Information and education only. Every page, figure and calculator on this site exists to explain how financial instruments work. Nothing here is investment, tax or legal advice, a recommendation, or a valuation you can rely on. Full disclaimer

VolatilitySome background helps

Finance's stand-in for risk: measured from the past, traded in the present, feared about the future — and an asset class of its own.

What the number means

Volatility is the annualised standard deviation of returns — the market's unit for "how much does this thing move". An asset at 20% vol delivers one-year outcomes roughly ±20% around its trend two times in three, and ±40% (two standard deviations) in 19 years of 20. It is not risk itself — permanent loss is risk — but it is the part of risk you can measure daily, price, and trade.

$$ \sigma_{\text{ann}} = \sigma_{\text{daily}} \cdot \sqrt{252} \qquad\qquad \text{expected move over } d \text{ days} \approx \sigma_{\text{ann}} \sqrt{\tfrac{d}{252}} $$
What the symbols mean
  • sigmavolatility, the standard deviation of returns

Interactive: volatility converterPractitioner

The √time rule in one tool: turn an annualised vol into daily and monthly terms and into the expected move over any horizon.

Daily vol
Monthly vol
Expected move (1σ)
95% range (2σ)

√252 for trading days; the same rule scales any horizon. It assumes independent daily moves — volatility clustering (below) is exactly where that assumption bends.

Realised vs. implied: the two volatilities

  • Realised (historical) vol — measured from past returns: what the asset did.
  • Implied vol — backed out of option prices (try the IV solver): what the market pays for going forward. The VIX condenses 30-day S&P implied vol into one number — mechanically, a variance-swap strike (see variance swaps).
Implied usually sits above subsequent realised — the volatility risk premium. In shocks, realised briefly overtakes: the insurer's loss year.
ShockImplied (bought today)Realised (delivered later)TimeVolatility
  • The vol risk premium: implied exceeds subsequently-delivered realised most of the time — option buyers systematically overpay, like insurance customers. Harvesting that gap (selling options, short VIX futures) earns steady carry with rare violent losses: "Volmageddon" (Feb 2018) erased short-vol ETPs in one afternoon.
  • Volatility clusters: calm days follow calm days, wild days follow wild ones (the GARCH observation). Vol is far more forecastable than returns — the whole vol-trading industry lives in that gap.
  • The leverage effect: equity vol rises when prices fall — one driver of the skew and of why long-vol positions hedge equity portfolios.

Volatility as a control variable

  • Vol targeting: funds scale positions to keep portfolio vol constant — mechanically selling after vol spikes, buying after calm. Stabilises the ride; adds pro-cyclical flows the market now anticipates.
  • Position sizing: the honest use — at 60% vol (crypto) a "small" position delivers big-position outcomes; the risk-based sizing tool and the leverage-decay calculator both run on vol inputs.
  • Vol term structure: like the yield curve, implied vol has a curve across expiries — upward in calm (near-dated cheap), inverted in panic (near-dated explosive). VIX futures trade this curve, with the same roll costs commodities know.

Reading vol like a practitioner

  • Vol is a price, not just a statistic — ask what level you're implicitly buying or selling in any position with optionality (every structured product page in this atlas is an example).
  • High vol ≠ sell, low vol ≠ safe: the calmest markets breed the positions that detonate (2017 → 2018, 2006 → 2008). Vol mean-reverts, but from when and where is the entire question.
  • Annualised numbers hide daily reality: 80% annualised vol is ~5% per day — run the converter above before judging any crypto or single-stock position.

Trading the difference: gamma scalping

A delta-hedged option is a pure bet on volatility. Its P&L over a small move has two terms that pull in opposite directions — convexity pays, time decays:

$$ \text{P\&L} \approx \tfrac{1}{2}\,\Gamma\,(\Delta S)^2 \;+\; \Theta\,\Delta t $$
What the symbols mean
  • Gammahow fast Delta itself changes
  • Deltahow much a derivative moves when the underlying moves
  • Sthe price of the underlying today
  • ta point in time
  • Long gamma is long realised volatility. Every re-hedge locks in a small profit from the move; the position pays theta for the privilege of waiting.
  • The breakeven move is where the two terms cancel. That daily move is the realised volatility the option's implied volatility is charging you for.
  • Short gamma is the mirror image: collect the decay, pay on every move, and lose disproportionately in the large ones — the risk profile that ends careers when it is left unattended.

Interactive: gamma scalping P&LPractitioner

Move in dollars
Convexity gain
Time decay
Net
Breakeven move
Reading

The breakeven move is the honest translation of "implied volatility". The defaults are a long at-the-money option on 1,000 shares at 20% implied vol with a month to run — and the tool returns a breakeven of about 1.05%, which is exactly 20% ÷ √365. That identity is the whole trade: implied vol is the daily move you are paying for. Deliver more and long gamma wins; deliver less and the decay does. Gains here are quadratic in the move — the reason a single large day can pay for a month of small ones, and the reason the short side blows up rather than bleeds.

Barriers: the probability of touching

Knock-outs, knock-ins and stop levels all hinge on one question — will the price touch a level, not where it finishes. For a driftless random walk the reflection principle gives a clean answer:

$$ P(\text{touch}) = 2\,N\!\left(\frac{-|\ln(B/S)|}{\sigma\sqrt{T}}\right) $$
What the symbols mean
  • Pa price, or a present value
  • Nthe normal distribution, or a count
  • Sthe price of the underlying today
  • sigmavolatility, the standard deviation of returns
  • Tmaturity, in years
  • Touching is roughly twice as likely as finishing beyond — the single most under-appreciated fact about barrier products and stop-loss orders alike.
  • It scales with volatility × √time, not with distance alone: a barrier 10% away is safe over a week and a coin flip over a year.
  • Real markets add drift, skew and gaps; the formula understates risk in exactly the products that sell on the barrier looking far away. See knock-out certificates and reverse convertibles.

Interactive: probability of hitting a barrierPractitioner

Distance to barrier
In volatility terms
Probability of touching
Probability of finishing beyond
Touch vs. finish
Reading

A 20% buffer over six months at 25% vol sounds generous and is not. This is the arithmetic behind every "the barrier is very far away" sales conversation — and behind the 2018 and 2020 case studies, where barriers assumed to be unreachable were reached.

Interactive: what averaging is worthSpecialist

Many structured products settle on an average of observations rather than the closing level. Averaging cuts the effective volatility of the payoff, and a cheaper option means less value delivered to you for the same headline terms.

Effective volatility
Averaged (Asian) call
Ordinary call
The averaging discount
Why

The continuous geometric-average case has a closed form, which is why it is used here: no simulation, no random numbers, a number you can check. Arithmetic averaging is slightly more valuable but has no exact formula; the discount is of the same order. When a term sheet says "average of the final twelve monthly closes", this is the size of what that sentence did — see how to read a term sheet.

Interactive: the worst-of termSpecialist

A "worst-of two" pays on whichever underlying does worse. This is not diversification — it is selling correlation risk, and it is the least visible term in retail structured products. Move the correlation slider input and watch the breach probability move.

One asset alone breaches
Worst of the two breaches
The difference
If correlation were 1
If they were independent
Reading

At correlation 1 the worst-of is identical to a single asset — there is only one bet. At correlation 0 the breach probability is far higher, because either one failing is enough. A term sheet that adds a third or fourth underlying "for diversification" is doing the opposite, and the compensation for it is rarely visible in the coupon. Computed from the bivariate normal distribution by numerical integration, with no simulation.

Information and education only. This page explains a mechanism in general terms, using simplified textbook models and illustrative figures. It is not advice, not a recommendation, and not a valuation you can rely on. Conventions and rules differ by market and jurisdiction and change over time.