The gap is settled at 09:15, not discovered there

Your market is shut for three quarters of the day

The Indian cash market trades from 09:15 to 15:30 IST — six hours and fifteen minutes. It is shut for the other seventeen hours and forty-five minutes, and everything the world does in that stretch has to arrive at one price, at one moment. That arrival is the opening gap.

Prices carry information continuously. Trading does not. A company in Ohio reports after the US close, a central bank speaks at midnight IST, a tanker route closes at four in the morning — and the Indian shareholder who would have reacted to each of them had no venue in which to do it. The reaction does not disappear. It queues.

So a gap is the mechanical consequence of a closed market, not a behaviour of an open one. Nothing about it requires enthusiasm, panic, or any disposition at all. Close a market for seventeen hours in a world that keeps producing news and it will reopen away from where it closed, and the size of the jump measures how much the news mattered rather than how strongly anyone feels.

That framing is worth holding, because the ordinary way of describing the same event — "the market opened strongly" — smuggles in a claim about conviction that the arithmetic does not support. Strength is a statement about buying pressure. A gap is a statement about elapsed information.

The relay, session by session

The world's equity markets are not open at once; they hand off. Sitting in IST, the sequence looks like this.

Two features of that table do the work. First, the biggest single input — the US cash session — lands entirely inside India's closed window. There is no arrangement of Indian trading hours under which a domestic investor transacts on the US session as it happens. Second, Europe is the only bloc whose session begins while India is trading. Asia is already open at 09:15 and winds down through the Indian morning — Tokyo shuts before noon IST, Shanghai and Hong Kong during the early afternoon — while Europe arrives fresh after them. That is why the two halves of the Indian day have different characters: the morning digests sessions that are finishing, the afternoon meets one that is starting.

The pre-open row is the one people skip, and it matters for what follows. The Nifty 50's opening value is not a traded price. It is computed from the opening prices of its fifty constituents, and each of those is struck in a call auction between 09:00 and 09:15 IST, where orders accumulate and are matched at a single clearing price rather than against a continuously visible book. The index open is an arithmetic result of fifty auctions, not a print anyone bought at.

Why the US tiles are futures and the Asian ones are not

Because the US market is shut while India trades and the Asian markets are open. On our Macro page's Global Equities group the US tiles are labelled S&P 500 Fut, Nasdaq 100 Fut and Dow Fut; the Japanese, Chinese, Korean and European tiles carry no such label.

That asymmetry follows directly from the clock. A cash index has no price while its market is shut — the last print is a fossil, and it will still be sitting there tomorrow morning looking exactly as authoritative as it did at the close. A futures contract on the same index keeps trading. So through the Indian session, the only live read on the United States is the futures read, and a page that showed the S&P cash index during the Indian morning would be showing a number that had not been able to move for hours.

Asia needs no such treatment, because Asia is open while India is. The Nikkei, the Hang Seng, the Shanghai Composite and the KOSPI tiles are live cash indices during the Indian morning, which is why they carry information the futures tiles cannot: they are a record of actual money changing hands on the overnight news, not a quote on what it might be worth.

Europe sits in between, and the practical consequence catches people out. Before Europe opens, the DAX and FTSE tiles are showing yesterday's European close. The number is real and it is not new. A reader checking those tiles at ten in the morning IST and finding them unchanged has not found a broken feed; they have found a market that has not started yet. The tile becomes informative in the Indian afternoon and not before.

One convention to carry over from the macro pillar: a tile's colour shows the effect on Indian equities, not the direction of the number. Every weighted global-equity tile takes a positive sign, so a rising Nikkei shows green. That is the simple case. The awkward one on the same page is USD/JPY, where a rising number — a weakening yen — also shows green, because a weak yen means the borrow-cheap-yen-and-buy-risk trade is intact.

The gap is usually discovered before the cash market opens

Here is the part that changes how the open reads. An Indian index future has been trading through nearly all of that overnight window.

The Nifty futures contract that used to trade in Singapore as SGX Nifty moved to NSE's International Exchange at GIFT City in Gujarat in July 2023, and trades there as GIFT Nifty. It runs in two sessions: the first opens in the early Indian morning, ahead of the domestic cash market, and the second runs through the US session and into the small hours IST. Between them they cover most of the window in which the Indian cash market cannot trade.

So when a US jobs number lands at seven in the evening IST, there is a venue on which a view about Indian equities can be expressed immediately — and it is expressed. By the time the domestic pre-open begins at 09:00 IST, an Indian index future has been trading on the US session, on the Asian reaction to it, and on whatever arrived in between.

The cash open is therefore catching up to a price that already exists rather than establishing a new one. Price discovery about the overnight news did not happen at 09:15 IST. It happened between 15:30 and 09:15, somewhere the cash market was not, and 09:15 is where the cash market settles up. This is the single most useful thing to know about gaps, and it inverts the usual telling: the gap is not the market's verdict arriving, it is the market's verdict being copied across.

Two honest limits, because this gets over-read in both directions.

• A future is not the cash index. Its price carries basis — financing cost less expected dividends over the remaining life of the contract — so it does not equal the index and the difference is not a fixed offset. Reading a futures level as a prediction of the cash open confuses two related contracts for one.
• An existing price is not a settled one. Overnight liquidity is thinner than session liquidity, so a price discovered at three in the morning IST has been agreed by far fewer participants than one struck at eleven. It can be disagreed with, and the first hour of the Indian session is where that disagreement gets resolved.

Note also what is not on our Macro page: there is no GIFT Nifty tile. The US futures tiles are the closest thing on it, and they are a read on the input rather than on India's own overnight repricing.

A gap is information; a trend is momentum

The distinction the whole article turns on is between a move that was transacted and a move that was not.

An intraday trend is a sequence of trades. Price moved from one level to the next because buyers and sellers met at every level in between, and the volume at each of those levels is a record of how much conviction it took to get through. A gap has no such record, because nobody in the cash market was permitted to trade at any of the intermediate prices. The market jumped the interval. There is a level change and there is no volume underneath it.

The specific mistake, and it is extremely common: reading gap size as buying pressure. "The market gapped up strongly" describes a large repricing, not a large demand. A gap of that size means the overnight news was worth that much; it says nothing at all about how many rupees changed hands to get there, because in the cash market the answer is none. The two get conflated because both are drawn as a move up on the same chart.

The trade-off worth stating. Because the gap prices news rather than flow, the minutes after 09:15 IST are where the largest genuine disagreement about that news gets worked out — and the same uncertainty that makes those minutes informative is what a market maker is quoting against. Someone obliged to show a two-way price when fair value is least settled protects themselves by quoting wider, so the spread a market order crosses is typically at its worst exactly then. The information is concentrated there and so is the cost of acting on it. Those two facts are the same fact seen from two sides.

None of this says what to do with a gap, and this article does not attempt to. A wider read of how the whole handoff looks on a day when global risk appetite turns is in what a risk-off day looks like.

What the handoff does not carry

The relay is a real mechanism and it is nowhere near the whole gap. Four things cut across it.

Domestic news accumulates in the same window. An RBI decision, a quarterly result published after 15:30 IST, a regulatory order, a monsoon update — each lands in exactly the same closed period as the US session and arrives in exactly the same print. The gap is a net. Nothing about it separates a domestic cause from a global one, which is why attributing an entire gap to Wall Street is a guess dressed as an explanation.

News India has already traded does not arrive twice. If a US inflation release came out during the Indian session and Indian prices adjusted to it before 15:30 IST, the US market's own reaction hours later is largely a second market responding to the same information. Two markets moving on one release is not one leading the other — and the part of it India priced in the afternoon is not waiting in the overnight window to be priced again.

The transmission is a channel, not an index. The routes by which global conditions actually reach an Indian company — the dollar's effect on funding and flows, the discount rate applied to future earnings, the price of an imported barrel — run at their own speeds and are owned elsewhere: the dollar index and emerging markets, US yields and Indian share prices, crude oil and the Indian economy. A global index is a summary of how those channels were received in another country, one step removed from the channel itself.

The foreign-flow print is itself overnight information. Net institutional cash purchases for a session are published after the Indian market has closed, so the number describing today's flows lands in the window that sets tomorrow's open. It is not a live input to the session it measures.

There is a fifth, and it is the reason the Nikkei tile is doubly loaded. Japan is both an Asian equity market and the funding leg of the yen carry trade, so a falling Nikkei alongside a fast-strengthening yen is not simply Asia trading lower — it is the signature of a leveraged position being unwound, which is a different event with a different reach. That is a mechanism, and the whole of it is in the yen carry trade explained.

The correlation panel measures the other relationship

The Macro page carries a correlation panel: the Pearson correlation of Nifty's daily returns against each of a fixed list of ten drivers — the dollar index, the US 10-year yield, Brent, USD/JPY, gold, the US VIX, copper, the S&P 500, the Nasdaq and USD/INR — over a window of 30, 60 or 90 sessions. India VIX is not one of the ten.

Reach for it to measure the handoff this article describes and it will answer a different question than the one asked. The panel aligns the two series by calendar date. Nifty's return for a given date is paired with the S&P 500's return for that same date — a session that begins several hours after the Indian market has already closed and gone home.

So the coefficient is a same-date co-movement. The lead the article traces — a US session on one date reaching the Indian open on the next — is a lagged relationship, and the panel never computes a lagged one. Both relationships can exist at once and they are not the same number. A reader who takes the same-date coefficient as "how much India follows Wall Street" has read a real measurement of the wrong thing.

What the same-date number actually describes is the two markets responding to a shared information flow within one calendar day, with the Indian half of the day going first. That is worth having. It is not the handoff.

One more detail from the same panel, and it is not pedantry. The S&P and Nasdaq entries in the correlation list are the cash indices, while the tiles above them are futures. Same names on one screen, two different instruments, two different trading clocks.

And the standing warning, which the pillar owns in full: a correlation is an observation about a window, not a mechanism. Thirty sessions is short enough that one violent week can carry the whole coefficient, so a sign that flips between the 30-day and the 90-day window is usually telling you about the window. The discipline for using the panel at all is set out in reading macro signals.

How much of the composite the global indices actually are

Our Macro page compresses its tiles into one score, and it is fair to ask how much of that number the world's equity indices are. The answer is smaller than the screen space suggests.

The Global Equities group shows nine tiles. Four of them — Dow futures, the Shanghai Composite, the KOSPI and the FTSE — carry a weight of zero and do not enter the score at all. They are on the page as context. The five that score are S&P 500 futures at 0.06, Nasdaq 100 futures at 0.04, the Nikkei at 0.03, the Hang Seng at 0.03 and the DAX at 0.02, each with a positive sign.

Those five come to 0.18 between them. The score divides by the sum of the weights of every input that reported, which is 1.45 when all of them do — twenty weighted tiles summing to 1.33, plus foreign institutional cash flow, which is not in the tile table at all and gets folded in at scoring time at 0.12, saturating at ₹5,000 crore either way. So the whole global-equities block is 0.18 out of 1.45 — a little under an eighth. That arithmetic is illustrative of the design; the weights themselves are our judgement.

Two consequences follow, and the second is the useful one.

The divisor is not a constant. An input whose feed fails drops out of the numerator and the denominator both, so a day on which the credit-spread feed is late produces a score computed on fewer inputs against a smaller divisor — comparable, but not identical in construction to yesterday's. And 1.45 is the all-present ceiling, never a fixed number to divide by in your head.

The second: because every input is clamped before it is weighted, the score measures agreement across channels rather than severity within one. Push all five weighted global-equity tiles to the limit of their clamp and, with nothing else moving, the composite reaches 0.18 out of 1.45 — about 12 points on a scale running from −100 to +100, which does not reach the −20 that would read as risk-off at all. Falling twice as far does not move it further, because the clamp has already been hit. A dramatic overnight session in world equities and nothing else moving is, by construction, a modest score. The full formula, the clamp, the regime cut-offs at plus and minus 20, and the stress state — which a VIX spike or a fast yen move triggers on its own, whatever the average says — are the pillar's subject, not this article's: reading macro signals sets all of it out.

Every weight and cut-off named here is a design choice we made about what matters to Indian equities. None was estimated from a regression and none is a property of the world. They are published so that a reader who disagrees can see precisely what they are disagreeing with.

Where this sits in the app

The handoff is only readable if the sessions are on one screen in the right order.

FNOTrader's Options Analytics app carries the macro page described here. Its Global Equities group holds the US index futures alongside the live Asian cash indices and the European ones, each with a plain-English note on what it adds; the composite score names its largest contributors in words; the correlation panel runs across 30, 60 and 90-session windows; and every tile opens its own history chart from one month out to five years, so an overnight move can be read against that market's own range rather than against a memory of it. Timestamps on the page are IST.

The weights and cut-offs quoted in this article are the ones the page uses.

Common questions

Continue reading

More in Macro & Markets · App: Options Analytics · Definitions: glossary · Free tools: calculators · All: every article