Home / Market Regimes / Crisis Volatility
Market Regimes Advanced Volatility behaviour in a market dislocation Contemporaneous (confirmed only once underway)

Crisis Volatility

Positions sized for the calm period are, by construction, the wrong size for the ten days that follow.

Quick answer: Crisis volatility is the behaviour of a market in dislocation — the term structure inverts into backwardation, skew goes vertical, correlation across constituents goes to one, and margin rises mechanically — all driven by a sudden collapse in the willingness to sell insurance that rebuilds far more slowly than it evaporated.

In simple words

Volatility goes up by the escalator and down by the stairs. The rise takes days; the decay back toward normal takes months. Suppose India VIX sits near 13 for a long quiet stretch and then, over a single week, prints 30 and then 45. That jump is the escalator. Getting back down to 13 will not take a week — it will take the better part of a season, drifting lower a point at a time. The reason is not statistical, it is structural: a crisis is a sudden collapse in the willingness to sell insurance. Everyone wants to own protection and almost nobody wants to write it, so its price gaps up. Rebuilding the willingness to sell insurance — convincing dealers and funds that the danger has passed and it is safe to underwrite again — is slow, cautious human work that no single day can complete.

The practical consequence is brutal and simple. A position sized during the calm — a short strangle that collected premium every quiet day, a put-selling program that looked effortless for a year — is, by construction, the wrong size for the ten days that follow. Nothing about it changed except the world it lives in. The same number of lots that felt conservative at India VIX 12 is reckless at 40, and the market does not give you a session to adjust between the two. This is why the danger of a short-volatility book is never visible in its track record. The track record is the calm. The risk is the transition, and the transition is not in the sample until it is the only thing in the sample.

Not to be confused with: A merely stressed or elevated reading. India VIX at 20 is stressed by convention but the machinery of a crisis — backwardation, vertical skew, correlation at one, a live margin feedback loop — need not be present. A crisis is a change in the market's plumbing, not just a high number on the volatility index. You can have a high reading without the plumbing failing, and, more dangerously, the plumbing can begin to fail while the headline number still looks merely elevated.

The escalator up and the stairs down

The rise takes days; the decay takes months

A stylised volatility path through a dislocation — a near-vertical spike followed by a long, halting descent.

daysmonths10%20%30%40%50%0d50d100d150d200dthe crisis beginsvolatility goes up by the escalator and down by the stairsTrading dayImplied volatility
The asymmetry in this picture is the whole concept. The up-move and the down-move are not mirror images and were never going to be, because the thing that collapses in a crisis — the willingness to write insurance — evaporates in a day and rebuilds over a season. Any risk model that assumes volatility rises and falls symmetrically is calibrated to a market that does not exist.

Professional explanation

The asymmetry is structural, not statistical

The single most important fact about a crisis is that volatility rises far faster than it falls, and the reason is mechanical rather than a quirk of the data. A crisis is a sudden collapse in the willingness to sell insurance. When a shock hits, everyone who is short protection wants to be flat and everyone who is unhedged wants to be hedged, so demand for options spikes while the supply of new sellers vanishes — nobody wants to write a fresh put into a falling, accelerating market. Price is the only mechanism that clears that imbalance, so implied volatility gaps higher in days. The reverse is slow because it requires the opposite: it requires participants to become willing to underwrite risk again, and that willingness returns cautiously, position by position, desk by desk, as confidence rebuilds. Demand can vanish in an afternoon. Supply has to be coaxed back over months. That is the escalator and the stairs, and it is why every symmetric mean-reversion assumption about volatility is wrong in exactly the regime where being wrong is expensive.

The term structure inverts into steep backwardation

In an ordinary market the volatility term structure is in contango: near-dated options imply a slightly lower volatility than far-dated ones, because on any given quiet day not much is expected but the future holds unknown events. A crisis flips this. The danger is immediate, so the front expiry — the one that has to survive the next few sessions — is bid to a far higher implied volatility than the deferred expiries, and the curve slopes downward from front to back. This is backwardation, and it is steep. The practical reading is uncomfortable for anyone who trades on the rule of thumb that high implied volatility means options should be sold: in backwardation the front-month volatility you would be selling is high precisely because the realised movement over that front month is likely to be enormous, and the curve is telling you the market agrees the near term is the dangerous part. Backwardation is also one of the few crisis diagnostics available in real time rather than after the fact — the inversion is visible the moment it happens.

Skew goes vertical and liquidity leaves the wings

In calm, the volatility skew slopes upward toward the downside — out-of-the-money puts imply higher volatility than at-the-money strikes, because index crashes are one-sided. In a crisis that slope goes vertical: the put wing is bid at almost any price, because the marginal buyer is not expressing a view, they are covering a liability or meeting a mandate, and price sensitivity disappears when you are forced to buy. At the same time — and this is the part that traps people — liquidity leaves the wings entirely. Market makers widen or pull their quotes on far strikes because they cannot hedge them, so the screen shows a price for a deep put that no size will actually trade at. An implied volatility extracted from such a quote is meaningless: it is the inversion of a number that is not a real, transactable price. The cruel timing is that this happens exactly when a risk manager most wants to read the wings, to gauge how much tail the market is pricing — and the wings have stopped carrying information at the moment they matter most.

Correlation goes to one and diversification stops working

Diversification and dispersion strategies rest on the constituents of an index not all moving together. In a crisis they do. When the shock is systemic — a liquidity event, a funding squeeze, a forced-deleveraging spiral — every name is sold at once because the selling is not about any individual company, it is about raising cash, and the cross-sectional correlation across constituents runs toward one. The mathematics is unforgiving: portfolio volatility that was suppressed by low pairwise correlations reverts to the simple weighted average of the individual volatilities, so the diversification benefit that made a book look safe evaporates in the same session the book most needs it. A dispersion trade — short index volatility, long single-name volatility, betting the components move independently — is a short correlation position, and it takes its largest loss at the exact moment correlation goes to one. Diversification and dispersion both stop working precisely when they are needed, which is not a coincidence but the definition of the regime.

The margin feedback loop is a real mechanism, not a metaphor

Exchange margin — SPAN and the exposure/ELM add-ons in the Indian framework — is computed from recent volatility, so when volatility rises, the margin required to hold the same position rises with it, mechanically and automatically. A trader who was fully deployed in the calm now faces a margin call they did not choose and cannot always meet, and the only way to reduce the requirement is to cut the position. But cutting the position means buying back short options and selling underlying into a market that is already falling, which pushes volatility higher still — which raises the margin again, on everyone holding similar positions, at the same time. This is a genuine feedback loop, not a figure of speech: forced exits beget higher volatility beget higher margin beget more forced exits. It is why crises accelerate rather than merely happen, and why the worst of the move often comes after the initial shock, as the deleveraging spiral feeds on itself. Position sizing done against calm-period margin quietly assumes this loop will never start.

Implied volatility can trade below realised — and that is when the year's gains disappear

In normal conditions implied volatility sits a little above subsequently realised volatility; that gap is the volatility risk premium and it is why writing options is compensated across a full cycle. In a crisis the relationship can invert. When the market is moving several percent a day, the volatility actually being realised over the next few sessions can exceed even the elevated implied volatility on the screen — implied volatility trades below realised. This is the precise moment a short-volatility position that looked safe for a year loses most or all of what it made, because the option it is short is not merely expensive, it is cheap relative to the movement now arriving. The seller is being paid a crisis-level premium that is nonetheless insufficient for the crisis-level movement. A year of collecting small, reliable premiums is arithmetically undone by a handful of sessions in which the premium collected was, in hindsight, a fraction of the loss incurred — and no amount of prior track record changes that arithmetic, because the prior track record was compiled in a different regime.

The term structure flips upside down

Near-dated versus far-dated at-the-money implied volatility, calm regime versus crisis.

15%20%25%30%7d30d60d90d120d180dpanic is priced into the FRONTthe market expects calm to returndownward slope = backwardation = the market is frightened NOWDays to expiryImplied volatility
In calm the curve slopes gently upward — the front month is cheaper than deferred months, ordinary contango. In a crisis it inverts: the front month is bid far above the back, because the danger is now, not later. That inversion — backwardation — is one of the few crisis signals that is contemporaneous rather than lagging, and it is why selling 'expensive' front-month volatility into a dislocation is selling the steepest, most dangerous part of the curve.

Formula

Why diversification fails: portfolio variance as correlation goes to one

σ²_p = Σ wᵢ² σᵢ² + Σᵢ≠ⱼ wᵢ wⱼ ρᵢⱼ σᵢ σⱼ → as ρᵢⱼ → 1, σ_p → Σ wᵢ σᵢ

In calm markets the cross terms are damped by low pairwise correlations ρᵢⱼ, so portfolio volatility σ_p sits well below the weighted average of the constituent volatilities — that gap is the diversification benefit. In a crisis every ρᵢⱼ runs toward one and the whole expression collapses to the weighted average σ_p → Σ wᵢ σᵢ, meaning the diversification benefit disappears entirely at the exact moment it is needed. The reduction in risk you were relying on was a function of correlations that a crisis removes.

  • σ_pVolatility (standard deviation) of the whole portfolio or index.
  • wᵢWeight of constituent i in the portfolio; the weights sum to one.
  • wⱼWeight of a different constituent j; the double sum runs over all pairs i ≠ j.
  • σᵢVolatility of constituent i.
  • σⱼVolatility of constituent j.
  • ρᵢⱼPairwise correlation between the returns of constituents i and j. Low in calm markets, close to one in a crisis.

The escalator-and-stairs asymmetry, stated as a rule of thumb

time to spike ≈ days time to decay back to normal ≈ months

Not an equation to solve but the asymmetry that defines the regime. The rise is fast because demand for insurance can spike in an afternoon; the fall is slow because the willingness to supply insurance rebuilds over a season. Any model that treats the up-path and down-path of volatility as symmetric is calibrated to a market that has never existed.

How to read whether a market is in a genuine crisis regime

  1. Check the term structure, not just the level. Compare front-month at-the-money implied volatility with the next two expiries. A high level with the curve still upward-sloping is stress; a high level with the front month bid above the back — backwardation — is a crisis signal.
  2. Look at the skew's steepness, not only the at-the-money number. When the downside put wing is bid vertically relative to at-the-money, the marginal buyer is hedging a liability rather than expressing a view, which is the signature of forced demand.
  3. Sanity-check the wings for real liquidity. If far-strike quotes are wide and no size trades at them, treat any implied volatility extracted from those quotes as meaningless — it is the inversion of a price that does not exist.
  4. Watch cross-sectional correlation. If constituents that normally move somewhat independently are all falling together, diversification and dispersion positions are losing their basis for existing, right now.
  5. Recompute your margin at current, not trailing, volatility. Ask what SPAN and the exposure add-ons require on your book today, and whether a further doubling of volatility would force you to reduce — because if it would, the market may reduce it for you.
  6. Compare implied against the volatility actually realising over the last few sessions. If realised has caught up to or exceeded implied, the premium being collected by short options is no longer compensating for the movement, and 'high IV, therefore sell' has stopped being true.

Practical example

NIFTY worked example

Take NIFTY at 24,000 with India VIX around 13 in a calm stretch: the 30-day at-the-money implied volatility is roughly 13%, which prices a daily move of about 24,000 × 0.13 ÷ √252 ≈ 196 points. Now the market dislocates and the same 30-day at-the-money implied volatility gaps to 40%. That now prices a daily move of 24,000 × 0.40 ÷ √252 ≈ 605 points — roughly three times as large — and the term structure has inverted so the front-week implied volatility is even higher than the 30-day. A short strangle that was collecting a few hundred rupees a day against 196-point sessions is now exposed to 605-point sessions, and if the market actually realises 45% over the coming week while implied sits at 40%, the position is short an option that is cheap relative to the movement arriving. Interpret it this way: nothing about the strangle changed — the strikes are where they were and the number of lots is unchanged — but the world it lives in now moves three times as far each day, and the margin required to hold it has expanded in lockstep. The position did not become risky. It was always this risky; the calm was hiding it.

BANKNIFTY worked example

BANKNIFTY teaches the correlation lesson more sharply than NIFTY, because it is a concentrated single-sector index. With BANKNIFTY at 52,000 and a crisis implied volatility of 50%, the priced daily move is 52,000 × 0.50 ÷ √252 ≈ 1,638 points. But the deeper point is what happens across its constituents. In calm, the individual banks in the index do not move perfectly together, so the index volatility sits below the weighted average of the single-bank volatilities — that gap is the diversification benefit baked into every BANKNIFTY risk estimate. In a funding or liquidity crisis, all the banks are sold at once for the same reason — raising cash — so their pairwise correlations run toward one and the index volatility jumps toward the full weighted average of the constituents. A dispersion trader who was short BANKNIFTY volatility and long the single-bank options, betting the banks would move independently, is short exactly the correlation that a crisis sets to one, and takes the largest loss at the worst possible moment. The lesson NIFTY teaches through the term structure, BANKNIFTY teaches through correlation: the safety was borrowed from a regime, and the crisis calls the loan.

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. Crisis volatility is routinely used to justify the most dangerous trade in the book: selling options because implied volatility is high. In a dislocation, high implied volatility is not evidence that options are overpriced — the term structure is in backwardation and implied volatility can trade below the volatility actually being realised, which means the 'expensive' front-month option you would sell is cheap relative to the movement coming. Short-volatility positions carry theoretically unlimited loss, margin expands mechanically against you as volatility rises, and the liquidity you would need to exit is gone from the wings at exactly the moment you want it. A high reading in a crisis is the market pricing danger correctly, not mispricing it in your favour. Nothing on this page is a recommendation to sell volatility into a dislocation.

Advantages & limitations

What it is good for

  • Backwardation in the term structure is one of the few crisis diagnostics that is contemporaneous rather than lagging — the front-month-above-back-month inversion is visible the moment the regime changes, not confirmed weeks later like a realised-volatility number.
  • Understanding the mechanism lets you size in advance for the ten days that matter, rather than discovering the wrong size mid-dislocation. The concept's value is that it is knowable before the crisis, even though the timing is not.
  • The margin feedback loop is predictable in direction even when unpredictable in timing, so a book can be pre-funded, pre-hedged, or held below the size at which a volatility doubling would force an exit — the loop is a design constraint you can respect ahead of time.
  • Correlation-going-to-one tells you which hedges will actually work in a dislocation (broad index puts, which pay off precisely because everything falls together) and which will not (dispersion and relative-value hedges, which rely on the independence that vanishes).
  • The realised-versus-implied crossover is a concrete, checkable test of whether 'high IV means sell' has stopped being true — it turns a vague sense of danger into an arithmetic comparison a trader can actually run.

Where it breaks down

  • A crisis regime can only be confirmed once it is already underway. Backwardation, vertical skew and correlation-to-one are contemporaneous signals, not forecasts, so the concept describes the storm but does not ring the bell before it — the transition is precisely the part you cannot time.
  • The wing implied volatilities you would most want to read in a crisis are the least reliable, because liquidity has left the far strikes and the quotes there are not transactable. The measurement tools degrade in exactly the regime they are meant to measure.
  • 'Crisis' is a label applied after the fact. A stressed reading of 20 is not proof that the plumbing has failed, and the plumbing can begin to fail while the headline volatility index still looks merely elevated, so the boundary is genuinely fuzzy and any threshold is a convention.
  • Reasoning by analogy to previous crises misleads, because each dislocation has different plumbing — a funding squeeze, a policy shock and a forced-deleveraging spiral produce different term-structure and correlation signatures, so a playbook tuned to the last one can be exactly wrong for the next.
  • The mechanism describes index and liquid-underlying behaviour. In illiquid single stocks the same shock can produce gapped, discontinuous prices where no continuous volatility number describes what happened at all, and the framework quietly assumes a market that is at least trading.

Common mistakes

  • Selling front-month volatility because it looks 'expensive' in a dislocation. In backwardation the front month is high because the near-term movement is expected to be enormous, and selling it is selling the steepest, most dangerous part of the curve into forced demand — the consequence is a short option that is cheap relative to what actually arrives.
  • Trusting Greeks or implied volatilities computed from far-strike quotes. In a crisis the wings are wide and untradeable, so a delta, a vega or an implied volatility extracted from those prices is the inversion of a number that is not real — hedging on it puts on a hedge sized to a fiction.
  • Assuming diversification and hedge ratios estimated in calm still hold. Correlations that made a book look safe run toward one in a crisis, so the diversification benefit and the relative-value hedge both evaporate in the same session — the trader discovers the portfolio was one position wearing many tickers.
  • Sizing against calm-period margin. When volatility rises, SPAN and the exposure add-ons rise mechanically, and a fully deployed book gets a margin call it did not choose — being forced to exit at the worst price is a direct, avoidable consequence of sizing to the calm.
  • Treating 'implied volatility exceeds realised' as a law rather than a tendency. In a crisis it can invert, and a short-volatility position built on the average relationship gives back a year of premium in a handful of sessions where the relationship reversed.
  • Rolling short puts 'down and out' to repair a losing position. Adding strikes and duration in a falling market increases the very exposure that is bleeding, and feeds the deleveraging loop — the repair is often a larger version of the original mistake, executed with less margin to survive it.

Professional usage

Volatility desks do not wait for a crisis to think about crisis volatility; they carry the regime in their risk from the start. A dealer running a short-gamma book from customer flow knows that the escalator-and-stairs asymmetry means their worst day is a gap, so they hold long-tail protection — deep out-of-the-money puts or variance — as a standing cost of doing business, accepting a small persistent bleed to avoid being the forced buyer when the wings go bid. A risk manager stress-tests the book with correlations pinned to one rather than to their calm-period estimates, because a value-at-risk figure built on historical correlations is, by construction, blind to the regime in which those correlations disappear. And a treasury or margin desk models the feedback loop explicitly: they ask not 'what is my margin today' but 'what is my margin if volatility doubles and I am one of many being asked to reduce at once', and they hold the book below the size at which that scenario forces a liquidation.

On the trading side, the backwardation of the term structure is itself the signal that repositions a book. When the curve inverts, a relative-value desk that was long the roll — collecting the contango carry of selling front and buying back — is now facing the opposite structure and either flattens or reverses, because the roll-yield mechanics that were a tailwind for months become a headwind in days. Dispersion desks treat correlation-to-one as their defining tail: they know the trade is a short-correlation position and they size the single-name-long, index-short structure so that a correlation shock to one is survivable rather than terminal, because the entire edge of the strategy is collected in calm and paid back in exactly the dislocation this page describes.

Key takeaways

  • Volatility rises by the escalator and falls by the stairs: the spike takes days and the decay takes months, because the willingness to sell insurance collapses in an afternoon and rebuilds over a season. The asymmetry is structural, not statistical.
  • In a crisis the term structure inverts into backwardation, skew goes vertical, liquidity leaves the wings, and correlation across constituents runs toward one — diversification and dispersion stop working at the exact moment they are needed.
  • Margin (SPAN and the exposure add-ons) rises mechanically with volatility, so forced exits push volatility higher, which raises margin again — a genuine feedback loop that makes crises accelerate.
  • Implied volatility can trade below realised in a dislocation, which is the moment a short-volatility position that looked safe for a year loses most of what it made. High IV in a crisis is a correct price for danger, not a mispricing in your favour.
  • A position sized for the calm is, by construction, the wrong size for the ten days that follow. The risk of a short-volatility book lives in the regime transition, which is never in the track record until it is the only thing in it.

Crisis volatility is the part of the subject that a track record cannot teach, because the track record is the calm and the lesson is the transition. If you take one thing from this page, take the escalator and the stairs: the market can triple the daily move it prices in a week and then take a season to give it back, and any position, model or margin assumption that treats those two paths as symmetric is calibrated to a market that has never existed. You cannot time the transition — backwardation confirms the crisis only once it has begun — but you can size for it in advance, hold the tail protection that costs you a little every quiet day, and refuse the size at which a doubling of volatility would let the market liquidate you at the worst price. The uncomfortable truth a marketing department would cut is that the strategies which look most effortless in the calm are the ones built to fail here, and their success right up to the edge is not evidence against that — it is the mechanism of it.

Frequently asked questions

What is crisis volatility?
Crisis volatility is how a market behaves in a dislocation: implied volatility spikes over days, the term structure inverts into backwardation, skew goes vertical, correlation across constituents runs toward one, and margin rises mechanically. It is driven by a collapse in the willingness to sell insurance, which rebuilds far more slowly than it vanished.
Why does volatility rise faster than it falls?
Because a crisis is a sudden collapse in the willingness to write insurance, and demand for protection can spike in an afternoon while the supply of new sellers has to be coaxed back over months. Clearing the demand imbalance requires an immediate price jump; rebuilding the willingness to underwrite risk is slow, cautious human work. That is the escalator up and the stairs down.
What is backwardation in the volatility term structure?
Backwardation is when near-dated implied volatility is bid above far-dated implied volatility, so the curve slopes downward from front to back. It is the opposite of the usual contango and signals that the market sees the danger as immediate rather than distant. It is one of the few crisis diagnostics visible in real time.
Why does skew go vertical in a crisis?
Because the marginal buyer of downside puts is covering a liability or meeting a mandate rather than expressing a view, and price sensitivity disappears when you are forced to buy. The put wing is bid at almost any price, so the skew — already sloped toward the downside in calm — steepens toward vertical.
Why does an implied volatility from a far strike become meaningless in a crisis?
Because liquidity leaves the wings: market makers widen or pull quotes on far strikes they cannot hedge, so the screen shows a price no real size will trade at. An implied volatility inverted from an untradeable quote is the inversion of a number that is not a real price, and using it to compute Greeks hedges a fiction.
Why does correlation go to one in a crisis?
Because systemic selling is not about any individual name — it is about raising cash — so every constituent is sold at once and their pairwise correlations run toward one. Diversification benefit is a function of low correlations, so it evaporates in the same session the portfolio most needs it.
What is the margin feedback loop?
Exchange margin is computed from recent volatility, so when volatility rises the margin to hold the same position rises mechanically. Traders forced to cut buy back short options and sell underlying into a falling market, pushing volatility higher, which raises margin again on everyone at once. Forced exits beget higher volatility beget higher margin — a genuine loop, not a metaphor.
Can implied volatility be below realised volatility?
Yes, and a crisis is the classic case. When the market is moving several percent a day, the volatility actually realising over the next sessions can exceed even the elevated implied volatility on the screen. That is the moment a short-volatility position is short an option that is cheap relative to the movement arriving.
Why is a short-volatility book most dangerous after a long calm?
Because position size tends to grow in inverse proportion to measured volatility — every quiet day makes yesterday's size look timid — so the book is largest exactly when realised volatility is lowest and the regime is closest to turning. The calm builds the exposure; the transition expresses it.
Does high implied volatility in a crisis mean options are overpriced?
No. In a dislocation the term structure is in backwardation and implied volatility can sit below realised, so the high front-month volatility you would sell is high because the near-term movement is expected to be enormous. It is a correct price for danger, not a mispricing you can harvest.
Why do crises accelerate rather than just happen?
Because of the margin feedback loop and forced deleveraging: the initial shock raises volatility, which raises margin, which forces exits, which raises volatility further. The move feeds on itself, which is why the worst of a dislocation often comes after the initial trigger rather than at it.
How do I tell a crisis from a merely elevated reading?
Look at the plumbing, not just the level. A high India VIX with the term structure still upward-sloping is stress; a high reading with the front month bid above the back — backwardation — plus a vertical skew and constituents all falling together is a crisis. The number can look merely elevated while the plumbing is already failing.
Why does diversification fail exactly when it is needed?
Because the diversification benefit comes entirely from low pairwise correlations, and a crisis sets those correlations toward one. Portfolio volatility then reverts to the weighted average of the constituent volatilities, so the risk reduction you relied on disappears in the same session the shock arrives. It is the definition of the regime, not bad luck.
What happens to a dispersion trade in a crisis?
A dispersion trade is short index volatility and long single-name volatility — structurally a short-correlation position. When correlation goes to one, the index moves as much as the names, and the trade takes its largest loss. Its entire edge is collected in calm and paid back in the dislocation.
Why can't crisis volatility be timed?
Because its signals are contemporaneous, not predictive. Backwardation, vertical skew and correlation-to-one confirm a crisis that is already underway; none of them ring a bell in advance. You can prepare for the regime, but the transition itself is precisely the part that is not forecastable from within the calm.
How should a book be sized given crisis volatility?
Below the size at which a doubling of volatility would force a liquidation, because margin rises mechanically and you may be one of many asked to reduce at once. Sizing against calm-period margin quietly assumes the feedback loop will never start, which is the one assumption a crisis is guaranteed to break.
Is holding tail protection worth the cost?
That depends on the book, but the standing trade-off is real: deep out-of-the-money puts or variance bleed a little every quiet day and pay off precisely when the wings go bid and you would otherwise be the forced buyer. Many dealers accept the persistent cost as the price of not being liquidated in the transition. It is a choice with an explicit price, not a free hedge.
Why does reasoning by analogy to past crises mislead?
Because each dislocation has different plumbing — a funding squeeze, a policy shock and a forced-deleveraging spiral produce different term-structure and correlation signatures. A playbook tuned to the last crisis can be exactly wrong for the next, even though both show a high volatility index.
What does the escalator-and-stairs metaphor actually mean?
It means the up-move and down-move in volatility are not mirror images: the spike takes days and the decay back toward normal takes months. The asymmetry is structural because demand for insurance spikes in an afternoon while the willingness to supply it rebuilds over a season.
Does India VIX capture crisis volatility well?
India VIX captures the level of near-dated implied volatility, so it rises clearly in a crisis, but a single index number does not show the term-structure inversion, the vertical skew, or the correlation shift that define the regime. The level is necessary but not sufficient — the plumbing signals live below the headline.
Why is the risk of a short-volatility strategy invisible in its track record?
Because the track record is compiled in the calm and the risk lives in the transition, which is not in the sample until it dominates it. A long string of small positive days is the mechanism of the eventual loss, not evidence against it, because it is what grows the position to the size that fails.
What is the single most dangerous trade justified by crisis volatility?
Selling options because implied volatility looks high. In a crisis that reasoning ignores backwardation, the possibility that implied is below realised, mechanical margin expansion, and the absence of exit liquidity in the wings. Short-volatility positions carry theoretically unlimited loss, and a dislocation is where that theory becomes arithmetic.

Voice search & related questions

Natural-language questions people ask about crisis volatility.

Why does volatility spike so fast in a crash?
Because everyone wants protection at once and almost nobody wants to sell it, so the price of insurance gaps up in days. Demand for options can spike in an afternoon, but the supply of new sellers only returns slowly, over months, which is why the fall back to normal is so much slower than the rise.
What is the escalator and the stairs?
It is the shape of crisis volatility: it goes up by the escalator — fast, in days — and comes down by the stairs — slowly, over months. The two paths are not symmetric because the willingness to write insurance collapses far quicker than it rebuilds.
Why did my hedges stop working when the market fell?
Most likely because they relied on things not all moving together, and in a crisis correlation runs toward one so everything falls at once. Diversification and relative-value hedges lose their basis in the same session the shock arrives — that is a defining feature of the regime, not bad luck.
Why did my margin blow up when I did nothing?
Because margin is computed from recent volatility, and when volatility rises the requirement to hold the same position rises with it, automatically. You did not change the position; the volatility input to your margin changed, and it can force you to reduce at the worst possible price.
Is it smart to sell options when volatility spikes?
It is far more dangerous than it looks. In a crisis the front-month volatility is high because the near-term movement is expected to be enormous, implied can sit below realised, and the exit liquidity you would need is gone from the wings. High volatility here is a correct price for danger. Short-volatility positions carry theoretically unlimited loss, and nothing here is advice to sell them.
How long does crisis volatility take to fade?
Far longer than it took to appear — typically the better part of a season rather than a week. The decay is slow because it requires participants to become willing to underwrite risk again, and that confidence rebuilds cautiously, one desk and one position at a time.
Can I trust the option prices I see on far strikes in a crash?
Often not. Liquidity leaves the wings, so far-strike quotes get wide and no real size trades at them. An implied volatility or a delta computed from those prices is built on a number that is not a real, transactable price, so hedging on it can size a hedge to a fiction.
Why does the worst of a crash often come after the first shock?
Because of the feedback loop: the initial move raises volatility, which raises margin, which forces exits, which raises volatility further. The deleveraging feeds on itself, so the acceleration frequently comes in the sessions after the trigger rather than at the trigger itself.

Sources & references

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

← Previous
Range-bound Markets
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.