Volatility Strategies Advanced Index volatility against constituent volatility Relative value / correlation

Dispersion Trading

The trade that is short exactly the thing that stops working when you need it most.

Quick answer: Dispersion trading is selling volatility on an index and buying it on the index's constituents, a position that profits when the individual names move more than the index does and that is therefore fundamentally a short bet on correlation — which rises toward one in precisely the crash the trade cannot survive.

In simple words

An index moves less than the average of its members, because when some stocks rise while others fall the moves partly cancel inside the index. Dispersion trading is built on that fact. You sell options on the index — betting it stays relatively calm — and buy options on the individual stocks — betting they move around a lot. If the stocks jump about while their ups and downs cancel out at the index level, the trade works. The hidden variable is correlation: how much the stocks move together. The whole position is really a bet that correlation stays low, and the uncomfortable truth is that correlation does not stay low in a crash. When markets fall hard, every stock falls at once, correlation shoots toward one, the index moves as much as its members, and the cancellation the trade relied on stops happening.

Put the arithmetic to it, because it is exact. If an index of 50 names has an average member volatility of 24% and an average pairwise correlation of 0.30, the index volatility is not 24% — it is about 13.5%, far below its members, because the idiosyncratic moves cancel. That gap between the members at 24% and the index at 13.5% is what a dispersion trade is trying to capture. But the gap exists only because correlation is 0.30. Push correlation to 1.0, as a crash does, and the same formula gives an index volatility equal to the members' — the gap closes completely, and a trade that was short that gap takes the full loss at the worst possible moment. Diversification stops working exactly when it is needed. That is not a side effect of the trade; it is the trade, written backwards.

Not to be confused with: A simple short-volatility position on the index. A dispersion trade is not just short index volatility — it is short index volatility and long constituent volatility, and the two legs are deliberately balanced so that the net exposure is to correlation rather than to the level of volatility. Confusing the two hides the real risk: a plain index short loses when volatility rises anywhere, while dispersion loses specifically when correlation rises, which is a narrower and more treacherous thing because correlation rises fastest exactly when it matters most.

Why the index prints far below its constituents

Correlation sets the gap the trade lives in

Index volatility as a function of average pairwise correlation, for 50 names each realising 24% volatility, via σ_index = σ_avg × √(ρ + (1−ρ)/n).

At low correlation the index volatility sits far below the 24% of its constituents, because idiosyncratic moves cancel; as correlation rises toward one the curve climbs to meet the constituent level and the gap the dispersion trade harvests closes entirely. The steepest part of the climb is in a falling market, which is why a short-correlation position loses fastest exactly when correlation is rising fastest.

Professional explanation

The mathematics that makes an index calmer than its members

For an index of n names with average constituent volatility σ_avg and average pairwise correlation ρ, the index volatility is approximately σ_index ≈ σ_avg × √(ρ + (1−ρ)/n). The formula says something intuitive precisely. Inside the square root, the ρ term is the part of each stock's variance that moves together with the others and therefore survives into the index; the (1−ρ)/n term is the idiosyncratic part, which is divided by n because independent moves across many names largely cancel. Put the site's numbers in. With ρ = 0.30, n = 50 and σ_avg = 24%, the bracket is 0.30 + 0.70/50 = 0.30 + 0.014 = 0.314, its square root is about 0.5604, and σ_index ≈ 24% × 0.5604 ≈ 13.5%. The index prints roughly 13.5% while its members average 24%, and the entire difference is the idiosyncratic movement cancelling out. That 13.5% is not an approximation of something messier — it is what the correlation and the constituent volatility jointly require, and the dispersion trade is a wager on that arithmetic holding.

The trade is short correlation, and that is the whole risk

A dispersion trade sells options on the index and buys options on the constituents, sized so the two volatility exposures roughly offset and what remains is exposure to correlation. Look again at the formula to see why. The trade profits when the index realises less volatility relative to its constituents than the options implied — which happens when correlation is low, so the idiosyncratic moves cancel and σ_index stays far below σ_avg. It loses when correlation rises, because then σ_index climbs toward σ_avg, the gap the short-index and long-constituent legs were straddling collapses, and the short-index leg takes a loss the long-constituent leg does not fully offset. So the position is short correlation, plainly. And correlation has a property that makes short-correlation positions uniquely treacherous: it is not stable. In calm markets stocks wander on their own news and correlation is low; in a crash every stock is sold at once for the same reason and correlation goes to one. The diversification that the trade sells stops working at exactly the moment the market most needs it, which is exactly the moment the trade is most short of it. Say that plainly, because no amount of position sizing changes it.

Why index implied volatility sits above its correlation-implied fair value

There is a structural reason a dispersion trade has anything to harvest at all. Index implied volatility tends to trade above the level that the constituents' implied volatilities and a reasonable correlation would justify — the index options are, in that sense, structurally rich relative to the basket. The cause is hedging demand: large institutions hold diversified equity portfolios and buy index puts to protect them, because one index put is cheaper and more liquid than fifty single-stock puts. That persistent one-sided demand for index downside props up index implied volatility, and by extension the implied correlation embedded in it. The dispersion trade is, in part, a bet that this implied correlation is higher than the correlation that will actually realise — that the index options are richer than the basket deserves. That is a real and studied phenomenon, and it is also the reason the trade is not a free lunch: the premium exists because index puts are insurance, and the buyer of that insurance is paying a spread to a seller who will, eventually, be on the wrong side of a crash.

In India this is a professional trade, effectively unavailable to retail

The elegant version of dispersion trading assumes you can trade liquid, European, cash-settled options on both the index and each of its constituents, and rebalance a fifty-leg book cheaply. In India that assumption breaks on the constituent side. NIFTY index options are European and cash-settled and deeply liquid — the index leg is fine. But single-stock options in India are American-style and physically settled, and many of the fifty names are not liquid enough to trade options in size without paying a punishing spread. Physical settlement means an in-the-money single-stock option can deliver actual shares you must fund and clear, which is an operational burden a fifty-name book multiplies fifty times. American exercise adds early-assignment risk to every short leg. The result is that a strategy which is capital-intensive, execution-heavy and margin-hungry even in a market built for it becomes, in India, effectively out of reach for retail participants and demanding even for institutions. This page explains the concept; it is not a route map, because for almost every reader the route does not exist.

The uncomfortable symmetry with diversification itself

Dispersion trading is worth understanding even by people who will never place it, because it is the clearest possible statement of a fact that governs every diversified portfolio. Diversification works because correlations are below one, and its benefit is exactly the gap between the average constituent volatility and the portfolio volatility — the same gap the dispersion trade harvests. When a dispersion trade loses because correlation spiked to one, a diversified investor is losing for the identical reason: the protection they thought they had, the cancellation of idiosyncratic moves, evaporated when every asset fell together. The dispersion trader has simply made that dependency explicit and levered it. So the trade's central risk is not exotic; it is the ordinary risk of every portfolio that assumes tomorrow's correlations will look like yesterday's, concentrated into a single position and made impossible to ignore. That is the sentence a marketing department would cut: the strategy is a bet against the one thing every investor is quietly relying on, and it pays until the day that reliance fails for everyone at once.

Formula

Index volatility from constituent volatility and correlation

σ_index ≈ σ_avg × √(ρ + (1 − ρ)/n)

The index realises less volatility than its members because idiosyncratic moves cancel. The ρ term is the common variance that survives aggregation; the (1−ρ)/n term is the idiosyncratic variance, divided by n because independent moves across many names largely cancel. As ρ rises toward 1, the bracket rises toward 1 and σ_index rises toward σ_avg — the diversification benefit, and the dispersion trade's edge, disappears together.

σ_indexThe volatility realised (or implied) by the index itself. For NIFTY with the site's numbers this is about 13.5% when ρ = 0.30.
σ_avgThe average volatility of the individual constituents, assumed equal across names in this simplified form; taken as 24% in the worked case.
ρThe average pairwise correlation between the constituents' returns — the single variable the dispersion trade is really exposed to. Low in calm markets, rising toward one in a crash.
nThe number of constituents in the index; the idiosyncratic variance is divided by n because independent moves across many names cancel. Taken as 50 in the worked case.

The correlation implied by option prices

ρ_implied ≈ (σ²_index − (1/n)·σ²_avg) / (σ²_avg · (1 − 1/n))

Rearranging the first identity for ρ gives the correlation that the market's index and constituent implied volatilities jointly imply. A dispersion trade is, in effect, a bet that this implied correlation is higher than the correlation the market will actually realise — the index options being structurally rich because of put-hedging demand.

How a dispersion trade is structured, conceptually

Form a view on realised correlation — whether the constituents will move more independently than the option market's implied correlation suggests.
Estimate the implied correlation embedded in current index and constituent option prices using the rearranged identity, to see what the market is charging.
Sell volatility on the index, typically through options or a variance instrument, to be short the index's relatively low expected movement.
Buy volatility on a representative set of the constituents, sized so the constituent volatility exposure roughly offsets the index volatility exposure and the net position is exposure to correlation.
Delta-hedge each leg to remove directional exposure, recognising that a fifty-name book multiplies the hedging cost and operational burden of the delta-hedging page.
Recognise the position for what it is — short correlation — and stress-test it against a scenario in which correlation goes to one, because that scenario is not a tail curiosity but the trade's defining risk.
Understand before proceeding that in India the constituent leg faces American, physically settled, often illiquid single-stock options, which makes the clean version of this trade effectively unavailable to retail participants.

Practical example

NIFTY worked example

Take a NIFTY-style basket to see the arithmetic the trade rests on. Suppose the 50 constituents each realise about 24% volatility and their average pairwise correlation is 0.30. The index volatility is σ_index ≈ 24% × √(0.30 + (1 − 0.30)/50) = 24% × √(0.30 + 0.014) = 24% × √0.314 = 24% × 0.5604 ≈ 13.5%. So the index prints about 13.5% while its members average 24% — the index is far calmer than any of its parts, entirely because idiosyncratic moves cancel. A dispersion trade sells the index's 13.5%-implied options and buys the constituents' 24%-implied options, harvesting that gap if correlation stays near 0.30. Now interpret the danger, not just the number. Recompute with ρ = 0.90, a stressed but not extreme crash correlation: the bracket becomes 0.90 + 0.10/50 = 0.902, its root is about 0.9497, and σ_index ≈ 24% × 0.9497 ≈ 22.8%. The index volatility has jumped from 13.5% to 22.8% toward the constituents' 24% while the members barely moved — the entire gap the trade was short has almost vanished, and the short-index leg has taken the loss. The example is not that the trade earns 13.5% versus 24%; it is that the difference between those two numbers is made of correlation, and correlation is the one input that betrays you in a crash.

BANKNIFTY worked example

BANKNIFTY sharpens the point because it is a narrow, sector-concentrated index, and concentration is correlation's natural habitat. BANKNIFTY holds a dozen banking and financial names whose fortunes are driven by the same interest-rate, credit and regulatory forces, so their baseline pairwise correlation is structurally higher than a broad index's — perhaps 0.55 rather than 0.30 in ordinary conditions. Put that into the formula with, say, n = 12 and σ_avg = 28%: the bracket is 0.55 + 0.45/12 = 0.55 + 0.0375 = 0.5875, its root is about 0.7665, and σ_index ≈ 28% × 0.7665 ≈ 21.5%. The BANKNIFTY index prints much closer to its constituents than a broad index would, because with high correlation and few names there is little idiosyncratic movement to cancel. The lesson is different from the NIFTY one: a dispersion trade on a concentrated sector index starts with a smaller gap to harvest and a correlation already elevated and quicker to reach one, so the same trade that looks marginal on a broad index looks worse on a narrow one. Concentration is not a place to find a bigger dispersion edge; it is a place where the trade's central risk is closer to the surface at all times.

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. Dispersion trading is a short-correlation position, and correlation goes to one in exactly the market crash the trade is least able to survive — every stock falls together, the index moves as much as its members, and the gap the trade was short collapses at the moment the loss is largest. Diversification stops working precisely when it is needed, and a dispersion trade is short that failure. The strategy is capital-intensive, execution-heavy and margin-hungry, its losses are fat-tailed and correlated with market crashes, and in India its constituent leg relies on American, physically settled, often illiquid single-stock options that make it effectively unavailable to retail participants. Nothing on this page is a route map to placing it.

Advantages & limitations

What it is good for

It isolates a specific and studied phenomenon. Rather than a vague bet on volatility, dispersion expresses a precise view on correlation — that the constituents will move more independently than the index option market implies — which is a cleaner and more analysable exposure than a directional or level-of-volatility bet.
It has an identifiable structural source of premium. Index implied volatility trades above its correlation-implied fair value because institutions buy index puts for portfolio protection, so the edge the trade targets has a real economic cause rather than resting on a claim that the market is simply wrong.
It teaches the anatomy of diversification. The same formula that prices the trade prices every diversified portfolio's benefit, so understanding dispersion is understanding exactly what a diversified investor is relying on and exactly when it fails — knowledge that is valuable even to someone who never trades it.
Its risk is nameable and stress-testable. Because the position reduces to a single exposure — correlation — its worst case can be modelled directly by pushing correlation to one, which makes the danger explicit rather than hidden inside a complex multi-leg payoff.
The index leg, in India, is liquid and clean. NIFTY index options are European and cash-settled, so the short-index side of the trade is easy to express precisely, even though the constituent side is not.

Where it breaks down

It is short correlation, and correlation rises toward one in a crash, so the trade takes its largest loss in exactly the market environment that produces the most damage everywhere else.
The formula assumes a single average constituent volatility and a single average pairwise correlation, but real constituents have widely different volatilities and correlations, so the clean identity is an approximation that mis-prices the trade when the dispersion of correlations is itself large.
It is capital-intensive and execution-heavy. A well-hedged dispersion book has one index leg and many constituent legs, each needing its own delta hedge, so transaction and margin costs multiply and can consume the structural premium the trade targets.
In India the constituent leg is impractical for most participants, because single-stock options are American-style, physically settled and frequently illiquid, adding early-assignment risk, delivery obligations and wide spreads that the clean theory ignores.
The premium it harvests is an insurance premium, so the steady gains in calm markets are compensation for bearing crash risk, not evidence that index options are simply mispriced — and the compensation is paid back when the insured event occurs.
Implied correlation can stay elevated or rise further for long periods, so even a correct long-run view on realised correlation can produce sustained mark-to-market losses that force a leveraged book to unwind before the view pays off.

Common mistakes

Thinking of it as a bet on volatility rather than on correlation. The two legs are balanced to remove the level of volatility and leave correlation, so a trader who watches volatility and ignores correlation is watching the wrong variable and will be surprised by the loss that arrives when correlation spikes.
Ignoring that correlation goes to one in a crash. Sizing the trade on calm-market correlation and treating a spike to one as a remote tail is the error that defines the strategy's blow-ups, because the spike is not a tail event for this position — it is the position's central risk.
Underestimating the cost of the constituent leg. A fifty-name long-volatility book carries fifty spreads, fifty hedges and, in India, fifty physically settled American options, and a trader who models the clean formula without those frictions has modelled a trade that does not exist.
Assuming the index-versus-basket premium is a mispricing to be arbitraged away. It is largely a structural insurance premium from put-hedging demand, so harvesting it means underwriting index crash risk, and treating that premium as free is the same mistake as treating any insurance float as profit before the claims arrive.
Using a single average correlation when the constituents' correlations vary widely. The approximation breaks when the correlation structure is heterogeneous, so the trade's true exposure can differ materially from what the simple formula suggests, especially in a sector-concentrated index.
Applying the broad-index intuition to a concentrated index like BANKNIFTY. A narrow index starts with higher correlation and fewer names, so the gap to harvest is smaller and the correlation is closer to one, and a trader who expects the NIFTY-style edge on BANKNIFTY has misjudged both the reward and the proximity of the risk.

Professional usage

Dispersion trading is a professional volatility-desk strategy, run by institutions with the capital to margin a large multi-leg book, the execution infrastructure to rebalance fifty legs at low cost, and the risk systems to monitor a correlation exposure in real time. Such desks trade the difference between the correlation implied by index and single-stock option prices and their own forecast of realised correlation, often through variance swaps on the index against a basket of single-stock variance swaps, which express the correlation view more cleanly than options-plus-hedges. The core of the professional discipline is not the trade structure, which is well known, but the risk management: sizing so that a jump in correlation to one is survivable, because that jump is not a possibility to be hedged away but the defining feature of the exposure, and the desks that treat it as a remote tail rather than the main event are the ones that become case studies.

Correlation itself is traded directly at the institutional level through correlation swaps and through the implied-correlation indices that some markets publish, which lets a desk take a view on average pairwise correlation without assembling the full dispersion book. These instruments are over-the-counter and institutional, and they do not exist in accessible form for Indian single-stock names, which is one more reason the strategy remains the preserve of well-capitalised desks. For most market participants the value of understanding dispersion is not the ability to place it but the ability to recognise, in their own diversified holdings, the same short-correlation exposure the trade makes explicit — and to remember that it is largest exactly when markets fall together.

Key takeaways

Dispersion trading sells index volatility and buys constituent volatility, balanced so the net exposure is to correlation rather than to the level of volatility — it is fundamentally a short-correlation position.
The index realises less than its members via σ_index ≈ σ_avg × √(ρ + (1−ρ)/n); with ρ = 0.30, n = 50 and σ_avg = 24%, the index prints about 13.5% against members' 24%, and that gap is what the trade harvests.
Correlation goes to one in a crash, closing the gap and inflicting the trade's largest loss at the worst possible moment — diversification stops working precisely when it is needed, and the trade is short that failure.
Index implied volatility sits structurally above its correlation-implied fair value because of index put-hedging demand, which is the source of the trade's premium and also the reason that premium is compensation for crash risk, not a mispricing.
It is capital-intensive, execution-heavy and margin-hungry, and in India the constituent leg relies on American, physically settled, often illiquid single-stock options, making the clean trade effectively unavailable to retail participants.
Understanding dispersion illuminates every diversified portfolio, because the gap the trade harvests is the same diversification benefit every investor relies on and that fails for everyone at once in a crash.

Dispersion trading is the honest name for a bet every diversified investor makes without noticing: that tomorrow's correlations will resemble today's, and that the cancellation of idiosyncratic moves will keep the whole calmer than its parts. The formula σ_index ≈ σ_avg × √(ρ + (1−ρ)/n) puts a number on that bet — 13.5% against 24% at a correlation of 0.30 — and the same formula shows the number collapsing as correlation climbs to one. The trade earns while correlation stays low and pays it all back in the crash where every stock falls together. It is capital-intensive, operationally forbidding, and in India effectively closed to retail. The reason to learn it anyway is that it names, precisely, the risk hiding inside every portfolio that mistakes calm-market diversification for a permanent property.

Frequently asked questions

What is dispersion trading?

Dispersion trading is selling options on an index and buying options on its constituents, sized so the net position is exposed to correlation rather than to the level of volatility. It profits when the individual names move more independently than the index option market implied, and loses when correlation rises.

Why is dispersion trading a bet on correlation?

Because the index volatility depends on correlation through σ_index ≈ σ_avg × √(ρ + (1−ρ)/n), and the trade balances index and constituent volatility exposure so that only the correlation term is left. When correlation is low the index stays calm relative to its members and the trade gains; when correlation rises the gap closes and it loses.

What is the formula linking index and constituent volatility?

It is σ_index ≈ σ_avg × √(ρ + (1−ρ)/n), where σ_avg is the average constituent volatility, ρ the average pairwise correlation and n the number of names. The ρ term is the common variance that survives aggregation and the (1−ρ)/n term is the idiosyncratic variance that cancels across names.

Why does the index move less than its constituents?

Because when some constituents rise while others fall, the moves partly cancel inside the index, leaving only the common component. That cancellation is the diversification effect, and it is why an index of 50 names each at 24% volatility can print about 13.5% when their average correlation is 0.30.

What happens to a dispersion trade in a crash?

It takes its largest loss, because in a crash every stock falls together, correlation goes toward one, and the index volatility rises to meet the constituent volatility. The gap the trade was short collapses at exactly the moment the loss is largest, which is the defining risk of the strategy.

Why does correlation rise in a crash?

Because in a sell-off investors sell broadly for the same macro reason — liquidation, deleveraging, fear — rather than reacting to individual company news, so stocks that normally move on their own move together. Idiosyncratic drivers are overwhelmed by a single common driver, and average pairwise correlation shoots toward one.

What is the 13.5% figure in the worked example?

It is the index volatility implied by 50 constituents each at 24% volatility with an average correlation of 0.30: 24% × √(0.30 + 0.70/50) = 24% × √0.314 = 24% × 0.5604 ≈ 13.5%. It shows the index printing far below its members purely because idiosyncratic moves cancel.

Why is index implied volatility above its fair value?

Because institutions buy index puts to hedge diversified portfolios, since one index put is cheaper and more liquid than fifty single-stock puts. That persistent one-sided demand props up index implied volatility and the implied correlation inside it, which is the structural premium a dispersion trade targets.

Can retail traders in India do dispersion trading?

Not practically. The index leg is fine, since NIFTY options are European and cash-settled, but the constituent leg relies on single-stock options that in India are American-style, physically settled and often illiquid, adding early-assignment risk, delivery obligations and wide spreads that make the clean trade effectively unavailable to retail.

What is implied correlation?

Implied correlation is the average pairwise correlation that the index and constituent option prices jointly imply, found by rearranging the index-volatility identity for ρ. A dispersion trade is a bet that this implied correlation is higher than the correlation the market will actually realise.

Why is BANKNIFTY a harder index for dispersion?

Because BANKNIFTY is narrow and sector-concentrated, so its constituents share the same interest-rate and credit drivers and start with higher baseline correlation and fewer names. That leaves a smaller gap to harvest and a correlation already closer to one, so the trade's central risk is nearer the surface.

Is dispersion trading the same as shorting index volatility?

No. Shorting index volatility loses whenever volatility rises anywhere, while dispersion is balanced against a long constituent-volatility leg so it loses specifically when correlation rises. The distinction matters because correlation rises fastest in a crash, giving dispersion a sharper and more crash-linked loss profile.

What instruments do professionals use for dispersion?

Often variance swaps on the index against a basket of single-stock variance swaps, which express the correlation view more cleanly than options plus delta hedges. These are over-the-counter and institutional, and they do not exist in accessible form for Indian single-stock names.

How does the number of constituents affect the trade?

The idiosyncratic variance is divided by n, so more names mean more cancellation and a larger gap between the index and its constituents at a given correlation. A broad index like NIFTY therefore offers a wider dispersion gap than a concentrated one, though the correlation risk remains the binding constraint.

Why is the premium in dispersion not free?

Because it is largely an insurance premium: index puts are rich due to hedging demand, and collecting that richness means underwriting index crash risk. The steady gains in calm markets are compensation for bearing that risk, and they are paid back when the crash the trade is short of finally arrives.

Does the dispersion formula assume all constituents are identical?

The simple form assumes a single average constituent volatility and a single average pairwise correlation, which is an approximation. Real constituents have widely different volatilities and correlations, so the identity mis-prices the trade when that heterogeneity is large, especially in a concentrated index.

What does dispersion trading teach about diversification?

It shows that a portfolio's diversification benefit is exactly the gap between average constituent volatility and portfolio volatility, and that this gap is made of correlation being below one. When a dispersion trade loses to a correlation spike, a diversified investor loses for the identical reason, which makes the trade a lens on every portfolio's hidden dependency.

Can a dispersion trade lose even if the long-run correlation view is right?

Yes. Implied correlation can stay elevated or rise further for extended periods, producing mark-to-market losses that force a leveraged book to unwind before the realised correlation confirms the view. Being right eventually does not help a position that is closed out first.

How is a dispersion trade delta-hedged?

Each leg — the short index options and the long constituent options — is delta-hedged in its own underlying to remove directional exposure, leaving the correlation exposure. A fifty-name book therefore carries fifty-one hedging programmes, which multiplies the transaction and operational cost relative to a single-name trade.

Is a higher constituent volatility good for the trade?

It widens the potential gap between constituents and index at a given correlation, but it is not automatically favourable, because the trade's outcome depends on realised versus implied correlation, not on the level of constituent volatility alone. High constituent volatility with high correlation offers little to harvest.

Why is dispersion described as capital-intensive?

Because a hedged dispersion book holds one index position and many constituent positions, each requiring margin and its own hedge, so the capital and margin requirements scale with the number of legs. The structural premium the trade targets is thin relative to those costs, which is why it is the preserve of well-capitalised desks.

Voice search & related questions

Natural-language questions people ask about dispersion trading.

Why is dispersion trading short correlation?

Because it sells index volatility and buys constituent volatility, and the only thing left after balancing those is exposure to how much the stocks move together. When correlation is low the index stays calm and the trade gains; when correlation rises the index catches up to its members and the trade loses.

What goes wrong with dispersion in a market crash?

In a crash every stock falls at once, so correlation jumps toward one and the index becomes as volatile as its members. The gap between the two that the trade was harvesting disappears, and because the position is short that gap, it takes its worst loss exactly when the market is worst.

Why does an index move less than the stocks in it?

Because the stocks do not all move the same way at the same time in normal conditions — some rise while others fall, and those moves partly cancel inside the index. Only the part that moves together survives, so the index is calmer than the average of its members.

Can I try dispersion trading with NIFTY stocks?

In practice, no. The NIFTY index options are easy to trade, but the single-stock leg in India uses American, physically settled options that are often illiquid, so you would face wide spreads, delivery of actual shares and early-assignment risk across dozens of names. It is a professional, capital-heavy strategy rather than a retail one.

Is dispersion trading a way to earn steady income?

No, and thinking of it that way is the trap. The steady gains in calm markets are an insurance premium for underwriting a correlation spike, and that spike arrives in a crash and pays back the calm-market gains in one move. Nothing here is a recommendation to run it.

How is dispersion different from just selling index options?

Selling index options alone loses whenever index volatility rises, for any reason. Dispersion adds a long constituent-volatility leg so that the level of volatility is offset and what remains is a bet on correlation, which loses specifically when stocks start moving together rather than whenever volatility goes up.

What number should I remember from dispersion trading?

That 50 stocks each at 24% volatility, correlated at 0.30, make an index of only about 13.5% — and that pushing the correlation toward one drags the index straight back up toward 24%. The whole trade lives in that gap, and the gap is made entirely of correlation.

Sources & references

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