Implied vs Realized Volatility
Implied volatility is the market's forecast; realized volatility is what actually happens. They're consistently different — and the gap (the variance risk premium) is what most options strategies harvest or pay.
Two volatility numbers traders use. Implied volatility is the market's forecast — derived from option prices. Realized volatility is what the underlying actually does — measured from price history. They're different by construction, and the relationship between them is what most options strategies are quietly betting on.
This article covers what each one measures, why they tend to diverge, what their gap (the variance risk premium) represents, and how the two play out for 0DTE specifically.
What realized volatility is
A measurement of how much the underlying actually moved over a past window, expressed as an annualized standard deviation of log returns. Backward-looking. Different windows (10-day, 20-day, 30-day) produce different RV numbers; there's no single "the" realized volatility.
Realized vol is computed from price history. The standard recipe:
- Take the log returns of the underlying over a window — daily for most retail calculations, but tick-by-tick or per-minute for higher resolution.
- Compute the standard deviation of those returns.
- Annualize by multiplying by √252 (for daily returns) or the appropriate factor for other frequencies.
A concrete example: SPX daily log returns over the past 30 trading days. If their standard deviation is 0.65%, the annualized realized vol is 0.65% × √252 ≈ 10.3%.
Realized vol is a measurement, not a forecast. It tells you exactly what happened over the chosen window, in volatility units. Two traders looking at the same underlying can compute different RV numbers if they choose different windows or different sampling frequencies — there's no canonical answer.
What implied volatility is
The σ that, plugged into Black-Scholes with the other known inputs (underlying, strike, expiration, rate), reproduces the option's market price. Forward-looking — it's what the market is implicitly forecasting for volatility over the option's remaining life. Derived from price by inverting BSM; not directly observable.
Where realized vol is a measurement of the past, implied vol is a forecast of the future. The market quotes option prices; we use BSM in reverse to extract the volatility consistent with those prices.
A few wrinkles:
- Different strikes give different IVs. The 30-delta put and the 30-delta call at the same expiration usually have different IV numbers. That asymmetry is the volatility skew.
- Different expirations give different IVs. 7-day SPX IV is usually different from 30-day SPX IV. This is the term structure of IV.
- IV is model-dependent. BSM-derived IV is the standard; other pricing models would produce slightly different IV numbers from the same observed prices.
Some traders speak of "the IV" of an underlying — usually referring to either ATM IV at a specific expiration or to a composite measure like VIX (which is for SPX, 30-day horizon, derived from a strike-weighted basket). When precision matters, specify which IV you mean.
They measure different things
Realized vol is backward-looking; implied vol is forward-looking. Both are expressed as annualized percentages. Both look like the same kind of number on a screen.
That symmetry of units is what makes the comparison meaningful — and also what trips new traders into treating them as the same quantity.
The two are linked but not equal. Implied is what the market thinks volatility will be; realized is what it turned out to be over a past window. Comparing them gives you a way to ask: was the market's previous forecast too high, too low, or about right?
Most options strategies are mechanically a bet on what that comparison will look like over the strategy's life.
The relationship: IV typically exceeds RV
Empirically, for broad equity indices and over long periods, implied volatility tends to run higher than realized volatility. This isn't a coincidence and it isn't a mispricing — it's the variance risk premium.
The systematic tendency for implied volatility to exceed subsequently realized volatility, especially for index options. Represents the compensation options sellers collect for bearing the risk of large moves — particularly left-tail (sharp downside) events. VRP is what makes index premium-selling strategies profitable on average over long samples — and it's also why those strategies have steep drawdowns during the rare periods when realized vol exceeds implied.
The intuition: options sellers are taking on the tail risk of large moves. If they were paid only the average expected volatility, there'd be no incentive to take that asymmetric risk — large losses on the rare bad day with no compensation on the average day. So the market prices options at IVs above the realized vol that typically unfolds, which:
- Pays sellers a premium for the tail risk they're underwriting.
- Costs buyers a fee for the tail-risk protection they're buying.
This isn't a free lunch for sellers. The variance risk premium pays out specifically in scenarios that don't happen often (catastrophic vol spikes), but when they do, the realized-greater-than-implied gap can eat through years of accumulated premium. Premium-selling is a strategy with positive expected value and a fat-tailed loss distribution — a tradeoff, not a free edge.
Why this matters for strategies
Most short-premium strategies (selling theta) are mechanically a bet that IV > realized over the trade's life.
- A short SPX iron condor entered at 18% IV: profitable if SPX's realized vol over the trade's life is less than what the entry IV priced in (translated to dollars via the gamma curve).
- A long SPX call: profitable if delta gains exceed theta paid, which usually requires realized vol high enough to push the underlying meaningfully in the trader's direction.
The variance risk premium also explains why long-premium strategies are structurally harder over long horizons. Buying options at IV > realized means paying more for the option than the underlying's actual movement will compensate, on average. You can win on specific trades — when the underlying moves more than the IV predicted — but the structural headwind is real.
IV and RV on 0DTE
For 0DTE specifically, two things to know:
VIX1D. Cboe publishes a 1-day-horizon version of VIX called VIX1D, constructed similarly to VIX but at the daily timescale. It measures the market's implied volatility for SPX over roughly the next trading day. When you look at SPX 0DTE option prices, the IVs you see are consistent with this short-horizon implied vol number.
Intraday realized. SPX's realized vol can be computed over any window — including intraday ones like "10:00am to noon today." That intraday RV is directly comparable to the IV being quoted on 0DTE options earlier the same day. The two measure the same horizon (the rest of the trading day) at the same timescale, just from opposite directions (forecast vs measured).
The 0DTE theta-collection strategies (credit spreads, iron condors, short strangles) are betting that realized SPX volatility over the rest of today < implied volatility priced into 0DTE options. When the bet works, premium decays into the seller's pocket. When realized comes in higher than implied — a sharp midday rally or sell-off — gamma overwhelms theta and the seller loses.
The bet works often enough that the strategies are popular. It loses badly enough on bad days that sizing is the difference between long-term profitability and account drawdown.
How to read the IV-RV gap
A practical heuristic for the IV-RV spread:
- IV well above RV: typical conditions for premium-selling. Sellers are being paid the variance risk premium; long-premium traders face a structural headwind.
- IV near RV: the premium has thinned. Sellers are taking the same tail risk for less compensation; buyers are paying closer to what the underlying's recent realized movement was costing.
- IV below RV: unusual. Either the market is underpricing future volatility (rare for broad equity indices) or recent realized vol was unusually high relative to what the market expects to continue.
Most retail platforms show some version of 30-day IV alongside historical vol (often labeled "IV" with an "HV" comparison column). Some show the IV-vs-HV ratio directly. Knowing where you sit in the spectrum tells you whether your strategy choice is fighting the structural premium or aligned with it.
Key takeaways
- Realized volatility measures how much the underlying actually moved (backward-looking, computed from price history). Implied volatility is the market's forecast (forward-looking, derived from option prices).
- Both are quoted as annualized percentages and look like the same kind of number, but they measure different things.
- IV typically exceeds RV for broad equity indices over long periods. The gap is the variance risk premium — compensation paid to options sellers for bearing tail risk.
- Premium-selling strategies are mechanically a bet that IV > realized over the trade's life. They work on average because of VRP, but the loss distribution is fat-tailed.
- For 0DTE, VIX1D measures the implied vol over roughly the next trading day. Comparing it to intraday realized vol tells you whether premium-sellers or buyers are getting the better of the trade structurally.