

This excerpt taken from the NMR 6K filed Aug 28, 2009. 3. Market Risk Market risk refers to the potential loss in the value of an asset resulting from changes in market prices, rates, indices, volatilities, correlations or other market factors. We are exposed to this type of risk primarily in connection with our trading activities. Effective monitoring and management of this risk requires an ability to analyze a complex and constantly changing capital market environment worldwide and to highlight any problematic trends quickly. We measure and manage market risk using VaR on a daily basis. The Risk Management unit reports ongoing market risk status and the results of its analysis to senior management. VaR The statistical technique known as ValueatRisk, or VaR, is one of the tools we use to assess market risk exposure of our trading portfolio. VaR is the potential loss in the value of our trading positions due to adverse movements in markets over a defined time horizon with a specified confidence level. We estimate VaR using a 99% confidence level and a oneday time horizon for our trading portfolio. Market risks that are incorporated in the VaR model include equity prices, interest rates, foreign exchange rates, and associated volatilities and correlations. The historical data to calculate volatilities and correlations are weighted to give greater importance to more recent observations. VaR Methodology, Assumptions and Limitations. We make a number of assumptions and approximations in connection with the modeling of the risk characteristics of our trading positions. Different assumptions, approximations or a combination of them could result in a materially different VaR. We believe that the assumptions and approximations we use are reasonable. Trading Portfolio Risk BackTesting We compare VaR values with the actual profit and losses in trading portfolio and verify model’s accuracy used in risk measurement. We count the number of actual times that VaR is exceeded and verify whether the number of times is within a predetermined range. If the number of exceptions is greater than the number predicted by the confidence level used for VaR then we examine any necessary adjustments to the VaR parameters and VaR methodology. Other Measures In some business lines or portfolios we may use additional measures to control or limit risk taking activity. Measures include sensitivities to small moves in key market parameters, credit portfolio risk measures or the financial impact of large market movements on certain portfolios. Metrics and limits of this type are typically specific to asset types, businesses or strategies and are used to compliment VaR and economic capital measures.
37
Table of ContentsStress Testing We also carry out stress testing and scenario analysis. We assess the impact of the occurrence of shock events such as the U.S. subprime loan problem on our profit and loss. We verify losses which are over VaR and which are equivalent to the “distribution base”, and verify if the amount of economic capital by division is appropriate. Scenario analysis or limits may also be used at more granular business levels to assess the impact of certain scenarios or limit risk taking within the businesses. Nomura continues to invest in the development of tools to analyze the impact of market stresses on the value of our portfolios. Model Review We use pricing models when some of the financial instruments cannot be valued based upon quoted market prices. Such models are used for management of risk positions, such as reporting against limits, as well as for valuation. The global risk management unit reviews the models and assesses model appropriateness and consistency independently of the front office. The model reviews consider a number of factors about the model’s suitability for valuation and risk management of a particular product. Nontrading Risk A major market risk in our nontrading portfolio relates to equity investments held for operating purposes which we hold on a longterm basis. Our nontrading portfolio is exposed mainly to volatility in the Japanese stock market. One method that can estimate the market risk in the portfolio is to analyze market sensitivity based on changes in the Tokyo Stock Price Index, or TOPIX, which is a leading index of prices of stocks on the First Section of the Tokyo Stock Exchange. We use regression analysis covering the previous 90 days which tracks and compares fluctuations in the TOPIX and the market value of our operating equity investments held for operating purposes. Our simulation indicates that, for each 10% change in the TOPIX, the market value of our operating equity investments held for operating purposes can be expected to change by ¥18,528 million as of March 31, 2008 and ¥11,983 million as of March 31, 2009. On March 31, 2008, the TOPIX closed at 1,212.96 points and on March 31, 2009, the TOPIX closed at 773.66 points. This simulation analyzes data for our entire portfolio of equity investments held for operating purposes. Therefore, it is very important to note that the actual results differ from our expectations because of price fluctuations of individual equities. This excerpt taken from the NMR 6K filed Jul 25, 2008. 3. Market Risk Market risk refers to the potential loss in the value of an asset resulting from changes in market prices, rates, indices, volatilities, correlations or other market factors. We are exposed to this type of risk primarily in connection with our trading activities. Effective monitoring and management of this risk requires an ability to analyze a complex and constantly changing capital market environment worldwide and to highlight any problematic trends quickly. We measure and manage market risk using VaR on a daily basis. The Risk Management unit reports ongoing market risk status and the results of its analysis to senior management. VaR The statistical technique known as ValueatRisk, or VaR, is one of the tools we use to assess market risk exposure of our trading portfolio. VaR is the potential loss in the value of our trading positions due to adverse movements in markets over a defined time horizon with a specified confidence level. We estimate VaR using a 99% confidence level and a oneday time horizon for our trading portfolio. Market risks that are incorporated in the VaR model include equity prices, interest rates, foreign exchange rates, and associated volatilities and correlations. The historical data to calculate volatilities and correlations are weighted to give greater importance to more recent observations. VaR Methodology, Assumptions and Limitations. We make a number of assumptions and approximations in connection with the modeling of the risk characteristics of our trading positions. Different assumptions, approximations or a combination of them could result in a materially different VaR. We believe that the assumptions and approximations we use are reasonable. Trading PortFolio Risk BackTesting We compare VaR values with the actual profit and losses in trading portfolio and verify model’s accuracy used in risk measurement. We count the number of actual times that VaR is exceeded and verify whether the number of times is within a predetermined range. If the number of exceptions is greater than the number predicted by the confidence level used for VaR then we examine any necessary adjustments to the VaR parameters and VaR methodology. Other Measures In some business lines or portfolios we may use additional measures to control or limit risk taking activity. Measures include sensitivities to small moves in key market parameters, credit portfolio risk measures or the financial impact of large market movements on certain portfolios. Metrics and limits of this type are typically specific to asset types, businesses or strategies and are used to compliment VaR and economic capital measures.
31
Table of ContentsStress Testing We also carry out stress testing and scenario analysis. We assess the impact of the occurrence of shock events such as the U.S. subprime loan problem on our profit and loss. We verify losses which are over VaR and which are equivalent to the “distribution base”, and verify if the amount of economic capital by division is appropriate. Scenario analysis or limits may also be used at more granular business levels to assess the impact of certain scenarios or limit risk taking within the businesses. Nomura continues to invest in the development of tools to analyze the impact of market stresses on the value of our portfolios. Model Review We use pricing models when some of the financial instruments cannot be valued based upon quoted market prices. Such models are used for management of risk positions, such as reporting against limits, as well as for valuation. The global risk management unit reviews the models and assesses model appropriateness and consistency independently of the front office. The model reviews consider a number of factors about the model’s suitability for valuation and risk management of a particular product. Nontrading Risk A major market risk in our nontrading portfolio relates to equity investments held for operating purposes which we hold on a longterm basis. Our nontrading portfolio is exposed mainly to volatility in the Japanese stock market. One method that can estimate the market risk in the portfolio is to analyze market sensitivity based on changes in the Tokyo Stock Price Index, or TOPIX, which is a leading index of prices of stocks on the First Section of the Tokyo Stock Exchange. We use regression analysis covering the previous 90 days which tracks and compares fluctuations in the TOPIX and the market value of our operating equity investments held for operating purposes. Our simulation indicates that, for each 10% change in the TOPIX, the market value of our operating equity investments held for operating purposes can be expected to change by ¥16,942 million as of March 30, 2007 and ¥18,528 million as of March 31, 2008. On March 30, 2007, the TOPIX closed at 1,713.61 points and on March 31, 2008, the TOPIX closed at 1,212.96 points. This simulation analyzes data for our entire portfolio of equity investments held for operating purposes. Therefore, it is very important to note that the actual results differ from our expectations because of price fluctuations of individual equities. This excerpt taken from the NMR 6K filed Jul 27, 2007. Market Risk
Market risk refers to the potential loss in the value of an asset resulting from changes in market prices, rates, indices, volatilities, correlations or other market factors. We are exposed to this type of risk primarily in connection with our trading activities. Effective monitoring and management of this risk requires an ability to analyze a complex and constantly changing capital market environment worldwide and to highlight any problematic trends quickly.
VaR. The statistical technique known as ValueatRisk, or VaR, is one of the tools we use to assess market risk exposure of our trading portfolio. VaR is the potential loss in the value of our trading positions due to adverse movements in markets over a defined time horizon with a specified confidence level.
For our VaR, which we report below, we use a oneday time horizon and a 99% confidence level. This means that, statistically, there is one day out of every 100 days on which the actual trading loss exceeds the VaR.
VaR Methodology, Assumptions and Limitations. We make a number of assumptions and approximations in connection with the modeling of the risk characteristics of our trading positions. Different assumptions, approximations or a combination of them could result in a materially different VaR. We believe that the assumptions and approximations we use are reasonable.
We estimate VaR using a 99% confidence level and a oneday time horizon for our trading portfolio. Market risks that are incorporated in the VaR model include equity prices, interest rates, foreign exchange rates, and associated volatilities and correlations. The historical data to calculate volatilities and correlations are weighted to give greater importance to more recent observations. Given our reliance on historical data, VaR is most effective in estimating risk exposures in markets in which there are no sudden changes in market environments. An inherent limitation of VaR is that past changes in market risk factors, even when weighted toward more recent observations, may not produce accurate predictions of future market risk. Also, VaR using a oneday time horizon may not capture the market risk of positions that cannot be liquidated or hedged within one day.
There are other limitations of VaR. For example, our VaR computation assumes normal distribution for the returns on trading portfolios, while nonlinear risk exposures on options would likely produce a nonnormal distribution for the returns on those portfolios. Different distributional assumptions could produce a materially different VaR.
32
Table of ContentsNontrading Risk. A major market risk in our nontrading portfolio relates to operating equity investments held for relationship purposes which we hold on a longterm basis. Our nontrading portfolio is exposed mainly to volatility in the Japanese stock market. One method that can estimate the market risk in the portfolio is to analyze market sensitivity based on changes in the Tokyo Stock Price Index, or TOPIX, which is a leading index of prices of stocks on the First Section of the Tokyo Stock Exchange.
We use regression analysis for a period of the past 90 days between fluctuation in TOPIX and the market value of our operating equity investments held for relationship purposes. Our simulation indicates that, for each 10% change in TOPIX, the market value of our operating equity investments held for relationship purposes can be expected to change by ¥21,774 million as of March 31, 2006 and ¥16,942 million as of March 31, 2007. The difference in simulated numbers between March 31, 2006 and March 31, 2007 is due mainly from market value decrease in our operating equity investments held for relationship purposes. On March 31, 2006, TOPIX closed at 1,728.16 points and on March 31, 2007, TOPIX closed at 1,713.61 points. This simulation analyzes data for our entire operating equity investments held for relationship purposes. Therefore, it is very important to note that the actual results differ from our expectations because of price fluctuations of individual equities.
This excerpt taken from the NMR 6K filed Jul 27, 2006. Market Risk
Market risk refers to the potential loss in the value of an asset resulting from changes in market prices, rates, indices, volatilities, correlations or other market factors. We are exposed to this type of risk primarily in connection with our trading activities. Effective monitoring and management of this risk requires an ability to analyze a complex and constantly changing capital market environment worldwide and to highlight any problematic trends quickly.
VaR. The statistical technique known as ValueatRisk, or VaR, is one of the tools we use to assess market risk exposure of our trading portfolio. VaR is the potential loss in the value of our trading positions due to adverse movements in markets over a defined time horizon with a specified confidence level.
For our VaR, which we report below, we use a oneday time horizon and a 99% confidence level. This means that, statistically, there is one day out of every 100 days on which the actual trading loss exceeds the VaR.
VaR Methodology, Assumptions and Limitations. We make a number of assumptions and approximations in connection with the modeling of the risk characteristics of our trading positions. Different assumptions, approximations or a combination of them could result in a materially different VaR. We believe that the assumptions and approximations we use are reasonable.
We estimate VaR using a 99% confidence level and a oneday time horizon for our trading portfolio. Market risks that are incorporated in the VaR model include equity prices, interest rates, foreign exchange rates, and associated volatilities and correlations. The historical data to calculate volatilities and correlations are weighted to give greater importance to more recent observations. Given our reliance on historical data, VaR is most effective in estimating risk exposures in markets in which there are no sudden changes in market environments. An inherent limitation of VaR is that past changes in market risk factors, even when weighted toward more recent observations, may not produce accurate predictions of future market risk. Also, VaR using a oneday time horizon may not capture the market risk of positions that cannot be liquidated or hedged within one day.
There are other limitations of VaR. For example, our VaR computation assumes normal distribution for the returns on trading portfolios, while nonlinear risk exposures on options would likely produce a nonnormal distribution for the returns on those portfolios. Different distributional assumptions could produce a materially different VaR.
Nontrading Risk. A major market risk in our nontrading portfolio relates to operating equity investments held for relationship purposes which we hold on a longterm basis. Our nontrading portfolio is exposed mainly to volatility in the Japanese stock market. One method that can estimate the market risk in the portfolio is to analyze market sensitivity based on changes in the Tokyo Stock Price Index (“TOPIX”) which is a leading index of prices of stocks on the First Section of the Tokyo Stock Exchange.
We used regression analysis for the period of the past 90 days between fluctuation in TOPIX and the market value of our operating equity investments held for relationship purposes. Our simulation indicates that, for each 10% change in TOPIX, the market value of our operating equity investments held for relationship purposes can be expected to change by ¥15.4 billion as of March 31, 2005 and ¥21.8 billion as of March 31, 2006. The difference in simulated numbers between March 31, 2005 and March 31, 2006 is due mainly from market value increase in our operating equity investments held for relationship purposes. On March 31, 2005, TOPIX closed at 1,182.18 points and on March 31, 2006, TOPIX closed at 1,728.16 points. This simulation analyzes data for our entire operating equity investments held for relationship purposes. Therefore, it is very important to note that the actual results differ from our expectations because of price fluctuations of individual equities.
This excerpt taken from the NMR 6K filed Jul 26, 2005. Market Risk
Market risk refers to the potential loss in the value of an asset resulting from changes in market prices, rates, indices, volatilities, correlations or other market factors. We are exposed to this type of risk primarily in connection with our trading activities. Effective monitoring and management of this risk requires an ability to analyze a complex and constantly changing capital market environment worldwide and to highlight any problematic trends quickly.
VaR. The statistical technique known as ValueatRisk, or VaR, is one of the tools we use to assess the market risk exposure of our trading portfolio. VaR is the potential loss in the value of our trading positions due to adverse movements in markets over a defined time horizon within a specified confidence level.
For our VaR analysis, which we report below, we use a oneday time horizon and a 99% confidence level. This means that, statistically, there is one day out of every 100 days on which the actual trading loss exceeds the VaR.
VaR Methodology, Assumptions and Limitations. We make a number of assumptions and approximations in connection with the modeling of the risk characteristics of our trading positions. Different assumptions, approximations or a combination of them could result in a materially different VaR. We believe that the assumptions and approximations we use are reasonable.
We estimate VaR using a 99% confidence level and a oneday time horizon for our trading portfolio. Market risks that are incorporated in the VaR model include equity prices, interest rates, foreign exchange rates, and associated volatilities and correlations. The historical data to calculate volatilities and correlations are weighted to give greater importance to more recent observations. Given the reliance on historical data, VaR is most effective in estimating risk exposures in markets in which there are no sudden changes in market environment. An inherent limitation of VaR is that past changes in market risk factors, even when weighted toward more recent observations, may not produce accurate predictions of future market risk. Also, VaR using a oneday time horizon may not capture the market risk of positions that cannot be liquidated or hedged within one day.
There are other limitations of VaR. For example, our VaR computation assumes a normal distribution for the returns on trading portfolios, while nonlinear risk exposures on options would likely produce a nonnormal distribution for the returns on those portfolios. Different distributional assumptions could produce a materially different VaR.
Nontrading Risk. A major market risk in our nontrading portfolio relates to operating equity investments held for relationship purposes which we hold on a longterm basis. Our nontrading portfolio is exposed mainly to volatility in the Japanese stock market. One method that can estimate the market risk in the portfolio is to analyze market sensitivity based on changes in the Tokyo Stock Price Index, or TOPIX, which is a leading index of prices of stocks on the First Section of the Tokyo Stock Exchange, Inc.
We used regression analysis for a 90day period between fluctuation in TOPIX and the market value of our operating equity investments held for relationship purposes. Our simulation indicates that, for each 10% change in TOPIX, the market value of our operating equity investments held for relationship purposes can be expected to change by ¥12.5 billion as of March 31, 2004 and ¥15.4 billion as of March 31, 2005. The difference in the simulated amounts between March 31, 2004 and March 31, 2005 is due mainly from market value increase in our operating equity investments held for relationship purposes, despite the fact that the number of shares for those investments has been decreased. On March 31, 2004, TOPIX closed at 1,179.23 points and on March 31, 2005, TOPIX closed at 1,182.18 points. This simulation analyzes data for our portfolio of operating equity investments held for relationship purposes as a whole. Therefore, it is very important to note that the actual results will differ from our expectations because of price fluctuations of individual equities.
 EXCERPTS ON THIS PAGE:
