Jane Street, Bank Nifty, and the Math of Expiry-Day Manipulation

The setup

Indian index options, Bank Nifty and Nifty 50 in particular, are the most heavily traded options in the world by contract count. The National Stock Exchange of India clears more options volume than the CBOE, Eurex, and CME combined. Most of that volume is retail, and most of it sits in weekly contracts that print a final value on a Thursday afternoon and then disappear forever.

On expiry day, each contract settles to a single number: a weighted average of the index level over the final 30 minutes of trading. Whoever holds the option at 3:30 pm gets paid (or pays) based on where the underlying basket closed during that window. Up to that point, near-the-money options are extraordinarily sensitive to small index moves. After it, they are zero or fixed.

That asymmetry, where an option's payoff is entirely determined by where the underlying sits during a tiny terminal window, is the structural feature Jane Street is alleged to have exploited.

What they actually did

The playbook SEBI describes is a cash-and-carry against options with an active leg on the underlying.

On the morning of expiry, Jane Street accumulates a large long position in Bank Nifty futures and the cash basket of constituent banks (twelve names make up the index). The buying is aggressive enough that the index itself moves up. As the index moves up, near-the-money calls become more valuable and near-the-money puts become less valuable.

Now they have an elevated index level. They sell call options at that level (richer than the calls would otherwise be) and buy put options (cheaper than the puts would otherwise be). The notional on the options leg is many multiples of the cash leg.

In the last few hours before expiry, they unwind the underlying. The futures and cash basket they bought in the morning get sold aggressively into the close. The index drops back down, often below where it started. The calls they sold expire worthless. The puts they bought pay off. The cash leg loses money on the round trip (bought high, sold low) but the loss is smaller than the options gain by an order of magnitude or more. The cash loss is the cost of admission for the options profit.

The math

A near-the-money index call has a delta of around 0.5 and a gamma that spikes as expiry approaches. Call the index level S, the strike K, and time to expiry τ. As τ → 0, the gamma of an at-the-money option diverges. On expiry morning you might reasonably be running gamma of a hundred or more on a single ATM contract.

If the index gets pushed up by some amount ΔS during the morning, the option's value moves by roughly:

Sell that option while ΔS is at its peak, then push S back down to its original level by expiry, and you keep essentially the entire premium you collected. The premium is bounded below by the linear delta term, but in practice it is larger because implied vol also rises during one-way buying. Dealers reprice when they see a flow imbalance, and that vega kick tends to stick even after the spot reverts.

Now scale this up. Bank Nifty options notional around expiry runs into tens of trillions of rupees. The amount of cash and futures you need to push the index by 0.3 to 0.5 per cent is small relative to that. The leverage ratio is what makes the trade work. A fairly cheap shove on the cash market produces a payoff on the options book several multiples larger.

To make this concrete with stylised numbers: suppose on a given expiry you spend roughly ₹500 crore of directional capital across cash and futures to push the index up 0.4 per cent. You sell ATM calls and buy ATM puts on a notional of, say, ₹50,000 crore at the elevated level. You reverse the cash leg and lose around ₹150 crore on the round trip. The options leg, moving on delta plus gamma plus the IV repricing, pays roughly ₹350 to 500 crore. Net P&L for the day comes in somewhere between ₹200 and 350 crore, with the inventory flat by the close.

Those numbers are illustrative; SEBI's order has the actual ones. The shape is what matters. A 1.4 to 2× ratio per expiry, repeated week after week across several years, is how you get to ₹36,500 crore. The cumulative number is large because the trade was run roughly two hundred times, not because any single afternoon was extraordinary.

Why India

In spirit this is the classic "marking the close" pattern that has been illegal in US and European markets for forty years. The reason it worked at this size in India has nothing to do with the sophistication of the trade itself, and everything to do with the structural features of the Indian options market.

The volume ratio is the first piece. On expiry days, options notional in the constituent names runs many multiples of the cash equity volume. The options market has grown so large relative to the cash market it depends on that the underlying is comparatively tiny. Moving the cash market half a per cent costs less than the gamma exposure of the options that depend on it. That is not an arbitrage in price. It is an arbitrage in market design.

Retail dominance is the second piece. The vast majority of weekly options buyers are retail traders, often at the edge of their understanding of what they are actually doing. They are providing a one-way flow of insurance premium to professional desks. When you sell calls into a manipulated peak, the buyers are largely retail accounts that have no idea what is happening to them.

The regulatory model is the third. SEBI has historically been more comfortable supervising the cash market than the derivatives market. The PFUTP regulations cover this kind of behaviour on paper, but enforcement has been thin. Jane Street is not the first firm to notice the gap. They may simply be the largest.

Was it manipulation?

In pure mechanical terms, the trade Jane Street is alleged to have run is what every options market-maker does on a smaller scale every day. You hedge gamma, you unwind delta, you manage your book around expiry. The question is one of intent and proportionality. Pushing the underlying because that is where you wanted to be flat anyway is hedging. Pushing the underlying specifically to move the settlement of options you have already positioned yourself in is manipulation.

SEBI's order alleges the latter. The intraday patterns it cites, sustained one-way buying in the morning followed by sustained one-way selling in the afternoon, do not match the firm's stated risk-management story. The order also points to options positions that were built ahead of the cash leg. That is the closest thing to a smoking gun in the document. Hedgers do not pre-position options.

The market-maker defence is that someone has to be on the other side of all that retail demand, and price discovery requires inventory adjustment. That argument is true in general. It is not a defence against pre-positioning options against your own intended direction in the underlying.

The trade is also path-dependent in a way ordinary hedging is not. A genuine hedger does not care whether the index closes up or down. They care about being flat. The alleged Jane Street trades were profitable specifically when the index reverted by 3:30 pm, which is closer to a directional view than a hedge.

What this tells us

Derivatives markets are only as honest as the underlying they reference. When the cash market is small relative to the options market, the options market is no longer pricing the underlying. It is pricing whatever can move the underlying. Indian regulators have started moving on this with lot size hikes, contract consolidation, and a longer settlement window, but the structural imbalance is still there.

"Marking the close" never went away. It relocated. US regulation crushed the practice in the 1990s. Variants ran through Asian markets in the 2010s. India is where the mechanics line up most sharply right now. There will be another jurisdiction in five years.

Speaking as someone who has traded options for a living, the line between aggressive market-making and manipulation in expiry-day mechanics is genuinely thin. A market designed with cash settlement at a single instant invites behaviour designed to influence that instant. The cleanest fix is not enforcement, it is design. Settle to a longer VWAP, or settle by a different methodology, or settle physically. Wherever a single point determines a billion-rupee payoff, someone with the balance sheet is going to push that point.

Whether Jane Street actually crossed the legal line is for SEBI's tribunal to decide, and ultimately the courts. The disgorgement order, roughly ₹4,840 crore, is a fraction of the alleged profit and will probably be litigated for years. The structural lesson holds regardless of the verdict. The trade was sitting there in plain sight for anyone with the balance sheet to run it. The interesting question is not why Jane Street ran it. It is why nobody at SEBI was watching the index-to-options leverage ratio for half a decade.