Volatility
What is Volatility?
Volatility measures the dispersion of investment returns, commonly as the annualized standard deviation of periodic returns.
Volatility charts should identify return frequency, annualization, lookback, weighting, currency, and treatment of missing data. Realized volatility is backward-looking and model-dependent. It should be paired with drawdown, downside measures, liquidity, concentration, and scenarios so smooth returns are not mistaken for safety.
Annualization assumes a scaling relationship that can break when returns have autocorrelation, volatility clustering, or valuation smoothing.
For multi-asset portfolios, component contribution should use an internally consistent covariance estimate and show whether diversification changed because weights, individual volatility, or correlations moved. Forecast and realized figures should remain separate and should not be blended into one unexplained risk number.
Calculation
Compute returns at a stated frequency, estimate their standard deviation, and annualize with a factor consistent with the data, such as the square root of 252 for daily observations under common assumptions. Sample versus population formulas, log versus simple returns, and weighting choices create differences. Annualization is an approximation, not a law.
What it captures
Higher volatility means returns varied more around their average. It treats upside and downside deviations symmetrically and does not describe loss shape, timing, liquidity, or permanent impairment. Two portfolios with equal volatility can have very different drawdowns, skew, tail risk, and investor outcomes, especially when withdrawals occur.
Realized and expected volatility
Realized volatility uses historical returns. Implied volatility is inferred from option prices under a model and includes market pricing of uncertainty and supply-demand effects. Forecast volatility can use historical, conditional, or market-based inputs. These measures serve different purposes and should not be substituted without explanation.
Portfolio applications
Volatility supports risk comparison, position sizing, Sharpe ratios, optimization, simulations, and risk targets. Diversification can reduce portfolio volatility when correlations are imperfect. Volatility-targeting strategies change exposure as estimates move, which can force selling after shocks and produce path-dependent results. Constraints and turnover belong in evaluation.
Limitations and interpretation
Returns often cluster, jump, have fat tails, and change regime. Illiquid and appraisal-valued assets can appear artificially smooth. A low number can reflect stale data rather than safety. Report lookback, frequency, annualization, currency, and missing-data method, and pair volatility with drawdown, concentration, liquidity, and scenario loss.
Sources and further reading
- Risk Management: An Introduction, CFA Institute