Implementation of a Value-at-Risk system

In this blogs, we discuss the implementation of a value-at-risk (VaR) system. The
focus will be on the practical nuts and bolts of implementing a VaR system in
software, as opposed to a critical review of the financial methodology. We have
therefore taken as our primary example a relatively simple financial methodology, a
first-order variance/covariance approach. The prototype of this methodology is the
basic RiskMetricsTM methodology developed by J. P. Morgan [MR96].1
Perhaps the main challenge in implementing a VaR system is in coming up with a
systematic way to express the risk of a bewilderingly diverse set of financial instruments
in terms of a relatively small set of risk factors. This is both a financial-engineering
and a system-implementation challenge. The body of this chapter will focus on
some of the system-implementation issues. Some of the financial-engineering issues
are discussed in the appendices.