Intraday strategy for the Nifty: a decision structure, not a setup

Start with the confusion that quietly wastes most of the effort retail traders put into this. Ask an intraday trader for their strategy and you will almost always be handed a trigger: a break of the opening range, a pullback to the volume-weighted average price, a moving-average cross. That is a setup. It is a hypothesis about pressure, a claim about who is trapped and who is absorbing, and it is worth knowing. But it is one component of a plan, and by some distance the component least responsible for the outcome. Two traders can hold the identical trigger and end the month in completely different places, and nothing about the trigger explains the gap.

A plan is the larger thing the trigger sits inside. It is a decision structure: the complete set of answers, written down in advance, to every question the session is going to ask you while you are least equipped to answer. Which conditions constitute a trade at all. Which part of the day they count in, and which parts you will sit out regardless of what the chart appears to offer. Where the position stops being an idea and becomes a mistake. How many units that stop permits. What ends the day early. Each of those is a decision that will get made either way. The only question is whether you make it in the quiet before the open or whether the market makes it for you at 14:50 with money on the line.

Framed that way, the reason a setup is so weak on its own becomes obvious: a setup is a statement about price, and price is not the thing that changed. The trigger you drew at 09:20 and the identical trigger at 12:40 are the same shape on the same instrument, and yet one of them is standing in the deepest participation of the day and the other in the thinnest. The rule cannot tell the difference. Only the structure around it can, and only if somebody built that structure deliberately. This is why the durable object here is the plan and not the setup, and why the rest of this guide is about the session first and the trigger last.

The market will ask you every one of these questions anyway. A plan is simply the decision to answer them while you are calm, flat, and not yet wrong.

Everything downstream depends on this, so it is worth being precise about the machinery. NSE does not simply switch the market on at 09:15. It runs a pre-open call auction from 09:00 to 09:15, which collects orders, matches them, and prints a single opening price, so that the day begins on a considered level rather than in a scramble. That auction is a genuinely different mechanism from the continuous trading that follows: it is one clearing price computed from a whole book, not a sequence of bilateral fills. Continuous trading then runs from 09:15 to 15:30. Those timings are published by the exchange and are the skeleton every intraday plan hangs on.

What happens on that skeleton is not uniform, and the shape is one of the most reliably documented regularities in market microstructure. Volume and price variability concentrate heavily at the open and again into the close, and sag through the middle of the session. The first stretch after the bell is the market repricing whatever the auction concluded, against overnight news, against foreign closes, against every order that waited. It carries the day's heaviest volume and its widest spreads at the same time, which is exactly why it looks so tempting and reads so badly. Then participation decays. By the midday trough, roughly 11:30 to 14:00, the volume that makes a break mean anything has largely gone home, and ranges compress to match. Late in the day it returns as positions are squared into the close.

The chart below draws one session with both halves of that story on the same clock: the price path above, the participation beneath it. Read them together rather than separately, because the argument of this entire guide lives in the mismatch between them.

Look at the two panels together and the practical rule falls out on its own. Price is not equally informative at every hour, because the participation that makes a price mean anything is not equally present at every hour. A level broken in the 09:15 bucket has been broken through the deepest book of the day. The same level broken at 12:35 has been broken through a book carrying a sixth as much interest. Those are not the same event, and no amount of care with the trigger will make them the same event. The session, not the setup, is the thing you are actually trading, and the next section makes that concrete.

The cleanest way to see why the window dominates is to hold everything else still. So take a single, entirely ordinary rule, the kind that appears in every intraday guide ever written: mark the high and low of the first fifteen minutes of the window, and treat a break of the high as a long trigger with the stop at the range low. Now run that identical rule twice on the same instrument, on the same day, with an identical range, an identical trigger level, and an identical break. The only thing allowed to differ is which window of the session the clock happens to be in.

This is the honest experiment, because it isolates the variable. If the rule were doing the work, two identical pictures would have to produce two similar outcomes. They do not, and the figure below shows why: the mechanism the rule is implicitly betting on is participation, and participation is the one thing the picture does not show you.

The lesson is not that midday breaks always fail, which would just be another rule and would be false in the same way. It is that the rule was never the complete claim. Written out honestly, the trigger is saying: enough interest has crossed this level that the balance has genuinely shifted. That claim is testable, and its truth depends almost entirely on how much interest was available to cross. The window is not context around the rule. It is a load-bearing term inside it, silently omitted. This is exactly why a regime filter belongs upstream of every setup rather than beside it: the filter is what puts the omitted term back.

Which produces the first real specification. A plan does not say trade the opening range break. It says trade the opening range break in this window, and not in that one, and it says so before the bell, in writing, when the 12:35 chart is not in front of you looking exactly like the 09:20 one.

The year in the title of this page earns its place in exactly one way: the rule environment around intraday Nifty has been rewritten over the last few years, each change is dated and verifiable, and each one moves a parameter that a plan has to respect. What the year does not license is a prediction. Nothing below says what the index will do. It says what the arithmetic of a position now costs and permits, which is a different and far more useful kind of fact.

Four changes stack up. Peak margin arrived with SEBI's framework of 20 July 2020, phased from December 2020 in steps of 25, 50 and 75 per cent and reaching 100 per cent upfront margin by September 2021, verified at random intraday snapshots rather than only at the close. The consequence is blunt: the broker-granted intraday multiplier that once let a small account carry a large position for a session is gone, and you now fund the full margin for whatever you hold. If your mental model of intraday still assumes that multiplier, start with what leverage in trading actually is. Next, the index-derivatives framework of 1 October 2024 did two things that matter here. It raised the minimum contract value so that lot sizes are set to keep the contract inside a ₹15 to ₹20 lakh band on review, effective for new contracts from 20 November 2024. And it limited each exchange to weekly expiries on a single benchmark index; NSE kept the Nifty 50, which is why it is that exchange's one surviving weekly while Bank Nifty and the other index contracts moved to monthly and quarterly cycles.

Then the two most recent. From 1 September 2025 NSE moved its entire derivatives expiry cycle to Tuesday, with monthlies and quarterlies settling on the last Tuesday, while BSE moved its own to Thursday. This is not cosmetic for an intraday trader, because the expiry tape has its own anatomy layered on top of the ordinary one: the pinning toward heavily traded strikes, the accelerating decay, the late-session volatility. All of that now arrives on Tuesday, ending roughly twenty-five years of a Thursday habit that a great deal of published intraday material still silently assumes. And from the January 2026 series, after the December 2025 expiry, the Nifty lot was reset from 75 to 65, with parallel cuts elsewhere, keeping the contract inside the same band as the index level drifted. The number matters less than the mechanism, and section five shows exactly where it bites.

Read down the right-hand column and notice what these have in common. Not one of them forbids anything, and not one of them is about direction. Every single one moves a parameter of the decision structure: how much capital a position demands, how finely you can size it, which weekday carries the event risk, how much slack an error leaves. That is the whole reason a plan is the durable object. A setup written for the 2019 environment could still be drawn on today's chart and would look identical. The plan around it would be wrong in four separate places.

Here is the core of it. Strip an intraday plan down and it has to answer three questions before the market opens, and they are not three independent choices. They chain. What qualifies determines where you would enter. Where the stop lives is then read off the same structure, because the stop is the level at which the idea you just described stops being true. And what size that stop implies is not a decision at all once the first two are fixed: it is division. Your rupee risk budget divided by the distance to the stop gives a number of units, and the instrument's lot then rounds that number down.

That last step is where most retail plans quietly break, and it is worth being exact about why. The temptation is to run the chain backwards: to decide the size first, from what feels affordable or from what the margin permits, and then place the stop wherever leaves that size intact. This inverts the logic completely. A stop placed to justify a size is not a stop; it is a wish about how far price will refrain from going. The stop has to come from the structure, and then the size has to accept whatever the arithmetic returns, including the frequent answer that the arithmetic returns nothing you can trade. Deciding where the level of invalidation actually sits, rather than where it would be convenient, is the whole subject of stop-loss placement, and it is the hinge the other two specifications turn on.

Two things follow that are easy to miss. The first is that the grid cuts both ways, and the direction it cuts is set by the stop. A tight stop leaves you close to your intended risk; a wide one leaves you at roughly two-thirds of it. So the stop distance quietly determines not just your loss if you are wrong but how much of your risk budget you are able to deploy at all, which means a plan that ignores the lot is a plan whose real risk per trade is a number it has never calculated. The second is that this arithmetic is exactly why the 2026 rule stack is not trivia. With no broker-granted multiplier left to paper over an inconvenient answer, and with each lot committing capital in the ₹15 to ₹20 lakh band, the sum above is the whole of your position sizing. There is no longer anything else in the system to absorb a bad one. Building that chain until it runs automatically, before the bell rather than during the session, is the substance of the method we teach.

Now the wall that every version of this plan has to clear, and the one place where intraday is genuinely different in kind rather than in degree from slower trading. It is not that any individual charge is large. It is that you pay the entire stack again on the next round trip, and the one after that, and intraday is defined by doing that many times a day. Turnover is the multiplier, and intraday is the setting where the multiplier is largest.

It is worth doing the arithmetic once, properly, rather than waving at it. Take a single Nifty index futures round trip on a contract worth about ₹16 lakh, which is illustrative but sits inside the ₹15 to ₹20 lakh band the October 2024 framework mandates. Six separate charges apply, on top of slippage. Securities transaction tax at 0.05 per cent of the sell-side value. Exchange transaction charges at roughly 0.00183 per cent of turnover. The SEBI turnover fee at ₹10 per crore. Stamp duty at 0.002 per cent on the buy. GST at 18 per cent, charged on brokerage plus the exchange charge plus the SEBI fee, but not on the tax lines. And brokerage, at a representative Indian retail broker, of the lower of about 0.03 per cent or ₹20 an order. Add them up and the figure below is what you get.

The right-hand panel is the part worth sitting with, because it converts the cost into the only unit an intraday trader actually thinks in. A round trip costs about 14.6 index points in explicit charges before slippage. That is the hurdle every single trade starts behind. Four round trips a day for twenty sessions is roughly 1,172 Nifty points of cost in a month, and that number is entirely indifferent to whether you were right. It is charged on the trades that worked and on the trades that did not. Any plan that does not have an answer to it is not a plan; it is a hope with an entry rule attached. You can run your own segment, notional and frequency through the cost estimator and watch the same wall assemble itself from your own numbers.

And notice the second-order effect the figure quietly proves, because it is the single most useful thing on this page. The cost per round trip is fixed. So the only two levers that move your cost base are the size of the move you are trying to capture and how often you go. Nothing else. A trader chasing a cheaper broker is negotiating over the ₹40 line in a ₹952 bill. A trader who halves their frequency has halved the bill. This is also the honest structural argument for why the same idea can be perfectly viable at a slower cadence and hopeless intraday, which is precisely the trade-off examined in intraday versus delivery trading: the edge did not change, the number of times you paid for it did.

Put the last two sections together and the base rate stops being a scary statistic and becomes a mechanism you can watch operate. Here is the crucial move: losing money net of costs does not require being wrong. It requires only being less right, on average, than the cost of trading is wide. That is a completely different and much lower bar, and it is the reason the honest framing of intraday outcomes is about arithmetic rather than about skill at reading charts.

Watch it happen on a trade the trader got right. Same instrument as before, one Nifty futures lot on a ₹16 lakh contract. Suppose the direction call is correct by 18 index points, which is a genuine, substantial, entirely respectable result on an intraday move. The gross is ₹1,170. Then the stack arrives.

That is the mechanism. Now the number, which should be read as the base case rather than as a warning: about 93% of individual traders in equity derivatives made net losses over FY22 to FY24, aggregate net losses exceeding ₹1.8 lakh crore (SEBI, September 2024). Intraday sits at the highest-frequency, highest-cost end of that segment, which is precisely where the waterfall above is steepest. The figure is not evidence that the participants were unusually foolish. It is what you would predict from the arithmetic alone once you accept that the average retail trader is somewhere near ordinary at calling direction and is paying a stack like the one above several times a day.

There is one more subtraction most intraday plans never write down, and it belongs here because it is the last step between gross and net. Tax. As of 17 July 2026, intraday equity, meaning a position squared off without delivery, is speculative business income under section 43(5) of the Income-tax Act, 1961, taxed at your slab rate rather than at capital-gains rates, and its losses are ring-fenced so that a speculative loss can only be set off against speculative gains. Exchange-traded derivatives are carved out by the proviso to that section and are non-speculative business income instead. The distinction turns on delivery, not on holding period, which surprises people constantly, and the ring-fencing means a losing intraday year is worth considerably less against your other income than most traders assume. The full treatment, including set-off and carry-forward, is in trading taxation in India. Tax rules change with each finance act, so verify the current position at source before you rely on it.

Everything above points at one question that a serious intraday plan has to be able to answer: how would I know? Not how would I feel confident, and not how would I find a chart where it worked, but by what procedure would this plan be shown to be worth running. And the session anatomy makes that procedure specific, because it tells you exactly where the usual test goes wrong.

The usual test aggregates. It runs the rule across every day and every hour, produces one number, and reports it. But we have already established that the same rule is a different rule in each window. A rule that works in the dense windows and bleeds in the trough does not report as two facts; it reports as one mediocre average, and the average conceals both. You would conclude the idea is marginal and discard it, when what you actually had was one good rule and one bad habit wearing the same name. So the first requirement of an honest test is that it partitions by window before it aggregates, because the window is a term in the rule and you cannot average over a term.

The second requirement is that the cost wall goes in at the start, not at the end. A gross result is not an early estimate of the net result; on the numbers in the last section, it is not even the same sign. Any test that reports gross points captured is reporting a quantity that has never been shown to survive contact with the ₹952 that follows it. And the third is that the test has to be run on the rules you actually specified, which means the rules have to have been specified before you looked, or you will simply discover the window in which your idea already worked and call it a finding. The table sets out the failures this produces and the test that catches each.

Read the right-hand column and the pattern from the three specifications repeats exactly. Not one of these tests asks you to predict anything. Each simply refuses to let a claim through until the term that was quietly omitted has been put back: the window, the cost, the honesty of the specification, the arithmetic of the size, the count of the trades, the calendar. A plan that survives all six is not a plan that will make money, and nothing here promises it will. It is a plan whose failures, when they come, will be legible, which is the only property that lets you improve anything.

So put it together into the artifact itself, because the entire argument of this page collapses if the output is a feeling rather than a document. Everything above reduces to a single page written before the bell, in which every row is settled while you are calm and flat, and none of them is reopened while the market is open. It is deliberately unglamorous. That is the point: there is nothing in it you could not have written, and almost nobody writes it.

Notice what is absent from that sheet. There is no forecast in it, no view on the index, and no claim about what any of it will earn. It is a structure for making decisions under pressure, and it is agnostic about whether the idea inside it is any good, which is exactly why it is worth building first. A good idea inside a bad structure is destroyed by the structure. A weak idea inside a good structure is at least visible, which means it can be found and dropped rather than repeated for a year.

And that is the honest end of an intraday page. The session will keep its shape: an auction, a burst of repricing, a long thin middle, a busy close. The rule environment will keep moving, as it did in 2021 and 2024 and 2025 and again this January, and every change will land on the parameters of your plan rather than on your setup. The costs will arrive on every round trip whether you were right or wrong, and turnover will keep multiplying them. Against all of that, the only durable asset is the structure you built before the bell, when nothing was at stake and you could still think clearly. Master it and the question of whether any particular idea is worth trading answers itself, which is the whole of the skill and the reason this page never told you what to buy.

• SEBI equity index derivatives framework. Circular SEBI/HO/MRD/TPD-1/P/CIR/2024/132, dated 1 October 2024, which raised the minimum contract value so that lot sizes keep the contract inside the ₹15 to ₹20 lakh band (effective for new contracts from 20 November 2024) and limited each exchange to one weekly-expiry benchmark index, leaving the Nifty 50 as NSE's surviving weekly. sebi.gov.in
• The base rate for individual traders. The Securities and Exchange Board of India study of profit and loss of individual traders in the equity derivatives segment (September 2024) is the source for the figure quoted verbatim in this guide: about 93% of individual traders in equity derivatives made net losses over FY22 to FY24, aggregate net losses exceeding ₹1.8 lakh crore. sebi.gov.in
• Session structure, expiry and lot sizes. NSE's published equity market timings are the source for the pre-open call auction (09:00 to 09:15) and continuous trading (09:15 to 15:30); NSE circulars are the source for the move of derivatives expiries to Tuesday from 1 September 2025 and for the January 2026 index lot-size revisions setting the Nifty 50 lot to 65. All as of 17 July 2026; verify the current position at source. nseindia.com
• The intraday pattern in volume. Anat R. Admati and Paul Pfleiderer, A Theory of Intraday Patterns: Volume and Price Variability (The Review of Financial Studies, 1988), together with the earlier transaction-level work of Wood, McInish and Ord (Journal of Finance, 1985), established the concentration of trading volume and price variability at the open and the close with a trough between. The profile in the first figure is shaped on that documented pattern and normalised; it is illustrative and is not a measured NSE session. jstor.org
• Peak margin, and the tax treatment of intraday. SEBI's peak-margin framework, circular of 20 July 2020, phased from December 2020 in steps of 25, 50 and 75 per cent to 100 per cent upfront margin by September 2021, is the source for the removal of broker-granted intraday leverage. Section 43(5) of the Income-tax Act, 1961, defines a speculative transaction as one settled otherwise than by actual delivery, with the proviso carving out exchange-traded derivatives; this is the basis for intraday equity being speculative business income while index futures and options are not. As of 17 July 2026. incometaxindia.gov.in