Value at Risk

Value at Risk became the dominant risk measurement tool on Wall Street during the 1990s after J.P. Morgan popularized the concept through its RiskMetrics framework. Regulators including the Basel Committee subsequently embedded VaR into bank capital requirements, cementing its role in the global financial system.

A VaR statement is always defined by three parameters: a portfolio (or position), a time horizon, and a confidence level. For example, a one-day 99% VaR of $1 million means that on 99 out of 100 trading days, losses are expected to be less than $1 million. Equivalently, there is a 1% probability of losing more than $1 million in any given day. The VaR itself does not say anything about the magnitude of losses in that worst 1% of scenarios — that is addressed by a related metric called Expected Shortfall (also called CVaR or Conditional VaR).

VaR can be calculated using several methods. The historical simulation approach simply ranks past returns and reads off the appropriate percentile. The parametric (variance-covariance) approach assumes returns follow a normal distribution and uses estimated means and standard deviations. Monte Carlo simulation generates thousands of hypothetical scenarios using statistical models. Each approach has different assumptions and is more or less reliable in different market environments.

VaR's most significant limitation, exposed dramatically during the 2008 financial crisis, is that it relies on historical return distributions which may systematically underestimate the probability and severity of rare extreme events. Black swan events by definition fall outside the range of recent historical experience, meaning the model's confidence interval provides false precision precisely when accurate measurement matters most.

Despite its limitations, VaR remains ubiquitous in professional risk management because it aggregates diverse risks into a single comparable number. Investors evaluating funds or managed accounts should look beyond VaR to metrics like maximum drawdown, expected shortfall, and stress-test results to get a more complete picture of tail risk.