ING Direct 2008 Annual Report Download - page 222

2.1 Consolidated annual accounts
The first aggregation step is between ING Bank and ING Insurance for each major risk type. All risk capitals, except for credit risk that
is already aggregated for ING Bank and ING Insurance, are delivered on a standalone basis for ING Bank and ING Insurance. These risk
capitals are aggregated between ING Bank and ING Insurance using a variance-covariance approach. Depending on the accounting
treatment the Bank – Insurance correlation factors used for EaR may differ from CaR correlation factors (e.g. for interest rate risk).
The result of this aggregation step are Group diversified EaR and CaR figures for each major risk type.
Major risk types distinguished:
Risk type Distribution used
Credit and transfer risk (2) KMV distribution
Market risk (8)
– Interest rate risk Europe, Asia and America
– Equity risk Europe, Asia and America
– FX risk
– Real estate risk Normal distribution
Insurance risk (1) Normal distribution
Business risk (1) Normal distribution
Operational risk (1) Empirical distribution
(Note numbers in parentheses indicate the number of risk types distinguished (total of 13)).
A second aggregation step exists between these major risk types at an ING Group level. The Group diversified EaR and CaR figure for
each major risk type are aggregated using a Monte Carlo simulation in combination with an inter-risk correlation matrix to obtain the
overall EaR and CaR figures for ING Group. The outcomes of the simulation represent the potential losses arising from the major risk
types, which are summed together to derive the aggregate potential losses. The diversified Group EaR or CaR is then calculated as
the 90th percentile of the simulated aggregate potential losses.
Principal assumptions of EaR and CaR measurement
CaR and EaR figures should always be viewed in the context of principal assumptions made to enable both comparability and updated
measurement of ING Group’s risk profile:
Risk dynamics are based on historic observation; historical events are used as a proxy for future risk estimates e.g. price changes, •
defaults, dependencies of markets;
Point-in-time risk profile of in-force business is presented; in general risk measurement does not include future volumes and margins;•
Discretionary management interventions are not explicitly modelled unless their measurement can be based on historical performance •
tracking (e.g. regular or planned actions);
Correlation factors between risk types used for diversification are based on best estimate assumptions supported by statistical analysis •
of historical data, ING risk expert judgement, external benchmark studies and common logic;
Behavioural assumptions for clients are included in risk measurement where applicable e.g. variable savings, embedded mortgage •
options or lapse ratios.
Reporting Framework
All data for each risk type and business line, as well as the empirical Group risk distributions, are uploaded to a web-based risk dashboard
program. The aggregation and simulation steps, as described above, are performed in a secure server based environment.
ECONOMIC CAPITAL ING BANK
Economic Capital is defined as the amount of capital that a transaction or business unit requires in order to support the economic risks
it originates. In general Economic Capital is measured as the unexpected loss above the expected loss at a given confidence level. Specific
measurement by risk type is described in greater detail in the separate risk type sections; i.e. credit and transfer and operational risk as
well as market and business risk bank.
This Economic Capital definition is in line with the net market value (or surplus) definition. The process of Economic Capital modelling
enables ING Bank to allocate Economic Capital to the business units and support risk-adjusted performance measurement (RAROC). By
comparing Economic Capital figures with ING’s available financial resources, adequate capital buffers can be ensured.
The following fundamental principles and definitions have been established for the model:
ING Bank uses a one-sided confidence level of 99.95% – consistent with ING’s target debt rating (AA) – and a one-year time horizon •
to calculate Economic Capital;
It is assumed that all currently known measurable sources of risk are included;•
The best estimate risk assumptions are as objective as possible and based on proper analysis of statistical data. There is one set •
of best-estimate assumptions for each risk type to be used at ING Bank;
The Economic Capital calculation is based on fair value principles. Where complete and efficient markets exist, fair value is equal •
to market value;
Risk management (continued)
ING Group Annual Report 2008
220