The implied-versus-realized comparison is the foundational lens of all volatility trading. Implied volatility (IV) is forward-looking: it is the volatility backed out of an option's market price, representing what buyers and sellers collectively expect the underlying to move between now and expiration — it is the price of volatility. Realized volatility (RV) is backward-looking (when measured over the past it is also called historical volatility): it is the annualized standard deviation of the underlying's actual returns over a window — what the market actually did. The core tension of the entire field lives in the gap between the two: an option's payoff is determined by how IV (what you paid) compares to the RV that subsequently occurs (what you got). Option buyers want realized to exceed implied; option sellers want the reverse. That gap — not market direction — is the thing volatility traders trade.

The core distinction

• Implied volatility — a single forward number per expiry, inverted from option prices via a pricing model (Black-Scholes / binomial). It embeds not just an expectation of future movement but also the premium the market charges for bearing jump and crash risk. It is a price, set by supply and demand for options.
• Realized (historical) volatility — computed directly from price data: the standard deviation of log returns over N days, annualized (×√252). No model, no expectations — just the delivered quantity. "20-day HV" describes the last 20 sessions; future realized volatility is the same calculation applied to the period that has not happened yet, and is only knowable in hindsight.

The clean mental model (Macroption): IV is "price," future RV is "value." A volatility trade is profitable when you buy the cheaper of the two and sell the richer. Note that IV is not one number — every strike and expiry can carry a different IV (the volatility surface) — whereas RV is a single realized series for the underlying.

How the gap is the basis of volatility trading

A delta-hedged option strips out direction and leaves a pure bet on IV vs RV. If you buy an option (pay IV) and delta-hedge, you profit if subsequent RV exceeds the IV you paid; if you sell (collect IV) and hedge, you profit if RV comes in below implied. The same logic drives undefined- and defined-risk premium structures (straddles, strangles, condors) and variance/volatility swaps. Reading IV against RV is therefore the trader's first question: are options rich (IV high relative to what the stock is realizing, favoring selling premium) or cheap (IV low relative to RV, favoring buying)?

The general tendency of IV to exceed RV

Over long samples, implied volatility tends to sit above subsequently realized volatility — especially on broad equity indices. This persistent wedge is the volatility risk premium (VRP): option sellers, on average, collect more than the eventual movement justifies, compensation for underwriting crash insurance. Commonly cited figures put IV above subsequent RV on the S&P 500 roughly 80–90% of the time (figures of ~84–85% are widely repeated; the exact share is sample-, window-, and underlying-dependent, so treat any single number as approximate rather than a constant). The average overstatement is often quoted around 2–4 vol points on indices — again sample-dependent.

Two honest caveats matter at the headline level:

• It is not free money. The VRP is the price of insurance, and short-vol losses cluster precisely in crashes when the rest of a portfolio is also losing (Feb 2018 "Volmageddon," March 2020). It is a real, repeatedly documented economic premium — but it is regime-dependent and crash-correlated, not a smooth arbitrage.
• It is an index-average effect. The robust academic VRP (Carr & Wu 2009; Bollerslev, Tauchen & Zhou 2009) is strongest at the index level. On individual names, idiosyncratic earnings and event jumps make the IV-over-RV tendency far noisier and less reliable. Do not let the index evidence vouch for "sell the high-IV stock."

This overview only summarizes the premium; the Volatility Risk Premium child node carries the full evidence, sign conventions, and the harvestability debate.

How a swing trader reads IV vs RV

For an equity swing trader, the IV-vs-RV read is mostly context and timing, not a standalone signal:

• Rich vs cheap options. When IV is elevated relative to the stock's own recent RV, options are expensive — favoring credit/premium-selling structures or simply not buying options. When IV is depressed relative to RV, options are cheap — favoring long-option/debit expressions if a catalyst is expected. The IV Rank & IV Percentile child node normalizes "is IV high or low for this name?" onto a 0–100 scale; raw IV alone is uninterpretable across symbols.
• Earnings vol. Ahead of a scheduled report, IV inflates to price the binary event, then collapses the moment results land — vol crush. A swing position carried through earnings is exposed to a gap whose width the option-implied expected move (from the ATM straddle) estimates better than the headline IV number. The Vol Crush Around Earnings child node covers this event case and the conditional, decayed nature of the buy/sell-straddle edge.

The decision is never "IV is high, so sell it." High IV is correct if the stock genuinely moves; the real question is whether implied is rich or cheap relative to that name's own history of realized moves.

Adoption, debate & evidence

The IV/RV distinction is universal and uncontested as a framework — every options desk, platform, and retail tool surfaces it. What is genuinely contested is the exploitable edge: the VRP's existence at the index level is one of the better-supported findings in empirical option pricing, but its cause (rational crash compensation vs. frictions) is unresolved, its harvestability net of tail risk and transaction costs is debated, and its reliability on single names is weak. The honest summary: the direction of the IV-over-RV tendency is robust on indices; the profitability of trading it is conditional, regime-dependent, and negatively skewed. See the three child nodes for the measured detail.

Sources

• Macroption — Difference between Implied, Realized and Historical Volatility (definitions; IV forward-looking, HV/RV backward, future RV): https://www.macroption.com/implied-vs-realized-vs-historical-volatility/
• MenthorQ — Implied vs. Realized Volatility Guide: https://menthorq.com/guide/implied-vs-realized-volatility/
• Macrosynergy — Realistic volatility risk premia (magnitude/share of premium, sample dependence): https://macrosynergy.com/research/realistic-volatility-risk-premia/
• Quantpedia — Volatility Risk Premium Effect (IV>RV frequency, premium framing): https://quantpedia.com/strategies/volatility-risk-premium-effect
• Carr, P. & Wu, L. (2009), "Variance Risk Premiums," Review of Financial Studies (index-level VRP; see child node for full cite).
• Child nodes: Volatility Risk Premium; IV Rank & IV Percentile; Vol Crush Around Earnings.

Disputes / flags: The "~84–85% of the time IV > RV" and "2–4 vol points" figures are commonly cited but sample-/window-/underlying-dependent — stated as approximate, not constants. The VRP is robust at the index level but weak and contested on single names; its cause and net-of-tail harvestability are genuinely unresolved. This is an overview — all depth and primary citations live in the three child nodes.