RiskBased Solvency Requirements

1
Risk-Based Solvency Requirements

• Hansjörg Furrer, Swiss LifeConference in Recent
  Developments in Financial and Insurance
  Mathematics and the Interplay with the Industry
• Oberwolfach, 18-24 February 2007

2
Agenda - Overview

• 1. From Solvency 0 to Solvency II
• Standard models and internal models
• Risk measurement in a multi-period framework
• Conclusions

3
Agenda - Overview

• 1. From Solvency 0 to Solvency II
• Standard models and internal models
• Risk measurement in a multi-period framework
• Conclusions

4
What is Solvency?

• Solvency
• the ability of an insurer to fullfil its
  obligations under all insurance contracts under
  all reasonable foreseeable circumstances
• How to meet this requirement?
• insurers establish over and above the technical
  provisions a buffer of freely available capital
  against adverse business fluctuations to meet
  their underwriting liabilities
• Question size of the buffer? time horizon?

5
Solvency in Europe Solvency 0

• 1973 / 1979 publication of the first non-life
  and life directives of the EEC (European Economic
  Community)
• mainly based on the work by Campagne 5
• Campagne recognizes that
• Investment risk is the most important risk for a
  life company
• Technical provisions are the most important
  invested amount
• Solvency margin as a percentage of the technical
  provisions
• the more prudent a company is, the more it has
  to pay for the solvency

6
Required solvency margin

• Required solvency margin for life companies (EEC,
  1979)
• 1st Result 4 of the mathematical reserves (?
  investment risk)2nd Result 3 capital at risk
  (? technical risk)
• Early warning signal, based on fixed ratios
  (wind-up barrier guarantee fund ? 1/3 of the
  solvency requirement)
• Rationale for the 4 solvency margin figure?

7
Pearson type-IV distribution
The Solvency I capital requirement (required
solvency margin) is defined via the ?-quantile
(one-year VaR) of the loss ratio distribution,
where the loss ratio X is defined as
Here L denotes the one-year loss and R the
technical provisions. Campagne 5 assumed that X
has a Pearson type-IV distribution
The required solvency margin qa is thus given by
Density function of the Pearson type-IV
distribution with parameters
For the data that were used, this led to qa 4
(a 95).
8
Solvency in Europe Solvency I

• 1992 2nd and later 3rd directive solvency
  margins left unchanged
• 1994 Insurance Committee asks the European
  supervisory authorities (from 2004 CEIOPS) to
  establish a working group to investigate solvency
  issues
• Chair of the group Helmut Müller (BAV,
  Bundesaufsichtsamt für das Versicherungswesen)
• 1997 presentation of the Müller Report 8
• current solvency margin structure satisfactory
• amount of the minimum guarantee fund needs to be
  increased (inflation)
• identification of three risk groups (technical,
  investment, non-technical)
• 2000 Based on the Müller Report and further
  preparatory work, new life and non-life
  directives were proposed
• 2002 New life and non-life directives were
  adopted by the European Parliament
  Solvency I directives

9
The Solvency I Life Directive

• Approved in March 2002
• Set into force as of Jan 2004

Timeline

• Available solvency margin covering the technical
  provisions must be of good quality
• Guarantee fund may not be less than a minimum of
  EUR 3m
• Solvency requirements for unit-linked contracts

Requirements

• Simple, robust
• easy to understand and use
• inexpensive to administer
• rule-based, and not explicitly risk-based (e.g.
  differences between asset and liability profiles
  are neglected)

Characteristics

• have worked well over the years
• have significantly increased the protection of
  the policyholders
• BUT Since the creation of these rules,
  significant changes have taken place in the
  insurance industry need to adapt the rules

Experience
10
Life insurance environment since the 90ies
In 1980 the life insurance industry was 150
years old, in 2000 it was 20 years old

• Equity markets experienced a strong bull run from
  1996-2000
• severe underwriting losses could be disguised
• Equity and corporate bond markets suffered falls
  in 2001-2002
• Interest rates stabilized on a low level
• problems with (high) guaranteed returns
• Increasing life expectancy / costs
• More frequent extremes / catastrophes (e.g.
  09/11, Tsunami)

11
Solvency in Europe towards Solvency II

• 1999 At a meeting of the Insurance Committee
  (IC) it was agreed that a more fundamental review
  of the overall financial position of an insurance
  company should be done. This review should
  include previously neglected risk classes (e.g.
  ALM risk, OpRisk,)
• 1999 Solvency II project was named for the first
  time in a paper entitled The Review of the
  Overall Financial Position of an Insurance
  Undertaking (Solvency II Review). Aspects
  mentioned
• Technical provisions (non-life) - Reinsurance
• Asset / investment risk - Solvency margins
  (reflecting the true risks)
• ALM - Accounting systems
• 2001 Launch of the Solvency II project by the
  European Commission (EC)
• CEIOPS was asked to provide input and
  recommendations

12
The three phases of the Solvency II project
13
Phase 1
KPMG Report, Sharma Report

• KPMG Report 7 criticizes that even though
  assets, provisions and solvency margins are
  connected, firm-specific risks are not explicitly
  taken into account
• CEIOPS set up a working group of insurance
  supervisors, known as LWG (London Working Group)
  as it was chaired by Paul Sharma from the FSA
• 2002 Publication of the Sharma Report 10
• Risk classification and causal chain mapping
• Survey on actual failures and near misses from
  1996-2001
• Discussion of 21 case studies in 17 European
  countries
• in each case, the causal chain began with
  underlying internal causes (e.g. incompetence,
  operating outside area of expertise, inadequate
  internal controls etc.)

14
Current Status of the Solvency II Project

• February 2007 Solvency II Framework Directive at
  a drafting phase
• to be published in July 2007
• Based on the various answers on the Calls for
  Advice that took place between 2004-2006
• During the same time period, CEIOPS Quantitative
  Impact Studies QIS (three so far, more to come?)
  provided insight mainly into the Pillar 1
  standards
• July 2007 Publication of the draft Solvency II
  Directive
• Political decision making process commences in
  the European Parliament and European Council of
  Ministers

15
Objective of Solvency II
Securing the benefits of the policyholders

• Note this does not necessarily require the
  continued existence of a company. Zero-failure
  will not be the aim of prudential supervisory
  systems. In a free market, failures will occur!

16
Agenda - Overview

• 1. From Solvency 0 to Solvency II
• Standard models and internal models
• Risk measurement in a multi-period framework
• Conclusions

17
European standard models (1/2)
From formula to spreadsheet solutions

• Major risk categories
• Market - Credit
• Insurance - Operational
• Most of these main risks can be subdivided
• into sub-risks (e.g. insurance risk into
• u/w risk, mortality risk, sickness risk,
• surrender or lapse risk, and expense risk)
• Notation Capital requirement for each risk
  category
• Market, insurance and credit risk are measured in
  the following way
• For OpRisk, the Basel II Standardized Approach is
  being used

18
European standard models (2/2)
From formula to spreadsheet solutions

• Depending on the correlations used (usually 1 for
  high correlation and 0 for low correlation),
  the solvency capital requirements result in
  factor-based formulae like the following
• Quantitative impact studies (QIS) serve to
  calibrate data from different countries

Insurance risk
Credit risk
19
The Swiss Solvency Test

• Definition The SST is a principle-based
  framework aiming to assess an insurance companys
  financial soundness in order to ensure the
  fulfillment of its (long-term) obligations
  towards the policyholders by taking into account
  the true risks the company is exposed to
• Risks entering the capital requirement
• market
• insurance
• credit
• emanating during a one-year time horizon
• Note OpRisk does not enter the SST capital
  requirement!

20
Timeline of the SST development
Market crash, deterioration of insurers solvency
H Lüthy becomes new director of FOPI,
reorientation to prudential supervision
2003
Start of Swiss Solvency Test project Mai 2003
with participation of industry, actuarial and
insurance association, consulting companies and
others, conceptual work finished end of 2003
2004
Field test 2004 with 10 insurers, supported by
consulting companies
2005
Further development underway on requirements on
internal models, group effects and intervention
levels
Field test 2005 with 45 insurers covering
approximately 90 of the market
2006

• Insurance supervision act implemented 01/01/2006

Field test 2006, mandatory for all large life
nonlife companies
Small companies, reinsurers and groups prepare
for full SST calculations in 2008
2007
Field test 2007, mandatory for all large life
nonlife companies
2008
As of 2008 all companies have to implement the
SST, and as of 201, target capital requirement
will be in force
21
SST and Solvency II timelines in comparison
SST

• Phase III
• Implementing phase
• New insurance supervision act AVO set into force
  on 01/01/2006
• SST for groups and financial conglomerates

• Phase II
• Two field test runs
• Calibration of the model, parameters

• Phase I
• Conceptual work
• Establishment of a standard model

Solvency II

• Phase III
• Implementing phase
• Modeling, standard models, calibration of models
  and parameters
• Implementation in national law

• Phase II
• Development of detailed rules
• Three waves of Calls for Advice giving structure
  of the framework
• QIS1, QIS2, Impact assessments,

• Phase I
• Gathering knowledge
• General design of the system (e.g. 3-pillar
  approach)
• KPMG report, Sharma report, .

22
The SST principles

• All assets and liabilities are valued market
  consistently
• Risks considered are market, credit and insurance
  risks
• Risk-bearing capital is defined as the difference
  of the market consistent value of assets less the
  market consistent value of liabilities, plus the
  market value margin
• Target capital is defined as the sum of the
  Expected Shortfall of change of risk-bearing
  capital within one year at the 99 confidence
  level plus the market value margin
• The market value margin is approximated by the
  cost of the present value of future required
  regulatory capital for the run-off of the
  portfolio of assets and liabilities
• Under the SST, an insurers capital adequacy is
  defined if its target capital is less than its
  risk bearing capital
• The scope of SST is legal entity and group /
  conglomerate level domiciled in Switzerland
• Scenarios defined by the regulator as well as
  company specific scenarios have to be evaluated
  and, if relevant, aggregated within the target
  capital calculation
• All relevant probabilistic states have to be
  modeled probabilistically
• Partial and full internal models can and should
  be used. If the SST standard model is not
  applicable, then a partial or full internal model
  has to be used
• The internal model has to be integrated into the
  core processes within the company
• SST Report to supervisor such that a
  knowledgeable 3rd party can understand the
  results
• Disclosure of methodology of internal model such
  that a knowledgeable 3rd party can get a
  reasonably good impression on methodology and
  design decisions
• Senior Management is responsible for adherence to
  principles

23
The revised insurance supervision act
Old act
New act

• Rule based
• Product and tariff approval
• Restrictions on products, investments and
  pricing
• Solely Solvency I capital requirements, no
  explicit consideration of investment risk
• Consequences
• Overexposure to risky assets
• Under-priced long-term guarantees
• Accounting and regulatory arbitrage
• Compliance culture
• Often underdeveloped risk management
• No link between regulatory requirements and
  companies internal risk measurements

• Principle based
• Review of provisions
• No restrictions on products (with the exception
  of mandatory group life and health insurance
  products)
• Less restrictions on investments
• Solvency I and risk-based capital requirements

24
The SST concept economic balance sheet
Valuation reserves
Free capital
Credit risk capt
Risk bearing capital
Market insurance risk capt (incl. scen)
Target capital
Statutoryassets
Market Value Margin
Market value of assets
Best-estimate Expected value of liabilities,
taking into account all up to date information
from financial market and from insurance. All
relevant options and guarantees have to be
valued No explicit or implicit margins Discounting
at the risk-free interest rate
Best- estimate liabilities
Market value of liabilities
25
The SST standard model

• Mixture between factor- and scenario-based model
• One-year time horizon for SCR, Cost-of-capital
  approach for MVM
• Major risk categories
• Market - Credit - Insurance
• Probabilistic approach the model output shall be
  a probability distribution function from where
  risk measures can be derived thereof
• Variance-covariance approach for market and life
  insurance risk
• Non-life different probability distributions for
  reserve risk and current year risk. The current
  year risk is further subdivided into a large
  claims risk and a small claims regime)
• SST already implemented in the law (Revised
  Insurance Supervision Act, set into force on
  01/01/2006)

26
Risk measurement framework
Risk models
Scenarios
Valuation models
Market risk
Market value assets
Insurance risk
Best estimate liabilities
Credit risk
Market Value Margin
Output of analytical model (probability
distribution)
Results from analytical model combined with
weighted average of scenario impacts
Target capital and Risk-bearing capital
27
Risk measurement in the SST
Average Value-at-Risk (Tail Value-at-Risk) as
risk measure
Probability distribution of the change in risk
bearing capital RBC1-RBC0
RBC1(?)
MVM1(?)
RBC0
Change in market value of assets
MVM0
BEL0
BEL1(?)
Capital requirement
Claims
Calamity
Target capital
Economic balance sheet at the valuation date t 0
Realization of theeconomic balance sheet at
time t 1
28
Market value margin MVM

• Rationale A buyer needs to put up regulatory
  capital aside during the run-off period of the
  portfolio of assets and liabilities
• The buyer needs to be compensated for the cost of
  having to put aside regulatory capital
• Market value Margin present value of future
  regulatory risk capital costs associated with the
  portfolio of assets and liabilitieswhere
  is the annually-compounded
  yield-to-maturity and the cost of capital
• In practice, a proxy such as the best-estimate of
  future liabilities can be used to calculate the
  future one-year risk capitals

29
SST standard model for market risk (1/2)

• The SST standard market risk model is a
  RiskMetrics type model with given risk factors
  Zi such as e.g.
• interest rates - exchange rates -
• equity prices - implied volatilities
• commodity prices - credit spreads
• Assumptions
• The risk factor changes X(t) Z(t) Z(t 1)
  are assumed to have a multivariate normal
  distribution with covariance matrix ?
• The change in risk bearing capital is a linear
  function of the risk factor changes
• The change in risk-bearing capital is univariate
  normally distributed

RBC(t) f (Z(t))
30
SST standard model for market risk (2/2)

• 73 Risk Factors
• 413 interest rate
• 1 Credit spread
• 4 FX
• 6 Shares
• 4 Real Estate
• 1 Hedge Fund
• 1 Private equity
• 1 Participations
• 3 Implied vola

Interest rate time buckets 1,,10, 15,20,30
years
CHF
EUR
USD
GBP
FX
Equity

• Shares
• CHF
• EUR
• USD
• GPB
• Japan
• Asia ex Japan
• EMU SmallCap
• Real Estate
• IAZI
• Commercial
• Rüd Blass
• WUPIX A

• Hedge Funds
• Private Equity
• Participations

• FX
• EUR USD
• GBP JPY

31
SST standard model for life risk (1/2)

• Analytical model with the following risk factors
• Mortality - Recovery - Expenses
• Longevity - Surrender / Lapse
• Morbidity - Capital option exercise
• Assumptions
• Risk factor changes are multivariate normally
  distributed with mean zero and covariance
  structure ? (based on expert opinion)
• Biometric risks are assumed to be independent
  from market risk factors
• Scenarios
• Pandemic (Spanish flu 1918 translated to the
  present)
• Disability scenario (short term increase
  systemic increase)
• Mortality long term changes (to take into
  account of systemic over- or underestimation),

32
SST standard model for life risk (2/2)
33
SST standard model for credit risk

• Definition of credit risk within the SST
• risk of default or change in the credit quality
  of issuers of securities to whom the company has
  an exposure
• Distinction between
• default risk
• downgrade or migration risk
• Quantification
• Basel II (standardized) approach
• Internal models
• Basel II IRB
• Credit risk portfolio models provided that they
  capture the credit migration riskCreditRisk
  would not be permissible since only the default
  risk is modeled

34
SST standard model for scenarios (1/3)

• Motivation during normal years, the market and
  insurance risk capital is determined on the basis
  of a variance-covariance type model, i.e.
• the risk factor changes are assumed to follow a
  multivariate normal distributionwith mean 0 and
  covariance matrix ?
• The change in risk bearing capital is a linear
  function of the risk factor changes (1st order
  approximation)
• Poor description of extreme tail
  events(distribution of risk factor changes have
  more extreme events than would be predicted by
  the normal law)

35
SST standard model for scenarios (2/3)
36
SST standard model for scenarios (3/3)

• The impact of a number of scenarios (both
  pre-specified and company specific) on the risk
  bearing capital has to be assessed
• Assumptions scenario i, i 1,,m occurs with
  probability pi and causes an additional loss of
  ci (cilt0)
• The probability of a normal year is
• The probability distribution of the change in
  risk bearing capital during one year with
  aggregated scenarios is

37
Scenario aggregation an illustrative example
? 10 ? 40 c1 - 50 c2 - 150 p1 0.04 p2
0.05 p0 1 (p1 p2)
38
Pros and cons of standard models
Pros
Cons

• Easy to implement in a spreadsheet
• Allows for fast calculations
• Easy to amend / extend
• Suitable for management purposes
• Easy to verify for the supervising
  authorities/auditors

• Too simplistic
• Hidden model assumptions
• Possibly inconsistent
• Model mixtures (stochastic ? volume-based)
• Valuation and risk measurement
• Aggregation of risk classes
• Expert opinion required at various levels
• Subjective parameter estimates such as
  correlation coefficients between the four main
  risk categories
• Incomplete
• Options and guarantees not priced

39
Characteristics of a good risk model

• An economic capital model should be consistent
• between the valuation of assets and liabilities
• between the valuation and the risk measurement
• between the risk assessments of different time
  periods in a multi-period setup
• With an internal model, one should be able to
  determine a level of capital that closely
  reflects the actual risks borne by the company in
  terms of
• Exposure to non-linearities, in particular
  (embedded) options and guarantees
• Volatilities of securities
• Underwriting results
• An economic capital model should be
• clearly and comprehensively documented
• integrated into the regular risk reporting
  process (i.e. no ad-hoc solutions for external
  parties such as regulators)
• equipped with clear responsibilities, reviewed by
  internal and external parties

40
The demand for internal models

• Regulatory perspective In October 2006, the
  Swiss regulator (FOPI) sent out a letter to all
  Swiss life insurance companies asking them to
  abstain from simplified economic capital models
  (such as variance-covariance type models) for SST
  purposes. Rather, companies should develop and
  use internal models. Companies have until 31
  March 2007 to present a project plan to the FOPI
  regarding the implementation of such a model.
  Rationale
• standard model not appropriate for large
  companies, re-insurers, insurance groups and
  financial conglomerates
• large volumes of the business may be written
  outside of Switzerland
• there may be substantial amounts of embedded
  options and guarantees, i.e. non-linearities
• Rating agencies SP rewards innovation within
  their ERM assessments

41
What makes a good model?
Internal model requirements
Technical
Process related
Corporate governance

• Consistency (all levels)
• Appropriate modeling of non-linearities, in
  particular (embedded) options and guarantees
• Multi-period stochastic simulation approach

• No ad-hoc solutions (e.g. models must be
  integrated into internal risk management
  reporting)
• Comprehensive documentation
• Regulatory assessment / review

• Clear responsibilities

?
42
Modeling and implementation issues

• Time discretization (continuous daily, monthly,
  quarterly, ?)
• In a parametric framework model choice, i.e.
  tractability versus correctness (Example
  dependency structures)
• Granularity of the asset and liability universe /
  number of model points
• Number of simulated sample paths / scenarios
  efficiency versus statistical significance of the
  output. Confidence intervals, rare event
  simulation, .
• Calibration
• One-period versus multi-period approach?
• Choice of risk measure

Complicated models will always be based on
subjective esti-mates which make the outputs at
best difficult to verify or compare across the
industry
43
Agenda - Overview

• 1. From Solvency 0 to Solvency II
• Standard models and internal models
• Risk measurement in a multi-period framework
• Conclusions

44
Motivation

• Belgian life insurance companies are given
  exemption fromthe strengthening of their
  reserves if they can demonstratethat they have
  an excellent asset-liability management in place.
  In this regard, the Belgian control authority
  CBFA recently released a document including the
  following request
• Lentreprise dassurances fournit la Value at
  Risk (VaR) et la Tail Value at Risk (TailVaR) à
  un horizon de 1 an, de 5 ans et de 10 ans, ainsi
  que la probabilité de ruine sur ces même horizons
  de temps.
• Is this a well-posed problem?

45
One-period vs multi-period risk measurement
Risk evolving over several periods of uncertainty
?? one-period risk
Multi-period
One-period

• Availability of information
• Possibility of taking intermediate actions
  (willingly or unwillingly)
• rebalancing the portfolio according to a
  predetermined investment strategy
• Bonus policy
• Raising capital
• Selling off subsidiaries (legal structure)
• Exercising embedded options
• Appropriate risk measure choice
• To be applied to the terminal values or the
  entire stochastic process?
• Properties such as time-consistency

• Object of interest financial position X at the
  end of the period under consideration
• Absence of inflow of information
• Buy and hold strategy, i.e. no intermediate
  actions are permissible

46
Time-consistent dynamic risk measures
cf Delbaen et. al.

• Key question how are the risk assessments of
  financial positions interrelated in different
  periods of time?
• Definition a dynamic convex risk measure
  is called time consistent iffor all
  financial positions X, Y ? L? and t 0, 1, ,
  T-1
• Interpretation acceptability tomorrow in any
  state of the world ensures acceptability
  today. Put another way, capital requirements
  should not contradict each other as time evolves.

47
Tail-VaR is not time-consistent

• One can show that Tail-VaR is not time
  consistent!
• More general risk measures that solely depend on
  the law of the terminal value can not be
  time-consistent
• The Belgian control authority should be careful
  when asking for multi-period Tail-VaRs this can
  lead to contradictory capital requirements.
  Assuming a holding strategy, then,
• at time 0, the regulator may accept a companys
  risk capital
• at time 1, however, the regulator might be forced
  to refuse!
• Different (investment) strategies lead to
  different capital requirements (regulatory vs
  internal view)

48
Future challenges

• and the role of academia
• Modeling of group and cross-sectoral issues (no
  longer on solo level), in particular
• Risk aggregation / diversification
• Legal structure
• Appropriate modeling of all the CRTIs (capital
  and risk transfer instruments such as
  reinsurance, guarantees, )
• Fungibility of capital

49
Agenda - Overview

• 1. From Solvency 0 to Solvency II
• Standard models and internal models
• Risk measurement in a multi-period framework
• Conclusions

50
Conclusions (1/2)

• Rapid changes in the design of insurance
  supervisory frameworks in the wake of the stock
  market crash at the beginning of the noughties
• Key driver towards model choice as simple as
  possible, but not simpler
• Standard models often not appropriate to capture
  all the risks taken by a company (too simple)
• Standard models rather designed for small
  companies than large ones
• Development of internal risk models very costly
• Any financial model is a simplified and thus
  imperfect representation of the economic world
  and the ways in which managers perform
  (financing) decisions, investment, or trading.
  That is, be aware of the model risk!
• Alignment of the different capital requirements
  (regulators, rating agencies)

51
Conclusions (2/2)

• Complicated models always be based on subjective
  (parameter) estimates.
• At best estimates difficult to verify (by
  investors / auditors) or compare across the
  industry
• At worst manipulation of models to flatter the
  bottom line
• Multi-period risk model approaches desirable, but
  leave the company with a myriad of decisions /
  assumptions to be taken
• Lack of capacity for regulators / auditors to
  verify, validate and approve the companies
  internal risk-calculation models
• Lack of convergence among regulatory regimes.
  Large and internationally active insurance
  companies may be faced with the problem that some
  of their (overseas) subsidiaries will be
  regulated by other, often cruder measures (A
  battle over Basel II. The Economist, 4 November
  2006)

52
References

• Basel Committee on Banking Supervision. (2005).
  International Convergence of Capital Measurement
  and Capital Standards. A Revised Framework (Basel
  II). Bank for International Settlements (BIS),
  Basel. www.bis.org/publ/bcbs118.pdf
• Bundesamt für Privatversicherungen (2004).
  Weissbuch des Schweizer Solvenztests.
  www.bpv.admin.ch/themen/00506/00552/index.html?lan
  gde
• Bundesamt für Privatversicherungen. (2005).
  Verordnung über die Beaufsichtigung von privaten
  Versicherungsunternehmen. Aufsichtsverordnung,
  AVO, 961.011. BPV Bern. Available under
  www.admin.ch/ch/d/sr/9/961.011.de.pdf
• Bundesamt für Privatversicherungen (2006).
  Organizational Requirements for Model Use in SST.
  BPV Bern, 12 January 2006. www.bpv.admin.ch/themen
  /00506/00530/index.html?langde

53
References (contd)

• Campagne, C. (1961). Standard minimum de
  solvabilité applicable aux entreprises
  dassurances. Report of the OECE (Organisation
  Européenne de Cooperation Economique).
• International Actuarial Association (IAA).
  (2004). A Global Framework for Insurer Solvency
  Assessment. Research Report of the Insurer
  Solvency Assessment Working Party. Available at
  www.actuaries.org/LIBRARY/Papers/Global_Framework_
  Insurer_Solvency_Assessment-public.pdf
• KPMG (2002). Study into the methodologies to
  assess the overall financial position of an
  insurance undertaking from the perspective of
  prudential supervision.

54
References (contd)

• Müller, H. et al. (1997). Report of the Working
  Group Solvency of Insurance Undertakings. Set
  up by the Conference of the European Union Member
  States. Available under www.ceiops.org
• Sandström, A. (2005). Solvency. Models,
  Assessment and Regulation. Chapman Hall.
• Sharma, P. (2002). Prudential Supervision of
  Insurance Undertakings. Paper presented at the
  Conference of Insurance Supervisory Services of
  the Member States of the European Union (now
  CEIOPS) http//europa.eu.int/comm/internal_market/
  insurance/docs/solvency/solvency2-conference-repor
  t_en.pdf