Market risks arise in both trading and non-trading businesses. We take risks by market-making and taking positions in debt securities, equities, foreign currencies, other securities and commodities, and in related derivatives.

We use a combination of risk sensitivities, value-at-risk, stress tests and economic capital measures to manage market risk and set limits. Economic capital is a measure of how we describe and aggregate all market risks in both the trading and non-trading portfolios. In addition, value-at-risk is a common way to manage our trading market risk.

The Board of Directors and the Risk Executive Committee, supported by Market Risk Management as part of our independent risk and capital management function, set a Group-wide value-at-risk limit for market risk in the trading book. Market Risk Management divides this total limit into the divisions. Below this, the limits are further subdivided into the subordinate business lines and trading portfolios as well as geographical regions.

Our value-at-risk reporting for trading transactions is based on our own internal value-at-risk model. In October 1998, the Federal Banking Supervisory Office, one of the predecessor institutions of the Federal Financial Supervisory Authority, approved our internal value-at-risk model to calculate capital requirements for general and specific market risk. The model has since been periodically reviewed and approval has been maintained.

Our value-at-risk disclosures are designed to ensure a consistent presentation of market risk for both internal risk management and external disclosure and regulatory purposes. The value-at-risk limit for the Corporate and Investment Bank Corporate Division was € 90 million in 2006 and the value-at-risk limit for the overall Group’s trading positions was € 92 million (based on a confidence level of 99%) %, as described below, and a holding period of one day). Both limits were unchanged compared to the previous year. SPECIAL FEATURES OF GERMAN SUPERVISORY MARKET RISK MESSAGE

According to German supervisory law, market risk reporting is governed by specific rules, which in particular concern the consolidation of companies, the calculation of the overall market risk position and the definition of trading and non-trading assets.

– CONSOLIDATION. For German regulatory reporting purposes, companies that are not credit institutions, financial service providers, financial companies, ancillary banking services or certain investment companies are not consolidated. However, some of these companies, which we do not consolidate for German regulatory reporting purposes, are consolidated under US GAAP. These mainly include variable interest entities.

– TOTAL MARKET RISK POSITION. We exclude positions in foreign currencies in our market risk report which, under German supervisory law, may be excluded from the calculation of the overall currency position. These are currency items that are deducted from equity or fully backed by equity, as well as investments, including shares in affiliates in foreign currencies, that are measured at historical cost (structural currency items). Our largest structural currency positions stem from our holdings in companies in the United States.

– DEFINITION OF COMMERCIAL AND NON-TRADE ACTIVITIES. The regulatory definition of the trading book and investment book essentially corresponds to the definition of trading and non-trading assets according to US GAAP. However, due to specific differences between the regulatory and accounting rules, certain assets are assigned to the trading book for market risk reporting purposes, although they are classified as non-trading assets under US GAAP. Conversely, certain assets are assigned to the banking book even though they are trading assets under US GAAP.


The value-at-risk approach is used to derive a quantitative measure of our market risk in the trading book under normal market conditions. An estimate of the potential future loss (in terms of market value) will be made that will not be exceeded over a given period of time and with a given confidence level. The value-at-risk measure allows us to apply a consistent and consistent measure of risk to all trades and products. It also allows a comparison of the identified market risks both over a given period of time and with the actual daily trading results.

We calculate the value-at-risk for both internal and external reporting purposes with a confidence level of 99%, in accordance with the provisions of the Bank for International Settlements (BIS). For internal reporting purposes, we use a holding period of one day. For regulatory reporting purposes, the holding period is ten days.

We believe that our value-at-risk model takes all material risk factors into account assuming normal market conditions. Examples of such risk factors include interest rates, stock prices, exchange rates and commodity prices, as well as their implied volatilities. The model considers both linear and, in particular for derivatives, non-linear influences of the risk factors on the value of a portfolio. The statistical parameters required for the value-at-risk calculation are determined on the basis of an observation time series over the last 261 trading days (which corresponds to at least one calendar year), with each observation being weighted equally. In general, we use the Monte Carlo simulation method for value-at-risk calculations, assuming that changes in the risk factors follow a normal distribution or logarithmic normal distribution. However, in some portfolios, such as credit trading, we continue to use a variance-covariance approach to calculate specific interest rate risk. To determine our aggregate value-at-risk we use historically observed correlations between the various general market risk factors. By contrast, when aggregating general and specific market risks, we assume that both risk components are uncorrelated.


We use a back-testing process in our trading units to test the predictive power of our value-at-risk calculations. In this approach, we focus on comparing the hypothetical on a daily basis gains and losses after the buy-and-hold assumption (in accordance with German regulatory requirements) with the values projected by the value-at-risk model.

A committee headed by Market Risk Management, which includes market risk operations and finance, discusses quarterly the back-testing results for the Group and for individual businesses. The committee analyzes the earnings fluctuations that have occurred and examines the forecasting quality of our value-at-risk model. This in turn allows us to further develop the risk measurement process.


While the value-at-risk calculated on a daily basis provides a forecast for potential large losses under normal market conditions, we also perform stress tests in which our trading portfolio is subject to extreme market scenarios that are not covered by the confidence interval of our value-at-risk model will be evaluated.

The results of these extreme stress scenarios form the basis for the calculation of economic capital, which in our estimation is needed to cover the market risks of all positions held by us. In doing so, the underlying risk factors that affect the various products are extremely distracted, that is, subjected to a sudden change according to predefined scenarios. We derive stress scenarios from historical worst-case scenarios, but also take into account structural changes in markets and liquidity.

For example, we calculate country-specific event risk scenarios for all emerging markets and check the results of these event risk analyzes on a daily basis. In addition, a specialist committee reviews country ratings and scenario loss limits on a monthly basis. Ad hoc reviews are conducted as needed.

In addition to the country-specific event risk scenarios for emerging markets, the positions of all major portfolios are subject to market stress scenarios at regular intervals. This is done on a weekly basis for the trading portfolios and on a monthly basis for the non-trading portfolios.

Our stress test scenarios include:

– Price and volatility risks for interest rates, stock prices, foreign exchange rates and commodity prices for developed countries. In doing so, we take into account both trading and non-trading portfolios of securities and investments as well as trading portfolios in derivatives. In addition, numerous basis risks are recorded;

– Risks in emerging markets, including falling equity prices, rising interest rates and currency devaluations;

– Risks arising from changes in issuer-related yield differentials for bonds, credit derivatives and tradable loans from developed and emerging market countries;

– Underwriting risks in the bond and equity issue business in industrialized countries.

We calculate the economic capital by aggregating the losses from these stress scenarios. We use correlations that reflect extreme market conditions (rather than correlations that apply in normal market conditions and are used in our value-at-risk model). Our economic capital requirements for market risk in our trading operations were € 1.6 billion Year-end 2006 and was thus almost unchanged at the end of 2005.