ESGs & Solvency II
Elliot Varnell – KPMG LLP
01 February 2010
Society of Actuaries, Dublin
Agenda





Need for this paper
Market consistent valuation
Solvency capital requirement
ESG governance
Discussion
Errata
 5.5.2
 “FSA Pillar 1 Peak 1” should read “FSA Pillar 1 Peak
2”
Motivation for ESGs & Solvency II
Past Challenges
 ESG Models have been widely used in UK & European
Insurance
 Commercial solutions are widely used
 Models are used in various areas of the insurance business
Today’s Challenges
 making market consistent valuation models reflect micro-market
features such as illiquidity and transaction costs
 making the real world ESG fit your company’s in-house view
 understanding the model risk
 getting the ESG model embedding and understood in the
company including within the governance structure
Uses of ESG Models
 Prudential Supervision
 FSA Pillar 1 Peak 2 / Solvency II / PGN-110 / SST
 Financial Reporting
 EEV / MCEV




Asset Liability Management
Dynamic Hedging (inc VA Business)
Product Design
Product Communication
Types of ESG Model
 Risk Neutral
 Designed for market consistent valuations
 Objective = infer a (quasi) market price for insurance
liabilities.
 Real World
 Designed for future economic projections (what-if
scenarios)
 Objective = capture true dynamics of market prices
in order to understand the risks to the insurer.
Market Consistent Valuation
Frequent Misconceptions
1. An arbitrage free ESG model will by itself give a market
consistent valuation
2. An ESG model calibrated to deep and liquid market
data will give a market consistent valuation
3. Market consistent valuation gives the right valuation
4. Market consistent valuation gives the amount a 3rd
party will pay for a business
5. Market consistent valuation is no more objective than a
traditional Discount Cash Flow (DCF) technique using
long term subjective rates of return.
Market Consistent Valuation
Criticisms
 Pro-cyclicality
 In a crisis asset values fall – liabilities value rise as own funds
are squeezed from both sides
 Marginal traded prices on stressed assets are imported to
insurers balance sheets
 Asset fire sales to reduce risk capital further depress markets
 Liquidity Premiums
 Liquidity Premiums and other micro-market features are not
reflected in market consistent ESG models.
 Other models are needed to calculate Liquidity Premiums and
their results are exogenous inputs to the ESG.
Market Consistent Valuation
Perils of Marginal Valuation
Volkswagen Share Price
Share Price
1000
750
500
250
01
/2
00
3
01
/2
00
4
01
/2
00
5
01
/2
00
6
01
/2
00
7
01
/2
00
8
01
/2
00
9
0
Date
Market Consistent Valuation
Coping with volatility
VIX (Volatility Index)
100
50
25
Date
09
20
08
20
07
20
06
20
05
20
04
20
03
20
02
20
01
20
00
20
99
19
98
19
97
19
96
19
95
19
94
19
93
19
92
19
91
19
90
0
19
Index
75
Market Consistent Valuation
Why Use An ESG
 Closed Form Solutions
 Formulae are often too simple
 Underlying models can be too simple
 Replicating Portfolios
 Good for fast recalibrations / optimal hedging
 Stochastic ESGs Best Solution When ..
 It matters how the markets moved during the life of the contract
not just where they ended up. (Path-Dependency)
 The policy payout depends on many economic variables (HighDimensionality)
 There are feedback loops through policyholder behaviour or
management actions.
Market Value Balance Sheet
 ESGs lend themselves to valuation of ..
 With-Profits
 Continental Participating Products
 Variable Annuities
 ESGs lend themselves less to valuation of ..
 Pension Products
 Unit Linked Products
 General Insurance Products
 Asset Liability Coherence
 ESG Valuation of Derivatives vs. Actual Valuation
 Approximations
 Swap Assets vs. Gilt Liabilities (former CP40)
 Historic vs. Market Implied Volatility (former CP39)
CP39 Final Advice
ESG Calibration for Technical Provisions
 3.257.Further guidance on the following areas of the
calibration (of an ESG) may be provided at Level 3:
 The types of assets which reflect the nature and term of
different liabilities and to which the asset model may be
calibrated.
 The appropriate derivation of correlation assumptions.
 The appropriate volatility measure including how volatility
may be estimated in cases where there is limited market data.
 Interpolation or extrapolation of market data, provided that
according this advice there are sufficient reliable points, to base
this calculation (i.e. intermediate volatilities, credit derivatives
spreads...).
 Calibration in cases where market volatilities and market
prices are not consistent.
Solvency Capital Requirement
Where Real-World ESG Models Get Used
 Real World ESGs Used
 Risk Premiums, Realistic Volatilities
 perhaps also Tail Correlations and Fat-Tails
 Standard Formula
 Recalibration of MC ESG under single stresses
 Calibration challenges for ESG models
 (Partial) Internal Model - Balance Sheet Projection
 Surface Fitting (Sensitivities)
 Replicating Portfolio Fitting
 Full (or Partial) Nested Stochastic
ESG Oriented Partial Internal Model
COM PANY
Strategic
Business Unit A
Strategic
Business Unit B
SCR
SCR nl
Adj
BSCR
SCR op
SCR mkt
SCR health
SCR def
NL pr
Mkt fx
Health LT
NL cat
Mkt prop
Accident &
Health ST
Mkt int
Health WC
Mkt conc
SCR
Mkt eq
Mkt sp
SCR life
SCR nl
Adj
BSCR
SCR op
SCR mkt
SCR health
SCR def
SCR life
Life mort
NL pr
Mkt fx
Health LT
Life lapse
Life long
NL cat
Mkt prop
Accident &
Health ST
Life lapse
Life long
Life exp
Life cat
Mkt int
Health WC
Life exp
Life cat
Life dis
Life rev
Life dis
Life rev
Mkt conc
Mkt eq
Mkt sp
Life mort
Solvency Capital Requirement
Passing the Tests







Use Test
Documentation Test
Statistical Quality Test
Calibration Test
Validation Test
External Models and Data
Profit & Loss Attribution
Solvency Capital Requirement
Internal Model Tests
 Use Test
 Use Test vs. Validation / Statistical Quality Tests
 Is it understood? vs. Is it accurate?
 Foundation Principle : Pressure to Improve Model
 Statistical Quality
 Consideration of all ESG risks, including validation and
documentation of the choice of data, distribution and use of
expert judgement.
 Outsourcing doesn’t waive the responsibility.
 Validation
 Back-testing a key requirement.
 How do you back-test an ESG model?
 Reverse Stress Test
 Understanding the Path to Ruin as well as the Stress to Ruin.
Solvency Capital Requirement
Documentation Standards
 Methodology





Mathematical basis
Empirical basis
Assumptions
Application of expert judgement
Where it doesn’t work
 Formulas & Parameters
 Method for estimating parameters
 Data policy
 Source code
 Future Developments
 IT Integration
Calibration Test Challenges
CP69 Equity Stress Test Dampener
Retained as is in
Final CEIOPS
Advice issued
29/01/2010!
SCR – Standard Formula – Equity Risk
Effect of the Equity Implied Volatility Stress Test (CP70)
Percentage Increase in an 10 Year ATM Put
Option Value at 31/12/2008
% Increase in Option Value
250%
200%
150%
100%
50%
0%
Vol Up
Index
Index
Down-Max Down-Mid
Not adjusted for Final CEIOPS
Advice issued 29/01/2010.
Index
Index
Index
Down-Min Down-Max Down-Mid
& Vol Up
& Vol Up
Index
Down-Min
& Vol Up
Governance
Comparing ESG Governance with CP33
ESGs & Solvency II
CP33
Actuarial & Risk
Function
Actuarial Function
Risk Function
Asset Management
Function
Economist Function
Internal Audit
IT Function
Internal Control
Sales & Marketing
Function
Outsourcing
Finance Function
Senior Management /
Board
Governance
ESG Governance Roles
Producer Functions
Consumer Functions
Economist Function
Senior Management /
Board
Risk & Actuarial
Function
Finance Function
IT Function
Sales & Marketing
Function
Governance
Key Relationships – Financial & Risk Reporting
Key Relationship (1)
Senior Management /
Board
Economist Function
Risk & Actuarial
Function
Finance Function
Governance
Key Relationships – Manufacturing Process
Contact
 Elliot Varnell FIA
 [email protected]