RiskMetrics is a methodology that contains techniques and data sets used to calculate the value at risk (VaR) of a portfolio of investments. RiskMetrics was launched in 1994, and the technical document outlining the methodology was released in October 1994. J.P. Morgan and Reuters teamed up in 1996 to enhance the methodology and make data widely available for practitioners and the general public. The aim of RiskMetrics is promoting and improving the transparency of market risks, creating a benchmark for measuring risk, and providing clients with advice on managing market risks.

RiskMetrics Methodology

The RiskMetrics methodology for calculating the VaR assumes that a portfolio or investment's returns follow a normal distribution. Following J.P. Morgan's and Reuters' release of its volatility and correlation data sets, the variance-covariance method used to calculate the VaR became an industry standard, as RiskMetrics intended.

For example, the profit and loss distribution of an investment portfolio is assumed to be normally distributed. Using the RiskMetrics methodology to calculate the VaR, the portfolio manager must first select the confidence level and lookback period. Suppose he chooses a lookback period of 252 days or one trading year. The chosen confidence level is 95%, and the corresponding z-score value must be multiplied by the standard deviation of the portfolio. The actual daily standard deviation of the portfolio over one trading year is 3.67%. The z-score for the 95% is 1.645. The VaR for the portfolio, under the 95% confidence level, is -6.04% (-1.645 * 3.67%). Therefore, there is a 5% probability that the loss of the portfolio, over the given time horizon, will exceed 6.04%.

Using RiskMetrics

Another method to calculate the VaR of a portfolio is to use the correlations and standard deviations of stock returns provided by RiskMetrics. This method also assumes that stock returns follow a normal distribution. The first step to calculating the VaR is taking the square of the allocated funds for the first asset, multiplied by the square of its standard deviation, and adding that value to the square of the allocated funds for the second asset multiplied by the square of the second asset's standard deviation. Then that value is added to two multiplied by the allocated funds for the first asset multiplied by the allocated funds for the second asset multiplied by the first and second asset's volatility and the correlation between the two assets.

For example, suppose a portfolio has two assets with \$5 million allocated to stock ABC and \$8 million allocated to stock DEF. Respectively, the price volatility of stock ABC and stock DEF is 2.98% and 1.67% for a one-day period. The correlation between the two stocks is 0.67. The value at risk is \$258,310.93 (sqrt((\$5 million)^2 * (0.0298)^2 + (\$8 million)^2 * (0.0167)^2 + (2 * \$5 million * \$8 million * 0.0298 * 0.0167 * 0.067)).