Value at risk (VAR) has gained rapid acceptance as a valuable approach to risk management. Not all VARs are equal, however. A study of VAR techniques used by dealers and end-users reveals that VAR calculations differ significantly for the same portfolio. VARs are extremely dependent on parameters, data, assumptions, and methodology. Calculation of eight common VARs for three hypothetical portfolios demonstrates the potentially seductive but dangerous nature of any single approach to risk management. In sum, although VAR and other quantitative techniques are necessary aspects of an effective risk-management program, they are not sufficient to control risk.