Economic capital for market risk measures the amount of capital needed to absorb very severe, unexpected losses arising from our exposures over the period of one year. “Very severe” in this context means that the underlying economic capital is set at a level which covers, with a probability of 99.98 %, all unexpected losses over a one year time horizon.
Our economic capital model comprises two core components, the “Common Risk” component covering risk drivers across all businesses and the suite of Business Specific Stress Tests (BSSTs) which enriches the Common Risk component. Both components are calibrated to historically observed severe market shocks.
The Common Risk component of the traded market risk economic capital model is based on an enhanced version of our regulatory stressed value-at-risk approach. Hence the economic capital model benefits from the value-at-risk model taking into account the longer liquidity horizons.
The calculation of economic capital for market risk from the trading units is performed weekly. The model incorporates the following risk factors: interest rates, credit spreads, equity prices, foreign exchange rates, commodity prices and correlations.
We also continuously assess and refine our BSSTs in an effort to promote the capture of material risks as well as reflect possible extreme market moves. Additionally, risk managers use their expert judgment to define worst case scenarios based upon the knowledge of past extreme market moves. It is possible however, for our market risk positions to lose more value than our economic capital estimates since all downside scenarios cannot be predicted and simulated.
Economic capital for traded default risk represents an estimate of the default and migration risks of credit products at a 99.98 % confidence level, taking into account the liquidity horizons of the respective sub-portfolios. It covers the following positions:
- Fair value assets in the banking book;
- Unsecuritized credit products in the trading book excluding correlation trading portfolio;
- Securitized products in the trading book excluding correlation trading portfolio and
- Correlation trading portfolio.
The traded default risk economic capital for the correlation trading portfolio is calculated using the comprehensive risk measure. For all other positions the calculation of traded default risk economic capital is based on our credit portfolio model. Traded default risk captures the credit exposures across our trading books and it is monitored via single name concentration and portfolio limits which are set based upon rating, size and liquidity. Single name concentration risk limits are set for two key metrics: Default Exposure, i.e., the P&L impact of an instantaneous default at the current recovery rate (RR), and bond equivalent Market Value (MV), i.e. default exposure at 0 % recovery. In addition, a traded default risk economic capital limit is set within the Market Risk economic capital framework while the incremental risk charge monitors the regulatory capital requirements associated with these positions. In order to capture diversification and concentration effects we perform a joint calculation for traded default risk economic capital and credit risk economic capital. Important parameters for the calculation of traded default risk are exposures, recovery rates and default probabilities as well as maturities. Exposures, recovery rates and default probabilities are derived from market information and external ratings for the trading book and internal assessments for the banking book as for credit risk economic capital. Rating migrations are governed by migration matrices, which are obtained from historical rating time series from rating agencies and internal observations. The probability of joint rating downgrades and defaults is determined by the default and rating correlations of the portfolio model. These correlations are specified through systematic factors that represent countries, geographical regions and industries.
Validation of the market risk economic capital model is performed by an independent team. The regular review covers, but is not limited to, the appropriateness of risk factors, the calibration techniques, the parameter settings, and model assumptions.