Developing an internal model to determine the Solvency ... - CNSF
Developing an internal model to determine the Solvency ... - CNSF
Developing an internal model to determine the Solvency ... - CNSF
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
<strong>Developing</strong> <strong>an</strong> <strong>internal</strong> <strong>model</strong><br />
<strong>to</strong> <strong>determine</strong><br />
<strong>the</strong> <strong>Solvency</strong> Capital Requirement<br />
Paolo Cadoni<br />
CEIOPS Internal Models Expert Group Chair,<br />
FSA<br />
XXI I t ti l S i I d S t<br />
XXI International Seminar on Insur<strong>an</strong>ce <strong>an</strong>d Surety<br />
Mexico City 18-19 November 2010
Outline<br />
• Internal Models under <strong>Solvency</strong> II: setting <strong>the</strong><br />
scene<br />
• What is <strong>an</strong> <strong>internal</strong> <strong>model</strong><br />
• What are <strong>the</strong> expected benefits<br />
• Approval process <strong>an</strong>d requirements<br />
• Questions
Internal Models under <strong>Solvency</strong> II:<br />
setting <strong>the</strong> scene…..<br />
Assets<br />
Liabilities<br />
Capital<br />
Requirements<br />
Excess of<br />
Assets Over<br />
Liabilities<br />
SCR<br />
Total assets<br />
at<br />
market<br />
value<br />
Subordinated<br />
t valuation<br />
le risk<br />
nts<br />
Marke et consisten<br />
for hedgeabl<br />
compone<br />
risk<br />
margin<br />
best<br />
estimate<br />
Technical<br />
Provisions<br />
Total<br />
Liabilities<br />
MCR
<strong>Solvency</strong> Capital Requirement (SCR)<br />
• “The SCR should deliver a level of capital that enables <strong>an</strong><br />
insur<strong>an</strong>ce undertaking <strong>to</strong> absorb signific<strong>an</strong>t unforeseen losses<br />
over a specified time horizon <strong>an</strong>d gives reasonable assur<strong>an</strong>ce <strong>to</strong><br />
policyholders that payments will be made as <strong>the</strong>y fall due”<br />
• Derived using ei<strong>the</strong>r <strong>an</strong> approved <strong>internal</strong> <strong>model</strong> or a ‘st<strong>an</strong>dard<br />
formula’<br />
– Partial <strong>internal</strong> <strong>model</strong><br />
– Full <strong>internal</strong> <strong>model</strong><br />
– St<strong>an</strong>dard formula<br />
• 99.5% confidence level over 1 year<br />
• As a minimum <strong>to</strong> cover insur<strong>an</strong>ce, market, credit <strong>an</strong>d operational<br />
risks<br />
• Part of supervisory review process (SRP)
The different approaches <strong>to</strong> calculate<br />
<strong>the</strong> SCR<br />
Full Internal<br />
Model<br />
Ri isk Sens sitivity<br />
Partial Internal<br />
Models<br />
St<strong>an</strong>dard Formula<br />
with USPs<br />
St<strong>an</strong>dard<br />
Formula<br />
St<strong>an</strong>dard Formula<br />
with simplifications<br />
Complexity
The starting point:<br />
what is <strong>an</strong> <strong>internal</strong> <strong>model</strong><br />
• What is <strong>an</strong> <strong>internal</strong> <strong>model</strong> according <strong>to</strong> <strong>the</strong> CEA/GC<br />
“A risk m<strong>an</strong>agement system developed by <strong>an</strong> insurer <strong>to</strong> <strong>an</strong>alyse <strong>the</strong><br />
overall risk position, <strong>to</strong>qu<strong>an</strong>tify risks <strong>an</strong>d <strong>to</strong> <strong>determine</strong> <strong>the</strong><br />
economic capital required <strong>to</strong> meet those risks”<br />
• What is <strong>an</strong> <strong>internal</strong> <strong>model</strong> according <strong>to</strong> <strong>the</strong> IAA<br />
“Ma<strong>the</strong>matical <strong>model</strong> of <strong>an</strong> insurer’s operations <strong>to</strong> <strong>an</strong>alyse its<br />
overall risk position, <strong>to</strong> qu<strong>an</strong>tify risks <strong>an</strong>d <strong>determine</strong> <strong>the</strong> capital<br />
<strong>to</strong> meet those risks”<br />
• What is <strong>an</strong> <strong>internal</strong> <strong>model</strong> under <strong>Solvency</strong> II<br />
No definition provided by <strong>the</strong> <strong>Solvency</strong> II Framework Directive…..<br />
But…
The Internal Model is more th<strong>an</strong> simply a<br />
calculation l engine…<br />
– Should be <strong>an</strong> integral part of <strong>the</strong> risk m<strong>an</strong>agement system<br />
– Meet <strong>the</strong> tests <strong>an</strong>d st<strong>an</strong>dards for <strong>internal</strong> <strong>model</strong> approval<br />
Source – CEIOPS-DOC-07/05
What are <strong>the</strong> expected benefits<br />
• Improved risk sensitivity of SCR related <strong>to</strong> <strong>the</strong> insurer’s specific<br />
profile leading <strong>to</strong> a more adequate <strong>model</strong>ling of non-st<strong>an</strong>dard,<br />
especially non-linear, contracts,<br />
• Better alignment of regula<strong>to</strong>ry capital requirements with economic<br />
capital<br />
• Encouragement of innovation in risk m<strong>an</strong>agement methodology<br />
leading <strong>to</strong> higher competitiveness through better risk m<strong>an</strong>agement <strong>an</strong>d<br />
hence lower costs of capital,<br />
• More effective pillar 2 discussion <strong>an</strong>d familiarity of <strong>the</strong> supervisor<br />
with more detailed exposure data th<strong>an</strong> is generally available in<br />
accounting records,<br />
• Cost efficiencies through re-use of risk <strong>model</strong>ling infrastructure for<br />
discussion with supervisors, rating agencies, <strong>an</strong>alysts <strong>an</strong>d<br />
shareholders.
The Approval Process<br />
Approval process ≠ Validation<br />
Pre-application Application * Assessment Decision<br />
* Includes <strong>the</strong> policy for <strong>model</strong> ch<strong>an</strong>ges<br />
9
The Approval Process<br />
Pre-application Application * Assessment Decision<br />
* Includes <strong>the</strong> policy for <strong>model</strong> ch<strong>an</strong>ges<br />
10
Pre-application<br />
• Recommended by CEIOPS<br />
NB: Pre-application does not me<strong>an</strong> pre-approval<br />
• Beneficial both for supervisors <strong>an</strong>d undertakings<br />
• Allows supervisors <strong>to</strong> better pl<strong>an</strong> <strong>the</strong>ir resources<br />
• Undertakings engage with supervisors while developing <strong>an</strong>d implementing <strong>the</strong>ir Internal Model<br />
• Early identification <strong>an</strong>d communication of concerns<br />
• The pre-application pp process is going g <strong>to</strong> be iterative:<br />
• Undertakings will not have all elements of <strong>the</strong>ir IM ready at <strong>the</strong> outset As undertakings get<br />
feedback from supervisors <strong>the</strong> IM may ch<strong>an</strong>ge<br />
• The depth of review will vary within <strong>the</strong> IM<br />
• Information <strong>to</strong> be reviewed is linked <strong>to</strong> <strong>the</strong> application (approval process)<br />
• Regular feedback <strong>to</strong> undertakings on <strong>the</strong> IM<br />
• Supervisors give <strong>the</strong>ir view on how prepared <strong>the</strong> undertaking is <strong>to</strong> submit <strong>the</strong> application 11
The Approval Process<br />
Pre-application Application * Assessment Decision<br />
* Includes <strong>the</strong> policy for <strong>model</strong> ch<strong>an</strong>ges<br />
12
Application<br />
• Formal application<br />
• Documentary evidence that requirements are met (art. 112.3)<br />
• The administrative or m<strong>an</strong>agement bodies have overall responsibility for<br />
application – role of <strong>internal</strong> committees (art.116.1)<br />
• Systems must be in place <strong>to</strong> ensure <strong>internal</strong> <strong>model</strong> operates properly on<br />
a continuous basis (art. 116.2)<br />
• Internal <strong>model</strong> govern<strong>an</strong>ce policies (e.g. policy for ch<strong>an</strong>ging g <strong>the</strong> full <strong>an</strong>d<br />
partial <strong>internal</strong> <strong>model</strong> (art. 115))<br />
• Supervisors shall decide on <strong>the</strong> application within six<br />
months (art. 112.4)<br />
• reason for rejection must be provided (art. 110.6)<br />
• Supervisor may require <strong>the</strong> firm <strong>to</strong> provide <strong>an</strong> estimate of<br />
<strong>the</strong> SCR <strong>determine</strong>d in accord<strong>an</strong>ce with <strong>the</strong> st<strong>an</strong>dard<br />
approach (art. 112.7)
Undertakings need <strong>to</strong> evidence…<br />
‣ Use test (art. 120)<br />
‣ Statistical quality st<strong>an</strong>dards (art.121)<br />
‣ Calibration st<strong>an</strong>dards (art. 122)<br />
‣ Profit <strong>an</strong>d Loss Attribution (art.123)<br />
‣ Validation st<strong>an</strong>dards (art. 124)<br />
‣ Documentation st<strong>an</strong>dards (art. 125)<br />
‣ An adequate system for identifying, measuring, moni<strong>to</strong>ring,<br />
m<strong>an</strong>aging <strong>an</strong>d reporting risk<br />
• Requirements shall not be considered in isolation –<br />
interrelations shall not be ignored<br />
• Use of external <strong>model</strong>s <strong>an</strong>d data does not exempt<br />
insurers from <strong>an</strong>y of <strong>the</strong> requirements (art. 126)
What does this me<strong>an</strong> in practice<br />
• Approval is not based on undertaking’s chosen<br />
software platform<br />
– Nor is it is simply based on <strong>the</strong> number generated by <strong>the</strong><br />
undertaking’s <strong>model</strong><br />
• Approval is required at <strong>the</strong> level of <strong>the</strong> <strong>model</strong> itself<br />
– i.e. its scope, coverage, design, build, integrity<br />
<strong>an</strong>d application<br />
– To embed <strong>the</strong> <strong>model</strong> in<strong>to</strong> <strong>the</strong> business, undertakings first<br />
have <strong>to</strong> embed <strong>the</strong> business in<strong>to</strong> <strong>the</strong>ir <strong>model</strong>s
Use test (art. 120)<br />
• Internal <strong>model</strong> is widely used <strong>an</strong>d plays <strong>an</strong> import<strong>an</strong>t role in:<br />
– risk m<strong>an</strong>agement <strong>an</strong>d decision-making<br />
– economic <strong>an</strong>d solvency capital assessment<br />
– system of govern<strong>an</strong>ce<br />
• Must be fully embedded in <strong>the</strong> operational <strong>an</strong>d org<strong>an</strong>isational<br />
structure of <strong>the</strong> insurer<br />
• Demonstrate that <strong>the</strong> <strong>model</strong> remains useful <strong>an</strong>d is applied<br />
consistently over time<br />
• Board is responsible for ensuring <strong>the</strong> ongoing appropriateness<br />
of <strong>the</strong> <strong>model</strong>
Use Test<br />
Principle 9<br />
Principle 1<br />
Principle 8<br />
recalculation<br />
integration in<strong>to</strong> <strong>the</strong> riskm<strong>an</strong>agement<br />
system on a<br />
consistent basis for all uses<br />
Principle 7<br />
improve <strong>the</strong><br />
undertaking’s riskm<strong>an</strong>agement<br />
system<br />
Foundation Principle:<br />
Inherent desire <strong>to</strong><br />
improve quality<br />
underst<strong>an</strong>ding of <strong>the</strong><br />
<strong>internal</strong> <strong>model</strong><br />
Principle 2<br />
fit <strong>the</strong> business<br />
<strong>model</strong><br />
Principle 3<br />
support <strong>an</strong>d verify<br />
decision-making<br />
Principle 6<br />
widely integrated with<br />
<strong>the</strong> risk-m<strong>an</strong>agement<br />
system<br />
Principle 5<br />
design <strong>to</strong><br />
facilitate <strong>an</strong>alysis<br />
of business<br />
decisions<br />
Principle 4<br />
cover sufficient risks:<br />
useful for risk<br />
m<strong>an</strong>agement <strong>an</strong>d<br />
decision-making<br />
i
Statistical quality st<strong>an</strong>dards (art. 121)<br />
• Probability distribution forecast:<br />
– based on sound actuarial/ statistical techniques<br />
– current, credible info/ realistic assumptions<br />
– broadly consistent with approach <strong>to</strong> technical provisions<br />
– assumptions justified by firms<br />
– data appropriate <strong>an</strong>d accurate<br />
– no particular method, provided IM meets criteria of use test<br />
• All material risks <strong>to</strong> be covered including all those<br />
• All material risks <strong>to</strong> be covered, including all those<br />
addressed in St<strong>an</strong>dard Formula
Statistical Quality St<strong>an</strong>dards<br />
What is all about<br />
Risks<br />
Expert<br />
Judgment<br />
Options &<br />
Guar<strong>an</strong>tees<br />
Assumptions &<br />
Methodology<br />
Output:<br />
Probability Distribution<br />
Forecast<br />
Diversification &<br />
Risk Mitigation<br />
Techniques<br />
M<strong>an</strong>agement<br />
Actions<br />
Data<br />
Expected<br />
Payments
Calibration st<strong>an</strong>dards (art. 122)<br />
• Where feasible, use VaR at 99.5% over 1 year confidence level<br />
• Different risk measure/ time horizon permitted provided<br />
policyholders’ ld protection ti equivalent <strong>to</strong> St<strong>an</strong>dard d Formula i.e.<br />
99.5% over 1 year, VaR<br />
• Approximations permitted where firm demonstrates approach<br />
provides equivalent protection<br />
• Supervisors may require undertakings <strong>to</strong> run <strong>the</strong>ir IM on<br />
relev<strong>an</strong>t benchmark portfolios <strong>an</strong>d using assumptions based<br />
on external data ra<strong>the</strong>r th<strong>an</strong> <strong>internal</strong> data <strong>to</strong> verify <strong>the</strong><br />
calibration of <strong>the</strong> IM <strong>an</strong>d <strong>to</strong> check that its specification is in<br />
line with generally accepted market practice
Calibration St<strong>an</strong>dards - Example<br />
• Example of <strong>the</strong> direct derivation of <strong>the</strong> SCR from <strong>the</strong> IM’s PDF<br />
TailVaR<br />
99%<br />
Supervisory <strong>to</strong>ols<br />
VaR<br />
99%<br />
VaR<br />
99,5%<br />
• benchmark portfolios<br />
• test alternative assumptions based on external data<br />
21
Validation st<strong>an</strong>dards (art. 124)<br />
• The undertaking has a primary responsibility<br />
for validation<br />
• Validation is <strong>an</strong> iterative ti process<br />
• There is no single validation method<br />
• Validation should encompass both qu<strong>an</strong>titative<br />
<strong>an</strong>d qualitative elements<br />
• Validation should be subject <strong>to</strong> independent<br />
review<br />
• Validation policy
Validation st<strong>an</strong>dards (art. 124)<br />
• Regular cycle of <strong>model</strong> validation shall include:<br />
‣ Moni<strong>to</strong>ring <strong>the</strong> perform<strong>an</strong>ce of <strong>the</strong> <strong>model</strong><br />
‣ Reviewing <strong>the</strong> ongoing appropriateness of its specification<br />
‣ Testing its results against experience<br />
‣ Analysis of stability of <strong>the</strong> <strong>model</strong><br />
‣ Testing <strong>the</strong> sensitivity of <strong>the</strong> results <strong>to</strong> ch<strong>an</strong>ges in key underlying<br />
assumptions<br />
‣ Assessment of <strong>the</strong> accuracy, completeness <strong>an</strong>d appropriateness of data<br />
used by <strong>the</strong> <strong>internal</strong> <strong>model</strong><br />
‣ A process <strong>to</strong> evidence that <strong>the</strong> resulting capital requirements are<br />
appropriate
Validation policy<br />
Scope<br />
Tools used<br />
Frequency<br />
Govern<strong>an</strong>ce<br />
Validation Policy<br />
Limitations<br />
Documentation<br />
Independence<br />
- Firms should have a documented Validation Policy that<br />
sets out how <strong>the</strong> <strong>internal</strong> <strong>model</strong> is validated <strong>an</strong>d why it is<br />
appropriate.<br />
- Set of <strong>to</strong>ols <strong>to</strong> gain confidence in <strong>the</strong> <strong>model</strong>: testing<br />
against experience; sensitivity testing; reverse stress<br />
testing
Profit & Loss Attribution (art.123)<br />
• Analysis of profit <strong>an</strong>d loss by cause/source for each major<br />
business unit, at least <strong>an</strong>nually<br />
‣ Need <strong>to</strong> be able <strong>to</strong> identify <strong>an</strong>d qu<strong>an</strong>tify sources of profit <strong>an</strong>d<br />
losses<br />
• Demonstrate how categorisation of risk chosen explains<br />
<strong>the</strong> causes/sources of P&L<br />
• Two main purposes: validation <strong>an</strong>d strategic<br />
‣ Need <strong>to</strong> consider how <strong>the</strong>se will be used (link with Use test) <strong>an</strong>d<br />
reflect in gr<strong>an</strong>ularity of assessment<br />
‣ Need <strong>to</strong> underst<strong>an</strong>d/reconcile differences <strong>to</strong> profits <strong>an</strong>d losses on<br />
accounting basis<br />
‣ Should be <strong>an</strong> import<strong>an</strong>t validation <strong>to</strong>ol
Documentation (art. 125)<br />
• It must be thorough, detailed <strong>an</strong>d complete enough <strong>to</strong> allow<br />
third parties <strong>to</strong> underst<strong>an</strong>d <strong>an</strong>d replicate <strong>the</strong> <strong>model</strong><br />
• It shall provide a detailed outline of:<br />
– The <strong>the</strong>ory<br />
– Assumptions<br />
– Ma<strong>the</strong>matical <strong>an</strong>d empirical basis underlying <strong>the</strong> <strong>model</strong><br />
• It is a crucial <strong>to</strong>ol for <strong>the</strong> firm <strong>to</strong> demonstrate <strong>the</strong> supervisor<br />
that <strong>the</strong> undertaking really underst<strong>an</strong>ds <strong>an</strong>d has mastered <strong>the</strong><br />
<strong>internal</strong> <strong>model</strong> it is using<br />
• It shall indicate <strong>an</strong>y circumst<strong>an</strong>ces under which <strong>the</strong> <strong>internal</strong><br />
<strong>model</strong> does not work effectively (weaknesses)<br />
• It must be revisited <strong>an</strong>d, if necessary, updated in line with <strong>the</strong><br />
complexity <strong>an</strong>d stability of <strong>the</strong> risk profile of <strong>the</strong> insurer
External <strong>model</strong>s <strong>an</strong>d data (art. 126)<br />
• The use of a <strong>model</strong> or data obtained from a third party shall not be<br />
considered <strong>to</strong> be a justification for exemption from <strong>an</strong>y of <strong>the</strong><br />
requirements for <strong>the</strong> <strong>internal</strong> <strong>model</strong><br />
• Firms must be able <strong>to</strong> document <strong>an</strong>d explain <strong>the</strong> role of vendor<br />
products <strong>an</strong>d <strong>the</strong> extent <strong>to</strong> which <strong>the</strong>y are used within <strong>the</strong>ir <strong>internal</strong><br />
processes<br />
• Firms must be able <strong>to</strong> demonstrate a thorough underst<strong>an</strong>ding of<br />
vendor products used in <strong>the</strong>ir <strong>internal</strong> processes<br />
• Vendor products should be appropriate <strong>to</strong> <strong>the</strong> nature, scale <strong>an</strong>d<br />
complexity of <strong>the</strong> risks incorporated within <strong>the</strong> firm’s risk strategy <strong>an</strong>d<br />
business objectives<br />
• Firms must have clearly articulated strategies for regularly reviewing <strong>the</strong><br />
perform<strong>an</strong>ce of vendor <strong>model</strong> results <strong>an</strong>d <strong>the</strong> integrity of external data<br />
used in <strong>the</strong>ir risk qu<strong>an</strong>tification process
‘Partial <strong>internal</strong> <strong>model</strong>s’ (PIMs)<br />
• Insur<strong>an</strong>ce <strong>an</strong>d reinsur<strong>an</strong>ce undertakings may use <strong>the</strong>m for<br />
<strong>the</strong> calculation of one or more:<br />
– risk modules or sub-modules of <strong>the</strong> SCR<br />
– operational risk<br />
– adj. for <strong>the</strong> loss-absorbing capacity of TPs <strong>an</strong>d deferred taxes<br />
• In addition <strong>to</strong>:<br />
– <strong>the</strong> whole business of insur<strong>an</strong>ce <strong>an</strong>d reinsur<strong>an</strong>ce undertakings, or<br />
– only one or major business unites<br />
• Approval subject <strong>to</strong> <strong>the</strong> same requirements as full IM, plus<br />
additional conditions:<br />
– Reason for <strong>the</strong> limited scope of application<br />
– IM SCR reflects more appropriately <strong>the</strong> risk profile of <strong>the</strong> undertaking
Example 1 - Modelling one risk sub-module
Example 2 - Modelling one risk module
Example 3 - Modelling two (or more) risk sub-modules<br />
within <strong>the</strong> same risk module jointly
Example 4 - Modelling two (or more) risk sub-modules from<br />
different risk modules jointly
Example 5 - Modelling two (or more) risk sub-modules from<br />
different risk modules jointly
Integrating <strong>the</strong> results of <strong>the</strong> PIM with <strong>the</strong><br />
st<strong>an</strong>dard formula<br />
(1) Use <strong>the</strong> SF correlation matrix<br />
Options Available <strong>to</strong><br />
integrate <strong>the</strong> PIM & SF<br />
(2) Use one method listed<br />
In CEIOPS level 3 guid<strong>an</strong>ce<br />
• Each of <strong>the</strong> options must<br />
meet:<br />
– Appropriateness <strong>an</strong>d<br />
– Feasibility tests<br />
(3) Use firms own method
The Approval Process<br />
Pre-application Application * Assessment Decision<br />
* Includes <strong>the</strong> policy for <strong>model</strong> ch<strong>an</strong>ges<br />
35
Assessment<br />
• Supervisors may require additional information<br />
• The assessment may be iterative with feedback <strong>to</strong> undertakings resulting in<br />
modifications <strong>to</strong> <strong>the</strong> IM<br />
• The supervisors are expected <strong>to</strong> assess:<br />
<br />
<br />
<br />
<br />
Scope <strong>an</strong>d <strong>model</strong> coverage<br />
Methodology <strong>an</strong>d documentation<br />
Data quality<br />
Qu<strong>an</strong>titative <strong>an</strong>d qualitative procedures<br />
<br />
Technological l environment<br />
• The assessment may include desk-based review, on-site inspections, requests for<br />
fur<strong>the</strong>r information <strong>an</strong>d o<strong>the</strong>r contacts<br />
Specificities for partial <strong>internal</strong> <strong>model</strong>s (PIM)<br />
• Justification of <strong>the</strong> limited scope; better reflection ect of <strong>the</strong> risk profile th<strong>an</strong> <strong>the</strong> st<strong>an</strong>dard<br />
d<br />
formula (SF) <strong>an</strong>d integration of <strong>the</strong> PIM’s <strong>an</strong>d SF’s results<br />
36
The Approval Process<br />
Pre-application Application * Assessment Decision<br />
* Includes <strong>the</strong> policy for <strong>model</strong> ch<strong>an</strong>ges<br />
37
Decision<br />
Approval<br />
Decision<br />
Rejection<br />
• SCR calculation using <strong>the</strong><br />
IM<br />
• May include terms <strong>an</strong>d<br />
conditions<br />
Level 3 Guid<strong>an</strong>ce on terms<br />
<strong>an</strong>d conditions<br />
Limited<br />
approval<br />
• Reasons for rejection <strong>to</strong><br />
be communicated <strong>to</strong> <strong>the</strong><br />
undertaking<br />
• Application’s withdrawal is<br />
• Rejection of a broader scope IM<br />
possible<br />
• SCR calculation using <strong>the</strong><br />
• Simult<strong>an</strong>eous approval of a reduced scope<br />
PIM<br />
SF<br />
• Only conceded if <strong>the</strong> approved parts<br />
indeed function as a PIM<br />
• SCR calculation using <strong>the</strong> PIM <strong>an</strong>d <strong>the</strong> SF<br />
(o<strong>the</strong>r parts)<br />
38
What next<br />
Is that all
On – Going Compli<strong>an</strong>ce<br />
• Firm’s responsibility <strong>to</strong> implement a regular cycle of validation<br />
<strong>an</strong>d moni<strong>to</strong>r on-going compli<strong>an</strong>ce.<br />
• Assessment of <strong>the</strong> <strong>internal</strong> <strong>model</strong> should include:<br />
– Appropriateness of <strong>the</strong> design <strong>an</strong>d operations of <strong>the</strong> <strong>internal</strong> <strong>model</strong><br />
– Internal <strong>model</strong> continues <strong>to</strong> appropriately p reflect <strong>the</strong> risk profile<br />
– Up keep of all tests <strong>an</strong>d st<strong>an</strong>dards<br />
• Supervisors will review <strong>the</strong> outcomes of <strong>the</strong> firm’s reviews <strong>to</strong><br />
• Supervisors will review <strong>the</strong> outcomes of <strong>the</strong> firm s reviews <strong>to</strong><br />
assess on-going compli<strong>an</strong>ce with <strong>the</strong> Internal Model<br />
requirements
Reporting <strong>an</strong>d Disclosure<br />
• Method for calculating <strong>the</strong> SCR<br />
• Scope of <strong>the</strong> (Partial) Internal Model<br />
• Internal Model govern<strong>an</strong>ce <strong>an</strong>d risk m<strong>an</strong>agement<br />
arr<strong>an</strong>gements<br />
• Equivalent SCR for <strong>the</strong> firm if <strong>the</strong>y were using <strong>the</strong><br />
St<strong>an</strong>dard Formula<br />
• Validation <strong>an</strong>alysis<br />
• And more….
Summary<br />
• Internal <strong>model</strong>s are not only for large <strong>an</strong>d<br />
sophisticated firms….<br />
• However, <strong>to</strong> develop <strong>an</strong>d maintain <strong>an</strong> <strong>internal</strong><br />
<strong>model</strong> four main conditions need <strong>to</strong> be satisfied:<br />
– Good underst<strong>an</strong>ding of <strong>the</strong> business<br />
– Commitment of senior m<strong>an</strong>agement<br />
– Knowledge, expertise <strong>to</strong> create <strong>an</strong>d maintain <strong>model</strong><br />
– Good, deep <strong>an</strong>d rich data sets
Th<strong>an</strong>k you!