September 6, 2011
description
Transcript of September 6, 2011
Enterprise Risk Management
“All of life is the management of risk, not its elimination” Walter Wriston, former chairman of Citicorp
September 6, 2011
Stuart Wason FSA, FCIA, CERA
2
Useful Background Material for this presentation includes:
IAIS ICP 16 on Enterprise Risk Management (e.g. ERM including ORSA)
IAIS ICP 17 on Capital Adequacy (e.g. internal models)
IAA Practice Note on Enterprise Risk Management for Capital and Solvency Purposes
IAA Practice Note on the Use of Internal Models for Risk and Capital Management Purposes by Insurers
IAA Practice Note on Stress and Scenario Testing
3
Key topics for this session
Introduction
Setting the Scene
Governance & ERM Framework
Risk Management Policy
Risk Tolerance
Risk Responsiveness & Feedback Loop
Own Risk and Solvency Assessment
Economic and Regulatory Capital
Continuity Analysis
Role of Supervision in Risk Management
Stress Testing
Internal models
4
Introduction
Today, ERM is increasingly regarded as an appropriate response or indeed a solution to managing risk in today’s more complex and interdependent markets and operating environments.
Insurance regulators have also played a leading role in setting standards and providing guidance to insurers on implementing appropriate frameworks for the management of risks faced by insurance companies.
The IAIS ICP 16 describes eight Key Features. The IAA Practice Note ‘unpacks’ each of the ‘Key Features’ by explaining them in more detail, thereby assisting insurance executives address strategic and operational issues associated with implementing an ERM framework in their insurance business.
5
Feedback LoopFeature 4
Risk Management PolicyFeature 2
Own Risk and Solvency Assessment (ORSA)Feature 5
Economic & RegulatoryCapital
Feature 6
Continuity AnalysisFeature 7
Role of supervisionFeature 8
Feedback LoopFeature 4
Governance and Enterprise Risk Management FrameworkFeature 1
Risk Tolerance StatementFeature 3
6
Setting the Scene
ERM is a logical and evolutionary response to growing complexity, uncertainty and ambiguity associated with 21st century corporate life. Now all management is risk management.
ERM involves identifying, assessing, mitigating and, if necessary, transferring risk.
In reality, risk involves a complex interplay of dynamic external influences and (unpredictable) human behaviour - ‘traditional’ or silo risk management is not enough to sustain a 21st century insurance business.
7
Setting the Scene
Risk management is commonly viewed through a lens of avoiding ‘bad’ things happening and limiting the downside. The more enlightened view is one of connecting risk to value maintenance and creation.
Effective ERM is inextricably linked with strategic planning for a business.
Effective ERM requires new investments in modelling and analytical capabilities, a different way of looking at risk and capital, and cultural changes to embed risk management in all activities of a corporation.
Regulators and rating agencies increasingly expect insurers to apply its techniques for managing their business on a day-to-day basis.
8
Evolution of Enterprise Risk Management
Risk models:Economic capital models Other models
Today
Link with strategy
High
Medium
Low
Balance sheet protection
Industry standard in the last 5-10 years
Industry standard in the next 5-10 years
Risk control Risk/return optimisation
Value creation
Compliance
Loss minimisation
Risk management
Risk measurement
Strategic integration
Return optimisation
‘The Role of ERM in Ratings’, Mark Puccia, Managing Director, Standard & Poor’s, March 30, 2007
9
Setting the SceneWhat is ERM?
ERM is concerned with,– All risks faced by insurers– Creating value for the owners of an insurance enterprise whilst ensuring
that promises made to policyholders are met.
Specifically, ERM– Considers the totality of systems, structures and processes within an
insurer that identify, assess, treat, monitor, report and/or communicate all internal and external sources of risk that could impact on the insurer’s operations
– Implies a common risk management ‘language’ across the insurer– Involves systematic organisation of and coordination between risk
functions– Includes both the management of ‘downside’ as well as ‘upside’ risks– Seeks to quantify all risks but not all risks can be quantified– Is concerned with both behaviours and risk control processes– Involves consideration of risk information relating to past events (e.g.
losses), current performance (e.g. risk indicators) and future outcomes (e.g. the risk profile or risk assessment).
10
Setting the SceneWhat is ERM?
Strong enabling conditions must exist for ERM to take hold, namely:– Demonstrable executive management support is critical– Strong and direct linkages must be made between ERM and the
insurer’s business strategy and its day-to-day operations– The insurer must establish clear accountabilities for the various
aspects of risk management, distinguishing between those in line management roles and those in risk management
Many of the insurers who have developed advanced practices describe ERM as a ‘journey’ implemented in waves
11
Setting the SceneWhere does one begin?
Key to implementation is buy in and support from the Board. For this to occur, ERM needs to inform the board about issues they want and need to know about.
Key Lessons1. ERM is one of the few truly enterprise wide business capabilities
that both provides an opportunity to change the way an organisation does business, but also can be ‘used’ to drive certain agendas that may not be aligned to the business imperatives, and stakeholder needs.
2. The output of ERM may not suit all stakeholders, so Board buy-in with management is critical to ensure needs and expectations are met and the ERM investment delivers maximum return and minimises any agency/stakeholder bias.
3. The Board is well placed to take a strategic and holistic perspective to ensure long term sustainability of the ERM investment
12
Governance & Enterprise Risk Management Framework
ERM framework must be appropriate to the nature, scale and complexity of insurer’s business and risks.
ERM framework should be fully integrated with (and embedded in) the insurer’s business operations.
ERM framework should be led and overseen by the insurer’s board and senior management.
For capital management and solvency purposes, the framework should include provision for the quantification of risk for a sufficiently wide range of outcomes using appropriate techniques.
Measurement of risk should be supported by accurate documentation providing appropriately detailed descriptions and explanations of risks.
Key Feature 1
13
Governance & Enterprise Risk Management Framework
The role of an insurer board with respect to risk management is broadly well understood and reflects an ‘ultimate responsibility’ for the insurer’s risk management framework. Stakeholders, including regulators, interpret this ultimate responsibility to mean, amongst other things:– Approving the insurer’s overall risk management strategy and/or
policy– Overseeing the process of ensuring the insurer’s ‘responsible
persons’ are fit and proper– Setting the risk appetite of the insurer– Monitoring key risks by ensuring the implementation of a
suitable risk management and internal controls framework.
14
Governance & Enterprise Risk Management Framework
TIPS FOR AN EFFECTIVE RISK COMMITTEE– Diverse member background with appropriate qualities such as
inquisitive/questioning minds, objectivity & relevant experience.– Ensure RC “ask questions” of the reports submitted and of
management rather than apply the “tick the box” approach. – Ensure RC directives have support of Board and the appropriate
level of management “buy in”.– Appropriateness of level & volume of reporting to RC - ensure
the right information is being communicated. – Responsible for keeping track of leading practices & trends.– Have an appropriate SMART self-assessment program.
15
Governance & Enterprise Risk Management Framework
Board versus management accountabilities
Management commitment and leadership
Establishing and developing an enterprise risk function
Importance of a common risk language
Risk management culture
Developing a risk behavior model
Developing an implementation plan
Upside risk management
Performance measurement and reward systems
16
Governance & Enterprise Risk Management Framework
Reporting and monitoring: At the highest level risk reporting should seek to identify the following (for example):– Current and emerging key risks in the business and within
the wider environment, and changes over time (the risk profile of the insurer)
– Changes in risk indicators (measures influencing risk likelihood and/or impact)
– Capability for identifying and managing risks
17
Governance & Enterprise Risk Management FrameworkExample: Anything to report?
Many stakeholders rely on quality risk information: – Audit Committees – Monitoring material financial risks and their
mitigation– Executives - Reviewing risk information for business strategy– Managers - Reviewing risk information for completeness and
changes in risk profile or control effectiveness– Risk Owners - Updating risk information and escalating changes
in likelihood, impact or control effectiveness as required– Control Owners - Updating status of treatments for controls that
they are responsible for– Internal Audit - Reviewing the effectiveness of internal control
measures – External Stakeholders – Reviews by regulatory bodies.
18
Governance & Enterprise Risk Management FrameworkExample: Anything to report?
A succinct dashboard is the most effective way to report so the information can be assessed at a glance. Supporting information can be attached for those who require more detail. Some of the key categories of a dashboard may include: – Top 10 residual risks– Key risk indicators– Scoring chart for risk severity and control effectiveness– Heat map of all substantial inherent and residual risks– An additional commentary section– Significant project progress.
19
20
Risk Management Policy
An insurer should have a risk management policy which outlines the way in which the insurer manages each relevant and material category of risk, both strategically and operationally. The policy should describe the linkage with the insurer’s tolerance limits, regulatory capital requirements, economic capital and the processes and methods for monitoring risk.
Key Feature 2
21
Risk Management PolicyAspects to consider
A clear risk management philosophy – for example outlining why risk management is important and the linkages with value creation
The relationship between risk management and the insurer’s purpose or mission, values and strategic objectives
How risk management is embedded in the related processes of capital management, pricing, reserving and performance management
Scope of activities to which the policy applies. For example, the policy should be sufficiently flexible to cater for multiple ownership structures (e.g. wholly-owned, majority-owned, joint venture etc.)
Appropriate regulatory requirements and considerations
Requirements with respect to acquisition of new businesses e.g. time frame for integration with the insurer’s ERM framework
Categories of risk and risk definitions and how these map to internationally accepted categories/definitions
In addition to risk categories, the policy should define risk ‘terminology’ used e.g. ‘risk’, ‘risk management’, risk management framework’
22
Risk Management PolicyAspects to consider
Most importantly, the insurer’s risk appetite should be set forth in the policy
Governance and oversight aspects
Board, board committee structures, responsibilities
Management structures, roles, responsibilities
Roles and responsibilities of the various corporate and business unit risk functions
Role of internal and external audit
Compliance aspects, including consequences associated with policy breach
Behavioural expectations of all staff
Minimum process-level requirements that apply across the insurer
Requirements risk assessment processes (e.g. stress and scenario testing, future financial condition and ‘Own Risk and Solvency Assessment’ testing
The process for reviewing and updating the policy.
23
Risk Tolerance
An insurer should establish and maintain a risk tolerance statement which sets out its overall quantitative and qualitative tolerance levels and defines tolerance limits for each relevant and material category of risk, taking into account the relationships between these risk categories.
The risk tolerance levels should be based on the insurer's strategy and be actively applied within its ERM framework and risk management policy. The defined risk tolerance limits should be embedded in the insurer’s ongoing operations via its risk management policies and procedures.
Key Feature 3
24
Risk Tolerance
Establishing an insurer’s risk tolerance involves making strategic choices.
The process must be connected with setting strategy and longer term direction.
While top-level management may be heavily involved in debating the appropriate risk tolerance to match a given strategic direction, it is the Board who must decide on risk tolerance and the insurer’s strategy.
The CRO should be involved in but not responsible for defining the insurer’s risk tolerance.
25
Risk Tolerance
Organisation PurposeOrganisation Purpose
Strategy
Business Unit Plans
Risk Tolerance
Limits
26
Risk Tolerance
For an insurer, the following parameters are typically used to articulate risk tolerance across financial and non–financial risks:– Lines of business that the insurer will/will not accept– Earnings volatility– Requirements to meet regulatory criteria (including allowance
for unexpected events)– Desired capital ‘strength’, usually by reference to a defined
rating level of a recognised credit rating agency – Maintaining levels of economic capital by reference to a
specified chance of meeting policyholder obligations or target return periods for ‘risk of ruin’
– Maximum exposure to aggregation of risk– Dividend paying capacity (for listed company insurers)– The maximum net loss the insurer is prepared to accept in any
given year in the event of a catastrophic loss (general insurers)
27
Risk Tolerance
Limits, being narrower in scope, tend to operate at the risk category level. Examples of risk limits include:– Establishing counterparty credit limits for investments and
reinsurers– Setting an overall target for credit quality for a reinsurance
buying program, usually by reference to credit rating– Establishing concentration limits for lines of business/products,
geographies and counterparties– Maintenance of underwriting and pricing principles and limits– Setting liquidity benchmarks by reference to the amount of
investment assets to be held in ‘highly liquid’ assets– Investment mandates setting limits for the investment of
policyholder and shareholder funds in traded instruments– Limits on the use of financial derivatives– Establishing operational risk policies that include limits for
outsourcing, business interruption, fraud etc.
28
Risk Responsiveness and Feedback Loop
The insurer's ERM framework should be responsive to change.
The ERM framework should incorporate a feedback loop, based on appropriate and good quality information, management processes and objective assessment, which enables the insurer to take the necessary action in a timely manner in response to changes in its risk profile.
Key Feature 4
29
Risk Responsiveness and Feedback Loop
An effective feedback loop is underpinned by:
Establishment of thresholds for reporting significant issues
Protocols for escalation of issues to various levels and management and, if necessary, regulators
Reporting of risk aggregations to identify where limits (and potentially risk tolerance) may have been exceeded.
30
Risk Responsiveness and Feedback Loop
Emerging risks are developing or already known risks which are subject to uncertainty and ambiguity and are therefore difficult to quantify using traditional risk assessment techniques.
Insurers are interested in emerging risks for a number of reasons including, whether emerging risks will:– Influence the organisation’s strategy– Impact the performance of the underwriting portfolios –
unexpected (latent) claims / claims frequency / claims costs– Impact on the operational risks facing the organisation– Present opportunities for new types of insurance products?
One way to evaluate high impact/low probability events is through scenario planning, which can augment statistical models and help companies prepare for specific events. Scenario planning is a powerful tool that helps executives assess the resilience of the organisation to internal and external shocks
31
Own Risk and Solvency Assessment (ORSA)
An insurer should regularly perform its own risk and solvency assessment (ORSA) to provide the board and senior management with an assessment of the adequacy of its risk management and current, and likely future, solvency position. The ORSA should encompass all reasonably foreseeable and relevant material risks including, as a minimum, underwriting, credit, market, operational and liquidity risks. The assessment should identify the relationship between risk management and the level and quality of financial resources needed and available.
Key Feature 5
32
ORSA
ORSA involves carrying out a combination of quantitative and qualitative techniques to identify, assess and manage risk.
The core process of risk management involves the systematic identification, analysis, evaluation and treatment of risks
Typically, the ‘context’ is framed around objectives of a business process or project or indeed the broader insurance enterprise.
The output of the risk management process is usually described as a ‘risk profile’, ‘risk register’, ‘heat map’ and/or ‘risk control self assessment’ (hereafter described as a risk profile).
The process of risk profiling can be applied at the insurance enterprise level, business unit, key business process level (e.g. underwriting, claims) or be applied in the management of projects. Risk profiling involves an assessment of risk at both the levels of ‘inherent risk’ and ‘residual risk‘.
33
ORSA
Inherent and residual risk highlight important management information not otherwise readily apparent:– Those risks whose management rely heavily on the continued
and effective operation of key controls (high inherent risk/low residual risk)
– Those risks whose nature does not significantly alter following the application of controls. This highlights that certain controls may be ineffective and that resources might be utilised better elsewhere, or that different controls are needed (high inherent risk/high residual risk)
– Those risks that may be over-controlled (low inherent risk/low residual risk).
34
ORSARisk profile elements
– Description of risks in enough detail for each risk to be understood in isolation
– Cause(s) or underlying conditions giving rise to a given risk– Consequence(s) of the risk - in both financial and non-financial
terms (e.g. loss of customers, regulatory sanction, cost over-runs etc)
– Categorisation of each risk - especially important where an insurer comprises multiple business units and risk aggregation is required at the enterprise level
– Inherent risk assessment that considers likelihood/frequency of risk occurrence and impact of the risk.
– Assessment of controls and/or risk mitigation strategies.– Residual risk assessment after taking into account the
effectiveness of controls– Action(s) to be taken to bring unacceptable residual risk within
appropriate limits.
35
Rare Unlikely Likely Probable Almost Certain
5 % 30 % 70 % 95 % 100 %
Descriptor
Probability
Lik
eli
ho
od
Control Effectiveness
High
Medium
Low
Opportunity
Risk Trend
Increasing risk
Decreasingrisk
Stable
Inherent to Residual Risk
= >$ 500 m
= >$ 250 m to < $ 500 m
Impact( Enterprise Value
Financial / Non- financial )
= >$ 100 m to < $ 250 m
= >$ 50 m to < $ 100 m
= >$ 20 m to < $ 50 m
= >$ 5 m to < $ 20 m
= >$ 500 , 000 to < $ 5 m
$ 0 to < $ 500 , 000
Risk Profile
Risk 2
Risk 1
Risk 3
Risk 4
Risk 5
Risk 6
Risk 7Risk 10
Risk 9
Risk 11
Risk 12
Risk 8
36
Risk Category Modelling Technique(s)
Enterprise /all risk categories Dynamic Financial Analysis
Underwriting (including reinsurance)
Financial Condition Report (FCR) and/or underwriting modelling or reviews
Market
Value at risk (VAR) or Tail VAR Interest rate models Scenario tests
Credit Credit risk models
Liquidity Asset/Liability modelling
Operational
Internal loss data External loss data Scenario analysis, simulations
37
ORSAThe ‘black swan’ dilemma – is ERM enough?
Nassim Taleb1 coined the phrase “black swan” to describe something that is a large-impact, hard-to-predict, and rare event beyond the realm of normal expectations. The metaphor here is that most people would expect a swan to be white (at least until black swans were discovered in the 17th Century in Australia) and therefore a black swan is a surprise.
Black swan events have occurred throughout history. More recently the events of 9/11 and the sub prime meltdown in the USA are examples.
But here is the dilemma. Since black swan event are surprises they cannot happen twice because once they have occurred they are within know experience. Planning to avoid repeated events of this nature is a good idea but cannot prevent further surprises. Even a forensic understanding of such events will do little to prevent the next black swan.
Good risk practices are our only real preventative measure – and honesty that surprises will happen. Through an appropriate ERM framework we can be well placed to manage surprising situations appropriately and decrease the impact.
So ERM is probably not enough to prevent all manner of risks, especially surprises, however it is a lot better than not having any preventative framework.1 Learning to Expect the Unexpected by Nassim Taleb, The New York Times, April 8, 2004
38
Economic and Regulatory Capital
As part of its ORSA an insurer should determine the overall financial resources it needs to manage its business given its own risk tolerance and business plans, and to demonstrate that supervisory requirements are met. The insurer's risk management actions should be based on consideration of its economic capital, regulatory capital requirements and financial resources.
Key Feature 6
39
Economic and Regulatory Capital
One of the basic principles behind capitalism is that the market will allocate capital to the most productive activities and organisations as measured by their ability to provide a return on that capital.
Owners of capital will assess proposals for the use of their capital based on their risk vs reward and provide their limited capital to the best available proposals.
A key component to managing these risks is to have a model that attempts to simulate the environment in which the insurer is operating.
Such models provide a guide to management of how specific decisions may impact the expected level and volatility of future profit. They can also provide indications of the risk of failure of the insurer. Referred to as Economic Capital Models, they are used by capital providers, regulators & companies.
40
Economic and Regulatory Capital
It is the ability of the ECM to allocate the capital down to the level of detail where ‘localised’ decisions can be made that is crucial to the success of the pricing function
Risk Class
Pricing Measure to Achieve X% RoC
Actual Pricing Measure
Rating Strength
Actual Business Volumes
(A) (B) (B / A)
X 10% 11% 1.10 100
Y 5% 4% 0.80 200
Z 7% 7% 1.00 70
Total 0.92 370
41
Economic and Regulatory Capital
Taking the example to a lower level of detail, if the ECM can provide capital requirements for Risk Class Y at a lower level of detail, i.e. Y1 and Y2, then more effective management decisions can be made by understanding the source of the underperformance of risk class Y.
Risk Class
Pricing Measure to Achieve X% RoC
Actual Pricing Measure
Rating Strength
Actual Business Volumes
(A) (B) (B / A)
X 10% 11% 1.10 100
Y1 5% 6% 1.20 67
Y2 5% 3% 0.60 133
Z 7% 7% 1.00 70
Total 0.92 370
42
Economic and Regulatory Capital
Regulatory capital requirements are just one input into capital requirements. There can be a multitude of others including:– Desired rating agency ratings– Desired earnings volatility– Desired shareholder return, dividend and capital growth– Market expectations
Key potential differences between a regulatory prescribed method and an ECM would often include:– The volatility of various classes of business– Different allowances for diversification (often performed by
correlation matrices, or sometimes via copulas) between risk types and within risk types
– Different focuses driving capital (i.e. different aims)
Capital management focuses on turning risk into shareholder value
43
Capital Management
Capital
Reserving
Performance Management
Pricing
Risk Management
44
Continuity Analysis
As part of its ORSA, an insurer should analyse its ability to continue in business, and the risk management and financial resources required to do so over a longer time horizon than typically used to determine regulatory capital requirements.
Such continuity analysis should address a combination of quantitative and qualitative elements in the medium and longer term business strategy of the insurer and include projections of the insurer's future financial position and modelling of the insurer’s ability to meet future regulatory capital requirements.
Key Feature 7
45
Continuity Analysis
An ECM allows an insurer to look further into the future than most regulatory prescribed methods are based on. This will require explicit decisions to be made regarding (amongst other things):– What time period of modelling should be used– Should the financial position of the insurer be assessed at a
future point in time, or once all relevant liabilities are modelled to have run-off
– What management actions are likely should results turn to the worst
– What capital reduction (e.g. dividend) / capital injection policy can be assumed
– How reliable are an insurer’s longer term forecasts and are they sufficient to form the basis of an ECM.
46
Continuity Analysis
A truly integrated ECM will be used for a wide range of purposes within an insurer. For example, it can used to provide analysis relating to:– Economic capital requirements– Investment strategy– Mergers, acquisitions and divestments– Capital allocation– Reinsurance programmes– Optimal business mix– Reserving volatility– Capital outflow / inflow– Financial Condition Report– Business Continuity Planning
47
Role of Supervision in Risk Management
The supervisor should undertake reviews of an insurer's risk management processes and its financial condition. The supervisor should use its powers to require strengthening of the insurer’s risk management, including solvency assessment and capital management processes where necessary.
Key Feature 8
48
Role of Supervision in Risk Management
Supervisors increasingly expect insurers to apply ERM as part of the on-going management of their business on a day-to-day basis. – ERM is consistent with the aims of risk-based supervision and
the protection of policyholders– Supervisors will wish to be kept informed in an appropriate and
regular manner of all the ERM Features noted in this Practice Note
– Supervisors will seek confirmation that ERM satisfies the “Use” test within the insurer.
– Supervisors have a range of interests in ERM from Board level governance to the technical specifications involved with internal model approvals for regulatory capital (for example)
– Insurers should aim to adopt ERM practices which are sound and forward-looking and be proactive in communications with their Supervisor.
49
Stress Testing
Lessons learned from financial crises
What is stress testing?
Supervisory expectations
50
Lessons learned from financial crises
Stress testing is a valuable risk management tool
Improved risk management is required
More up-to-date stress testing results are needed
Increasingly, crises are due to systemic inter-connected events
Greater consideration needs to be given to extreme events
Individual insurers can have difficulty identifying industry-wide contagion impacts.
51
Lessons learned from financial crises
Insufficient consideration of extreme events– “It is hard for us without being flippant, to even see a scenario
within any kind of realm of reason that would see us losing one dollar in any of those transactions.” August 2007, Joseph Cassano, a former AIG executive
– “Almost no one expected what was coming. It’s not fair to blame us for not predicting the unthinkable.” Daniel Mudd, former chief executive of Fannie Mae
52
What is stress testing?
Stress Testing: Generalized process of determining the impact of a change in one or more risk factors on the operations of an insurer.
Scenario Testing: Form of stress testing which uses a hypothetical future state of the world to define the changes in the risk factors affecting the insurer.
Sensitivity Test: Form of stress testing which typically involves an incremental change in a risk factor to determine the impact on an insurer. Improved risk management is required
53
Complexity
Severity
Stress Testing
Single risk/single time period
Many risks/interactions & time periods
Low stress
Severe stress
Scenario testing
Financial Condition testing
Sensitivity testing
54
Supervisory expectations
Strengthened risk management
Robust capital management plan
Scenario types considered
Stress testing used frequently
Timely future financial condition stress testing
Financial condition disclosure improved
Supervisory stress tests considered
55
Strengthened risk management
Stress testing is an important risk management tool
Supervisory expects stronger ERM:
– Demonstrate embedding of stress testing in ERM via “use test”
– Stress testing clearly documented & available for review
– Develop & implement stress testing controls
– Notify supervisor quickly of material adverse effects from stress testing
– Ability to run stress tests frequently through the year
56
Strengthened risk management
Supervisory future financial condition (FFC) expectations:
– Senior management engagement with FFC
– Risk management is more than risk identification
– FFC report experimentation
– Importance of actuarial function with FFC
57
Robust capital management plan
Stress testing (including FFC report) is an important part of internal capital management planning
Board sets risk appetite & senior management implements through internal target capital levels
Management should not rely exclusively on complex models
Stress testing should be used to test the vulnerability of assumptions
Management plans should anticipate both severe and more plausible stresses and scenarios
58
Scenario types considered
Scenarios enable a complete picture of a crisis to be constructed
Some types of scenarios:– Reverse Scenarios: start with a given loss and make assumptions about a
possible event that could give rise to such a loss– Macro Scenarios: Scenarios that are defined on a global and market wide
level Examples: Global downturn of financial markets, Pandemic– Historical Scenarios: Scenarios that are defined with reference to a past
observed event. Example: Influenza pandemic, Tokyo earthquake– Company Specific Scenarios: Scenarios that are tailor-made for the risk
exposure of a specific company– Narrowly Specified Scenarios: Scenarios that describe a very specific
event. It can be easily implemented consistently but might have a very small probability. Examples: A very specific earthquake, defined by location, magnitude etc.
59
Stress testing used frequently
Financial conditions can change rapidly - stress tests more than a few months old may not reflect the current economic reality
Prudent ERM practice dictates that insurers review their future financial condition more frequently than once a year
Supervisor expects insurers to prepare on a regular basis (as often as needed by prudent ERM) comprehensive “what if” stress tests
Supervisor should be informed immediately of a significant change in the condition of a company and an updated FFC report be prepared and filed with the Board and supervisor within short (45 days?) timeframe
60
Timely FFC testing
Annual FFC report expected to be integral part of risk management & strategic planning process
Supervisor should expect regular reports (including the FFC report), be presented to Board on a timely basis
Supervisor should expect FFC report to be used to identify risks that must be addressed in planning process - not the other way around
61
FFC disclosure
FFC reports should utilize better practices currently used by insurers, including:
– Displaying financial results for each year in the projection period, not just the beginning and end of period results
– Including revaluation of liabilities at end of projection period according to revised assumptions
– Displaying the results of the adverse scenarios both before and after relevant management reaction
– Reporting on regulatory capital breaches– Reporting on ERM implications (e.g. breach of risk limits,
necessary risk mitigation etc.)
62
Supervisory stress testing
Stress testing should be part of an on-going supervisory regime
Stress testing enables the supervisor to,– analyze insurer specific risk exposures– assess the effectiveness of risk management (especially under
simulated crisis)
Stresses or scenarios which impact all companies, can have industry-wide contagion impacts
63
Supervisory stress testing
Consistent application of these macro scenarios across all companies is the best way to identify these systemic impacts
Standardized stresses or scenarios enable supervisor to– analyze market-wide risk exposures in a jurisdiction or globally– assess the resilience of each insurer to similar stresses or
scenarios.
Supervisor will need to prepare its standardized stress tests, including underlying key assumptions, and revise them from time to time
In defining these tests, the supervisor may want to provide the industry relatively short notice of their design to simulate a crisis environment
64
Internal models
For technically sophisticated firms
Requires supervisory approval
Requires extensive risk management program
Subject a “use test ”
65
Internal models
“use test” implies a broader application of internal models
– Economic capital– ALM– Reserving– Pricing– Product design– Risk management– . . . . .
66
IAIS Solvency Sub-committee
Guidance Paper and Standard on the use of internal models for regulatory capital purposes
Internal models are seen as having an application to both Pillar I and Pillar II requirements
IAIS asked the IAA Solvency Sub-committee for technical help
67
IAA Internal Models Paper
1. Introduction
2. Model Fundamentals
3. Design Considerations
4. Construction of Model
5. Controls
6. Governance
7. Communication
8. Supervisory Approvals
68
2. Model Fundamentals
Financial Model
Modelling Process
Proportionality
Risk assessment framework
Real World versus Risk Neutral Probabilities
Managing Models
Types of Model
69
3. Design Considerations
Results
Order of Calculation
Control over Assumptions
Reproducibility
Flexibility
70
4. Construction of Model
Time Granularity
Population of the Model
Product Descriptions
In-force Data
Assets
Insurance Experience Assumptions
Insurance Assumptions for Projections
Assumptions Concerning the Insurer
Algorithms
On the Use of Random Numbers
The Number of Scenarios in a Stochastic Model
Extreme Values
Documentation
71
5. Controls
Sufficiency Test
Calibration Test
Use Test
Change Test
72
6. Governance
Approvals, Policies, Expertise, Tools, and Resources
Risk Management Policy
Audit
Review
Documentation
Compliance
73
7. Communication
Fundamentals
Identify Stakeholders of the use of the Model
Identify the Communication Requirements of each Stakeholder
Internal Management Communication Needs
Examiner Communication Needs
Public Communication Needs
74
8. Supervisory Approvals
Role of the Board of Directors and Senior Management
Risk Management Infrastructure
Corporate and Operational Limits
Model Integration
Stress Testing
Documented Policies
Internal Audit
Quantitative Model Standards
Modifications to Capital Models
75
76
Issues
Diversification
Limits to the use of and reliance on models
Scenarios and stress testing