Home / Core Volatility / What is Volatility?
Core Volatility Beginner Dispersion of returns

What is Volatility?

The one number that describes the journey and refuses to name the destination.

Quick answer: What is Volatility? At its core it is the annualised standard deviation of an asset's returns — a measure of how far price typically strays from its own path, which says how much a market moves and deliberately nothing about which direction it moves.

In simple words

Volatility measures how bumpy the ride is, not where the ride ends up. Picture two journeys that both start at NIFTY 24,000 and, a month later, both finish at 24,000. The first drifted there in tiny steps and never lost more than a few points in a day. The second lurched up 400 points, crashed 600, rallied 500, and only happened to land back at 24,000. Same start, same finish, same return of zero — and wildly different volatility, because volatility is a fact about the size of the daily steps, not about the destination. The second market is the volatile one, and if you had held a leveraged position in it you would have felt every one of those swings whether or not the round trip ended flat.

The number itself is a standard deviation: take each day's percentage change, measure how spread out those daily changes are, and that spread is the volatility. If NIFTY typically moves about 0.8% up or down on an ordinary day, that daily spread of roughly 0.8% is the raw material. Everything else — annualising it, comparing it, pricing options off it — is bookkeeping applied to that one idea: how big is a normal move.

Not to be confused with: Return, which is where price ended up, and risk, which is broader than either. Volatility is orthogonal to return by construction — it squares away the sign, so a market that fell 2% and one that rose 2% contribute identically. And it is only one component of risk: a position can have low volatility and still be dangerous through leverage, illiquidity or a gap that no standard deviation anticipated.

Two paths, one destination, two volatilities

Volatility measures the journey, not the destination

Two simulated NIFTY paths that start at 24,000 and end at 24,000 — one calm, one violent.

22,00024,00026,00028,00030,0000d30d60d90d120dsame startsame destinationTrading daysNIFTY level≈9% annualised volatility≈30% annualised volatility
Both paths deliver exactly the same return: zero. Yet nobody would call them the same market. The only thing that separates them is the size of the daily steps, and that is precisely what volatility measures. This picture is the whole concept: the return lives at the endpoints, the volatility lives in everything between them.

Professional explanation

Volatility is a property of the path, not the outcome

The most common misconception is that a volatile market is one that has fallen, or one that is about to. Volatility contains no such information. It is computed by taking a series of returns, measuring their dispersion around their own average, and reporting that dispersion. Two return series with identical dispersion have identical volatility even if one rose all year and the other fell all year, and two series with the same start and end point can have volatilities that differ by a factor of ten. This is why the dispersion-basic chart on this page shows two paths that begin and end at the same level: the entire difference between them — the thing a leveraged trader lives and dies by — is invisible if you only look at the endpoints, and it is exactly what volatility captures.

It is a standard deviation of returns, computed on returns and not on prices

Volatility is not the standard deviation of the price; it is the standard deviation of the returns. The distinction matters because a ₹100 move means something different at NIFTY 24,000 than it would at 2,400, whereas a percentage move is comparable across levels and across instruments. Practitioners use log returns — the natural logarithm of today's price divided by yesterday's — because they add up cleanly across time and treat a rise and the fall that undoes it symmetrically. You collect those daily log returns over some window, compute their sample standard deviation with an n−1 denominator, and you have a daily volatility. That single daily number is the atom of everything that follows.

Annualising by the square root of 252, and why it is a square root

A daily standard deviation is rarely quoted directly; the market convention is an annualised figure, so that a NIFTY volatility and a BANKNIFTY volatility and an option's implied volatility can all be compared on one scale. The conversion is σ_annual = σ_daily × √252, using 252 because that is roughly the number of trading days in an Indian market year — weekends and holidays are excluded, because prices do not move when the exchange is shut. The square root, rather than a plain multiplication, is the part beginners trip over. Variance — volatility squared — grows linearly with time when returns are independent from one day to the next, so a year's variance is 252 daily variances added together. Volatility is the square root of variance, so a year's volatility is √252 times a day's. The rule quietly assumes returns are independent across days; when volatility clusters, which it does, the assumption is imperfect and the annualised number is a convention more than a measurement.

It is sign-blind, and that is both its power and its blind spot

Because volatility squares the returns before averaging them, a +2% day and a −2% day are treated as identical contributions. This is a genuine feature: it lets one number summarise the magnitude of movement without the noise of direction, which is exactly what an option — a bet on movement rather than direction — needs. But it is also a blind spot. Volatility cannot distinguish a market that grinds steadily upward from one that lurches violently in both directions, if their dispersions happen to match, and it cannot see skew — the empirical fact that equity indices fall faster than they rise. Two markets can share a volatility and have entirely different shapes of risk. A trader who treats volatility as a complete description of risk has thrown away the sign of every move, and in equity indices the sign is where the danger concentrates.

The normal distribution assumed by the maths has thinner tails than reality

The convenient mathematics of volatility — the idea that one standard deviation captures about 68% of days and two standard deviations about 95% — rests on returns being normally distributed. They are not. Real return distributions are leptokurtic: a taller, narrower peak of quiet days and, decisively, fatter tails than the bell curve allows. Under a normal distribution a five-standard-deviation day should occur roughly once every several thousand years; equity markets deliver them within living memory, repeatedly. This is the sentence a marketing department would cut: the standard model of volatility is at its least reliable in exactly the conditions that matter most, because it systematically under-counts the extreme moves that cause the largest losses. Volatility remains the right first number to reach for — it is just not the last one, and anyone using it to bound a worst case is using the wrong tool for that job.

Volatility clusters and mean-reverts

One more empirical regularity separates real volatility from the tidy textbook version: it is autocorrelated. Calm days tend to follow calm days and violent days tend to follow violent days, so volatility arrives in regimes rather than as independent noise — the property that makes GARCH-family models and the whole apparatus of volatility forecasting possible. Over longer horizons it also mean-reverts: no level of volatility, high or low, persists forever, and an extreme reading tends to be pulled back toward a long-run average. Neither fact tells you the timing. Clustering means a quiet stretch understates the risk of the storm that ends it, and mean reversion does not promise that a high reading will fall before it first goes higher.

The bell curve the maths assumes is not the one the market delivers

A normal distribution overlaid on the actual distribution of daily NIFTY returns.

00.10.20.30.40.5-4σ-3σ-2σ-1σ0σ1σ2σ3σ4σthe tails thatmattera fat-tailed market has MORE quiet daysand MORE extreme days than the bell allowsStandardised daily return (in standard deviations)Probability densityNormal (Black–Scholes assumption)Fat-tailed (real returns)
The two humps agree in the middle and disagree where it matters. Real daily returns have a taller peak and, crucially, fatter tails — extreme days that the normal curve says should almost never happen arrive several times a decade. Volatility built on the normal assumption therefore under-prices exactly the moves that do the damage, which is the single most important caveat on this page.

Formula

Annualising volatility — the square-root-of-time rule

σ_annual = σ_daily × √252

Annualising rests on returns being independent across days, so that variance adds linearly with time and volatility scales with the square root of time. 252 is the approximate count of trading days in an Indian market year; weekends and exchange holidays are excluded because prices do not move when the market is shut. When volatility clusters — and it does — independence is only approximately true, so the annualised figure is best read as a convention that makes numbers comparable rather than as a literal forecast of a year's dispersion.

  • σ_annualAnnualised volatility — the standard, quotable figure, expressed as a percentage (0.13 = 13%).
  • σ_dailyDaily volatility — the sample standard deviation of one day's log returns, expressed as a decimal (0.0082 = 0.82%).
  • √Square root. Variance grows linearly with time under independence, so volatility, its square root, grows with the square root of time.
  • 252Approximate number of trading days in an Indian calendar year — the annualisation factor for volatility built on daily returns.

The daily standard deviation underneath it

σ_daily = √( (1 / (N − 1)) × Σ (r_t − r̄)² ), with r_t = ln(P_t / P_{t−1})

The sample standard deviation of daily log returns, using an N−1 denominator (Bessel's correction) so the estimate is unbiased on a finite sample. The mean return r̄ over a short window is usually tiny relative to the daily moves, which is why realised-volatility estimators often drop it entirely and assume it is zero.

How to compute a volatility from a price series

  1. Take a series of daily closing prices for the window you care about — say the last 20 NIFTY closes.
  2. Convert each pair of consecutive closes to a log return: r_t = ln(P_t ÷ P_{t−1}). A column of prices becomes a column of returns, one shorter.
  3. Compute the average of those returns. Over a short window it will be very small; you may assume it is zero, and most estimators do.
  4. For each day, subtract the average and square the result. Add all the squares together.
  5. Divide the sum by the number of returns minus one, then take the square root. This is the daily volatility.
  6. Multiply the daily volatility by √252 ≈ 15.87 to annualise it. Report it as a percentage.
  7. Sanity-check the scale: an annualised figure near 13% for NIFTY is ordinary; one near 30% signals genuine stress; a negative or zero figure means an error in the arithmetic, because a standard deviation cannot be negative.

Practical example

NIFTY worked example

Suppose over a recent stretch NIFTY's daily log returns had a sample standard deviation of 0.82%. Annualise it: 0.82% × √252 = 0.82% × 15.87 = 13.0%. That is a perfectly ordinary NIFTY reading, and it sits right on top of where India VIX typically lives, around 13. Now translate it back into something you can feel. A 13.0% annual volatility implies a one-day, one-standard-deviation move of 13.0% ÷ √252 = 0.82% of spot, which at 24,000 is about 197 points. Two standard deviations is about 394 points. If daily returns were normal, roughly one day in three would close outside the ±197-point band and about one day in twenty outside the ±394-point band. Here is the part to sit with: the 13.0% says nothing about direction. It is equally consistent with NIFTY drifting from 24,000 up toward 27,120 over a year and with it sliding down toward 20,880 — a one-standard-deviation annual band of ±3,120 points around 24,000, with the sign left entirely open.

BANKNIFTY worked example

Run the same machine on BANKNIFTY and the lesson is about comparability. Suppose BANKNIFTY's daily log returns show a standard deviation of 1.05%. Annualised, that is 1.05% × 15.87 = 16.7% — meaningfully above NIFTY's 13.0%. It is tempting to say BANKNIFTY is 'riskier', but the honest reading is narrower: BANKNIFTY is a concentrated basket of a single sector, so it genuinely realises larger daily moves, and 16.7% is simply the correct measurement of a more volatile thing. In points, a one-standard-deviation BANKNIFTY day is 1.05% of 52,000 ≈ 546 points, versus NIFTY's 197. The volatilities are directly comparable because both are annualised percentages of returns — that is the entire reason the annualisation convention exists — but a higher number does not by itself make one instrument a worse holding than the other. It makes it a bigger-stepping one.

Lot sizes used above (NIFTY 75, BANKNIFTY 30) are those in force at the time of writing; NSE revises them periodically. Figures exclude brokerage, STT, exchange charges, stamp duty and GST. Examples are teaching scenarios built on round numbers — they are not historical quotes, not backtests and not trade calls.

Risk note. Because volatility is annualised and normalised, it is easy to read a low reading as a statement that the market is safe. It is not. A low volatility says recent daily moves have been small; it says nothing about the size of the move that ends the calm, and the historical record is that the lowest-volatility regimes are the ones in which leverage and complacency accumulate before the largest reversals. The normal-distribution maths that produces tidy one- and two-standard-deviation bands under-counts exactly those extreme days, so any risk limit that trusts those bands as hard boundaries is under-provisioned against the events that actually break accounts.

Advantages & limitations

What it is good for

  • It reduces an entire price path to a single, comparable number. One annualised percentage lets you compare NIFTY against BANKNIFTY, this month against last, and a stock against an index, on the same axis.
  • It is directional-noise-free by design. By squaring away the sign, it isolates the magnitude of movement — precisely the quantity that options, which are bets on movement, are priced from.
  • It scales predictably across horizons under its own assumptions. The square-root-of-time rule lets you move between a daily, monthly and annual figure with one multiplication, which is what makes expected-move calculations possible.
  • It is the foundation every other volatility measure is built on. Historical, realised, implied, expected and forward volatility are all the same standard-deviation idea applied to different data, so understanding this one number unlocks the whole family.
  • It is computable from nothing but a column of closing prices, which are free and public, so anyone can reproduce it and check the site's arithmetic against their own.

Where it breaks down

  • It says nothing about direction. A volatility figure is equally consistent with a market about to rise and one about to fall, so using it as a bull or bear signal is a category error, not a subtle mistake.
  • It assumes a normal distribution that real returns violate. The one- and two-standard-deviation bands under-count extreme days, so the measure is least trustworthy in precisely the tail events that cause the largest losses.
  • It is window-dependent and backward-looking when computed from history. The same series gives different volatilities for a 10-day and a 60-day window, and none of them is a promise about tomorrow.
  • It treats all dispersion as equivalent and cannot see skew. A market that lurches symmetrically and one that falls faster than it rises can share a volatility while carrying entirely different shapes of risk.
  • Its annualisation assumes day-to-day independence, which volatility clustering breaks. When calm and storm arrive in regimes, the √252 scaling is a convention rather than an accurate extrapolation.
  • It is a measure of expected wobble, not of catastrophe. Leverage, illiquidity and overnight gaps can destroy a position whose measured volatility was low the whole way in.

Common mistakes

  • Reading a rising volatility as a bearish signal. In equity indices high volatility usually coincides with falling prices, but that correlation lives in the skew and in hedging demand, not in the volatility number, which is sign-blind and would report the same value in a violent rally.
  • Comparing a daily figure to an annual one without scaling. A 0.82% daily move and a 13% annual volatility are the same fact expressed on two horizons; treating 0.82% as 'low volatility' next to 13% double-counts the horizon and makes the market look calmer than it is.
  • Annualising by multiplying the daily figure by 252 instead of √252. That inflates the number by a factor of nearly sixteen, turning an ordinary 13% into an absurd 207%, and it is the single most common arithmetic error in volatility work.
  • Trusting the two-standard-deviation band as a hard boundary. Because real tails are fat, days outside the 95% band arrive far more often than the normal maths predicts, so a stop or a risk limit placed exactly there will be breached more than the model claims.
  • Treating a stretch of low volatility as evidence the market is safe. Low readings measure recent calm, not future safety, and volatility clustering means the quiet is exactly the setup for the storm that ends it.
  • Confusing the volatility of the price with the volatility of returns. Standard deviation of raw prices grows mechanically as the index level grows and is not comparable across time or instruments; volatility is always defined on returns for exactly that reason.
  • Assuming a single low-volatility asset makes a portfolio safe. Correlation, not just individual volatility, drives portfolio risk, and assets that look calm in isolation can move together violently when it matters.

Professional usage

On a trading desk, volatility is never the destination — it is the input to everything else. A market maker prices options off a volatility surface whose backbone is an at-the-money volatility of this kind, and a portfolio risk manager feeds volatility estimates into value-at-risk and margin models to size how much capital a book must hold against a normal day. The sophistication is in knowing the assumptions: professionals scale volatility by √252 fully aware that clustering makes the scaling approximate, and they overlay fat-tailed or regime-switching models precisely because they know the normal-distribution version under-provisions the tails. The number is treated as a well-understood convention rather than a truth, and the entire craft lies in remembering where the convention lies to you.

Quantitative research desks decompose realised volatility into its components — a slow-moving level, a clustering term and jumps — and forecast each separately, because a volatility that is high because a scheduled event sits in the window is a different animal from one that is high because the market has entered a stressed regime. Asset allocators use volatility to build risk-parity and volatility-targeting portfolios that lever up when measured volatility is low and de-lever when it rises; the uncomfortable footnote, which the best of them state openly, is that such strategies mechanically sell into the exact fat-tailed drawdowns that the volatility measure under-counted on the way in.

Key takeaways

  • Volatility measures the size of a market's typical move — the journey — and says nothing about the direction or the destination. Two paths with the same start and end can have completely different volatilities.
  • It is the annualised standard deviation of returns: compute a daily standard deviation from log returns, then multiply by √252 ≈ 15.87 to annualise. The square root, not a plain multiply, is what trips people up.
  • For NIFTY, an annualised volatility near 13% is ordinary and corresponds to roughly a 197-point one-standard-deviation day at 24,000. The figure is directly comparable across instruments because it is a percentage of returns.
  • The maths assumes a normal distribution, but real returns have fatter tails, so the one- and two-standard-deviation bands under-count the extreme days that cause the largest losses.
  • Volatility clusters and mean-reverts, so a calm reading understates the risk of the storm that ends it, and a high reading can still rise before it falls.

If you take one idea from this page, take this: volatility is the size of the step, not the direction of the walk. Every other volatility concept on this site — historical, realised, implied, the VIX, the skew, the whole surface — is this single standard-deviation idea applied to different data and dressed in different clothes. Master the front door and the rest of the building is furniture. And carry the caveat with you as carefully as the definition: the tidy bell-curve maths that makes volatility so usable is precisely what makes it under-count the rare, violent days, so treat the number as an indispensable first estimate and never as the last word on how bad things can get.

Frequently asked questions

What is volatility in simple terms?
Volatility is how much a market moves, expressed as an annualised percentage. If NIFTY typically swings about 0.8% on an ordinary day, that daily spread scaled up to a year is its volatility. It describes the size of the moves, not their direction, so a market that rises steadily and one that crashes and recovers can share the same volatility.
Is volatility the same as risk?
No — volatility is one component of risk, not all of it. It measures the size of ordinary price moves, but it cannot see leverage, illiquidity, correlation or an overnight gap that no standard deviation anticipated. A position can carry low volatility and enormous risk, which is why volatility is a starting point for risk work rather than the whole of it.
Does volatility tell me which direction the market will move?
No, and this is its defining property. The calculation squares each return before averaging, so a +2% day and a −2% day contribute identically. A volatility of 13% is equally consistent with a market about to rise sharply and one about to fall sharply — it quantifies magnitude and is completely silent on direction.
Why is volatility annualised?
So that different numbers can be compared on one scale. A daily figure, a monthly figure and an option's implied volatility would otherwise be incomparable. By convention everything is scaled to an annual standard deviation, which is why a NIFTY volatility of 13% and an implied volatility of 15% can sit side by side and mean the same kind of thing.
Why multiply daily volatility by the square root of 252 rather than by 252?
Because variance, not volatility, grows linearly with time when returns are independent. A year's variance is 252 daily variances added together, and volatility is the square root of variance, so a year's volatility is √252 ≈ 15.87 times a day's. Multiplying by 252 instead would overstate the figure roughly sixteen-fold.
Why 252 and not 365?
Because volatility is built from trading days, and markets are shut on weekends and holidays — prices do not move when the exchange is closed. An Indian market year contains roughly 252 trading sessions, so 252 is the correct annualisation count for a figure computed from daily returns. Note that time-to-expiry in option pricing uses calendar days over 365, a separate convention.
Is high volatility bad?
It depends entirely on your position. High volatility is good for an option buyer, who owns movement, and bad for an option seller, who is short it. For a leveraged holder it means larger swings in both directions. There is no version of 'is high volatility bad' that has a general answer, because volatility is directionless.
What is a normal level of volatility for NIFTY?
As a rough guide, NIFTY's annualised volatility spends much of its time between about 11% and 16%, dips under 11% in the calmest stretches, and exceeds 25% only in genuine stress. India VIX, which summarises the option chain over a 30-day window, tracks the same range. These are conventions drawn from history, not rules.
What is the difference between volatility and standard deviation?
Volatility is a standard deviation — specifically, the annualised standard deviation of returns. Standard deviation is the general statistical measure of dispersion around a mean; volatility is that measure applied to an asset's returns and scaled to a year. So every volatility is a standard deviation, but not every standard deviation is a volatility.
Why use log returns instead of simple percentage returns?
Because log returns add up cleanly across time and treat a rise and the fall that reverses it symmetrically. A +10% followed by a −9.09% returns you to the start, and only log returns make those two cancel exactly. Over short daily windows the difference from simple returns is tiny, but log returns are the cleaner foundation for the maths.
Can volatility be negative?
No. Volatility is a standard deviation, which is the square root of a variance, and neither can be negative. If a calculation returns a negative volatility, there is an arithmetic error somewhere — a standard deviation floors at zero, which is reached only if every return in the window was identical.
Does volatility measure how fast the price moves or how far?
How far, per unit of time. It captures the typical size of a move over a fixed interval — a day, then annualised — rather than the speed of any single tick. A market that makes one large jump and then sits still can have the same measured volatility as one that fidgets constantly, if their return dispersions match.
What does a 13% annualised volatility mean in NIFTY points?
It implies a one-standard-deviation daily move of 13% ÷ √252 = 0.82% of spot, which at 24,000 is about 197 points. Over a full year the one-standard-deviation band is 13% of 24,000, about 3,120 points, in either direction. The point figure makes the abstract percentage something you can actually feel on a position.
Why do the maths assume a normal distribution if returns are not normal?
Because the normal distribution is mathematically convenient and roughly right in the middle, where most days live. The problem is the tails: real returns produce extreme days far more often than the bell curve allows. The assumption is kept for tractability and corrected for separately, but it means the standard bands under-count exactly the moves that cause the biggest losses.
What are fat tails?
Fat tails describe a distribution in which extreme outcomes are more likely than a normal curve predicts. Under a normal distribution a five-standard-deviation day should almost never occur; equity markets deliver them within a single career. Fat tails are why volatility, which is built on the normal assumption, systematically underestimates the probability of a crash.
Is volatility constant over time?
No — it varies, and it varies in a structured way. Volatility clusters, so calm follows calm and turbulence follows turbulence, and over longer horizons it mean-reverts toward a long-run average. This is why the same asset can read 11% one month and 30% the next, and why forecasting volatility is a serious discipline rather than a lookup.
What is volatility clustering?
Volatility clustering is the empirical tendency of large moves to be followed by large moves and small moves by small moves, regardless of sign. It means volatility arrives in regimes rather than as independent noise, which both makes short-horizon volatility forecastable and undermines the independence assumption behind the √252 annualisation.
How is volatility different from beta?
Volatility measures an asset's total variability in isolation; beta measures how much of that variability moves with a benchmark like NIFTY. A stock can have high volatility and low beta if most of its movement is idiosyncratic. Volatility is a standalone standard deviation; beta is a relationship between two return series.
Can two assets with the same return have different volatility?
Yes, and it is the whole point of the concept. Two assets can both return 8% over a year while one glided there and the other whipsawed violently the entire way. Same return, different volatility — the dispersion-basic chart on this page shows exactly this with two NIFTY paths that share a start and an end.
Does higher volatility mean higher returns?
Not reliably. There is a long-run argument that riskier assets should offer higher expected returns, but volatility and realised return are separate quantities, and a high-volatility asset can and does deliver poor returns. Volatility is explicitly constructed to be blind to return, so it can never be read as a promise of one.
What is the difference between volatility and variance?
Variance is the average of the squared deviations of returns from their mean; volatility is the square root of variance. Variance is what adds linearly across time, which is why annualising uses variance internally, but volatility is the quantity actually quoted because it is in the same units as the returns themselves — a percentage.
Is volatility a percentage or a number of points?
It is quoted as an annualised percentage, but it converts directly into points once you fix an underlying and a horizon. A 13% NIFTY volatility becomes about 197 points for a one-day, one-standard-deviation move at 24,000. The percentage is the portable form; the point figure is what you feel on a specific position.

Voice search & related questions

Natural-language questions people ask about what is volatility?.

What does volatility actually measure?
Volatility measures how big a market's typical move is, scaled to a year and expressed as a percentage. It is a standard deviation of returns — the spread of the daily ups and downs — and it deliberately ignores which direction those moves went.
Why should I care about volatility if I only care about direction?
Because volatility decides how much your position swings while you wait to be right. Two markets can end at the same place, but the volatile one will have tested your stops, your margin and your nerve on the way there. If you use any leverage, the journey can end you before the destination arrives.
How do I turn a daily move into an annual volatility number?
Take the standard deviation of the daily log returns and multiply by the square root of 252, which is about 15.87. So a daily spread of 0.82% becomes an annual volatility of roughly 13%. The square root, not a plain multiply, is the step people get wrong most often.
Is a volatile market a falling market?
Not necessarily. Volatility is sign-blind, so a market rocketing upward in large daily jumps is exactly as volatile as one crashing in jumps of the same size. In equity indices high volatility does tend to coincide with falling prices, but that link lives in the skew and in hedging demand, not in the volatility figure itself.
If volatility says nothing about direction, what use is it?
It tells you how much to expect to be moved around, which is what you need to size a position, price an option, or set a realistic target and stop. Options in particular are bets on movement rather than direction, so a directionless measure of movement is precisely the right tool for pricing them.
Why does everyone use 252 days?
Because that is roughly how many days the Indian market is actually open in a year, once you strip out weekends and holidays. Volatility is built from trading days, so 252 is the right count for annualising a figure computed from daily returns. It is a market-days convention, distinct from the 365 calendar days used inside option pricing.
Can a calm market suddenly become volatile?
Yes, and the calm is often the setup. Volatility clusters, so quiet periods lull positioning and leverage into building up, and the eventual break is violent precisely because so much was resting on the calm continuing. A low volatility reading measures the recent past, not the safety of the near future.

Sources & references

Last reviewed 10 July 2026. Educational content only — not investment advice.

← Previous
Volatility Clustering
Educational content only — not investment advice. Every diagram on this page is generated from the site's own model, using illustrative inputs rather than live quotes. Options and futures carry substantial risk, including loss exceeding your deposit on short-volatility positions. See our Risk Disclosure and SEBI Disclaimer.