01.02.2015 Views

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

SHOW MORE
SHOW LESS

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!

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!