How to Read a Contract Note, Line by Line

The document that decides who is right

There is a hierarchy of records behind every trade, and most traders live at the wrong end of it. At the top of the hierarchy sits the position screen, which is fast, comforting and provisional. Below it sits the funds ledger, which is a cash record and knows nothing about what you still hold. Below that sits the contract note, which is neither a summary nor an interpretation. It is the statement of what was executed, in what quantity, at what price, at what time, and at what cost, issued by the member who executed it.

That last property is what makes it different in kind rather than in degree. A dashboard is a rendering. A contract note is an assertion made by a registered intermediary about specific transactions, with serial numbers attached, and it is the artefact that later processes ask you to produce. When a charge is queried, the query is answered from the note. When the character of a trade has to be established for tax, it is the time stamped note carrying your client identity and permanent account number that establishes it. When a grievance escalates beyond the member, every stage of the escalation opens by asking which note, which serial number, which trade number.

Set against that, the ordinary treatment of the document is odd. It arrives by email, usually as a password protected attachment, and it is filed unread on the reasonable assumption that a regulated intermediary computing a published rate against a printed base will get it right. That assumption is mostly correct and it is also beside the point. The reason to open the note is not suspicion. It is that the arithmetic is the only part of trading that is fully determined, fully disclosed and fully checkable, and a trader who will not check the determined part has no standing to complain about the undetermined part.

There is a second reason, which is quieter and probably more valuable. Reading the note forces you to look at costs on the same page as prices. On a screen, the fill price and the charge stack live in different places and are experienced at different moments, which is exactly the arrangement under which costs stop feeling real. On the note they are adjacent, in the same column, adding to the same total, and the effect of seeing them together is difficult to unlearn.

Five regions, and what each one is for

Contract notes are laid out differently by different members, but they contain the same regions in roughly the same order, because they are answering the same set of questions. The figure below is an illustrative note for a single day of trading, with each region labelled. Nothing in it is a real document, no member is named, and the instruments are placeholders, but the structure and every rupee figure follow from the trades and rates stated further down this page.

The identity region answers who is asserting this. It carries the member's name, the registration under which they operate, the exchange and segment, and three numbers that matter more than they look. The note serial number is the handle by which this specific document is cited in any later correspondence, so it is the first thing to quote and the first thing to keep. The trade date is the date the transactions happened, which is not the date money moves. The settlement number identifies the settlement cycle those trades belong to, and it is the link between this document and the day your funds and securities actually change hands, a timing question worked through in detail in our guide to what T plus one settlement actually means.

The client region answers who it is for. Your trading code, your permanent account number, and the depository account the securities settle into. The last of these catches people out, because a trading account and a demat account are different things held with different entities under different rules, and the note touches both. If the separation is not obvious, the three account architecture is worth reading first, since a charge you cannot place is often a charge that belongs to the other account. This region also carries the state used for the tax split on the service charges, which is why the same charge appears as two half sized lines rather than one.

The trade region is the evidentiary core. Every order appears with its own number, and beneath it every fill appears with its own trade number, its own time stamp, its own quantity and its own price. This granularity is not decoration. An order for 250 shares that filled in 3 pieces is 3 separate transactions in the exchange's record, and if a later question is ever asked about one of them, the trade number and the time stamp are the answer. The price shown against the order itself is the weighted average of its fills.

The charge region is the part this page exists for, and it is discussed at length below. Note only its shape for now: each levy on its own line, with a rate, a base and an amount, and no two lines necessarily using the same base.

The net region converts everything above into one number, then the document closes with the signature and the clause that names the route a dispute takes. The clause is the part nobody reads and the part that turns the document from a record into an instrument.

The trades, before a single charge is applied

The constructed day is deliberately ordinary and deliberately mixed, because the interesting behaviour only appears when more than one product type is present. There are 4 orders. Two of them are delivery, one a purchase and one a sale of a different instrument. Two of them are an intraday position in a third instrument, opened and closed inside the session. Between them they filled in 9 pieces.

Start with the arithmetic that happens before any charge exists. Order one is a delivery purchase of 250 shares that filled three times: 100 at 1,842.15, then 100 at 1,842.70, then 50 at 1,843.30. Those three fills are worth 1,84,215.00, 1,84,270.00 and 92,165.00 rupees, which sum to 4,60,650.00. Divide that by the 250 shares and you get a weighted average of 1,842.60, which is the price the note prints against the order. Multiply the weighted average back by the quantity and you recover 4,60,650.00 exactly, which is the property that makes the average useful rather than merely descriptive.

That property matters because the transaction tax is computed on the weighted average price rather than on individual fill prices. Charging it fill by fill and rounding at each fill would produce a slightly different number, and would produce a different number again depending on how the fills happened to break. Computing once on the order aggregate removes that noise entirely, and it means your check is a single multiplication rather than three.

Do the same for the other three orders and you get four traded values: 4,60,650.00 for the delivery purchase, 2,28,330.00 for the delivery sale, 2,96,860.00 for the intraday purchase and 2,99,560.00 for the intraday sale. Added together they come to 12,85,400.00 rupees, and that total is the base for two of the charges further down. Purchases across both products come to 7,57,510.00 and sales to 5,27,890.00, a difference of 2,29,620.00 rupees that has to be funded before any charge is added at all.

Nine charge lines and seven different bases

Here is the thing that makes contract notes harder to read than they should be. The rates are published, stable and easy to find. The bases are not uniform, and nothing on the document tells you that they are not uniform. Nine lines on the illustrative note draw on seven different bases. One charge applies to a purchase leg only. Another applies to a sale leg only, at a rate that depends on whether the position settled to delivery. Two apply to the whole day. One applies to a subtotal of the other charges. One is not a percentage of anything.

Work down the list. Brokerage is set by the member and is the only line that is genuinely negotiable, and on the illustrative plan used here it is a flat 20.00 rupees for each executed order, so 4 orders produce 80.00. It is charged per order, not per fill, which is why 9 fills do not become 9 brokerage lines.

The securities transaction tax appears three separate times, on three separate bases, and this is where most confusion lives. On a delivery purchase it falls at 0.100 percent, on the delivery sale at 0.100 percent again, and on the intraday sale at 0.025 percent. There is no intraday purchase line at all, because the levy on a non delivery transaction falls on the seller only. So the same instrument bought and sold in the same session pays the tax once, while the same value bought and sold for delivery pays it twice at four times the rate. That asymmetry produces 461.00, 228.00 and 75.00 rupees on the illustrative day, 764.00 in total.

The exchange transaction charge and the regulator's turnover fee are the two that apply to everything. Both are computed on the whole 12,85,400.00, at 0.00307 percent and 0.0001 percent respectively, giving 39.46 and 1.29. Stamp duty then goes the other way and falls on the purchase side alone, at 0.015 percent where the transaction settles to delivery and 0.003 percent where it does not, producing 69.10 on the delivery purchase and 8.91 on the intraday purchase.

Finally the goods and services tax, at eighteen percent, applied to a base that is not the trade at all. Its base is the service lines: brokerage plus the exchange charge plus the turnover fee, which here is 120.75 rupees, giving 21.74, split into equal central and state halves of 10.87 each. The two transaction taxes sit outside that base, and the reason is worth stating correctly because the usual explanation is wrong in a way that happens to reach the right answer. It is not that a tax is never charged on a tax. Other levies are included in the value of a supply by default, and the escape here is the pure agent treatment in the central tax rules, which applies because the legal liability for those levies is yours and the member is merely remitting them. That treatment carries a condition: a payment made as a pure agent has to be shown separately on the invoice, and a contract note is issued as a tax invoice. Separately from that, the format the exchange prescribes gives the transaction tax its own line anyway. Two requirements pointing the same way is why you will never find these levies folded into a single figure, and it is what makes the tax base checkable at all.

The table below sets out every line a retail equity or derivative note can carry, with the base it uses and the formula. The rate schedule itself, and the question of what a trade costs in total across segments, belongs to our reference on the real cost of an Indian trade, which this page deliberately does not restate. What is set out here is which number each rate is multiplied by, because that is the half you need in order to check a document rather than to estimate a cost.

Recomputing every line from the trade details

Now the part that gives the page its reason to exist. Every one of the nine lines on the illustrative note can be reproduced from two numbers that are printed on the same document: a rate and a base. Below, each line is recomputed and set beside what the note says.

Read the middle column as an instruction rather than a result. Brokerage is the plan rate times the order count. The three transaction tax lines are each a rate against a leg you can find in the trade region: 4,60,650.00 for the delivery purchase, 2,28,330.00 for the delivery sale, 2,99,560.00 for the intraday sale. The exchange charge and the turnover fee both take the full 12,85,400.00. The two stamp lines take the two purchase legs, 4,60,650.00 and 2,96,860.00. The tax line takes 120.75, which is the sum of three lines above it and nothing else.

Six of the nine agree to the paisa. Three do not, and the reason is worth knowing before you decide you have found an error, because it is the single most common way a correct recomputation looks like a failed one. Your arithmetic on the delivery purchase gives 460.65; the note prints 461.00. On the delivery sale you get 228.33 and the note prints 228.00. On the intraday sale you get 74.89 and the note prints 75.00. The transaction tax is rounded to the nearest rupee under rule 4 of the Securities Transaction Tax Rules 2004, notified by S.O. 1059(E) of 28 September 2004, which provides that where the amount contains part of a rupee, fifty paise or more rounds up and less than fifty paise is ignored. Nothing else in the charge block behaves this way. Across the three lines the effect here is 0.13 rupees, which is the entire difference between the 984.37 your calculator produces and the 984.50 the note prints.

The same rules contain the other half of the point made earlier about weighted averages. Rule 3 rounds the transaction value itself to the nearest paisa, and rule 4 rounds the tax to the nearest rupee, which is why an order is priced to two decimals and taxed to none. Members differ in whether they apply the rupee rounding line by line or to a consolidated figure for the segment, and some notes show one transaction tax line rather than three, so allow a rupee either way and compare the total when the breakdown is not printed. What that gives you is a working tolerance with a reason behind it: a rupee or two is the rule, tens or hundreds are a question.

The band at the foot of the figure is the reason the exercise is worth doing at all. Suppose the intraday sale had been billed at the delivery rate rather than the intraday one. The line would read 300.00 against the correct 75.00, a difference of 225.00 rupees on a total charge bill of 984.50. As a fraction of the day it is small. As a fraction of that line it is four times too much, and it is completely invisible unless you look at the line rather than the total. The same class of error on the stamp duty, the delivery rate applied to a non delivery purchase, adds 35.62 rupees, which nobody would ever notice once and everybody would notice across four hundred orders in a year.

That is the general shape of the thing. Charge errors, when they occur, are almost never large enough to be obvious in a total and almost always systematic enough to matter in aggregate. A total is one number and it looks plausible over an enormous range. A line has a rate and a base printed beside it, and it either matches or it does not.

The three checks that catch most of it

Nine multiplications is more than most people will do on a Tuesday evening. So here is the compressed version: three checks, none of which needs anything the document does not already give you, and between them they catch the error shapes that actually occur.

Check one asks whether the traded value adds up. Sum the fills under each order and confirm the total equals the order value, then sum the four order values and confirm the total equals the turnover the percentage charges are computed on. On the illustrative note that total is 12,85,400.00. This check fails when a fill is missing from one side of the record or when a quantity has been transposed, and it fails loudly, because a dropped fill of 50 shares at 1,843.30 moves the base by 92,165.00 rupees and every percentage charge with it. It is also the check that matters most when you are reconciling against your own trading records rather than against the note, since a trade that appears in one and not the other is a genuinely serious discrepancy rather than an arithmetic one.

Check two asks whether each rate hit its stated base. Take the rate printed beside a line, multiply it by the base printed beside it, compare with the amount. Do it for all of them, allow the rupee rounding on the transaction tax, and you have reproduced 984.50. The failure mode here is a rate applied to the wrong leg or the wrong product, which is exactly the 225.00 rupee error above, and it is the only one of the three checks that can tell you a specific line is wrong rather than that something somewhere is wrong.

Check three asks whether the note ties to the ledger, and it is the one that most often looks like a failure and is not. The illustrative note settles at 2,30,604.50 rupees. The trading account, on the same day, moves 2,30,620.43. Those are different numbers, and nothing is wrong. The difference is 15.93 rupees: a depository charge of 13.50 for the one security debited out of the demat account on the delivery sale, plus 2.43 of tax on it, billed to the ledger rather than shown on the note. The rule to hold is not that the two numbers should be equal. It is that every rupee of difference should have a name you can point at. A named difference is housekeeping. An unnamed one is a question. Extending that idea from one day to a whole year, across a trade record, a funds ledger and a tax computation that were never designed to agree, is a separate and harder exercise worked through in why your three profit numbers never match.

The depository line is worth a further sentence, because it is the clearest example of a charge that mixes a fixed component with a commercial one. Of the 13.50 rupees billed here, 3.50 is the depository's own flat fee to the participant for each debit, a published figure that does not vary with the value of the shares. The remaining 10.00 is the participant's own charge, which is a commercial decision and varies widely. Both halves are legitimate. Only one of them is a rate you can look up, which is a useful thing to know before deciding whether a depository charge is high.

The right rate on the wrong base

Almost every mistake made when checking a contract note is a correct rate applied to an incorrect base. This is a more dangerous class of error than getting a rate wrong, because the answer it produces is the right order of magnitude and looks entirely plausible. A wrong rate gives you an absurd number. A wrong base gives you a number you will believe.

The first example is the tax on services, and it is the error made most often, including by people writing about charges. The rate is eighteen percent and nobody disputes it. On the illustrative note the correct base is 120.75 rupees, giving 21.74. Apply the same rate to brokerage alone, which is a common shorthand, and you get 14.40, understating by 7.34 rupees. Apply it to the whole charge block of 962.76, which is what you naturally do if you add up the charges and then look for the tax, and you get 173.30, overstating by 151.56 rupees. That last figure is almost eight times the correct one, and it is wrong for a specific reason: it sweeps in the two transaction taxes, which are outside the base precisely because they are not services.

The second example is the more expensive one, and it concerns an option carried to exercise. When an option is sold or squared off, the transaction tax applies to the premium. When an in the money option is exercised, it applies to the intrinsic value instead, and it falls on the purchaser. Those two bases are not close to each other. On an illustrative hundred unit contract bought for a premium of 4,200.00 rupees and exercised with an intrinsic value of 31,000.00, the same 0.150 percent produces 6.30 rupees if you assume the premium base and 46.50 if you use the base the schedule actually specifies. That is 7.4 times the figure, from one word.

Both examples share a diagnostic. When a computed figure disagrees with a printed one, the productive question is not whether the rate is right, since rates are published and easy to confirm. It is what the rate was multiplied by. Ask that question first and most disagreements resolve in a sentence.

The same trade value, two products, a different document

One more comparison, because it isolates the base question cleanly. Take a round trip of 2,00,000.00 rupees on each leg and run it twice, once as delivery and once as an intraday position squared off the same session. Same instrument, same value, same member, same plan. The note that comes back is a different document.

Four of the six charge lines are identical to the paisa. The entire difference lives in two lines, and both differences are about a base rather than about a rate: which legs the transaction tax touched, and which rate schedule the stamp duty used. A third difference is not on the note at all, because in one case securities left a demat account and in the other they never entered it. Whether that difference should influence how you trade is a separate question, one that depends on horizon and tax treatment as much as on charges, and it is worked through in the comparison of intraday versus delivery trading. The point being made here is narrower and purely documentary: the same trade value produces a differently shaped document, and if you do not know which lines should be present you cannot tell whether one is missing.

What the document is for when something is wrong

Everything above treats the contract note as an arithmetic object. It is also a regulated one, and the two properties are connected: the reason the arithmetic is disclosed at that level of detail is that the document has to be capable of being disputed.

Start with the time limit for issue, which is real and less tidy than it is usually presented. The twenty four hour figure that circulates everywhere does have a primary source. It appears in a market regulator circular numbered SMD/SED/CIR/93/23321, dated 18 November 1993, which states that member brokers shall issue the contract note for purchase or sale of securities to a client within twenty four hours of the execution of the contract. It is still operative, restated at clause 24.4 of the Master Circular for Stock Brokers numbered SEBI/HO/MIRSD/MIRSD-PoD/P/CIR/2025/90 and dated 17 June 2025, and repeated in the investor charter attached to the same document.

Now the part that complicates it, included here because it is true rather than because it is convenient. That same master circular, in the annexure setting out the rights and obligations of stock brokers and clients, says at clause 29 that the broker shall send contract notes within one working day of the execution of the trades. Twenty four hours and one working day are not the same measure, and they come apart exactly when it matters: a trade executed on a Friday afternoon falls due on Saturday under one and on Monday under the other. Both formulations sit inside one instrument bearing one date. The instruments above it decline to fix a number at all. The current broker regulations, notified on 7 January 2026, require issue within a specified timeline in a specified form without saying what the timeline is, and the exchange bye laws require issue within such period as the relevant authority specifies. So the accurate statement is this: twenty four hours is the figure the regulator repeats and inspects against, one working day appears beside it in the same document, and anyone arguing about a matter of hours should read the current instrument rather than an article, this one included.

The form is prescribed too, which is why notes from different members look so alike. The current format on the National Stock Exchange is set out in circular NSE/INSP/61999 of 13 May 2024 and is titled contract note cum tax invoice, because the same sheet does two jobs at once. Its compliance date was deferred more than once and finally fixed at trade date 27 June 2025 by circular NSE/INSP/68618 of 18 June 2025. That form is why your note carries a contract note number, the member's registration number, order and trade numbers with their time stamps, brokerage on its own line, the transaction tax on its own line, the place of supply that drives the tax split, and a signature block. Every one of those fields exists so that something specific can be pointed at afterwards.

Electronic delivery has its own conditions, and one of them allocates a risk to you. Notes authenticated by digital signature were permitted by circular SMDRP/POLICY/CIR-56/00 of 15 December 2000, and the consolidated conditions require the member to hold a digital signature certificate from a certifying authority under the Information Technology Act 2000, to have your authorisation in writing, to send to an address you created rather than one created for you, and to publish the note securely on their own site as well. The risk allocation is this: where the delivery does not bounce, it counts as delivery. A full mailbox is not a defence, and a note you never opened is a note you received.

The reason the timing matters at all is not tidiness. It is that a short window between execution and documentation is what makes a dispute tractable. If the record arrives while you can still remember the day and still check it against your own records, a discrepancy is a conversation. If it surfaces weeks later, it is an archaeology exercise, and the burden falls on whoever has the worse records, which is almost never the member.

If a line does look wrong, the sequence is unglamorous and it works. Write down the specific line, the rate you expected, the base you expected it applied to, the figure you computed and the figure printed. Raise it with the member in writing, quoting the contract note number and the relevant trade numbers. Specificity does most of the work, because a query saying a charge looks high invites a general answer, while a query saying a rate appears to have been applied to the purchase leg as well as the sale leg invites a specific one. Keep the delivered note rather than a screenshot of a dashboard, because the note is the evidence and the screenshot is not.

If the member does not resolve it, two routes open and they are different things. The complaint route runs through the regulator's grievance platform, where a complaint must be lodged within one year of the cause of action and the entity is required to upload an action taken report within twenty one calendar days, timelines set by circular SEBI/HO/OIAE/IGRD/CIR/P/2023/156 of 20 September 2023. The adjudication route runs through the online dispute resolution portal established by circular SEBI/HO/OIAE/OIAE_IAD-1/P/CIR/2023/131 of 31 July 2023, which expects you to have approached the member first and which reckons its own time limits by the general law of limitation rather than by a fixed count of days.

One thing we could not establish, and would rather say so than fill in. There does not appear to be a deadline specific to raising a discrepancy on a contract note. The rights and obligations annexure imposes a reporting window on the statement of accounts and delegates its length to the exchange. The complaint route has its one year. Arbitration runs on the ordinary law of limitation. None of that produces a stated number of days in which a charge line must be queried. The practical implication is not that you have unlimited time; it is that the clock you are actually running against is a general one, which makes checking the note when it arrives a better plan than relying on a grace period that may not exist.

Which is the argument for reading it, in the end. The document is constructed to be disputable. It carries serial numbers so it can be cited, trade numbers and time stamps so individual transactions can be isolated, and separate lines for each levy so each can be examined on its own. All of that machinery exists on the assumption that somebody looks. A trader who files it unread is discarding the only leverage the system hands them for free.

What the habit is actually worth

It would be easy to finish a page like this with a warning about being overcharged, and that would be the wrong conclusion. Charge errors on contract notes are not epidemic. The statutory lines are computed by systems against published rates and they are usually right, and a reader who audits notes for a year will most likely find nothing except their own arithmetic mistakes.

The value is elsewhere, and it is larger. The first part of it is that a trader who has recomputed a charge block once knows, in a way that no summary can teach, which lines respond to which decisions. They know that the transaction tax is the dominant line on a delivery round trip and that changing member does not move it. They know that the tax on services is smaller than it looks because its base is narrower than the bill. They know that a depository debit lands on a sale and not on a purchase. Those facts stop being trivia and become inputs, because they were derived rather than read.

The second part is a matter of standing. Most of trading is uncertain, and a great deal of the effort in it goes into forming opinions that may turn out to be wrong. The charge block is the one region of the whole activity that is fully determined, fully disclosed, and checkable in ten minutes with a calculator. Choosing not to check it, while continuing to hold strong views about matters that cannot be checked at all, is a revealing allocation of attention, and it is one of the few habits in trading that costs nothing to change.

None of this is advanced. It is a rate, a base, a multiplication and the willingness to open an attachment. If working through the arithmetic on this page felt useful rather than tedious, that same insistence on deriving the checkable part before arguing about the rest is the method we teach.

Every rupee figure on this page comes from one deterministic script that constructs the trading day and computes each charge from a stated rate against a stated base. Nothing is typed twice: the figures in the prose, the tables and the diagrams are all read from the same computation, and the script asserts that the fills reconcile to the order values, that the order values reconcile to the turnover, that the charge lines sum to the printed total, and that the note total plus the separately billed depository line equals the ledger movement. If any of those had failed, the page would not have been produced.

The statutory and exchange rates used are those verified against primary instruments on 17 July 2026: the transaction tax rates from the enacted Finance Act schedule, stamp duty from the central schedule that has applied uniformly since July 2020, the exchange transaction charge from the exchange's own circular effective 1 March 2026, the regulator's turnover fee from the broker regulations notified on 7 January 2026, and the tax on services at eighteen percent on the service lines. The rounding of the transaction tax to the nearest rupee, and of the traded value to the nearest paisa, follow rules 4 and 3 of the Securities Transaction Tax Rules 2004. The issue deadline, the prescribed format, the electronic delivery conditions and the two dispute routes are each cited above to the instrument and date they come from. Rates and procedures are revised, sometimes more than once in a year, so verify each against its current instrument before applying it to a real document. Brokerage and the depository charge billed to a client are commercial rather than statutory, and the figures used for them here are stated as illustrative for that reason.

Two things on this page are explicitly not established. There appears to be no deadline specific to querying a discrepancy on a contract note, and rather than supply a plausible number the page says so. And while the pure agent treatment requires a separately indicated disbursement on an invoice, and the exchange format independently gives the transaction tax its own line, we found no ruling that names the first as the cause of the second, so the page presents them as two requirements pointing the same way rather than as one explaining the other.

The trading day, the instruments, the client and the member are all constructed. No member is named anywhere, including on the illustrative note itself, and the instruments are placeholders rather than securities. Every rupee figure is illustrative. Nothing on this page is a record of any actual trade, a representation of what any particular member charges, or advice about how to trade or how to treat anything for tax.

• Work the charge stack on your own numbers →
• Opening a demat account in India →
• Capital gains for the active trader →
• What the market regulator actually does →