Naming the risk before managing it
Standard deviation (volatility) is the most commonly cited risk measure because it is easy to calculate, but it treats a sharp rally and a sharp decline as equally ‘risky,’ which does not match how most people experience gains versus losses. Maximum drawdown is often more intuitive: it answers ‘what is the worst this has looked before, from a prior high?’ Value at Risk (a statistical estimate of the most a portfolio might lose over a set period at a given confidence level) is used more in institutional settings and comes with a caveat—it describes a typical bad outcome, not the worst possible one, which is why stress-testing against past crisis periods is often paired with it. Correlation—how closely two assets move together—is the mathematical basis for diversification, but correlation is not fixed. Assets that appear weakly correlated during calm markets frequently move together during a broad sell-off, precisely the moment diversification is needed most.