Financial Simulation Modeling in Excel

VALUE AT RISK (VAR)

Price forecasting has a major role in risk management. There are a number of methods that portfolio managers and regulators will use in order to track the risk a portfolio runs due to sudden market movements. One method is scenario modeling, in which managers will use a specific scenario, either imagined or from the past (such as the credit crunch in 2008 or the oil shocks of the 1970s) and estimate what the impact on their portfolio would be in such a case. This is typically used to track response to certain exogenous events, but doesn't give much information about how the portfolio is likely to perform during most days or months. For this, a set of stochastic processes often referred to as “value at risk,” or VaR, are often employed.

VaR is an estimation, over a certain time horizon and at a certain probability level, of what the minimum mark-to-market losses portfolio would be. A daily, 99 percent VaR of $1 million means that a portfolio manager would expect to have a 1 percent chance of losing $1 million or more in a day. It is important to notice that this does not say that in the worst 1 percent of performance that a portfolio will lose only $1 million. VaR is not a worst-case-scenario measure.

While there are an endless number of variations of how VaR can be calculated, a few conventions are generally followed in its reporting. Generally 95 percent VaR and 99 percent VaR are reported: what a portfolio's returns would be expected to exceed 19 out of 20 ...