Guide · Market structure · Updated 2026-09-19

Pin risk is not uncertainty about price, it is uncertainty about what you will own when the market reopens

Most descriptions of pin risk stop at the observation that an option sitting on its strike at expiry is hard to call. That is true and almost useless. The part worth knowing is what the ambiguity converts into, who resolves it, when, and what the person carrying the position holds in the hours between the last trade and the answer. In the Indian stock derivatives segment those four questions have unusually sharp answers, and the fourth has changed twice in the last thirteen months.

A strike is not a level, it is a boundary between two different positions

Every other price on the board behaves smoothly. A share that moves from 1,150 to 1,151 changes a position by a rupee a share, and a move back unwinds it. The strike does not behave that way at expiry. On one side of it a written call expires and leaves nothing behind. On the other side it is an instruction to deliver the full lot against the strike price. There is no intermediate state and no proportionality: the transition happens in the smallest price increment the exchange recognises.

That discontinuity exists in every options market. What differs is what sits on the far side. In a cash settled market the far side holds a debit equal to the intrinsic value, which for a contract pinned to its strike is small by construction. Where contracts settle by delivery the far side holds a share position, and the size of that position has nothing to do with how close the settlement price landed to the strike.

Timing is the second half of it. The exercise decision is not the holder's. Options that finish in the money here are exercised automatically, and the test of whether they did is applied to a settlement price published after the close. A writer cannot decline assignment, cannot pre empt it, and cannot learn the answer while there is still a market to trade in. So the position carried overnight on an expiry day is not one the trader chose. It is one of two positions, selected by a number that does not exist yet.

The settlement price is the whole of it, and it has just been re specified twice

Because everything turns on one number, the definition of that number is the most important fact on this page. For a single stock derivative the final settlement price is taken by reference to the closing price of the underlying share in the cash market on expiry day. Nothing in the derivatives contract computes a price of its own. It points at the cash market and adopts whatever the closing price mechanism produces.

That pointer is why a change to the cash market closing price rule is a change to derivatives settlement, whether or not anybody describes it that way. Until 2 August 2026 the closing price of a share was a volume weighted average of trades in the last thirty minutes of continuous trading. From 3 August 2026 shares carrying derivatives contracts stopped closing that way. Their closing price is now discovered in a closing auction: continuous trading in those shares ends, orders are pooled rather than matched on arrival, entry closes at a randomised moment rather than a published instant, and one equilibrium price is computed and becomes the close.

No derivatives circular was issued to produce that effect. It followed from the pointer. And because a broad index is computed from the closing prices of its constituents, most of which carry derivatives contracts, the index settlement value inherited the change one level up, again with no index rule amended.

The consequence was visible enough that the regulator reopened the question. On 12 September 2026 SEBI put out a consultation paper on the closing auction session, market timings and the expiry day settlement price methodology for index and single stock derivatives, setting out two options. The first is a blended volume weighted average across trades in the last thirty minutes of the continuous session and in the ten minute closing auction, each window contributing in proportion to its actual traded value rather than by a fixed weight. The second keeps the continuous session average alone as an interim arrangement, on the reasoning that participants and auction liquidity both need time to develop, with the blended method revisited after about a year. Comments were invited until 3 October 2026.

Two things follow. The practical one: almost every explanation of Indian expiry settlement published before August 2026 describes a rule that no longer operates, and its replacement is itself under consultation. Check the date on anything you read about this, including this page. The structural one does not depend on which option is adopted. Under every version of the rule the settlement price is an average or an auction clearing price, which is to say a constructed number rather than a price any single participant could transact at. That is why pin risk cannot be traded away at the close: no order exists that guarantees execution at the number which decides your position.

Cash settled and physically settled are two different risks wearing one name

Single stock futures and options in India are settled by delivery. The move away from cash settlement was directed in 2018 and phased across the segment through to October 2019, and it is the single fact separating the Indian version of this problem from the textbook one. Index derivatives remain cash settled. The difference is not a matter of degree. Take the written call below and settle it both ways at the same settlement price.

The final rows carry the point. An 80 paise band in the settlement price moves a cash settled outcome by ₹200. It moves a physically settled outcome by ₹6,00,200 of stock position, which is 3,001 times larger. Same contract, same strike, same settlement price. Only the settlement method differs, and it is the part most explanations of pin risk written for other markets leave out entirely.

What survives the close, computed for each outcome

Here is that strike carrying all four elementary positions, at both settlement prices: strike 1,200.00, lot 500 shares, settlement at 1,200.40 and at 1,199.60, which are 40 paise either side.

Read the table as four separate traders. Each holds a contract whose entire remaining value is a few hundred rupees at most. Each is 0.80 of settlement price away from a cash market position worth roughly six lakh. None can find out which side of the line they are on until the close has passed.

The ratios are why this is a risk category of its own rather than an ordinary market exposure. On the written call, the gross position created is 3,001 times the intrinsic value that created it, and 66.7 times the premium received for writing the contract. On the long call, the cash required to take delivery is 200 times the premium paid. No position sizing rule applied to the premium constrains either figure, because neither is a function of the premium.

The overnight exposure on a position nobody chose

Assignment is not the loss. Assignment creates a position, and the result arrives at the next open, after a full session gap. The writer assigned on the call above is short 500 shares. To be flat they must buy 500 shares, and the earliest they can is the next session, at whatever price is then available.

At a flat open the writer keeps ₹8,800, being the premium less the intrinsic value given up. All of it is gone at a gap of 1.47 per cent, and the trade turns into a loss of ₹9,206 at three per cent. A three per cent overnight move in a single share is not an extreme event. It is the ordinary consequence of a result announcement, an index review, a sector move or a block trade, and the position is exposed to all of them without ever having expressed a view on any.

Note what the table does not contain: any figure that depends on the trader's opinion. The exposure is ₹6,00,200 of stock, its direction is fixed by which contract was written, its size by the lot. The only free variable is the gap, and that belongs to the market.

Hedging into the close relocates the exposure, it does not remove it

The obvious response is to pre empt the assignment: if the call looks likely to be assigned, buy the 500 shares before the close so delivery can be met from what you hold. The logic is sound and the arithmetic does not support it.

Suppose the writer buys 500 shares at 1,200.60, the last price reachable in continuous trading. If the settlement price lands at 1,200.40 they are assigned, deliver what they bought, and finish flat, having paid ₹100 more than an unhedged writer buying back at the settlement price. If it lands at 1,199.60 they are not assigned, and hold ₹6,00,300 of stock bought for a delivery that never happened, already marked -₹500 against them, funded overnight, and exposed to a fall rather than to the rise they were worried about.

The hedge does not reduce the number of states in which the trader carries an overnight position. It changes which state carries it. Generalising across hedge ratios makes the point exactly.

The final column runs from ₹3,00,100 to ₹2,99,900: a range of ₹200 on about three lakh, or 0.067 per cent, across the entire span from no hedge to a full hedge. No ratio meaningfully reduces the expected exposure, because the hedge and the assignment offset perfectly in one branch and add perfectly in the other. Every rupee the hedge removes from the assigned branch it adds to the unassigned one.

That is the precise sense in which pin risk cannot be hedged. It is not that hedging is difficult or expensive. It is that the instrument you would hedge with resolves into the same binary as the thing being hedged, and the binary is resolved by a number neither leg can reach. Under the auction mechanism now in place the last price available in continuous trading is not the settlement price and is not intended to be, so even a trader willing to pay the spread cannot transact at the number that decides the outcome. A further cost sits on the hedged branch and the worked figures treat it as free: shares bought for a delivery that never happens must be funded and then unwound, at another spread, in a session nobody planned to trade.

Margin, delivery, and what happens if the obligation cannot be met

The exposure begins before expiry day. Positions likely to result in delivery attract a delivery margin on top of the usual derivatives margin, phased in over the final sessions before expiry in increasing steps, because what is being secured shifts from a derivatives exposure to a cash market delivery exposure. The schedule is published by the clearing corporation and is worth reading before the week rather than during it. The capital cost of a near the money position therefore rises as expiry approaches, whether or not it is ever assigned.

On assignment two obligations crystallise. A writer of a call must have 500 shares in the demat account on the settlement date. A writer of a put, or a buyer of a call, must have ₹6,00,000 of funds. Both are hard obligations against a clearing corporation rather than soft ones against a broker, and neither is proportionate to the premium at stake. One relief is worth computing rather than assuming: delivery obligations are netted at the client level across positions in the same underlying, so a writer who also holds a long future or long call on that share may owe far less than the gross figures suggest, and anyone carrying a spread should work out the net rather than plan for the worst leg.

If the shares are not there, the obligation becomes a failure to deliver, and the mechanism that follows is not one the failing party participates in. The clearing corporation sources the shares through a separate auction among other participants, or, where that auction cannot source them, applies a cash close out at a formula price carrying a premium over defined reference prices. The cost is charged back to the party that failed, who does not bid, does not set a reserve, cannot withdraw, and does not know the amount until the process has run. The full mechanism, including the case where a broker nets the shortfall internally and applies its own policy rate instead, is set out in the guide to short delivery and the auction.

Stacked, those facts give the honest shape of the exposure. A written contract whose remaining value is a few hundred rupees can produce, with no further decision by the writer, a six lakh position, an overnight gap exposure on it, a funding requirement, and in the worst case a settlement failure whose cost is set by a process they are not in. Each step follows mechanically from the one before, and the chain starts at a settlement price 40 paise from a strike. One qualification belongs here: exchanges have at times offered a facility letting the holder of a barely in the money option instruct that it not be exercised. It has been introduced and withdrawn at different points, it sits with the long holder rather than the writer, and if your expiry plan depends on it, confirm it against the framework in force before the week rather than on the day.

The only reliable defence is absence, and what absence costs

The risk above cannot be measured in advance, because its probability is the probability that a constructed number lands on one side of a line, and cannot be reduced by hedging, because the hedge inherits the same binary. What remains is not a technique but a decision: whether to hold the position at all when the settlement window opens.

Leaving costs something, and the cost is knowable. On the worked position, buying the written call back before the settlement window at 1.20 per share costs ₹600, which is 6.67 per cent of the premium received, leaving ₹8,400 of it intact. In exchange it removes an overnight exposure of ₹6,00,200, about 1,000 times the cost of leaving. Rolling to a later series has the same character: a definite, quotable cost paid now to remove an indefinite one later.

Those figures do not establish that the exit is always correct. They establish that the two things being compared are of different kinds. The cost of leaving is a price on a screen. The cost of staying is a distribution with an unbounded tail and no reliable way to sample it, and the premium that would have compensated for it was fixed weeks earlier, when nobody knew the contract would settle near its strike. A position taken for ₹9,000 was never sized for a ₹6,00,200 overnight exposure, and the exposure arriving without a further decision is what makes it dangerous rather than merely large.

The operational version is short. Decide before the expiry week at what distance from a strike you will not carry a physically settled contract into the settlement window, and apply it mechanically rather than by judgement on the day, because judgement on the day is being exercised on the one piece of information that is not yet available. Expiry now falls on a fixed weekday for each exchange, following the standardisation effective from 1 September 2025, so the session carrying this risk is a known date. The practical deadline inside it is earlier than it used to be, since continuous trading in derivative eligible shares now ends before the auction that sets the close, and a plan written against the old clock will be late. How a call auction clears, and how position limits bite in expiry week, are covered in the guide to call auction price discovery and the guide to the market wide position limit and ban period.

This is the kind of exposure that is invisible in a profit and loss statement until the one month it is not, which is why it belongs in a written process rather than a trader's memory. Building that process, and knowing which market structure facts it must be re checked against when they change, is method rather than opinion.