Implied vs realised volatility: what option prices say about expected risk versus what actually happens

An asset's volatility can be measured two ways. Realised volatility is the actual standard deviation of past returns—what already happened. Implied volatility is the volatility back-solved from current option prices—what option markets expect. The two are typically close but not identical, and the systematic gap between them encodes meaningful information about market sentiment, risk pricing, and the structural premium earned by sellers of options.

What implied and realised volatility are

Realised volatility is the standard deviation of an asset's actual returns over a defined window, typically annualised by multiplying by the square root of the number of periods per year. The metric is purely backward-looking: it summarises what has happened to the asset's price during the measurement window. The conventional measure uses daily log returns over a one-month or three-month window, annualised to a percentage.

Implied volatility is forward-looking. Option prices reflect the market's expectations about future volatility, and the implied volatility is the volatility input that, when fed into an option-pricing model (typically Black-Scholes), produces the observed market price for the option. The metric is therefore a market-derived expectation: it is what the marginal option buyer and seller agree on as a fair compensation for bearing or taking volatility risk.

The two metrics measure related but distinct things. Realised volatility is a historical fact about the asset; implied volatility is a market expectation about the asset's near future. They tend to track each other in long-run averages but can diverge meaningfully over shorter windows.

How they differ in practice

The most-watched implied volatility metric is the VIX index, which measures the implied volatility of S&P 500 options at a constant 30-day horizon. The VIX has averaged approximately 19–20 over the post-1990 period, while realised volatility on the S&P 500 over the same window has averaged approximately 15–16. The roughly 4-percentage-point gap is the volatility risk premium—the systematic difference between what option markets price and what actually happens.

The gap exists because option sellers demand compensation for bearing volatility risk. Like the equity risk premium, it is a positive expected return for those willing to take on a specific risk that others want to hedge against. The gap is unconditional in the long-run but time-varying: in calm regimes, the gap can widen as realised volatility falls below implied; in stress regimes, the gap narrows or even inverts as realised volatility spikes above implied.

The two metrics also differ in their dynamics. Realised volatility is by construction backward-looking and changes slowly as the rolling window shifts. Implied volatility is forward-looking and reacts immediately to events that change the market's volatility expectations—earnings announcements, central-bank decisions, geopolitical events all move implied volatility on the day of the news, while realised volatility responds only as the news event itself produces price movement.

What the evidence shows

Carr and Wu (2009), in Variance Risk Premiums, document the implied-versus-realised gap across major equity indices and currencies over multi-decade samples. The gap is consistently positive (implied above realised) and statistically significant at standard confidence levels. The same finding has been replicated in commodities, fixed income, and individual stock options, with magnitudes that vary by asset class but a consistent direction.

The gap's most visible expression is in the VIX-versus-realised-S&P gap. Studies by AQR, Goldman Sachs, and many academic researchers have produced consistent estimates: implied volatility has been approximately 4–5 percentage points above realised volatility on average over multi-decade samples for US equities, with the gap narrowing or inverting in approximately 20–30% of months—exclusively those including unexpected volatility spikes.

The implication for strategy is direct. Selling volatility—through short variance swaps, short straddles, or short volatility ETPs—captures the volatility risk premium over time. The structural profit comes from selling implied volatility above the realised volatility that follows. The structural risk is the months in which realised exceeds implied; these are rare but can produce losses many multiples of the average gain.

Limitations and trade-offs

Implied volatility depends on the option-pricing model used to back-solve from market prices. Black-Scholes assumes log-normal returns, which is empirically violated by most asset classes—equity returns have fatter tails than log-normal, with skewness toward the negative side. The implied volatility extracted from out-of-the-money put options is therefore systematically higher than from at-the-money options (the volatility smile), reflecting the option market's pricing of crash risk that the standard model does not capture.

Realised volatility is sensitive to the chosen window. A 30-day rolling realised volatility responds quickly to recent events; a 12-month window smooths out short-term spikes but lags substantively. The choice of window must match the use case—pricing 30-day options against 30-day realised volatility, year-long strategies against year-long realised volatility—to produce comparable measures.

The two metrics also disagree about asset-class-specific events. Earnings announcements typically produce a temporary spike in implied volatility before the announcement (the option market prices the uncertainty) and a corresponding drop after (the uncertainty is resolved); realised volatility responds only on the announcement day itself, with the rolling realised measure picking up the spike with a lag. Comparing the two across earnings windows requires care to align the time periods consistently.

Implied vs realised volatility in pfolio

pfolio reports realised volatility (and rolling realised volatility) for any asset or portfolio in the analytics suite. Implied volatility is not directly reported—the platform does not consume option-pricing data—but exposure to the volatility risk premium can be expressed through VIX futures and volatility ETPs available in the asset universe.

Related articles

• Volatility in investing: how to measure and manage portfolio risk
• VIX and volatility investing explained: alternative assets and their role in a portfolio
• The volatility risk premium: why selling volatility has historically been profitable
• Volatility as an asset class: how VIX products behave and when they diversify
• Options explained: calls, puts, and the basics of derivative contracts