India Built a Closing Auction. Then Discovered Who Was Supposed to Keep It Honest

On the first day India's stock market switched to a closing auction, the Nifty 50 jumped about 200 points in the final 20 minutes of trading to set a five-month high. The next morning, it gapped back down. A Dubai-based hedge fund trader described the closing spike as "essentially fake." He was probably right.

Here's the thing about closing auctions. They work great — on paper, and in markets that already have the plumbing to make them work. The auction itself is simple: instead of averaging trades over the last half-hour of the day, you pool all orders into a single window and find the price that maximizes matched volume. Nasdaq does it. Euronext does it. The London Stock Exchange does it. In those places, the auction is the cleanest price the market produces all day.

India decided it wanted that too, and on August 3rd the National Stock Exchange flipped the switch. The Closing Auction Session, as they're calling it, runs from 3:15 to 3:35 p.m., replacing the old 30-minute volume-weighted average price for stocks that have derivatives tied to them. The regulator, SEBI, said the goal is better price discovery, less manipulation, and alignment with global peers.

The problem, which the market discovered in roughly 200 Nifty points on day one, is that a closing auction only produces an honest price when both sides of the market show up. If only buyers are there, the "equilibrium price" is just the price at which the people who showed up happen to agree. That's not price discovery. That's a consensus hallucination.

And in India, the people who are supposed to show up on the sell side — the high-frequency traders, the algorithmic desks, the arbitrageurs who profit from correcting price dislocations — have largely stayed away from the auction window. Not because they dislike auctions, but because the rules around short selling make it too expensive to participate on the way down.

The short-selling hurdle is structural, not philosophical. India has a securities lending and borrowing framework, but it's shallow. Borrowing shares to sell them short is difficult and costly. Then there's the settlement layer: India requires physical settlement of cash-market trades. You can't just net out the difference and call it done. You have to actually deliver the securities. For a high-frequency strategy that runs hundreds of round trips a day, the friction of physically borrowing, selling, and covering stock in a 20-minute window is a nonstarter.

SEBI has also banned "naked" short selling — selling stock you don't own and haven't borrowed — and requires institutional investors to disclose at the moment of order placement whether they're shorting. Retail investors can wait until end of day, but institutions can't hedge without flagging their hand.

In practice, this means expressing a bearish view in the Indian cash market is effectively impossible for the kind of speed traders who normally provide two-sided liquidity at the close. The result is a natural upward bias in the auction. When the only people who can easily participate are buyers, the "equilibrium" drifts higher.

That's what happened on Monday, August 3rd. The Nifty surged during the auction to close at 24,774, up 1.6% for the day, with roughly 200 of those points coming in the auction itself. Meanwhile, the BSE Sensex — which runs a separate order book — barely moved. The divergence between the two exchanges was one of those details that tells you the auction wasn't discovering a price. It was revealing a participation gap.

There's a second layer of friction that makes the whole setup even less attractive for arbitrageurs who might have stepped in to keep things honest. India changed its securities transaction tax in April this year. Futures are now taxed on the entire contract value, not just the gains. Options are taxed only on the premium. The tax rate on futures is 0.05% of total contract value.

When you stack the higher tax on top of exchange charges, bid-ask spreads, and impact costs, cash-futures arbitrage — the basic mechanical trade of buying cheap and selling expensive across markets to close a gap — barely clears the bar. The arbitrageurs shifted toward options, where the economics still work. But options don't provide the same corrective pressure on the cash market. You can't really "sell" an ATM option the same way you sell a future to push a mispriced index back toward equilibrium.

So the machine has a feedback loop: thin short-side participation pushes the auction price up, the resulting dislocation should attract arbitrageurs who'd sell high and buy low to close the gap, but the cost structure makes the arbitrage unprofitable, so the dislocation persists. The auction is designed to centralize liquidity and produce a clean benchmark for index funds, ETF tracking, mutual fund NAVs, and derivatives settlement. Without two-sided participation, it's doing something closer to centralizing a one-sided bet.

SEBI has drawn a clear line: there are no flaws in the design and no immediate changes planned. The regulator views higher participation as the variable that matters and noted improvement from day one to day two. By Thursday, volatility had eased somewhat. Raj Deepak Singh, a derivatives researcher at ICICIDirect, said it's too early to judge the system, which is true. But "too early to judge" is also the exact condition under which these things either mature or harden into structural quirks.

The simplest model is this. A closing auction is just a mechanism for finding the price where the most volume clears. In New York, the market makers who show up on both sides are compensated by a deep securities lending pool, cash settlement, and a tax structure that doesn't punish the round-trip trade. In India, the same mechanism operates in an environment where shorting is expensive, settlement is physical, and the tax on futures has made the basic arbitrage that would normally police the auction price marginally profitable at best.

The auction itself isn't broken. It's exposing where the market's liquidity infrastructure is still underdeveloped. India has over 13 crore registered investor accounts, but only 20 to 30 lakh trade on any given day. That's a lot of paper capacity and relatively little active two-sided liquidity. The closing auction is the sort of mechanism that makes thin liquidity visually obvious.

What to watch, then, isn't whether the auction produces another spike. It's whether SEBI addresses the friction that keeps the sell side from showing up. If securities lending gets deeper, settlement gets lighter, or the futures tax structure shifts back toward making arbitrage economically viable, the auction will start behaving like the ones in London or New York. If not, the closing price on the NSE will remain what it essentially is right now: a consensus price among the people who find it easy to buy, with the arbitrage correction that should follow it priced in but unaffordable.

The funny thing about reforming a market is that the plumbing you fix today often reveals the plumbing that's still broken. India got its closing auction. Now it has to decide whether it wants the machines that make it honest.