Volatility

Formula

How it works

Start from a return series at a fixed frequency: daily or monthly for listed portfolios and most hedge funds, quarterly for funds whose net asset value is struck quarterly. Compute the mean, the deviation of each return from it, and the standard deviation of those deviations; then scale to an annual figure. Because the input is a periodic return, a private fund needs a time-weighted return series built from its NAVs and cash flows. A fund's IRR is a single money-weighted figure and has no volatility of its own. Two volatilities are comparable only if they use the same frequency, window and estimator.

Realised, implied and forecast volatility

Realised (historical, ex post) volatility is computed from past returns, as above. Implied volatility is the volatility input that makes an option-pricing model reproduce an option's market price, so it is a forward-looking market estimate for a listed underlying; private assets have none. Forecast (ex ante) volatility comes from a risk model that combines exposures with estimated factor volatilities and correlations. Risk reports should say which of the three a figure is.

Private assets: smoothed returns understate it

Getmansky, Lo and Makarov (2004) show that for portfolios of illiquid securities, reported returns tend to be smoother than true economic returns, which understates volatility and raises risk-adjusted measures such as the Sharpe ratio. Private equity, private credit and real estate funds report NAVs built from periodic valuations rather than market prices, so their reported return series are exposed to the same effect, and a lagged mark also delays losses into later quarters. Return smoothing covers the mechanism and the desmoothing methods. A low reported volatility for a buyout or real estate fund is therefore not evidence that its value moves less than listed equity of similar risk.

How LPs use it

Volatility is the denominator of the Sharpe ratio, the volatility of active returns is tracking error, and risk budgets are often set in volatility terms. Under the Global Investment Performance Standards (GIPS) for Firms 2020, a time-weighted composite or pooled fund report must show, where monthly returns are available, the three-year annualised ex post standard deviation, using monthly returns, of the composite or fund and its benchmark at each annual period end; once the composite or fund has at least three annual periods of performance, the report must disclose when the figure is not shown because 36 monthly returns are not available. In asset allocation, LPs often replace reported private-market volatility with desmoothed or listed-proxy estimates before mixing private and listed assets in one model.

Limits

Volatility treats an upside surprise as risk in the same way as a loss, which is why the Sortino ratio uses downside deviation instead. It summarises a whole distribution in one number, so it misses fat tails and option-like payoffs: a strategy that sells protection can show low volatility until a large loss. It says nothing about the path of returns, which drawdown captures, or about liquidity. Pair it with drawdown, scenario analysis and liquidity terms.

Worked examples

Examples are illustrative; figures are not market data.

Not the same as

• Drawdown (Peak-to-Trough): Drawdown measures the fall from a previous peak and depends on the order of returns; volatility measures dispersion around the mean and ignores order.
• Tracking Error: Tracking error is the volatility of the difference between a portfolio's and its benchmark's returns; volatility is measured on the portfolio's own returns.
• Beta: Beta measures sensitivity to a market's returns, the systematic part of risk; volatility measures total dispersion, systematic and specific together.

Common mistakes

• Annualising monthly volatility by multiplying by 12 instead of √12.
• Comparing the reported volatility of NAV-based private fund returns with that of listed indices without adjusting for smoothing.
• Comparing volatilities computed over different windows, frequencies or estimators.
• Reading low volatility as low risk of loss, for a strategy with option-like or illiquid exposures.

Edge cases

• Short histories give imprecise estimates; a GIPS report for a composite or pooled fund with at least three annual periods of performance must disclose when the three-year figure is missing because 36 monthly returns are not available.

Questions

External standards

Sources

1. An Econometric Model of Serial Correlation and Illiquidity in Hedge Fund Returns (NBER Working Paper 9571). Mila Getmansky; Andrew W. Lo; Igor Makarov, National Bureau of Economic Research, NBER WP 9571, March 2003; published Journal of Financial Economics 74(3), 2004, pp. 529-609 (per NBER page). Status: published (checked 2026-10-01). Abstract — supports: Reported returns of illiquid portfolios are smoother than economic returns, understating volatility and raising the Sharpe ratio
2. Global Investment Performance Standards (GIPS) for Firms 2020. CFA Institute, 2020 edition; effective 1 January 2020; required for GIPS Reports with periods ending on or after 31 December 2020. Status: Current (checked 2026-10-01). Provisions 4.A.1.j (p. 22), 4.C.36 (p. 30), 6.A.1.h (p. 44), 6.C.29 (p. 51) — supports: Three-year annualised ex post standard deviation using monthly returns, where monthly returns are available, in time-weighted composite and pooled fund reports; disclosure, for composites and pooled funds with at least three annual periods, when it is not presented because 36 monthly returns are not available

• Sharpe Ratio
• Sortino Ratio
• Drawdown (Peak-to-Trough)
• Tracking Error
• Beta
• Risk Budget
• Time-Weighted Return (TWR)
• Correlation
• Scenario Analysis