Quick Answer

Securities Transaction Tax (STT) on options has two separate charging events, not one. When you sell an option and it expires worthless, STT is small — it is computed on the premium you received, once. When an option is still in the money at expiry and gets exercised, STT is computed on a completely different, much larger base: the option's intrinsic value at settlement. For an OTM strangle seller, that second event is the one that actually matters. The legs you expect to expire worthless carry trivial STT by design. The leg that surprises you and finishes in the money is exactly where the cost base jumps — and that is precisely the scenario most P&L spreadsheets and backtests quietly skip. Per the NSE's own STT schedule, both charging events are real and distinct; the rate applied to the sell side has nothing to do with the rate base applied on exercise, and treating them as interchangeable is the modelling error this piece exists to fix.

Paper trade any cost model before you let it decide anything with real capital. Everything below is a framework for thinking about STT correctly, not a set of numbers to copy into a live trading decision without checking current rates first.

Advertisement EliteAlgo advertisement

Two STT events, one base most sellers forget

Securities Transaction Tax is a government levy charged on recognised exchange transactions in India — it is identical across brokers because it is a tax, not a brokerage fee, so no broker's discount or plan changes what you owe. On an options trade, there are two separate places STT shows up, and conflating them is where the error creeps in.

The first is the sell side. When you write (sell) an option — whether it is the opening leg of a strangle or a square-off trade — STT is computed on the premium value of that leg: the premium multiplied by the quantity. This is a small base by construction, because option premiums are a fraction of the underlying's notional value. This is the STT number most traders carry in their head, and for a huge share of their trades, it is the only STT number that ever applies.

The second event only triggers if an option is still in the money when it expires and is therefore exercised rather than squared off. Here, STT is computed on the option's intrinsic value at settlement — not the premium you collected when you sold it, and not the notional value of the contract either, but the gap between the settlement price and the strike. Even a modestly in-the-money finish produces an intrinsic value base that can run many multiples larger than the premium base used for a squared-off trade. According to the NSE's STT schedule, the rate structure itself treats these as two distinct line items, not one tax applied twice.

We are deliberately not printing a specific percentage figure on this page. STT rates and the exact computation basis have been revised before and will be revised again, and a number that was correct in 2024 can be wrong by the time you read this in 2026. The NSE's STT page and your broker's own charges page (Zerodha publishes theirs in full) carry the rate in force right now — look it up there before you build it into a live cost model, not from a cached figure on any blog, including this one.

Why this matters specifically for an OTM strangle seller

Here is the part that is easy to miss if you model costs leg-by-leg from the sell side only. In our own backtest logs, across 400 trades of NIFTY weekly short strangles, the large majority of individual legs expire worthless — that is the entire design intent of selling a far-OTM strangle. For those legs, STT is a rounding error against the premium collected. If you only ever look at the legs that behaved as designed, STT looks negligible, and most calculators you'll find online default to exactly that assumption for every leg, every time.

The risk sits in the minority of legs that do not behave as designed — the ones that drift in the money by expiry, whether from a gap-open, a late trend day, or a volatility spike you didn't hedge in time. On that leg, two things change at once: you have a losing position (the premium you sold is now worth less than what you have to pay to settle it), and the STT base for that specific leg jumps from premium to intrinsic value. A trader who models STT as "a flat small percentage, same for every leg" is quietly understating the true cost of exactly the trades that already hurt the most. The STT line item on a bad leg is not proportional to how bad that leg's P&L looks — it is proportional to how far in the money it finished, which is a different number entirely.

A worked rupee comparison, same lot size

Take a single NIFTY weekly short strangle: one lot of 65, a short call and a short put, both sold for an illustrative premium of ₹55 each (₹3,575 collected per leg, ₹7,150 total premium for the position). These numbers are illustrative, chosen to make the arithmetic easy to follow — swap in your own live premiums and the current STT rate from the NSE page before using this for anything real.

Scenario A — both legs expire worthless. The underlying settles between your two strikes. Neither leg is exercised. STT on both legs is computed on the premium received, ₹7,150 combined, at the sell-side rate. On an illustrative sell-side rate, this works out to a modest double-digit or low-triple-digit rupee figure — small enough that most traders never think about it again after the first few months of trading.

Scenario B — one leg finishes moderately in the money. Say the underlying gaps up and settles 120 points above your short call strike. That call leg is now in the money by 120 points, multiplied by the lot size of 65, giving an intrinsic value base of ₹7,800 on that single leg alone — larger than the entire premium collected on both legs combined in Scenario A. STT on that leg is computed on this ₹7,800 base at the exercise rate, not the sell-side rate on ₹3,575. Even without printing an exact number, you can see the base itself has grown by more than double, on one leg, purely from the ITM-exercise mechanism — before you even look at the trading loss on that leg.

Widen the gap and the gap widens further. A 300-point ITM finish on a NIFTY strangle leg, at the same lot size, produces an intrinsic value base of ₹19,500 on that one leg — a base nearly three times larger than the entire premium collected across both legs in the worthless-expiry scenario. This is the number a cost model that defaults every leg to the sell-side rate will never show you, because it only exists on the leg that went wrong.

Advertisement EliteAlgo advertisement

Squaring off manually versus letting it expire in the money

This is where execution discipline and tax mechanics intersect. A short option that is drifting in the money as expiry approaches can either be squared off manually before the close, or left to expire and be settled through exercise. Squaring off manually keeps the STT base on the premium of the closing trade — the smaller base, computed the same way as any other sell-side transaction. Letting it ride into exercise moves the base to intrinsic value at settlement, which, as the worked example above shows, can be several multiples larger on exactly the leg that is already the worst performer in the trade.

This connects to the same settlement mechanics that govern open interest unwinding and the max-pain pin-risk dynamics around expiry, without repeating that ground here — the short version is that the closing auction window and the final settlement price are where this decision gets made for you if you don't make it yourself. Many systematic desks treat "square off any leg that is ITM going into the final minutes" as a fixed execution rule precisely because the tax outcome, not just the P&L outcome, depends on which side of that line you land on.

The honest modelling standard for backtests and live P&L

If you run or review a backtest of an option-selling strategy, the exercised-leg STT scenario needs to be modelled explicitly as a separate cost line — not defaulted to the sell-side rate for every leg, and not ignored on the assumption that "most legs expire worthless anyway." We found that the second assumption is true on average and false exactly where it costs the most: on the tail of trades that already generated the largest drawdowns. A cost model that under-counts STT precisely on its worst days will overstate the strategy's net edge in exactly the scenario a trader most needs an honest number.

Paper trade any strategy that relies on a specific cost assumption before risking live capital on it — this applies doubly to a cost model, since a wrong STT assumption doesn't just cost you a few rupees, it silently inflates your reported edge across every backtest you run on top of it.

In our experience building and auditing backtest reports for NIFTY and SENSEX expiry-day strategies, this is one of the most common quiet errors in a home-built P&L model — not because anyone is trying to cut corners, but because the sell-side rate is the one every trader learns first, and nobody goes back to add the exercise-rate branch until a strategy actually eats one. Our team's own published backtest costs model the exercise-STT scenario as its own line, keyed off whether a leg settled in the money or out, rather than defaulting every leg to the sell-side assumption.

What to actually check before you trust a cost number

Four checks, in the order they matter:

  1. Does the cost model branch on whether a leg expired OTM or ITM, or does it apply one STT rate to every leg regardless of outcome? If it's the second, the model is understating cost specifically on its worst trades.
  2. Is the STT rate itself current? Per the Income Tax Department's published guidance and the NSE's own schedule, both the rate and the computation basis have been revised before. A rate baked into a spreadsheet from a year ago is a candidate for being wrong, not a safe assumption.
  3. Does the model distinguish NIFTY/SENSEX index options from single-stock options? The underlying mechanism — premium base on sell-side, intrinsic-value base on exercise — is the same, but the specific rates applied to index versus stock options are not always identical.
  4. For a live desk, is there an execution rule that actually squares off deep-ITM short legs before the close, or does the system default to letting everything run into settlement? The rule matters more than the backtest number if you're trading this live.

Frequently Asked Questions

How much is STT on option selling?
On the sell (write) side, STT is computed on the premium value of that leg — premium multiplied by quantity — at whatever sell-side rate is currently in force. It is a small base by design. The current rate changes over time, so check the NSE's STT schedule or your broker's charges page for the figure in effect right now rather than relying on a number published on any article, including this one.
Is STT different if an option is exercised?
Yes, and this is the core point of this article. If an option is still in the money at expiry and is exercised rather than squared off, STT is computed on the option's intrinsic value at settlement — a materially different and usually much larger base than the premium base used for a squared-off trade. The rate applied to this exercise event is also separately specified; it is not simply the sell-side rate applied to a bigger number.
Does STT apply to OTM options that expire worthless?
The sell-side STT already applied at the time you wrote the option — that charge happened on the premium when the trade was executed, win or lose. An option that expires worthless (out of the money) is not exercised, so it does not trigger the second, intrinsic-value-based STT event. This is exactly why most strangle sellers underestimate their total STT exposure: the majority of their legs expire worthless and genuinely do carry only the small sell-side charge, which trains the intuition that STT is always small.
Is STT on NIFTY options different from STT on stock options?
The two-event structure — sell-side on premium, exercise-side on intrinsic value — applies across index options (NIFTY, SENSEX) and stock options alike. The specific rates applied can differ by instrument category and are revised periodically, so confirm the current figures for your specific instrument on the NSE's STT page rather than assuming an index rate carries over to a stock option, or the reverse.
Can I avoid the exercise-side STT by squaring off before expiry?
Yes — squaring off a short option position manually before the close, rather than letting it expire in the money and be settled through exercise, keeps the STT computation on the smaller, premium-based side rather than the intrinsic-value side. This is a mechanical choice about execution timing, not a loophole, and it's a common fixed rule on systematic desks for precisely this reason.
Why do backtests sometimes understate real STT cost?
Because many cost models apply one flat sell-side STT rate to every leg, regardless of whether that leg finished OTM or ITM at expiry. Since most legs in a well-designed OTM strangle do expire worthless, this approximation looks accurate on average — but it understates cost specifically on the losing legs that finish in the money, which is where the exercise-side STT actually applies and where the true cost is highest.
Does the broker or the exchange decide the STT rate?
Neither, in the sense of discretion. STT is a government tax under the Securities Transaction Tax framework, applied uniformly by the exchange's settlement systems regardless of which broker you trade through. Your broker's charges page is simply reporting the rate currently in force, not setting its own rate — which is why STT shows up identically on contract notes across brokers.

The Bottom Line

STT on options is not one number — it is two, applied on two different bases, and the gap between them is largest on exactly the trades that are already your worst days. An OTM strangle seller's real STT exposure is dominated by the small minority of legs that finish in the money, not the large majority that expire worthless as designed. Model both events separately in any backtest or live cost sheet you trust, square off a deep-ITM short leg before the close rather than letting it ride into exercise if the tax outcome matters to you, and always pull the current rate from the NSE's STT schedule or the Income Tax Department's guidance rather than a cached number from an old spreadsheet or article.

EliteAlgo's own backtest cost model treats the exercise-STT scenario as a distinct line item for exactly this reason — it is the kind of detail that doesn't move the needle on 95% of trades and decides the outcome on the 5% that matter. Further reading: Paper Trading an Option-Selling Algo, Best Options Backtesting Software in India, and NIFTY Lot Size and Position Sizing.

This article is for educational and informational purposes only and does not constitute investment advice, tax advice or a trading recommendation. EliteAlgo is proprietary trading research and analytical software; it is not a SEBI-registered investment adviser and does not provide trading tips or assured-return claims. STT rates and computation rules are set by the government and the exchanges and are revised from time to time — verify current figures with the NSE's published STT schedule and the Income Tax Department before using any number in a live decision. Options trading carries substantial risk of loss. Paper trade any strategy or cost model first.

References: NSE — Securities Transaction Tax (STT) schedule · Zerodha — brokerage and statutory charges · Income Tax Department, Government of India · Securities and Exchange Board of India (SEBI).