrophe cases are rare but over time, there have been a number of
different events that together have caused billions of US dollars in insured losses. According to
Towers Watson, since 1950 there have been around 300 casualty catastrophes that have resulted
each in insurance industry losses of more than USD 100mn. Together these have generated more
than USD 500bn in total costs for re/insurers.
3
This section reviews some of these events.
Asbestos
Asbestos is the largest insurance industry loss event in history, with total losses in the US Property
& Casualty (P&C) sector alone estimated to be more than USD 85bn.
4
Asbestos falls under the
serial aggregation type of casualty catastrophe. The economic damages arising out of occupational
and general public exposure to asbestos and its resulting deadly disease, mesothelioma, consist of
many of the characteristics global re/insurers consider to be a worst-case scenario. For centuries,
asbestos was widely used in industrial and consumer applications until, and even after, scientific
studies proved that it was harmful to humans. Claims arose across a wide spectrum of industries,
with plaintiffs bringing litigation against multiple companies in many different industry sectors (eg,
asbestos fibre manufacturers, asbestos product manufacturers, asbestos product distributors). In
addition to the accumulation of losses across hundreds of companies spanning multiple industries,
claims also accumulated over time. Due to the cumulative, multi-year nature of the manifestation of
mesothelioma, courts awarded damages to plaintiffs spanning decades of products and policy years.
Even today, plaintiff lawyers continue to find new ways to litigate on asbestos claims.
5
3
Ball, M., Jing, Y, and Sullivan, L. (2011). “The Need for Casualty Catastrophe Models: A Way to Prepare for the ‘Next
Asbestos’” . Retrieved from https://www.towerswatson.com/en/Insights/Newsletters/Global/emphasis/2011/The-Need-for-
Casualty-Catastrophe-Models-A-Way-to-Prepare-for-the-Next-Asbestos
4
A.M. Best Special Report (2015). “U.S. Insurers Continue Funding of Asbestos & Environmental Liabilities Despite Elusive
End Game” Retrieved from http://news.ambest.com/presscontent.aspx?altsrc=14&refnum=22066
5
Swiss Re Webinar (August, 2013). "What’s Next For Asbestos“ Retrieved from
http://www.swissre.com/clients/newsletters/2013_08_claims_webinar.html
CRO Forum – October 2015 14
Cumulative Incurred Asbestos Losses – U.S. Property Casualty Insurers
Source: Towers Watson analysis of annual statement data compiled by A.M. Best and other industry data
The accumulation of losses across industries and time, the latency of mesothelioma and the long
time it took the insurance industry to react with policy exclusions, has caused an annuitized drag on
insurance industry profitability over the course of many years. As a result, highly accumulating mass
tort risks as exemplified by the case of asbestos should remain a top concern for re/insurers and
their Chief Risk Officers.
Deepwater Horizon oil spill
A blowout and explosion on the Deepwater Horizon drilling rig in the Gulf of Mexico on 20 April
2010 killed 11 workers and led to an ensuing 85 days of oil spillage, the largest marine oil spill event.
The accident was tragic from many different perspectives. For the insurance industry it forced a re-
think of the approach to managing sudden-event accumulation.
The explosion and spill generated claims across a wide suite of P&C products. For example:
1. commercial/marine property policies for property damage and business interruption;
2. workers' compensation cover for injuries/deaths
6
;
3. public/excess liability policies for bodily injury;
4. property damage liability, including damages arising out of pollution in certain cases;
5. economic damages to shareholders of directly and indirectly implicated companies prompted
D&O liability claims; and
6. Cameron, the manufacturer of the blowout preventer used on the rig was named in multiple
product liability lawsuits.
The spill attracted contractors from many industries to the Gulf-region to assist with the clean-up
effort, creating additional workers' compensation and public liability exposures. In addition, had the
oil slick moved more closely to the shore, contingent business interruption claims could have
emerged across a broad geographic area.
Tracking the linkages between geographical locations of casualty exposures is almost impossible
without overly-burdensome data collection, storage and analysis. Nonetheless, individual re/insurers
need to ensure they are not overly exposed to industrial explosion disaster events. Geographical
6
There were 11 deaths in the blowout and explosion, of BP and Schlumberger employees.
0
10
20
30
40
50
60
70
80
90
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
$ billions
Cumulative paid ($52 billion)
Case and IBNR reserves ($23 billion)
A.M. Best ultimate ($85 billion)
CRO Forum – October 2015 15
concentration is a key consideration in the management of classic clash losses, and geographic
information portals and maps can help ensure accumulation risk is properly reviewed.
The medical malpractice crisis in France
A recurring pattern in casualty insurance is that key legal decisions change the risk landscape, and
often come in times of economic stress. For victims, liability insurance could be a potential source
of revenue as the state legislates in favour of consumer rights in response to a particular situation,
and at the expense of insurance companies.
The French medical malpractice market was historically written on occurrence triggers with
unlimited coverage. In the early 2000s, this business was unprofitable and a landmark judgment
passed in November 2000 by the Cours de Cassation in plenary session, called "arrêt Perruche",
threatened to make medical malpractice uninsurable. The High Court awarded damages to a 19-year
old teenager born severely disabled because his condition had not been properly diagnosed prior to
birth. Payment for prejudice to be born disabled was thus acknowledged.
The whole market was hit by the decision and medical malpractice cover suddenly became very
expensive, if at all available. Insurers with high concentration in medical malpractice faced difficult
times and a series of reserve increases to cope with the new jurisprudence. To solve insurability
problems and to ensure cover was available, at the end of 2002 the French government introduced
several measures (laws called “Kouchner” and “About”), including establishing legal minimum
limits of insurance cover for practitioners, and a move to a claims-made trigger. Finally, pools
(Groupement Temporaire d'Assurance Médicale, GTAM, for insurance and Groupement Temporaire
de REassurance Médicale, GTREM, for reinsurance) were created to share the most delicate and
exposed risks. Since then obstetricians, anaesthetists and surgeons, those practitioners most
exposed, are required by law to be insured. Premiums remain high, anything up to EUR 10mn, given
that damages awarded in medical malpractice claims are significant.
Economic, societal and legal environments
Claims inflation
7
is one of the greatest risks for casualty insurers. Most re/insurers’ reserves assume
historical average inflation rates for the claims payments expected to be made in 10-, 20- or even 30
years or more into the future. Periods of high inflation, both medical and general, can have a direct
influence on claim outcomes. For example, the last period of prolonged high Consumer Price Index
(CPI) inflation in the US was from 1974 to 1982. Many factors, including the US legal environment,
influenced casualty incurred losses over that time period. However, in the late 1970s, there were
significant reserve movements across the industry in the US, which could be partially linked to high
CPI inflation.
Periods of high claims inflation (eg, medical, court rulings) can however also be disconnected from
CPI inflation, making future insurance claims even more difficult to predict. For example, in France
CPI inflation decreased over the period of 2002 to 2008 yet claims inflation for insurers was
increasing significantly due to favourable court rulings for victims in bodily injury cases. More
recently, most developed markets have gone through a period of low CPI inflation. However, cases
of claims inflation can still be observed, for example driven by increasing bodily injury costs. The
potential remains for a spike in costs that would cause accumulation of casualty reserve increases
across the industry and globally.
7
Claims inflation is not necessarily closely linked to widely published inflation indices like the CPI. Indeed, inflation of
liabilities is often more strongly impacted by legal, social, medical and fiscal changes, and portfolio effects than by pure CPI
inflation.
CRO Forum – October 2015 16
Casualty insurance products fundamentally cover litigation outcomes and costs. Therefore, casualty
insurance books are particularly exposed to systemic changes in the legal environments within the
countries in which their insureds do business. The time leading up to the mid-1980s was a
historically significant period of increase in tort costs in the US. Thereafter many US states enacted
tort reform laws which implemented various changes in their Justice systems to directly reduce tort
litigation and damages. There was a clear trend downward in market prices post-1986, when many
tort reform acts were passed.
8
Total Premium Share – U.S. Property Casualty
Source: Wang, Major, Pan, Leong (2010)
Similar trends, either upward or downward, will likely follow in other countries. There have already
been cases outside of the US where legal changes have impacted casualty business. These cases
have also demonstrated legal risk as a key driver of accumulation exposure. For example, the Courts
Act 2003 and the 2008 court decision in the Thompstone vs Tameside case in the UK changed the
landscape of compensation of bodily injuries entirely, moving away from structured settlements to
periodic payments without the consent of either party, and with the use of an adjustment index in
excess of the retail price index. The impact on the re/insurance industry has been significant and
long lasting.
2.3 Casualty accumulation matters in an interconnected and fast evolving society
As our environment changes and evolves, casualty accumulation risk is steadily growing. The
following Global Risks Interconnection Map visualizes the increasing cross-border interconnectivity
and interdependency due to globalization, technology advances, regulatory changes and
macroeconomic factors. These all increase potential accumulation risk.
8
Nockleby, J.T., & Curreri, S. (2005) “100 Years of Conflic: The Past and Future of Tort Retrenchment” Retrieved from
http://digitalcommons.lmu.edu/cgi/viewcontent.cgi?article=2467&context=llr
3.0%
3.2%
3.4%
3.6%
3.8%
4.0%
4.2%
4.4%
4.6%
4.8%
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010
Total Premium Share = Net Written Primium / Private Sector GDP
Calendar Year
CRO Forum – October 2015 17
Source: Global Risks Perception Survey 2013-2014
With more connections between industries, insureds and countries, there is larger potential for
casualty catastrophe events. Globalisation blurs the boundaries of countries and industries. For
example, companies span continents, and the rise or fall of one currency can be felt within a
company globally. Supply chains also span countries and companies and their growing complexity
and interdependency magnifies the impact of small disturbances. As claimants seek compensation,
they may try to sue every point in the supply chain, leading to an event which could spread across
industries and insurance lines. The continuous aim to optimize and shift business processes down
the value chain, including to outsourcing partners, can result in increased exposure for casualty
insurers.
Technological and scientific advances are changing at lightning pace. As technology develops, it
gives unprecedented access to information. Scientific advances also result in uncertainty around the
long-term effects on and potential for risk accumulation in liability insurance. Alongside these
increasingly interconnected risks, new forms of litigation make courts more accessible (eg, class
litigations actions outside of the US may rise), which can rapidly change the risk landscape.
Insurers have learnt from past experience and introduced some key mitigants against accumulation
risk in their products. Examples are the reduction of policy stacking risk, the use of claims-made or
occurrence-notified triggers, and the adaptation of exclusionary language. Yet, the risk of a casualty
CRO Forum – October 2015 18
catastrophe has not disappeared. On the contrary, the risk is ever heightened by the fast changes in
society and the way people do business and live their lives.
Asbestos as discussed earlier is the classic example of risk accumulation, but what could the next
big thing be? We do not have a crystal ball but we are convinced that Chief Risk Officers should
encourage their organizations to be opened minded about potential future scenarios. To spark
discussion we will now look at some current hot topics and scenarios that could potentially be
casualty catastrophe events in the years to come.
Cyber
Cyber-risk is defined as “the risk of doing business (in the widest sense) in the cyber environment
or internet. This risk is evolving rapidly with technological changes, particularly as organisations
compete to reach customers through the internet. As the environment becomes increasingly
interconnected and complex, so the tools and sophistication needed to exploit vulnerabilities
become simpler. Given the many ways that cyber-risk can affect the operations of a business, the
costs and impact are uncertain and increasingly substantial.”
9
The potential for accumulation in cyber-risk stems from the highly interconnected IT systems that
run the world today. This risk needs to be addressed if a successful insurance market to help
manage the exposure is to be developed. The interconnectivity of IT systems hinders the ability to
measure and monitor an insurer's cyber-risk accumulation exposure because a cyber-attack can
trigger several insurance products and independent policies in a chain mechanism, similar to
contingent business interruption. The challenge is further exacerbated by second- and third-order
linkages, which are particularly difficult to identify and analyse.
In the CRO Forum paper "Cyber Resilience: the Cyber Risk Challenge and the Role of Insurance"
written in November 2014 the accumulation potential of cyber risk was highlighted. The present
paper "Casualty accumulation risk" proposes re/insurance risk management tools to address such
accumulation challenges, therefore the cyber-risk case is highlighted again here. A key component is
the development of scenarios to help understand the accumulation exposure, keeping in mind the
factors that will influence the probability/severity of losses and accumulation potential. The diagram
below sets out a process that can be used to establish and maintain a cyber-risk exposure
accumulation framework.
9
CRO Forum (November, 2014).“Cyber Resilience: the Cyber Risk Challenge and the Role of Insurance” Retrieved from
http://www.thecroforum.org/cyber-resilience-cyber-risk-challenge-role-insurance.
CRO Forum – October 2015 19
Source: CRO Forum (2014)
10
Key steps in the process involve:
■ The development of realistic cyber-risk scenarios.
■ An analysis of which insurance products are affected by which scenario, and to what extent.
■ A catalogue of cause-effect impact maps to help insurers visualise the types of cyber-risks to
which they are exposed and the damage that can be caused.
The cyber-risk landscape is not static. Developments in IT, dependency on IT services, the
development of instruments and tools to identify and use system vulnerabilities, the motivation of
hackers, and legislation/litigation may all change the results of an assessment of accumulation
exposure. The framework must therefore be sufficiently dynamic to allow for regular scenario
updates and for cyber-risk accumulation exposure to be monitored on an ongoing basis.
10
CRO Forum (November, 2014).“Cyber Resilience: the Cyber Risk Challenge and the Role of Insurance” Retrieved from
http://www.thecroforum.org/cyber-resilience-cyber-risk-challenge-role-insurance.
Establish a network
of subject matter
specialists from
across the company
including IT, Risk
management,
Underwriting and
Claim departments
Set up cross functional
workshops with internal
subject matter specialists and
potentially external cyber risk
experts to collate information
on cyber risk exposure
Use information gathered in
the workshops to compile
databases of historical events,
underwriting and claims
information (eg limits,
attachment points, premiums,
losses, insurance wordings ad
endorsements)
Monitor the effects of the
changing cyber environment
on the cause-effects-impacts
landscape and underwriting
needs
Apply scenarios to insurance
portfolios so that it is clear
how and to what extent an
insurance portfolio is affected
by a scenario.
Develop realistic
cyber risk scenarios
and map cause
effect impact to
asset potential
accumulation of
losses
CRO Forum – October 2015 20
The diagram below illustrates the different lines of business that can be affected by cyber-risk, to
various degrees depending on the scenario.
Source: CRO Forum (2014)
11
Pandemic
Pandemics are also a potential source of casualty accumulation risk. They occur when a strain of
virus that can infect humans, cause serious sickness and is easily transmissible emerges or re-
emerges. In a hospital, accumulation risk can stem from the institution's suite of professional and
medical liability policies which are triggered if failures in protocol implementation result in new
contaminations to staff or patients.
A pandemic outbreak also leads to higher medical expenses and healthcare costs. Business
interruption and contingent business interruption policies can be indirectly affected by a pandemic
outbreak as transport, cross-border trade, supply chains and tourism could become disrupted.
Operational costs could rise to ensure business continuity, and a loss of new business could be
expected. Further casualty exposures could arise from accident & health policies with sickness
extensions or through workers' compensation, depending on the specific regulatory framework
around occupational diseases.
12
Nanotechnology
Nanotechnology is the engineering or manipulation of matter on a molecular scale. Products
manufactured with some form of nano-engineered materials are becoming increasingly common.
11
CRO Forum (November, 2014).“Cyber Resilience: the Cyber Risk Challenge and the Role of Insurance” Retrieved from
http://www.thecroforum.org/cyber-resilience-cyber-risk-challenge-role-insurance.
12
Boggs, C.J. (2014). “Is Ebola Compensable Under Workers’ Compensation.” Retrieved from
http://www.insurancejournal.com/news/national/2014/10/10/343250.htm.
CRO Forum – October 2015 21
Nano-manufacturing processes can improve structures and properties of materials, making them
stronger, water repellent, more durable, lighter, self-cleaning and more. Due to these properties, the
use of nanotechnology can be expected to grow significantly. There may in the future be some
similarities with asbestos, which for a long time was also considered "white gold" due to its superior
properties. No link between nanotechnology and a specific adverse health outcome has yet been
established, but should it turn out that some adverse health implications are connected with
nanotechnology once the long-term effects are better known, the potential accumulation loss could
be very significant.
E-cigarettes
E-cigarettes (as well as e-cigars and e-pipes) are a growing trend around the world. Available with or
without nicotine, e-cigarettes are marketed as a cigarette alternative, smoking cessation aid, and as
healthy and non-toxic. They are being used extensively across the US, Europe, and Asia. E-
cigarettes are not strongly regulated and laws governing use and sale vary widely. In some cases,
there are no regulations for product classification, product safety and quality, use in public places,
youth protection and advertising. There are two major risks for the insurance industry: the technical
component (battery, cartridge, etc.) and the chemical component (the liquids).
Currently, it is difficult to conduct a thorough risk assessment of e-cigarettes and the long-term
health impacts are unknown. A loss-accumulation scenario could arise though if e-cigarettes are
proven to be more harmful to health than presumed today. Respiratory diseases or other health
problems may increase, triggering liability claims similar to tobacco claims already seen.
Concussions
Concussions are traumatic brain injuries that occur when a blow to the head or body causes shaking
of the brain. Repeated concussions that are not identified and managed correctly have been linked
to Chronic Traumatic Encephalopathy (CTE), a degenerative brain disease. It has been estimated
that in the US, there are between 1.6 million and 3.8 million sports and recreation-related traumatic
brain injuries each year.
13
This number is increasing each year as the number of young people
playing contact sports grows and as many athletes become bigger and stronger.
In 2011 the first lawsuit was filed claiming that the National Football League (NFL) and Riddell – the
"official helmet of the NFL" were failing to protect players from brain injuries and concealing the
long-term dangers of concussions. Additional lawsuits have since come forward. On 22 April 2015,
the Federal judge approved a settlement agreement which could cost the NFL USD 1bn in
compensation payments to retired players.
14
The cases against Riddell are still pending.
Concussion safety laws related to youth sports have been passed in 49 states in the US. So, unlike
before, concussions are no longer seen as minor injuries. It might very well be that advances in
medical science may lead to new liabilities as the link between other injuries and concussions is
made. At present, this is more a hot topic in the US than elsewhere, but there could be a spill-over
to Europe and other regions. The large number of potential victims, including high-earning
professional athletes or young players, could create a perfect storm and casualty accumulation risk
scenario.
13
Center for Disease Control and Prevention (CDC). “Facts for Physicians” Retrieved from
http://www.cdc.gov/headsup/pdfs/providers/facts_for_physicians_booklet-a.pdf
14
Associated Press. “US NFL Concussion Lawsuit”. Retrieved
http://hosted.ap.org/dynamic/stories/U/US_NFL_CONCUSSION_LAWSUIT?SITE=AP&SECTION=HOME&TEMPLATE=DEFA
ULT
CRO Forum – October 2015 22
Climate change
According to the US Environmental Protection Agency (EPA) "climate change refers to any
significant change in the measures of climate lasting for an extended period of time. In other words,
climate change includes major changes in temperature, precipitation, or wind patterns, among other
effects, that occur over several decades or longer." Man-made climate change in simple terms
refers to the rise in the earth atmospheric temperature as a result of trapped CO2, methane and
nitrous oxide. For the past 10 years, carbon dioxide emissions have grown by an average of 2.5%
per annum.
15
This could amplify the risk for catastrophic weather events, like excessive heat,
drought, rains or storms, a rise in sea levels and a thawing of permafrost.
Categorization of climate risk
Source: Carbon Disclosure Project(2007)
16
To date there have been several climate-change related lawsuits in the US, but none have been
successful. Potential liability cases could include failure to build or maintain infrastructure (eg, dikes,
dams, well-functioning water ways, waste water/sewage systems, adequate staffing and material –
direct-link liability
17
) to mitigate the risks posed by severe weather events. There could also be
legislative actions to limit CO2 emissions (indirect-link liability), for which public and private sector
parties could be held liable. In the current legal environment, it seems unlikely that indirect-link
liability will be approved in court but if it is, the result would be a critical loss accumulation.
In the US, the direct-link liability is a reality and the potential for casualty accumulation after a natural
catastrophe can be dealt with by the insurance industry. For example, following floods and
hurricanes, pollution cases have been observed, and several cases of liability for construction or
utility companies have already been seen. Amongst other reasons a greater willingness to hold
15
Carbon Disclosure Project (2007). “Carbon Disclosure Project Report 2007” Retrieved from
https://www.cdp.net/CDPResults/CDP5_FT500_Report.pdf
16
Refer to footnote above.
17
Definition from Coping with climate change risks and opportunities for insurers, The Chartered Insurance Institute 2009
Direct risks
Physical risks –
extreme
weather events
Weather pattern
changes
Indirect risks
Regulatory
risks
Regulatory
risk
Litigation
risk
Other risks
Credit
risk
Market
price
risk
Reputa-
tion
risk
Operat-
ional
risk
Sector / Corporate
CRO Forum – October 2015 23
other parties liable following a natural catastrophe as well as better investigation of cause due to the
advances of science, increase the probability of liability claims after a natural catastrophe.
18
As an example of an extreme weather event, a large wildfire started on 20 October 2007 in
Southern California and continued for 19 days. Over 1200 homes were damaged. After
investigation, it was found that improperly maintained and designed power lines ignited the fire and
that the utility company was liable for the damages. The liability insurance tower was exhausted.
This is a typical example of a loss where liability insurance towers cover for mainly natural
catastrophes and where an accumulation potential is possible.
2.4 Embedding casualty risks in the overall accumulation risk
Casualty accumulation is rarely limited solely to casualty insurance losses. All the examples referred
to in this publication may have implications on other P&C lines, life/annuity products, investments
and to a lesser degree, operational impacts. These enterprise-wide implications require that casualty
accumulation risk is not managed within a silo. Business-unit executives should be held accountable
for considering firm-wide exposures in the context of their unit's business plans, portfolios and
individual account underwriting. Casualty accumulation quantification efforts should factor in all
material sources of risk across all liabilities and assets.
As described later in this paper, Enterprise Risk Management (ERM) plays a critical role in identifying
and quantifying firm-wide accumulation risks and concentrations. Most companies maintain
separate functions and management hierarchies for different insurance businesses (eg, P&C vs.
Life), business functions (eg, Asset Management vs. Underwriting), and regions. These silos are not
naturally focused on accumulation and correlations outside of their remit given incentive structures
and general areas of focus. The emergence of ERM functions in recent decades is evidence of
financial institutions’ recognition of risk correlation across disparate business units and regions, and
it is ERM’s remit to ensure firm-wide accumulations are well managed.
Crossing the lines of business
Casualty insurance payouts are often linked to multi-line events, and this dependence should not be
underestimated. A few examples of real and potential dependencies and correlations are:
■ Catastrophes involving property fire policies, marine and casualty insurance, as in the
Deepwater Horizon case.
■ Several very long tail lines of business, like workers' compensation in the US and motor third
party in countries which compensate bodily injuries by means of period payments, are exposed
to longevity risk factors, in a similar way to Life & Health (L&H) business. For example, in certain
European countries, the regulator mandates the use of specific mortality tables to the
capitalization of periodic payments. Changes in these actuarial assumptions occur regularly and
reflect changes in assumptions that spans across to L&H products.
■ Natural and man-made catastrophes result in damage to property and people. The damage is
covered through workers' compensation or accident & health policies, creating cross-line
accumulation.
18
Stevens, J., & Knuesli, D. “Property Events create emerging risks for casualty insurers” Retrieved from
http://www.carriermanagement.com/features/2014/05/18/123020.htm
CRO Forum – October 2015 24
Casualty risks are important on their own and as a first step, the exposure should be assessed on a
standalone basis. However, it remains essential that the Chief Risk Officer considers the
correlations and dependencies across all underwriting areas.
Accumulation with the asset side of the balance sheet
Casualty accumulation of liabilities can be correlated to risks on the asset side of the balance sheet.
These correlations can manifest through mutual sensitivities to macroeconomic factors or through
classic clash events.
The influence of externalities/macro-economic factors on asset prices has been studied for decades.
Macroeconomic factors which adversely impact asset valuations should be analysed for correlations
across all sources of risk, including casualty, as part of a strong asset liability management (ALM)
program. Selected key questions for re/insurers to address are:
■ Can a severe spike in general inflation impact the firm’s casualty reserves and fixed income
assets via rising interest rates?
■ Can equity market volatility or a rise in credit spreads lead to asset devaluation in concert with
an increase in D&O liability claims?
■ Do recessions or periods of high unemployment have an influence on workers' compensation or
other casualty line losses?
■ How do changes in gross domestic product (GDP) growth influence asset classes and top-/
bottom-line performance of the casualty portfolio?
■ How do real estate prices influence professional liability claims?
A detailed understanding of structural drivers of asset and liability risk can enable a re/insurer to
optimize its portfolio and facilitate resilience to financial crises.
In addition to macroeconomic factors, classic clash events can impact assets and casualty liabilities
simultaneously. As referenced earlier, the Deepwater Horizon event caused clash across a variety of
insurance contracts. Had the event been severe enough to result in a BP default, the losses could
have also crept into the asset portfolio via corporate bonds, equity, or other credit exposures such
as captive-fronting arrangements. Likewise, a cyber-attack could lead to financial distress in an
individual firm or industry sector, causing correlated losses across assets and liabilities. Finally, a
global pandemic would increase L&H and P&C insurance claims, and could also wreak havoc on the
financial markets.
Re/insurers need to comprehensively analyse correlated asset and casualty risks. There are perhaps
other lessons to be learnt from asset management within the casualty accumulation risk space. For
example, concentration risk and diversification are concepts well-known to financial risk managers,
but may not be properly deployed for casualty accumulation. Ensuring that an asset portfolio is not
overly concentrated in one industry or geography can help prevent outsized losses under adverse
circumstances. The same concept can be applied to casualty insurance. Firms should analyse
concentrations of exposure by geography, court jurisdiction, industry sector, insurance product line,
supply chains and other dimensions to provide protection against unforeseen catastrophes.
Analysing risk concentrations in this manner requires significant investment in data and technology,
but can yield very powerful tools with which to manage casualty accumulation.
CRO Forum – October 2015 25
3 Assessing the potential effects of uncontrolled casualty accumulation
The modelling and assessment of the casualty accumulation risk is complex and requires skills and
modelling techniques that go beyond the more-established means of assessing natural catastrophe
accumulation risk. In this section we set out the issues encountered when modelling extreme
events in long tail business and present some ways to go forward.
3.1 Why is casualty accumulation risk more challenging to assess and model than property
catastrophe accumulation risk?
Casualty accumulation risk differs from property catastrophe accumulation risk for the following
main reasons:
■ Its underlying drivers are intrinsically different. Property catastrophe risk is mainly driven by the
random occurrences of natural perils and the vulnerability of properties hit by an event. Hence
an analysis of past events, an assessment of how vulnerable a property might be and also of the
associated costs of repair allow the formulation of probabilistic laws for property catastrophe
losses in the future.
– Natural catastrophe property risks are fundamentally governed by physical laws. Casualty
accumulation risk, on the other hand, is linked to human behaviour, the political, legal and
economic environment, and social standards. Casualty lines are also usually characterized by
a far higher legal risk (eg, stacking or date of loss or applicability of exclusions). As a result,
the risk is constantly evolving. Because of its changing nature, the assessment of casualty
accumulation risk is difficult and more subjective. Lessons from the past are less directly
applicable to modelling the future than in natural catastrophe risk.
– As the past may be considered a partial indicator of the future, assessing the plausibility of a
specific type of casualty accumulation occurring in the future also relies on expert judgment.
It requires asking questions such as:
■ What could be the next products/substances to harm the environment or humans?
■ What is the likelihood or plausibility of these products/substances resulting in mass
litigation?
■ Which policies and particular lines of business of an insurance portfolio could be exposed
in such a scenario, and to what extent?
– Furthermore it requires knowing how different industries are connected, inasmuch as a
court ruling resulting in a mass litigation could potentially have an impact throughout the
value creation chain / lifecycle of an allegedly harmful product or substance. Understanding
the types of trading connections through which clients in an insurance portfolio do business
is hence also a key aspect of assessing more fully the potential for risk accumulation.
■ Casualty exclusively relates to longer-tail lines of business and it takes far longer for the risks to
be identified than in the natural catastrophe space. The long-term nature also means that
multiple underwriting years are likely to be affected in a severe accumulation scenario, and that
portfolio steering and actions to mitigate or manage the risk take time to become effective.
■ In the case of an extreme tail accumulation scenario, casualty may allow for less diversification
benefits than natural catastrophe cover. Geographical diversification, a key feature of natural
catastrophe risk, may not hold for casualty because the spread of an event underlying the
accumulation may not be confined to a particular area. It could even have a worldwide span.
■ Casualty catastrophes have a lesser impact on liquidity than property events. Extreme property
events are paid out quickly and require strong liquidity.
CRO Forum – October 2015 26
■ Under solvency regimes, property risks are mainly linked to premium charges. In casualty, the
exposures are covered through the reserve risk.
In the last two decades, the re/insurance industry has focused on developing increasingly
sophisticated stochastic models for property catastrophe risk. These models have traditionally
focused almost exclusively on earthquakes, windstorms and cyclones hitting so-called "peak
markets," in other words markets with high concentration of property assets/value and where
insurance penetration is high (the US, Europe and Japan). Today a number of new risk models that
cover wider geographical areas and a broader range of perils (eg, floods, tsunamis, droughts) are
being developed. The strong developments in property catastrophe risk modelling has made the
associated risk management easier and more systematic, and facilitated the development of
concepts now familiar to every stakeholder in the re/insurance sector, such as being able to set a
return period for a CAT event.
The same progress in modelling has not happened in casualty accumulation risk, even though the
loss potentials could be as high as or in excess of those resulting from severe natural catastrophe
events. The challenges to identifying and assessing casualty accumulation risk already discussed
likely explain why this is. The current approach to modelling casualty accumulation risk is usually
based on deterministic extreme-scenarios whose specifications reflect expert views on plausible tail
events (eg, the new asbestos, a financial crisis like in 2008). It does not use the probabilistic
dimensions common in property catastrophe modelling. Both the quality and availability of data are
key aspects for the impact assessment of any casualty accumulation scenario analysis.
Impact on insurers
Accumulation management is a driver of re/insurance company profitability. The volatility of
individual account-level risks tends to diversify as the size of a portfolio grows. In aggregate,
accumulation events and trends drive portfolio outcomes. Uncontrolled casualty accumulation can
impact re/insurers, as happened in the asbestos scenario described earlier, which resulted in a multi-
decade drag on industry profitability in the US and was a capital-level event for many insurers.
The unknown quantification of accumulation risk, due to long-tail and uncertain nature of casualty
can cause adverse outcomes for re/insurers in two ways:
■ It can cause casualty insurers to add more cumulative risk to their portfolio each year on an
occurrence policy form to finally generate potentially huge hidden accumulations.
■ Lack of quantification and understanding of accumulation can also lead to missed opportunities.
The re/insurance industry often reacts swiftly to issues in the public eye or that are trending in
scientific literature by placing exclusions on policy forms. In certain cases, the accumulation may
in fact be within the firm’s risk appetite, and there may be an opportunity to write the business
at the right price. But the reverse happens. For example, the liability policies for industries in
which people are exposed to electro-magnetic fields (EMF), such as telecoms, often exclude
this risk. Yet as of today, no causal link between bodily injuries and EMF has been proven.
Rather than exclude, an insurer could develop and leverage a deeper and more accurate
understanding of EMF exposure to write more telecoms business.
At moderate levels of severity, casualty accumulation events tend to result in earnings volatility.
Re/insurance industry stakeholders expect a certain degree of volatility for adverse events due to
the nature of the risks in a casualty insurance portfolio, so these moderate events do not have much
impact in the long run. However, if an individual re/insurer has a disproportionately large share of a
loss for an event relative to the rest of the market, the firm may well be subject to questions
regarding risk selection and accumulation management practices.
CRO Forum – October 2015 27
Very severe events threaten the solvency of re/insurers with lower levels of policyholder surplus and
more limited diversification across business lines. History has shown that casualty accumulation
events can be large enough to cause capital-reduction level losses. For this reason, casualty
accumulation management is critical for proper risk management in the eyes of all stakeholders.
3.2 Scenario-based modelling
Casualty accumulation risk is a concern for enterprise risk management. In an environment strongly
shaped by solvency frameworks and internal risk models, casualty accumulation requires a robust
approach for monitoring and quantification.
The unfolding of a liability disaster
The below graph is the authors' attempt to summarise the evolution of a potential liability disaster.
We shall use it to guide us through the sequence of steps to develop a quantitative risk model.
Unfolding of a liability catastrophe
Source: Swiss Re
A long period during which some harmful phenomenon might occur, followed by a number of years
during which the legal system determines liability and defines compensation, is characteristic of the
potential complexity of a casualty accumulation catastrophe event. Long after the original exposure,
effects will become manifest and some underlying causal link will be postulated. Over time, that link
will become more established, and eventually re/insurers will set an initial provision for future
claims. For a major catastrophe, there will likely be still much uncertainty in the form of pessimistic
estimates to counter optimistic denials of liability. A lengthy path through the courts will uncover
more evidence, leading to upward (or downward) changes in ultimate estimates.
A large portion of the loss to the insurer could be the legal defence transaction costs. The case of
asbestos in the US gives indication of the potential size of these: out of the total losses of USD
85bn, USD 21bn have been cited as being defence costs
19
.
19
Carroll, Stephen J. et al (2005). "Asbestos Litigation" Retrieved from
http://www.rand.org/content/dam/rand/pubs/monographs/2005/RAND_MG162.pdf
Initial reserve
is set
Court decisions lead to changed estimates
Effects
become
apparent
time
Payments (incl. Cost)
Policies affected
Exposure to harmful phenomenon
Impact duration
Loss emergence
CRO Forum – October 2015 28
The scenario approach to quantification
There are typically two ways to quantify a random event. The first is based on statistical analysis of
past experience. Standard actuarial reserving techniques fall into this category. This approach is
relevant also for casualty accumulation risk, especially for large and mature portfolios. However, the
unique nature and complexity of potential catastrophes, and the lack of history on large-scale
events, limits the effectiveness of this method in quantifying accumulation exposure.
Alternatively, a scenario-based approach can accommodate different views about potential future
developments. Underwriters, claims managers, lawyers and scientific experts use their knowledge
and imagination to construct specific hypothetical case studies. These scenarios provide a better
understanding of which components in the chain – from the initial exposure, to some harmful event,
to eventual settlement – might lead to a disastrous accumulation. They can also give indication of
which types of business are more dangerous, and of how expensive things may get.
The long time that an unfolding of a liability catastrophe takes exposes participants to changes in
legal and societal conditions. All risk models have to cover aleatory and epistemic uncertainty. In
casualty accumulation risk modelling, epistemic uncertainty is particularly relevant.
The split between these two ways of addressing the problem is not absolute. The loss experience
of the past can be adapted in some way to reflect more recent developments. Similarly, some
components of scenario may need to be quantified based on what has been observed before. Once
a set of scenarios is established and quantified, they need to be evaluated for their relevance.
However comprehensive it might seem to be, the highly unspecific nature of casualty accumulation
catastrophes leads us to suspect the existence of some unexpected and negative surprise events,
which could be characterised as “black swans”.
Mapping scenarios to portfolio exposures
The losses that the company has to bear should a scenario materialise depend on its portfolio of
re/insurance policies. To be useful, the link between scenario and policies has to be modelled.
Obviously, one could ask the experts defining the scenario to include an estimate of the company’s
share of the market loss into their considerations. However, a simple model based on market share
makes it difficult to track the evolution of the exposure to a scenario over time, as it does not
directly model the impact of the portfolio composition. We advocate the development of a clear
model that maps the scenario to the underlying portfolio and exposure based on clear metrics.
Since no (quasi-)standardized exposure measure for casualty catastrophe risk exists, additional
assumptions will have to be introduced to be able to construct such an exposure measure. In
general, it will depend on both the scenario to be assessed and the portfolio information available.
One of the challenges of a casualty catastrophe is the extent to which past underwriting years might
be involved in the accumulation of losses. Finality is a concept not well defined in this context. But
even besides this fundamental issue, there is the reduced availability of contract data from old
years, and also the possibility that certain soft properties of contracts (like the use of certain clauses)
not captured in the exposure measure underwent substantial changes.
3.3 A forward-looking approach
Limitations of current actuarial techniques
Liability exposure is a moving target. Risks are characterized by extremely heterogeneous
exposures, high uncertainty and a vast array of risk drivers. Changes in local legislation,
technological advances and societal trends such as urbanization, can fundamentally alter the risk
CRO Forum – October 2015 29
landscape, and have to be taken into account where historical data is available. For other casualty
lines, mostly motor, more data is available but the risk of change remains.
Current predictive modelling techniques address this problem only partially as their structure is
necessarily constrained. For example, they can quantify the impact of changes to the levels of
certain variables on a loss frequency basis. But they fail to cope where societal trends alter the
relationship between dependent and independent variables. In using exposure to predict loss
directly, traditional actuarial and predictive modelling techniques fail to fully utilize the information of
the loss-generating process. Differently put, they lack insight into the cause-effect chain from
exposure to loss.
Studying loss generation: from natural catastrophe modelling to liability exposures
The systematic integration of the cause-effect chain characterises the new era of natural
catastrophe modelling. It has allowed insurers to evaluate these risks with greater precision, transfer
findings from data-rich geographies to others with more sparse empirics, and to improve models
continuously as new events change and enhance understanding of loss-generating dynamics.
Whereas the loss- generating process for natural catastrophes such as an earthquake can be
defined in a relatively contained manner, for liability events the influencing factors – or risk drivers –
affecting loss frequency and severity are less distinct, more numerous and, consequently, less-
easily defined.
Risk Drivers in Property vs Casualty
Source: Swiss Re
While the challenge is substantial, given their global perspective re/insurers are well-positioned to
gain insight into the drivers of liability risk and to utilize their data sources, and also their expert
resources to identify and monitor relevant dynamics. Modelling the cause-effect chain and its driving
properties overcomes the direct dependency of predictive models on the availability of big data sets,
and also their limitations in the face of a changing risk landscape. These forward-looking models
(FLMs) often use a scenario-based approach, and their structural form allows for more complex
relationships between observable risk drivers and loss.
Liability event= rules of life (typically more individual company sources)
Hazard Vulnerability Values Conditions
Conditions
Legal
system
Social
standards
Macro
environment
etc.
Loss
Frequency
Earthquake
Loss
Frequency
Liability
?
?
?
+ + +
Cat event= laws of physics (typically more common to all firms)
….
CRO Forum – October 2015 30
Parameterization of FLMs is a difficult task as the number of parameters is high and not all can be
assessed based purely on data, either because data is not available or because the risk driver
concerned such as the general level of litigiousness within a jurisdiction, escapes objective
quantification. In principle, FLM could be set up in a way that each risk driver can be measured in an
objective observable. However, this would require data for initial calibration beyond what
re/insurers currently have at their disposal. As such, expert assessment still plays an important role
in FLM parameterization.
Back-testing remains a challenge for similar reasons. FLMs can be fully back-tested and anchored in
data-rich contexts (geographies, industries, severity ranges), but when transferring FLMs to a data-
sparse context like in some emerging markets, there is often no historic data available for back-
testing at all. For classical predictive models, this is a hard limitation. The structure of FLMs allows
the incorporation of predictive modelling techniques such as generalized linear models (GLMs)
where data availability allows, say to quantify the effects of some risk drivers, but to transcend their
limitations. FLMs base their key value proposition on the assumption that a cause-effect chain can
be established based on observations in a data-rich context and transferred to a data sparse one,
and that only the risk drivers having to be re-assessed for new geographies/industries.
FLM approaches are still nascent and more years of historic data are required before their
performance in adequately predicting loss frequency and severity can be judged. From a conceptual
standpoint, FLM approaches provide three new areas of opportunity in risk accumulation
assessment:
1. Their cause-effect approach could increase risk understanding and also help further improve
visibility and awareness.
2. As both data and expert judgment are used in calibrating their parameters allows for the findings
of FLMs to retain validity as they are transferred from data-rich to data-sparse contexts.
3. Their higher level of structural sophistication – which, granted, poses a risk of the models
becoming "black boxes" to many – allows FLMs to account for more complex trends. They can
preserve their predictive power also in light of changing legal, economic, societal and
technology dynamics.
Casualty is a long-term business. Once the FLM approach is developed and adapted, it should be
able to decompose the time steps from the first exposure or occurrence all the way to payout
patterns. It could allow the study of risk drivers working on these components which can reveal
accumulation risk not only for the current underwriting year, but also from past exposures for today
and in the future.
Considerations on modelling for solvency purposes
A quantitative model for casualty accumulation risk needs to be integrated in the economic one-year
framework of a re/insurer's internal solvency model. Despite this one-year view, re/insurers will take
into account the ultimate risk horizon when making management decisions (see chapter 4 on
effective management of the casualty accumulation risk). One of the characteristics of a liability
disaster is its gradual unfolding over a number of years. For a long time, neither a reliable estimate
of ultimate claims, nor of the timing of the payout, will be available. Asbestos is an obvious reminder
of how long this period of uncertainty might be. And again, it is clear that neither the initial reserve
estimate, nor some individual reserve strengthening will be a good proxy for the ultimate economic
impact of a liability disaster. Even before a court has established the existence of liability, market
perception could severely affect a company that is deemed to be involved.
Given the large uncertainty about the casualty scenario, no re/insurer will want to abandon the one-
year approach, just to be forced to introduce new unjustifiable assumptions that would be required
for quantifying the contingent evolution of the catastrophe over many years. Rather, a much-
CRO Forum – October 2015 31
reduced simple evaluation rule that assigns an estimated economic one-year impact to the scenario
will be employed. The reasoning behind such a rule will depend on the kind of scenario considered.
Solvency is determined on the legal entity level. Therefore the casualty scenario has to be evaluated
jointly with all other sub-models impacting the (economic) balance sheet. In other words, re/insurers
must think about whether potential dependencies with other risk categories have to be quantified
and modelled.
CRO Forum – October 2015 32
4 Towards an effective management of the casualty accumulation risk
In the view of the authors, managing casualty accumulation risk should remain a strategic priority for
every Chief Risk Officer. The impact of a mismanaged long-tail exposure can have dramatic
consequences.
20
The complexity of potential scenarios and the related modelling challenges are
reflected in the complexity of how to best manage the long-tail accumulation risk. In this chapter we
present some key aspects of the role of the Chief Risk Officer in this respect, and provide a way
forward based on increased visibility into the underlying exposures and risk metrics.
4.1 The role of Enterprise Risk Management
The Enterprise Risk Management (ERM) function in a re/insurance company articulates the firm’s
risk appetite and risk limits, identifies key risks (including emerging ones), quantifies/prioritises
material risks, and determines strategy on mitigating actions. Strong firm-wide risk management
requires an ability to continually identify and assess accumulation risks, including casualty and
exposure correlations with other risk factors, such as market and credit risk or the risks posed via
other Property & Casualty insurance products.
Data quality and risk identification
Identifying casualty accumulation risks can and should be approached in more than one way. A
re/insurer should utilise internally and externally-focused research to identify new and emerging
sources of systemic or event-based accumulation risk. At the same time, a re/insurer should
organize “exposure” data across various dimensions to identify areas of key concentration which
could create outsized accumulation losses. The ERM function can help coordinate the risk
identification efforts across all areas of the firm, and conduct concentration/exposure analysis across
both assets and liabilities.
Risk quantification / prioritization
ERM has a leading role in quantifying accumulation risks. Quantification efforts for casualty
accumulation will often require the construction of non-standard models in addition to dedicated
scenarios, and ERM functions are generally staffed with statistical or scenario-modelling subject
matter experts who are well-positioned to lead these development efforts. As already established,
casualty accumulation events are rarely isolated to casualty insurance claims. Quantification should
contemplate all material sources of risk across the firm, and ERM is a natural place to ensure
consideration of all the potentially correlated impacts of casualty accumulation.
Once quantified, these risks will need to be prioritized. ERM will have awareness of all of the firm’s
material risks across all sources. The firm’s full portfolio of risks and opportunities should be
weighed when considering mitigation action for casualty accumulation.
Governance
Understanding accumulation risk should manifest itself through a firm-wide governance structure.
Re/insurers should establish risk committees within each business unit and at the corporate level
with clear escalation thresholds for risk exposures. Risk aggregation should be a clear mandate for
20
Standard & Poor’s (June, 2013) “What May Cause Insurance Companies To Fail--And How This Influences Our Criteria”
Retrieved from http://www.standardandpoors.com/spf/upload/Ratings_EMEA/2013-06-
13_WhatMayCauseInsuranceCompaniesToFail.pdf
CRO Forum – October 2015 33
these committees to ensure that silo-based accumulations are reported and well-understood at the
senior management and board levels.
Incentives
Performance incentive-setting is an important topic for casualty accumulation risk, given the long-tail
nature of casualty business and the potential of organizations to inadvertently incentivise excessive
risk taking. It is possible that a firm’s growth strategy may directly increase accumulation risk. For
example, strategies to increase capacity within individual products or for individual accounts add
larger accumulation exposures to a firm’s balance sheet. “Cross-selling”, an industry term for
seeking to sell multiple insurance products to a single client, increases the likelihood of multi-policy
payouts from single events. New business growth targets can incentivize individual underwriters to
maximize short-term premium growth under the veil of hidden future adverse loss ratios within
casualty lines. ERM functions, human resources and senior management need to collaborate to
carefully optimize a firm's incentive structure to balance risk taking with the firm’s growth
aspirations.
Mitigating actions
ERM would generally not be accountable for decision-making on accepting casualty accumulation
risks or deciding on actions to be taken: the business owners are the decision-makers. However,
ERM checks that the business decision making remains within the risk appetite parameters of the
firm. Mitigating actions are discussed in greater detail in the next section.
4.2 Managing the assumed risk
This section discusses how a company can manage its assumed casualty risk. Identifying and
assessing the risk are prerequisites to managing risk. Thereafter different tools can be used and
actions taken to manage the risk.
Monitoring of risk exposures
The first natural tool to develop is a system of limits to ensure that exposure taken on a particular
risk driver or line of business remains within pre-assigned limits. The aim here is to avoid over-
concentration on a particular risk in the portfolio. For instance, casualty exposure may be limited in
terms of premium volumes and/or reserve amounts to a given percentage of available capital or
insured value. The design of a stochastic model may allow a system which controls exposure to
casualty accumulation at a given remote return period.
Once a system of limits has been defined, exposures may be managed upward or downward to
meet the assigned limit either directly through actions impacting the inward new business
(underwriting) or through outward risk transfer mechanisms (ie, reinsurance and retrocession).
Management actions on underwriting
New business may be limited by imposing caps on total policies’ aggregate limits. The wording of
contracts (eg, on specific exclusions) may also be adjusted to ensure scope of the cover is properly
understood. However, the long-term nature of casualty risk makes immediate management actions
far less effective and hence portfolio steering much more complicated.
Risk transfer solutions
Generally speaking, the retrocession market for casualty risk is less developed than for property
catastrophe risks. The long-term nature of casualty risk means that the buyer of cover is exposed to
a higher credit risk on the risk protection provider. Indeed, a company would expect a protection
provider covering long-tail adverse developments of its casualty business to be operating in good
financial health throughout the duration of the retroceded portfolio, which may be several years or
CRO Forum – October 2015 34
even decades. All things being equal, the selection of the risk transfer counterparty is therefore
more important for casualty than for a property catastrophe cover. The latter is usually set up over a
one-year time horizon and renewed annually. A collateralization of the contract may be also a way to
mitigate this increased counterparty risk.
In recent years, there has been a lot of focus on capital market solutions, notably Insurance-Linked
Securities (ILS), given the very strong influx of alternative capital into the reinsurance and
retrocession spaces. The capital has, however, remained largely confined to short-tail lines, almost
exclusively property catastrophe, for two main reasons. Firstly, modelling in property ILS relies on
physical external events understandable by anyone. Casualty risk, on the other hand, is more
challenging to identify and assess because it is not linked to easily understandable external events.
New entrants to the liability market feel less comfortable taking on this more ambiguous exposure.
Secondly, the alternative reinsurance capital providers often have to fully collateralize their maximum
liability. This full funding of the contract limit is far more adequate for short-tail lines like property
catastrophe, for which the collateral is unwound at the end of the year (if no event has happened).
For long-tail lines, however, the collateral needs to be frozen for a long period of time. Another likely
reason for the currently low observed appetite of capital markets for casualty lines is that the
(modelled) profitability margins are usually much higher in property catastrophe risk.
4.3 Increasing understanding to shape the industry
Following the elaboration on how to measure and manage accumulation risk internally, this section
will discuss how a better market understanding of casualty accumulation can help Chief Risk
Officers do their jobs. Building greater understanding in the casualty re/insurance market will lead to
better assessment and management of casualty tail risks. This will ultimately support additional risk
taking, which will be necessary to keep up with technological and societal progress.
Why is increased market transparency the future?
The need can be exemplified by an example from property re/insurance: Almost 50 years ago, an
earthquake in Mexico resulted in huge unexpected losses. The 1968 event demonstrated the
importance of accumulation control in property portfolios. Since then the natural catastrophe market
has undergone a "transparency transformation". From standardized data collection (CRESTA zones)
through to the founding of vendor models (eg, RMS) to the development of alternative capital
markets (ILS), re/insurers have come a long way from non-adequate accumulation monitoring to
systematically publishing exposure to natural catastrophe scenarios in their annual reports.
In comparison, to date understanding in casualty has been "reactive", that is post-event, as was the
case in the natural catastrophe market 50 years ago. The asbestos case is a good example: today,
most firms publish the amount and movement of their asbestos reserves in their annual reports.
The extent of information available in the property market does not yet exist on measures of
prospective casualty accumulation potential. How to achieve a better monitoring of such future
risks?
In recent years there has been an increased awareness of casualty catastrophe risks from both
internal and external stakeholders. Internally, better exposure data will be crucial in supporting risk-
taking activity. An increased focus on ERM will require re/insurers to better explain stakeholders the
risks the firm faces, including casualty accumulation and a disclosure of tail risks. The combination
of these internal and external factors will likely lead to the same process that the property market
went through, making casualty tail risks more transparent and easily understood.
CRO Forum – October 2015 35
Internal need for better data
Casualty accumulation monitoring is essential for optimal capital allocation within a company's risk
appetite, and for risk selection. The forward-looking modelling approach may be the solution to
improve reporting on casualty tail risks. Yet without good quality input data, the reliability of
sophisticated FLMs will be limited. To make use of FLMs to their full potential, re/insurers need to
collect better, more granular and standardized data on underwritten risks.
Furthermore, due to increased interconnectivity, it is increasingly difficult to understand and quantify
a re/insurance company's exposure to accumulation risk. Increased data quality can help overcome
this challenge.
However, one question remains as to what data standards should be used, and what the term
"better data" actually means. Creating a data warehouse for comprehensive information on every risk
underwritten by the firm may be the first step towards internally available standardized data.
Variables collected and stored for all liability risks could be, for instance, an insured company's size,
activity by industry code (eg, NAICS, SIC), turnover, insurance limits and sub-limits, and policy
exclusions or triggers. This, however only ensures internal standardization of data.
Industry data standards for measuring casualty accumulation
To cater to the potentially increased demand by external stakeholders for more disclosure and to
increase efficiency, industry-wide data format standards may be required. Much like in the natural
catastrophe sphere, where an unforeseen accumulation event triggered the re/insurance industry to
set up the CRESTA standards, the casualty insurance industry needs to find its own CRESTA zone-
type data format for coding casualty exposure information. This will enable, among other things, the
measuring of accumulation potential and thereby also increase market understanding. Industry
exposure data standards will help monitor accumulation risk in liability in the future, even if it is a
challenge to define those standards. An industry-wide agreement will be crucial. Associations like
the CRO Forum and re/insurance companies will have to work together to set the right standards.
There are well-known challenges in defining data standards for measuring casualty accumulation.
Challenges previously mentioned are that casualty events are not recurring and that the exposures
are more complex than in property. However, as the industry is moving towards the requirement to
better-monitor accumulation, the need for standardization is increasing. Along with the internal
motivations discussed above, industry data standards would benefit external stakeholders as well.
The increased understanding would allow for industry-wide comparisons, development of models,
more sophisticated scenarios and new risk-transfer markets. For instance, for insurance companies
not wanting to or able to invest in developing their own models for casualty accumulation, existing
vendor models could provide a less expensive solution. These models will require standardized data
inputs. Also, the potential development of more efficient risk-transfer solutions in the form of long-
tail ILS products calls for "CRESTA standard" data, much like on the natural catastrophe ILS market
where industry triggers (PERILS) are set up based on CRESTA loss data to facilitate payouts.
Possible external demand for more information and thorough understanding
Interest from external stakeholders in casualty tail risks is increasing. With the realization of the
complexity and possible downside potential of the risks, they will demand more information from
re/insurers.
4.4 External stakeholders
Risk transfer market
Over the past couple of years, large insurance companies have cut back their reinsurance cessions,
retaining more risks net. Along with the steady growth of accumulation risk described in Chapter 1,
CRO Forum – October 2015 36
growing retentions have contributed to the increased net exposure to casualty catastrophes. To
manage the increased accumulation potential, insurers have to find new solutions to keep this
exposure within their risk appetite limits. With better models and information, Chief Risk Officers
will be able to identify the right tail risks to transfer to the reinsurance market. At the same time, in
order to accept and cost for these risks, the reinsurers will require insurance companies to be
transparent on the accumulation potential in their reinsured casualty portfolios.
Along with traditional reinsurance solutions, the potential development of a casualty ILS market
would enable Chief Risk Officers to transfer casualty catastrophe risks more efficiently and to a
wider audience. Just like in the natural catastrophe alternative risk transfer market, a higher level of
information as well as advancements in understanding and modelling casualty catastrophe risks will
be crucial in developing a long-tail ILS market.
Shareholders
Some measurements on catastrophe risks are widely disclosed. For natural catastrophe and Life &
Health, prospective exposure measures for tail risks are published by numerous re/insurers in their
annual reports. This is not yet the case for casualty, where the only measure published is often the
amount and movement of asbestos reserves. This reflects on past losses, but does not give
shareholders a look into potential future casualty exposures.
Comparing the amount of asbestos reserves to the impact of certain natural catastrophe scenarios
shows that casualty catastrophes can influence the earnings and the balance sheet as much as
some property events. And, due to their long-tail nature, the effect can be span several years or
decades. It is in the best interest of shareholders to demand more information and visibility on
casualty risks.
Re/insurers could also increase shareholder understanding by publishing key metrics around
casualty accumulation. Some re/insurers already provide information on limits on total premium
deployed to long-tail exposures. Others disclose details on R&D activities to better assess casualty
accumulation risks. And some insurers publish their reinsurance structures to show how they
protect their portfolio. These are the first steps towards a comprehensive disclosure of casualty tail
risks.
Rating agencies & regulators
Rating agencies tend to focus on property catastrophe exposures when measuring tail risks. Their
capital models charge for exposure to specific natural catastrophe
21
scenarios but there is no such
for casualty catastrophes yet. Recently, the industry has seen some rating agencies demand more
information about the casualty sphere. For instance this year, for the first time, A.M. Best is asking
US-based insurers to provide details of their casualty catastrophe scenarios.
22
Insurance industry regulators require more disclosure not just for P&C but operational, investment
and other risks as well. Detailed reporting requirements for internal models' risk measures are
already in place but regulators will always strive to better understand risk management measures so
as to improve policyholder protection. Further, although not mandated yet, the spirit of new
regulations goes in the direction of more understanding on risk measures, including for casualty.
21
Standard & Poor's capital model charges for 250-year probable maximum loss for Nat Cat but no such scenario is defined
for casualty. AM Best is using a similar approach with defined return periods.
22
Insurance ERM (2015), “A new breed of casualty cat model” Retrieved from https://www.insuranceerm.com/analysis/a-
new-breed-of-casualty-cat-models.html
CRO Forum – October 2015 37
Casualty catastrophe losses can test the strength of a re/insurance company. Therefore, along with
shareholders, rating agencies and regulators are bound to be interested in a re/insurers' exposure to
accumulation risks. Coupled with the trend towards economic capital models, a requirement for full
disclosure of casualty tail risks is likely to be inevitable.
CRO Forum – October 2015 38
5 Conclusions
In the view of the authors, one key task of the Chief Risk Officer of a re/insurance firm is to increase
the understanding of casualty exposures, casualty accumulation risk and casualty catastrophic
scenarios. These risks pose a real threat to the financial security of the firm since:
■ In casualty in particular, past events might not be (fully) predictive of the future.
■ Our fast evolving and highly connected society increases the risk of severe casualty
catastrophes.
■ The slow emergence of adverse scenarios potentially exposes multiple underwriting years to
the same risk factor.
■ Management actions are not effective immediately as most of the risk is already on the firm's
books.
Casualty accumulation risk could be better assessed and monitored through the development of
more sophisticated and standardized scenarios and models:
■ We propose a systematic assessment by mapping catastrophic scenarios to the three classes of
losses: the classic clash, the serial aggregation and the systemic risks. Such scenarios should
cover well-known historical examples of casualty catastrophes, and also forward-looking
scenarios (ie, those that have not yet materialized but could happen). A core responsibility of the
Chief Risk Officer is to stress that future-loss scenarios could be materially different from
historical ones, and that it would be a mistake to base any consideration of casualty
accumulation on known losses only. Also, the impact of inter-line accumulation and of
accumulation with financial market exposures should be part of the assessment.
■ The quantitative modelling of the casualty accumulation risk of a portfolio has not yet reached
the level of development and standardization found in the management of natural catastrophe
exposure in property – examples were standardized data collection (CRESTA zones) and the
founding of vendor models in the property market. Casualty accumulation models, in contrast,
are mostly based on deterministic scenarios and statistical analysis of loss triangles. We
reviewed some key aspects of a casualty accumulation model and note that forward-looking
models could help better assess and monitor accumulation risk exposure.
A key enabler for the next generation of casualty accumulation models is not just data quantity, but
also data quality and a better understanding of the data. The absence of standards on exposures
demonstrates that the industry has not yet achieved a harmonized, more efficient understanding of
how to better manage casualty accumulation risk. The industry should strive to improve quality and
breadth of the casualty exposure information collected. It should also adopt data standards to
facilitate effective and efficient management of the risk and further development of risk transfer
solutions. The benefits would be great and support broader well-controlled risk taking activities and
better communication to all stakeholders.
It is key for any re/insurance company to actively manage casualty accumulation risk to reduce the
potential impact of casualty catastrophes. We hope that this paper will help raise awareness on this
important subject and shape the industry's approach to the next generation of risk management of
casualty catastrophes.
CRO Forum – October 2015 39
6 References
A.M. Best Special Report (2015). “U.S. Insurers Continue Funding of Asbestos & Environmental
Liabilities Despite Elusive End Game” Retrieved from
http://news.ambest.com/presscontent.aspx?altsrc=14&refnum=22066
Associated Press. “US NFL Concussion Lawsuit”. Retrieved
http://hosted.ap.org/dynamic/stories/U/US_NFL_CONCUSSION_LAWSUIT?SITE=AP&SECTION=HO
ME&TEMPLATE=DEFAULT
Ball, M., Jing, Y, and Sullivan, L. (2011). “The Need for Casualty Catastrophe Models: A Way to
Prepare for the ‘Next Asbestos’” . Retrieved from
https://www.towerswatson.com/en/Insights/Newsletters/Global/emphasis/2011/The-Need-for-
Casualty-Catastrophe-Models-A-Way-to-Prepare-for-the-Next-Asbestos
Boggs, C.J. (2014). “Is Ebola Compensable Under Workers’ Compensation.” Retrieved from
http://www.insurancejournal.com/news/national/2014/10/10/343250.htm
Carbon Disclosure Project (2007). “Carbon Disclosure Project Report 2007” Retrieved from
https://www.cdp.net/CDPResults/CDP5_FT500_Report.pdf
Carroll, Stephen J. et al (2005). "Asbestos Litigation" Retrieved from
http://www.rand.org/content/dam/rand/pubs/monographs/2005/RAND_MG162.pdf
Center for Disease Control and Prevention (CDC). “Facts for Physicians” Retrieved from
http://www.cdc.gov/headsup/pdfs/providers/facts_for_physicians_booklet-a.pdf
CRO Forum (November, 2014).“Cyber Resilience: the Cyber Risk Challenge and the Role of
Insurance” Retrieved from
http://www.thecroforum.org/cyber-resilience-cyber-risk-challenge-role-
insurance
Insurance ERM (2015), “A new breed of casualty cat model” Retrieved from
https://www.insuranceerm.com/analysis/a-new-breed-of-casualty-cat-models.html
Nockleby, J.T., & Curreri, S. (2005). “100 Years of Conflic: The Past and Future of Tort
Retrenchment” Retrieved from
http://digitalcommons.lmu.edu/cgi/viewcontent.cgi?article=2467&context=llr
Standard & Poor’s (June, 2013) “What May Cause Insurance Companies To Fail--And How This
Influences Our Criteria”
Retrieved from
http://www.standardandpoors.com/spf/upload/Ratings_EMEA/2013-06-
13_WhatMayCauseInsuranceCompaniesToFail.pdf
Stevens, J., & Knuesli, D. “Property Events create emerging risks for casualty insurers” Retrieved
from http://www.carriermanagement.com/features/2014/05/18/123020.htm
Swiss Re Webinar (August, 2013). "What’s Next For Asbestos“ Retrieved from
http://www.swissre.com/clients/newsletters/2013_08_claims_webinar.html
CRO Forum – October 2015 41
Disclaimer:
Dutch law is applicable to the use of this publication. Any dispute arising out of such use will be brought
before the court of Amsterdam, the Netherlands. The material and conclusions contained in this
publication are for information purposes only and the editor and author(s) offer(s) no guarantee for the
accuracy and completeness of its contents. All liability for the accuracy and completeness or for any
damages resulting from the use of the information herein is expressly excluded. Under no circumstances
shall the CRO Forum or any of its member organisations be liable for any financial or consequential loss
relating to this publication. The contents of this publication are protected by copyright law. The further
publication of such contents is only allowed after prior written approval of CRO Forum.
© 2015
CRO Forum
The CRO Forum is supported by a Secretariat that is run by
KPMG Advisory N.V.
Laan van Langerhuize 1, 1186 DS Amstelveen, or
PO Box 74500, 1070 DB Amsterdam
The Netherlands
www.thecroforum.org
