The Covered Call: Pricing the Upside You Are Selling

The payment arrives first, and the cost arrives later

Consider the sequence of events. On the day you write the call, money lands in your account. It is a real number, it has a date, and you can see it on a statement. Six weeks later the share has run past the strike and the position is closed at the strike price. Nothing has gone wrong. There is no loss to record, no red figure, no moment that registers as a cost. The cost is the difference between what you received and what you would have received, and that second quantity does not appear anywhere, because it never happened.

This asymmetry in visibility is the whole reason the structure is the most widely mis-sold arrangement in retail options. Both halves of the exchange are perfectly real, but only one half is observable. A person who writes calls for two years can point at every payment received and cannot point at a single instance of the thing they gave up, because the thing they gave up has no receipt. Human memory does not keep a ledger of counterfactuals, and no broker statement keeps one either.

The second reason is that the structure is genuinely comfortable to hold. It wins more often than it loses, which is a property of a truncated distribution and not evidence of an edge, and the months in which it disappoints are months in which the share went up, which is not a mood in which people audit their positions. A structure that makes you feel prudent most of the time and mildly foolish occasionally is one that survives a long time without ever being examined.

None of that tells you whether the exchange is a good one. To know that, you have to price the thing you sold, and pricing it is entirely possible: it is an option, it has a market, and the expected value of its payout can be computed directly. That is what the rest of this page does. The result is not the one a critic of the structure would want, and it is not the one a promoter would want either.

A note on what this page is not. It does not argue about whether a covered call is a hedge, or set it against a protective put, because our guide to hedging in trading already makes that qualitative case at length: that the payment cushions only a sliver of a decline, that the label is routinely misapplied, and that the direction the premium flows tells you which side of an insurance contract you are on. That argument is settled there and is not repeated here. This page takes the same structure and computes it.

The context in which any of this sits is worth stating plainly, once. About 93% of individual traders in equity derivatives made net losses over FY22 to FY24, with aggregate net losses exceeding ₹1.8 lakh crore (SEBI, September 2024). The covered call is among the most conservative arrangements available in that market, and it is still an options position with a capped outcome, an assignment mechanic and a charge stack, held against a single company.

The position, stated in full so it can be checked

Everything on this page comes from one configuration, run once, seeded so it reproduces. The underlying is a simulated share rather than a real one, deliberately: it keeps the assumptions visible and avoids implying that any particular Indian company would have produced any particular outcome.

Two rows in that table carry more weight than the rest, and it is worth being explicit about why. The assumed drift decides how valuable the surrendered upside is, because the upside is a call and a call is worth more when the underlying is expected to rise faster. The realised volatility decides the same thing from the other direction. Between them they determine the entire answer, and neither is knowable in advance. Everything else in the structure, the strike, the tenor, the lot, moves the result by far less.

The lot deserves a caution. Exchanges size derivative lots to keep contract value inside a prescribed band, and that band was raised in the 2024 revision of the derivatives framework. Lot sizes were increased in November 2024 and revised again from January 2026, in several cases downward, as prices moved. Any specific lot number is therefore a snapshot rather than a fact with a shelf life. The 1,500 shares used here follow from choosing a ₹15,00,000 contract value and a ₹1,000 share, both illustrative. Our note on lot sizes and the contract value band covers the mechanism.

Both payoffs, computed

Start with the expiry picture, which is the part everyone has seen, and read the three numbers on it that are rarely quoted.

The covered line sits above the outright line everywhere below ₹1,065.20 and below it everywhere above. That crossing point is the strike plus the payment, and it is the single most useful number in the diagram, because it is the threshold at which the structure stops helping. It sits 6.52 per cent above the entry price. Any month in which the share gains more than 6.52 per cent is a month in which you would have been better off doing nothing.

The maximum the position can make is ₹65.20 a share, or ₹97,801 on the lot, and it is a hard ceiling rather than a soft one: at ₹1,200 the position makes ₹65.20, and at ₹2,000 it makes ₹65.20. The breakeven moves down from ₹1,000 to ₹984.80, an improvement of exactly the payment and not one paisa more. The maximum loss is ₹984.80 a share, or ₹14,77,199 on the lot, which is what happens if the company goes to zero. That figure is not a quirk of the option. It is the shares, and it was there before the call was written.

What the diagram does honestly is show the shape. What it cannot do, and what almost every presentation of this structure relies on it not doing, is tell you how likely each part of that shape is. The flat section on the right looks like a small region of the picture. Whether it is a small region of your actual outcomes is a completely different question, and the diagram is silent on it. A payoff chart is a photograph of one day with the probabilities removed.

There is a related sleight of hand in how these charts are usually drawn. Extend the horizontal axis far enough to the right and the flat cap looks like a modest concession at the edge of the frame; truncate it just past the strike and the cap barely registers at all. The same position can be made to look generous or brutal purely by choosing where to stop drawing. Nothing about the position changes. Only the framing does, which is a reason to distrust the diagram as a summary and to ask for the distribution instead.

What the payoff diagram cannot show you

Run the same position across 200,000 simulated months and the missing dimension appears. This is the exhibit the whole page exists for.

Look at the left half first. The two distributions are the same shape. Not similar, the same: below the strike the covered position is arithmetically identical to holding the shares plus a fixed constant, so the entire left side is the outright distribution slid across by 1.52 percentage points. Among the simulated months that finished more than 10 per cent down, the covered position averaged a loss of 11.65 per cent against 12.51 per cent for the shares alone. The worst month of the 200,000 was a loss of 30.26 per cent against 31.78 per cent. The payment is a constant, and a constant is not a defence against a variable.

Now look at the right half, where there is nothing. 59,629 of the 200,000 months, 29.8 per cent of them, finished at exactly the cap and not a rupee above it. The green bar at the cap runs off the top of the chart and has to be drawn with a break in it. Everything to the right of that bar in the gold outline, the entire upper tail of holding the shares, has no counterpart in the covered distribution at all. Among the months that finished more than 10 per cent up, holding outright averaged 14.62 per cent and the covered position averaged 6.52 per cent, because 6.52 is all it can average there.

Read that table one column at a time and a pattern emerges that is worth naming. Every measure on which the covered position looks better is a frequency measure: how often it finished ahead, where the middle of the distribution sits, how tightly the outcomes cluster. Every measure on which it looks worse is a magnitude measure: the mean, the best case, the size of what is available in the top 5 per cent. That is not a coincidence, it is the definition of what the structure does. It buys frequency with magnitude.

This also explains why the structure is so persuasive in conversation. Frequency is what people notice and remember. Someone who has written calls for a year can truthfully say that most months worked, and they are describing a real property of the distribution. The mean, which moved the other way, is invisible to experience, because nobody experiences a mean. You experience a sequence of individual months, and in a clear majority of them the structure did what it said.

The difference in the means is small in percentage terms and specific in rupee terms. Across the 200,000 simulated months the covered position averaged 0.15 percentage points less than holding the shares, which on the illustrative ₹15,00,000 contract is about ₹2,250 a month. That figure is the same quantity, arrived at by an entirely separate route, as the pricing comparison in the next section, and the two agree to within the sampling noise of the simulation. It is worth stating clearly what it is: it is not a fee, and no one is charging it. It is the difference between what was received and what was given up.

What the upside was actually worth

The payment can be compared directly against the thing it bought, because the thing it bought is a call option and the expected value of a call's payout is a computable quantity. The comparison has an exact answer at one particular assumption and it is worth understanding why.

At an assumed drift equal to the risk free rate, the present value of the payout promised to the call buyer is ₹15.20, and the payment received is ₹15.20. They are identical, to the last decimal the computation carries. This is not a numerical coincidence and it is not a property of this particular strike. It is what an option price is. The price of the call already is the discounted expected cost of the obligation, computed under the assumption that the underlying drifts at the risk free rate. Selling it at that price is an exchange of exactly equal values, by construction.

Which means the whole question reduces to a single further one: does the share you hold drift faster than the risk free rate? If it does not, the payment is generous. At a zero drift assumption the surrendered upside is worth ₹13.62 and the payment covers 111.6 per cent of it. If it does, the payment is short. At 12 per cent the surrendered upside is worth ₹16.64 and the payment covers 91.3 per cent, a shortfall of ₹1.44 a share, about ₹2,162 on the illustrative lot. At 18 per cent the payment covers 82.9 per cent.

Stated plainly: the excess return that equities are held for in the first place is precisely the quantity a covered call sells. The option is priced as though the share drifts at the risk free rate, because that is how options are priced and it is the correct way to price them. The person writing the call is generally holding the share because they expect it to do better than that. Those two beliefs cannot both be acted on for free, and the covered call is the arrangement in which the difference between them is handed to somebody else in exchange for a fixed sum.

Now the complication, which pushes the other way and is roughly the same size. The argument above assumes the volatility that arrives equals the volatility the option was priced at. If the option is priced at a volatility higher than the one that materialises, the payout is smaller than the pricing implied and the writer keeps the difference. In this position that effect is powerful: dropping realised volatility from 28 to 24 per cent, four points, cuts the value of the surrendered upside from ₹16.64 to ₹12.56, a swing of ₹4.08 a share. That is more than a quarter of the entire payment, produced by a four point change in one assumption.

Put the two effects side by side and the honest answer emerges. At a 12 per cent drift the writer is behind by ₹1.44 a share. A gap of roughly 1.4 volatility points between the price and the outcome is enough to cancel that exactly. So the exchange is close to fair, and the direction it tips is decided by a quantity that is not observable in advance, is different for every underlying, and moves around over time. Anyone who tells you the structure reliably pays is asserting a volatility gap they have not measured. Anyone who tells you it reliably costs is asserting the same thing with the opposite sign.

One methodological confession belongs here, because it changed the conclusion. The first version of this model set realised volatility four points below implied as its base case, on the strength of the widely repeated claim that implied usually exceeds realised. With that assumption in place the covered position came out ahead in every regime tested, including a strongly rising one, and the page would have reported that as a finding. It was not a finding. It was the assumption, restated: four volatility points is worth ₹4.08 a share against a payment of ₹15.20, so the input had decided the output before a single path was drawn. The base case now sets realised equal to implied, which asserts nothing, and the volatility gap appears only as a labelled sensitivity, the dashed line in the figure above.

Doing it again, every month, for two years

A single write is not how anyone actually runs this. The structure is run as a programme: write a call, let it expire or get assigned, write another. Repetition changes the arithmetic, because the cap resets every month while the exposure to a decline never does.

In the rising regime the median sequence ends 3.7 percentage points behind simple holding after 24 months, about ₹55,716 on the illustrative contract, and the mean sequence ends 10.2 points behind. The mechanism is visible in the accounting: across those 24 months the position surrendered 44.1 points of value to the cap and collected 36.5 points in payments. The payments did not cover the surrender, and they could not, because in a market that keeps rising the call finishes in the money again and again. 7.7 of the 24 calls did.

In the flat regime the same programme ends 7.7 points ahead of holding, and in the falling regime 11.0 points ahead. This is not a small effect and it should not be buried: in two of the three regimes tested, the repeated covered call was the better outcome by a clear margin, and it was ahead in 71.3 and 86.7 per cent of individual paths respectively. In the rising regime it was still ahead in 44.8 per cent of paths, which is close to a coin toss rather than a rout.

What the three panels together show is not that the structure is bad. It is that the structure is a bet on the market not rising much, dressed as a way of being paid to hold shares you already own. Those are the same position with different labels, and the label matters because it determines who ends up holding it. A person who believes the share will drift sideways and writes calls against it has taken a coherent view and expressed it efficiently. A person who is bullish on the share and writes calls against it to be paid while they wait has taken two contradictory positions and will discover the contradiction in exactly the scenario they were hoping for.

There is a second effect in the repeated case that the single month picture cannot show. After each assignment the shares are gone and re establishing the position means buying them back at the new, higher price and writing a new call struck 5 per cent above that. The cap ratchets upward, which sounds like a benefit and is not, because the ratchet only ever happens after the move it would have captured has already been surrendered. You are permanently one step behind the price, and each step costs a full set of charges.

Assignment, and the bill for being right

Which brings the charge stack into view, and it lands in an unintuitive place. The option leg is almost free and the equity legs are not, so the cost of running this structure is concentrated entirely in the event of the share going up.

Writing the call costs the transaction tax on the sale of an option, 0.15 per cent of the premium, which the Finance Act 2026 raised from 0.10 per cent with effect from 1 April 2026. On a premium of ₹22,801 that is ₹34.20, or 0.15 per cent of the payment received. As a drag on the structure it is negligible, and if the call expires worthless that is the entire cost of the month.

Assignment is a different order of magnitude. A single stock option in India is physically settled, so an in the money call is not cashed out, it is delivered against: the writer hands over the shares at the strike. That delivery is an equity sale and pays the full equity charge stack. On ₹15,75,000 of shares delivered at ₹1,050 the computed cost is ₹1,638, being ₹1,575 of transaction tax at 0.1 per cent, ₹48.35 of exchange transaction charges at the verified NSE cash rate of 0.00307 per cent, ₹1.57 of regulator turnover fee at 0.0001 per cent, a ₹3.50 depository debit and ₹9.62 of goods and services tax at 18 per cent on the chargeable components. Buying the shares back to re establish the position costs a further ₹1,870, which includes ₹236.25 of stamp duty at 0.015 per cent on the purchase leg. The round trip is about ₹3,508, which is 15.4 per cent of a month's payment. Every figure here is illustrative.

Rolling instead, meaning buying back the written call before expiry and selling a later dated one, touches only the option leg. There is no transaction tax on buying an option back, and the new call sold attracts the same 0.15 per cent, so the computed cost is about ₹34 against ₹3,508. The difference, about ₹3,474, is roughly a hundredfold. In the rising regime, where 7.7 of 24 calls finished in the money, the assignment charges alone come to about ₹13,507 a year against ₹2,73,610 of payments collected, or 4.9 per cent of them.

There is a tax consequence too, and it runs the same direction. Being assigned is a disposal of the shares: it realises whatever gain has accumulated on them and resets the holding period. Rolling the call is not a disposal, because the shares never move. The rates that apply to a listed equity disposal were revised by the Finance (No. 2) Act 2024, which amended sections 111A and 112A for transfers on or after 23 July 2024, but this page does not reproduce the rates because the official source pages could not be retrieved directly for verification. The mechanism is what matters here and the mechanism is not rate dependent: assignment converts an unrealised position into a realised one at a moment chosen by the option, not by you.

Three components are missing from every figure above and all three push the same way, so the computed costs are a floor rather than an estimate. Brokerage is commercial, varies by provider and is excluded entirely. The exchange transaction charge on the equity derivatives segment is not in our verified set for the relevant exchange and is excluded, as is the regulator turnover fee on that segment. The treatment of transaction tax on a physically settled assignment specifically, as distinct from an ordinary delivery sale, is also unverified here and has been computed at the ordinary delivery rate. Anyone running this should read their own contract note rather than this page. The settlement mechanic itself is covered in our note on the risks of selling options in India.

The floor under the capital, and the concentration behind it

The structure has an entry requirement that has nothing to do with the option and everything to do with the word covered. To write one call against a single company's shares, you must hold a full lot of those shares outright.

That sets a floor, and the floor is the contract value, not the premium. At the illustrative ₹15,00,000 contract the payment is ₹22,801, which is 1.52 per cent of the capital it commits. At every contract size the proportion is the same, because both scale with the share price: the payment is always about a hundredth and a half of the money on the table. The most the position can make in the month is 6.52 per cent of it.

The part that is easy to miss is the concentration. This is not ₹15,00,000 of diversified equity with a call written over it. It is ₹15,00,000 in one company, because the option can only be covered by the specific shares it is written on. A holder who would never put that much of their capital into a single name may nonetheless arrive at exactly that position by starting from the question of how to write a call, and the lot requirement will have made the allocation decision for them. The capital floor is really a concentration floor wearing different clothes.

Index options avoid the physical settlement problem, because index derivatives are cash settled and nothing is delivered. They do not solve the covering problem, because there is no way to hold the index itself: covering an index call requires a basket that tracks it, and any tracking difference between the basket and the index is a risk the payoff diagram does not contain. That is a different structure with a different failure mode, and it is not the one computed here.

Where the structure behaves well, and where it quietly does not

Gathering the computed results into the conditions that produce them gives a compact summary. Nothing in this table is a recommendation about when to hold the position; it is a description of how the arithmetic responds to circumstances.

What the payment actually buys

The covered call is not a swindle and it is not free money, and the reason it is so persistently misunderstood is that people reach for one of those two descriptions when the truth is a third thing that is harder to hold in mind. It is an approximately fair exchange of one distribution for another. You give up the right hand tail and you receive a fixed sum, and at a fair price those two are worth about the same. What you get for the trouble is not value. It is shape.

Shape is a real thing to want. A narrower distribution is easier to live with, easier to size, and less likely to produce the kind of month that makes someone abandon a position at the worst moment. If a holder intends to keep the shares regardless and would genuinely prefer a tighter range of outcomes to a wider one with the same centre, the structure delivers precisely that, and the computed cost of the delivery is small. That is a legitimate use and this page has not found anything against it.

What the arithmetic will not support is the version in which the payment is treated as a return on the shares, arriving on top of whatever the shares do. It does not arrive on top of anything. It is the sale price of the best outcomes, collected in advance, and in the months when those outcomes would have arrived it is all you get. The distinction sounds pedantic until you have written calls through a rising market, at which point it becomes the only thing that matters.

The practical test that falls out of all this is a single question, and it is not the one most people ask. The usual question is how much the call pays. The useful question is what volatility the call is priced at, compared with the volatility that underlying actually tends to deliver, because that comparison is the entire margin in the structure and everything else is arithmetic around it. Our guide to implied volatility covers how to read the first half of that comparison; the second half you have to measure yourself.

None of this requires sophisticated mathematics. It requires computing the thing you are selling instead of describing it, keeping the counterfactual in the ledger next to the receipt, and being willing to publish the version of the answer that complicates your own argument. Those are habits rather than techniques, and they are the substance of what serious work on options actually consists of. If the arithmetic on this page was the interesting part rather than the tedious part, that is the method we teach.

Every figure comes from a single deterministic model, seeded so it reproduces identically on each run, written in Python with numpy and no other numerical library. The call is priced by the Black and Scholes formula at the stated volatility and rate. The present value of the surrendered upside is computed in closed form under a stated real world drift, and the closed form is checked two ways: at a drift equal to the risk free rate it reproduces the option price exactly, and an independent simulation of four million paths reproduces it to within 0.22 per cent. A third check reconciles the pricing result against the mean gap between the two outcome distributions, which is computed on a separate sample by a separate route; the two agree to within 3.5 per cent.

The outcome distribution is 200,000 simulated one month terminal prices under geometric Brownian motion at the stated drift and volatility. The repeated writing exercise is 20,000 sequences of 24 monthly writes per regime, re striking 5 per cent out of the money each month and assuming the shares are repurchased after assignment. Charges are computed from the site's verified research on Indian statutory and exchange charges, dated 17 July 2026; brokerage, the equity derivatives exchange charge, the derivatives segment regulator fee and the specific treatment of a physically settled assignment are excluded as unverified, so all charge figures are floors.

All results are illustrative and simulated. They are not a track record, they are not a forecast, and they are not an indication of what any position would produce in a live account. The purpose is to demonstrate the relationship between a premium and the distribution it is exchanged for, which is a property of the instrument rather than of any particular company or market. Contract specifications, lot sizes and statutory rates change; verify all of them at source before relying on any figure here.

• Hedging, and what each hedge really costs →
• Calls and puts, from the ground up →
• How the strike changes everything →
• Reading an option chain →
• A framework for options decisions →