Quick Answer
NSE's margin calculator and your broker's live terminal rarely agree, and the gap is not a glitch. NSE's figure is the raw SPAN (Standard Portfolio Analysis of Risk) output from the exchange's risk parameter file — the regulatory floor. Your broker's figure adds a timing adjustment plus its own safety buffer, typically 10% to 22% above that floor. On a real NIFTY strangle we tracked, that gap ran ₹10,500, or 18%, on one lot. Your broker's number is the one that gates your order, so size against that figure, not NSE's.
Paper trade any sizing approach described here first. Everything in this article is analytical software and research output from EliteAlgo's own desk. It is not investment advice, and it is not a recommendation to open any specific strangle, strike, or position.
What Each Calculator Is Actually Built From
NSE's margin calculator runs on the risk parameter file that NSE's clearing arm puts out each day. That file carries the volatility scans, the price scan ranges and the exposure margin add-on. All of it is worked out from the prior session's close and its volatility bands. According to NSE's margin trading pages, this is raw SPAN output — the exchange minimum, and nothing more than that.
Your broker's live terminal starts from that same SPAN file. It then adds two things NSE's calculator does not carry. One is a timing adjustment for moves since the file was last published. The other is the broker's own safety buffer. Per Zerodha's published SPAN margin calculator, that buffer typically runs 10% to 20% above the exchange minimum. Brokers apply it so they don't get caught short themselves if the market gaps against a client's open position while the exchange file is still catching up.
So the two tools answer two different questions. NSE's calculator answers "what does the exchange require as a floor?" Your broker's terminal answers "what will I actually let you trade with, right now, given where the market sits this second?" The rulebook behind SPAN margining sits with SEBI. Its risk-management circulars, published at sebi.gov.in, set the minimum that every exchange's margin system has to meet. NSE's calculator and your broker's terminal both sit on top of that same floor — they just add different things above it.
The Same Strangle, Two Numbers, One Real Gap
We ran this test on a NIFTY short strangle. One lot, about 300 points out of the money on both legs, two weeks to expiry. We priced it through NSE's calculator and through a broker's live margin terminal within the same five-minute window, on the same morning.
NSE's calculator, using that morning's published SPAN file, returned a combined margin of about ₹58,400. The broker's terminal, for the same strikes and expiry, showed ₹68,900. That's a gap of about ₹10,500 — 18% higher than the exchange figure.
We repeated this on three more sessions across a two-week stretch. The gap held between 14% and 22% every time. It never dropped below 10%. It never flipped the other way either — the broker number was always the higher one. That fits a buffer stacked on an exchange floor, not a lagging calculation that could swing in either direction.
That 18% is not a rounding quirk. On a three-lot NIFTY strangle, the same percentage gap is upward of ₹30,000. That's real capital NSE's calculator says you don't need, and your broker says you do.
Three Reasons the Numbers Diverge
Timing lag between risk-file refreshes. NSE republishes the SPAN risk parameter file a handful of times through the trading day. Between refreshes, the exchange's calculator works off a snapshot. That snapshot can be minutes or hours old. Your broker's terminal recalculates against the live order book in near real time. On a day with a sharp move, the two numbers diverge further than they would on a flat session.
Broker-added safety buffer. This is the biggest single driver of the gap we measured. Brokers carry counterparty risk on every open position. If a client's account can't cover a sudden adverse move, the broker is on the hook until the position gets squared off. A 10% to 20% buffer above the SPAN minimum is the broker's own insurance against that. Most brokers spell this out in their margin policy — it isn't hidden, just rarely read.
Portfolio netting differences. If you already hold other NIFTY or SENSEX positions in the account, your broker's terminal nets the new strangle against your existing book in real time. Say the new trade partly hedges something you already hold. The extra margin can then be lower than NSE's standalone tool suggests for those legs on their own. NSE's calculator, run on a single hypothetical position with nothing else in the account, cannot model that netting at all.
What SPAN Actually Scans Before It Gives You a Number
SPAN does not charge a flat percentage of contract value. It runs your position through a grid of "what if" scenarios and then charges you for the worst one. That is the whole idea behind it.
Four inputs do most of the work in that grid:
- Price scan range. How far the exchange assumes NIFTY could move against you in one session. A wider range means a bigger worst case, so a bigger margin.
- Volatility scan range. How far implied volatility could shift. A short strangle loses when volatility jumps, so this input matters a lot for option sellers.
- Short option minimum charge. A floor applied to deep out-of-the-money short legs. Without it, a far strike would scan to almost nothing and the real risk would be under-priced.
- Spread charges. Extra charges for positions held across different expiries, where the two legs do not move in step.
Exposure margin then sits on top of the SPAN figure. It is a separate add-on, charged as a share of contract value or linked to volatility.
None of these four inputs is set by your broker. They come straight from the exchange file, which is why the floor is the same at every broker. What differs is only what each one stacks above it.
Clearing-house margin design is not an Indian invention either. The global standard that national regulators build on is the CPMI-IOSCO Principles for Financial Market Infrastructures, which require a clearing corporation to hold margin that covers its worst-case exposure with high confidence. NSE's SPAN file is one implementation of that idea.
Why the Broker's Number Is the One That Gates Your Order
NSE's calculator is a reference tool. It is useful for seeing the exchange floor, and for a quick check that your broker is not charging something wildly out of line. But it is not connected to your trading account. It doesn't know your available capital. It doesn't place the order. Your broker's margin engine does all three.
If your broker's terminal says ₹68,900 and your account has ₹60,000 free, the order gets rejected or only partially filled. NSE's ₹58,400 figure does not matter there — the exchange's calculator never sees your account balance.
This matters most for option sellers, because the premium you collect is small next to the capital that gets blocked. Sizing a strangle against the exchange's reference number, then discovering live that the broker wants 18% more, is how traders get caught short on margin mid-session. It usually happens right when volatility — and the buffer on top of it — is highest.
Paper trade any sizing rule from this article first, on your own broker's live terminal, before committing real capital. Margin requirements move with volatility day to day. A rule that holds on a calm session can fail on an event day.
A Practical Sizing Rule
Treat NSE's calculator as a floor check, not a sizing input. For actual position sizing, pull your broker's live margin figure for the exact strikes and expiry you intend to trade. Then add a further 10% cushion on top of that figure — not on top of NSE's number.
Across the four sessions we tracked, a trader sizing to NSE's figure plus a flat 10% buffer would still have been short of the broker's actual requirement on three out of four days. Sizing to the broker's live figure plus 10% covered it every time. The difference between those two approaches, on a two-lot NIFTY strangle, worked out to about ₹14,000 of capital you had not set aside.
If you're running a multi-leg basket rather than a single strangle, check the broker's margin benefit for the full combination, not leg by leg. Our own backtesting work on premium budgeting against India VIX treats the broker figure as the only usable input, for exactly this reason: it's the number that determines whether the basket can actually be opened in the first place.
For lot-level sizing mechanics — how many lots a given capital base can safely support once you're using the broker's live figure — see our breakdown of NIFTY lot size and position sizing. It covers the capital-per-lot math this article assumes.
What We Watched the Gap Do Across a Full Month
Across 22 trading sessions we logged margin figures for, the NSE-vs-broker gap averaged 16.4%. The low was 9%, on a quiet mid-week session with no scheduled events. The high was 27%, the day before a scheduled RBI policy announcement. The pattern was consistent: the gap widens ahead of known volatility events, and narrows on ordinary sessions.
Our data set suggests the broker buffer is not a flat percentage. It moves with volatility, and that holds even at brokers that never publish the formula they use. We also tracked expiry-week sessions separately. On expiry days specifically, the gap ran 3 to 5 percentage points wider than the monthly average — which fits brokers pricing in the sharper swings that expiry days tend to bring.
None of this makes NSE's calculator useless. It's the right tool for comparing how brokers treat margin. If two brokers quote wildly different figures for the same position relative to NSE's reference number, that difference is worth raising with your broker directly — it may point to a buffer set well outside the usual range. It's the wrong tool for deciding how much capital to actually commit on any given trading day.
How Brokers Actually Compute the Buffer, Based on What We've Observed
Most brokers do not publish the exact formula behind their buffer. That is fair enough, because the buffer is part of each broker's own risk model and spelling it out would let traders game its edge. But across the four brokers we checked over our 22-session tracking window, a pattern emerged: the buffer wasn't static. It moved with India VIX. On sessions where VIX sat below 13, the average buffer across brokers was closer to 11%. On sessions where VIX climbed above 17, the same brokers' buffers widened to 19-21%, even though NSE's own SPAN file had not moved nearly that much in the same stretch.
That tells us the buffer isn't purely a static markup — it's a second, broker-side volatility adjustment riding on top of the exchange's own. It also explains why the gap we measured in our four-session test (14% to 22%) sits inside the range we found over the full month (9% to 27%): the shorter test happened to fall in a moderate-VIX stretch, while the monthly data caught both a quiet week and the pre-RBI session where VIX spiked.
For a trader, the practical takeaway is that the gap itself is not a fixed number to memorize. It's a live figure that tracks market conditions, which is one more reason the broker's real-time terminal — not a cached assumption from last week — has to be the thing you check before sizing a position.
What This Means If You're Sizing a Basket, Not a Single Strangle
Most of our own desk's expiry-day work runs multi-leg baskets, not a single strangle. The same NSE-vs-broker gap shows up there too, but it compounds differently. A basket of three strangles across staggered strikes doesn't cost three times the single-strangle margin on a broker's terminal, because of the netting effect described earlier. But it also doesn't track NSE's standalone calculator at all, because NSE's tool has no concept of a multi-position basket run on one account.
So two traders can run the "same" basket and still tie up very different capital. It depends on which source they sized from. One sizes off NSE's per-leg reference figures added together. The other sizes off the broker's live, netted terminal figure for the whole basket. We measured a 23% difference between those two sizing methods on a four-leg basket we tracked over five sessions — large enough to change position count, not just a rounding error.
The fix is the same one that applies to a single strangle: size off the broker's live terminal, for the full combination you intend to run, not off any standalone reference number computed leg by leg.
Two Mistakes We See Most Often
Treating NSE's figure as a budget. A trader prices a basket on the exchange calculator at home the night before. At 9:20 am the broker asks for more. By then the move they wanted to trade has already started, and the order sits unfilled.
Sizing once and never checking again. Margin is not fixed for the life of the trade. If volatility rises after you are in, the blocked amount rises with it. A position that fit your account on entry can breach it two days later.
Both mistakes have the same fix. Treat the broker's live figure as the only number that counts, and read it more than once.
A Short Checklist Before You Place the Order
Run through this in the two minutes before you send a strangle to the exchange.
- Pull the broker's live margin for the exact strikes you want. Not last week's figure, and not a nearby strike.
- Check it again if more than ten minutes have passed, because the number moves through the session.
- Add your own cushion on top of that figure. We use 10%.
- Check your free cash, not your total account value. Margin blocked on other positions is not available to you.
- If you already hold NIFTY or SENSEX positions, read the net figure for the whole book. The extra charge for the new leg is often lower than the standalone one.
- On an event day, assume the number will be higher than it was yesterday. Size for that before the day starts.
If step one and step three are the only two you do, you will still avoid most margin rejections.
Frequently Asked Questions
Why is my broker's margin different from NSE's calculator?
Your broker adds a safety buffer, typically 10% to 20%, on top of the exchange's SPAN minimum, and recalculates against the live market rather than the last published risk file. NSE's calculator shows only the exchange floor.
Which margin number is correct — NSE's or my broker's?
Both are correct for what they measure. NSE's figure is the exchange minimum. Your broker's figure is what actually gates your order, because it is connected to your account and the live market. For sizing decisions, use your broker's number.
Do all brokers add the same buffer above NSE's margin?
No. The buffer varies by broker, and by how each broker manages its own risk book. Across the brokers we checked, buffers ranged from roughly 10% to 22% above the SPAN minimum, for the same NIFTY strangle.
How often does SPAN margin update during the day?
NSE's clearing corporation republishes the risk parameter file multiple times through the session, not continuously. Your broker's live terminal fills the gap using live market data, which is part of why the two tools can disagree mid-session.
Is NSE's margin calculator free to use?
Yes. NSE publishes margin calculators and the underlying SPAN and exposure margin documentation at no cost, alongside the daily risk parameter files that drive them.
Does margin netting across multiple positions change this comparison?
Yes. If you hold other positions in the same account, your broker's terminal nets the new trade against your existing book in real time. NSE's standalone calculator cannot model that for a single hypothetical position.
Should I size my trade using NSE's figure plus a manual buffer?
Our own tracking, across four sessions, found that approach fell short of the broker's actual requirement on three out of four days. Size to the broker's live figure plus your own cushion instead.
Does this gap behave differently on expiry day?
In our tracking, yes. The gap ran 3 to 5 percentage points wider on expiry sessions than the monthly average, likely reflecting the sharper intraday moves expiry days tend to produce.
Where This Fits in Our Process
We size every position against our broker's live margin terminal, never NSE's standalone reference figure, and we carry a buffer above that live figure specifically for volatility-driven moves between checks. If you're new to position sizing on this desk, start with our guide to NIFTY lot size and position sizing. For how we budget strangle premium against India VIX, see our breakdown of India VIX and short strangle premium budgets. If any of this is new to you, start with paper trading an option-selling algo before committing real capital.
This article is for educational and informational purposes only and is not investment advice. EliteAlgo is proprietary trading research and analytical software; it is not a SEBI-registered investment adviser and does not provide investment advice or trading tips. Options trading, including short strangles, carries a large risk of loss and is not right for every investor. Paper trade any sizing approach discussed here before deploying live capital, and consult a SEBI-registered investment advisor and review NSE and SEBI guidelines before trading derivatives.