ING Direct 2008 Annual Report Download - page 226

2.1 Consolidated annual accounts
Loss Distribution approach
The main objective of the LDA approach is to derive an objective capital amount based on the size and the risk appetite of an institution
and its business units. This approach estimates the likely (fat-tailed) distribution of operational risk losses over some future horizon for
each combination of business line and loss event type. The main characteristic of the LDA is the explicit derivation of a loss distribution,
which is based on separate distributions for event frequency (Poisson) and severity (Inverse Gaussian). The model uses both external and
internal loss data above one million EUR.
The calculation of operational risk capitals for the units follows five basic principles:
Principle 1: If the world gets riskier, the business units need more Economic Capital;•
Principle 2: If a business unit’s size increases, so does its capital;•
Principle 3: If the business of a business unit is more complex, it needs more capital;•
Principle 4: If the level of control of a business unit is higher, it needs less capital;•
Principle 5: If the business units’ losses from internal incidents exceed the level of expected loss accounted for in the first four •
framework principles, it needs more capital.
The capital calculated according to the first three is ‘generic’: if two business units operate in the same markets and have the same size,
the resulting capital will be the same. The specific capital adjustments mentioned below adjust the generic capital of a specific institution
to its specific operational risk capital.
Scorecard approach (principle 4)
The scorecard adjustment reflects the level of quality of control in a specific institution. Scorecards aim to measure the quality of key
operational risk management processes. The scorecard procedure concerns questions that require quantitative data, qualitative
judgements or simple yes/no questions (e.g. indicating compliance with certain group policies). The scorecards are completed by all
business units using self-assessment and reviewed by an expert panel who determines the final score. The set of scorecards then leads
to an increase or decrease of the capital of the specific institution.
‘Bonus/Malus’ approach (principle 5)
Units are assigned additional capital in case losses from internal incidents exceed the level of expected losses that have been accounted
for in the LDA. When actual losses are lower than expected, the capital will be decreased. Only internal incidents above one million EUR
from the last five years are used. The Bonus/Malus adjustments are capped at + and – 20% to prevent large capital fluctuations in total
ING capital.
BUSINESS RISK BANK
The current calculation method applied within ING Bank defines business risk as the ‘residual risk category‘ that includes all risks that are
not covered by the explicitly defined (and managed/measured) credit/transfer, market and operational risk categories. In accordance with
the residual risk definition, the measurement of business risk capital is based on a single risk factor; i.e. the volatility of the ‘residual’ profit
and loss figures (for each BU) that are cleansed for the effects of other risk types. As a consequence there is no further gain in insight
regarding sub business risks.
The level of the business risk capital is linked to the volatility of (cleansed historical) profit and loss data taking into account observed
trends. In practice, this means that more stable earnings over time generally lead to less capital.
Using a T-distribution and the level of confidence, the volatility is then ‘capitalised’ to obtain a business risk capital. The T-distribution is a
theoretical probability distribution, is symmetrical, bell-shaped and similar to the standard normal curve. However, the T-distribution has
relatively more scores in its tails than the normal distribution.
As relatively short data series are available, a capital floor and cap are included in order to prevent the business risk capital from being
under- or overestimated. The minimum (floor: 20%) and maximum (cap: 80%) are specified as a percentage of the operating costs and
as such link business risk capital for units that operate at the floor to cost efficiency.
ECONOMIC CAPITAL ING INSURANCE
Economic Capital, ‘EC’, is defined by ING as the amount of assets that needs to be held in addition to the market value of liabilities to
assure a non-negative surplus at a 99.95% level of confidence on a 1 year time horizon. ING measures Economic Capital by quantifying
the impact on the market value surplus (MVS) as a result of adverse events that occur with a specified probability related to the AA
rating. Therefore ING’s Economic Capital model is based on a ‘Surplus-at-Risk’ concept. The confidence level consistent with an AA rating
has been defined as the 99.95% one-sided confidence level over a one-year horizon. The change in market value surplus (MVS) is the
combined effect of changes in Market Value of Assets (MVA) minus market value of liabilities (MVL) and an adjustment for illiquidity
spreads due to current dislocated asset markets. The MVS is adjusted to correct this asymmetry by applying an illiquidity spread to the
insurance liability cash flows.
Risk management (continued)
ING Group Annual Report 2008
224