The recent global financial crisis has highlighted the need for trading areas in banks to have access to the latest risk measures in a timely fashion to enable proper supervision of risk-taking and proper pricing of risk in dealing activities. Generally, these risk measures require high performance risk measurement systems that utilise the most modern technology to deliver simulation results in near real-time. BoundaryRider risk engine has been built using High Performance Computing technology including vectorisation and distributed processing to guarantee the fastest possible simulation calculations. The power of this technology opens up the opportunity to have the following risk measures at your finger-tips.

Market Risk Value at Risk (VaR)

At a global, regional or desk level using historic simulation or Monte Carlo simulation with results returned in a few minutes for a global VaR and seconds for desk level values. This enables desk managers and dealers to be confident they are within limits before going home for the day.

Market Risk Expected Shortfall (ES)

Rapidly calculate this statistic at global, regional or desk level. The ES helps identify cases where the tail of the loss distribution is very wide thus suggesting that stressed markets could lead to losses considerably higher than those suggested by the VaR.

Extreme Value Theory (EVT)

EVT covers a wide range of techniques but in its purest form it is useful in reducing the volatility of risk measures that are dependent on high percentiles of distributions. Often, especially with historic simulation, risk measures such as VaR are derived from a handful of points in very sparse distribution tails. The use of EVT to fit a more substantial part of the tail allows the estimation of high percentile values more reliably. It can also be applied to large Monte Carlo distributions to estimate very high percentile results as are often required for capital purposes.

Credit Risk Potential Exposure (PCE)

Widely used as the measure against which limits are set. This is a confidence interval measure of potential exposure during the life of a counterparty portfolio. Using BoundaryRider HPC technology it is possible to calculate the PCE using simulation to ensure that credit mitigation techniques such as close-out netting and collateral are included correctly. In addition, exotic products requiring numerical valuations; and products requiring path dependence to be modelled; can also be incorporated..

Credit Value Adjustment (CVA)

New credit measures such as CVA are used in the trading room to price credit risk properly. Based on expected exposures from simulation calculations combined with credit spreads or hazard rates, this is a measure of expected loss for proposed trades. In a bilateral form it gives a means of pricing the risk in the trade, in a unilateral form it allows for statistical provisioning for losses in near real-time. In the past, such measures have been generated from within finance departments at month end and have been little more than a black box to front office users. With access to CVA and statistical provisioning, business managers can rapidly determine the true value of new trading activity.

Effective Expected Potential Exposure (EPE)

As banks move to advanced internal models for regulatory credit capital calculations, the exposure at default for trading activity to be used is an EPE figure. This is based on an expected exposure profile generated using Monte Carlo simulation. BoundaryRider's HPC technology allows the calculation of EPE routinely as part of the daily risk management process, thus letting you meet the "use" test for these figures.

Summary

BoundaryRider is modern in design and utilises the latest technology innovations such as grid computing to achieve a performance that is superior to any known competitive solution. In addition Vector Risk has used it's in-depth knowledge of risk management and complex mathematics to design an engine that supports highly complex financial instruments. It is designed to allow new instruments to be created easily and quickly. Furthermore, BoundaryRider's componentized architecture allows financial institutions to tackle areas within their risk business infrastructure that might be causing issue on an as needed basis without the need to implement a complete new vendor package via the inherently risky and costly 'big bang' implementation approach required by most other vendor products in the market today.