Value at Risk (VaR) estimates the maximum loss a portfolio is expected to suffer over a defined time horizon at a given confidence level under normal market conditions.
Notation: '1-day 95% VaR of Rs.2 lakh' means there is a 95% probability the daily loss will not exceed Rs.2 lakh - and a 5% probability it will be worse. Three methods: historical simulation, variance-covariance, and Monte Carlo.
Example: An equity portfolio worth Rs.1 crore with 95% confidence 1-day VaR of Rs.3 lakh means: on a normal day, losses should stay under Rs.3 lakh; on roughly 1 in 20 trading days, they will exceed it.
When to use: Setting position limits, computing margin requirements, regulatory capital calculations (Basel III for banks), and risk-budgeting in multi-asset portfolios. Do not use VaR in isolation - it says nothing about how bad the loss is when the threshold is breached.
SEBI caveat: VaR assumes normal distributions and historical data repeats - both assumptions break in crashes (2008, March 2020). Always pair with Stress Test results and Downside Deviation. VaR is a regulatory tool, not a guarantee.
Related: Stress Test, Downside Deviation, Standard Deviation, Maximum Drawdown.