Guide · Costs and taxes · Updated 2026-09-18

Turnover for a derivatives book is not contract value, and the rule changed for this assessment year

There is one error in Indian trading tax that costs more than all the others combined, and it is arithmetic rather than judgement. A trader opens the broker statement, sees a figure in crores next to the word turnover, and concludes that an audit is mandatory. The figure is contract value. It is not turnover, and the two are not close.

The second error is subtler and newer. Turnover is computed under a guidance note rather than a section of the Act, and that guidance note changed on this exact point for the assessment year now being filed. An article written any time between 2022 and 2025, including several that still rank well, gives an answer that is no longer correct.

The category error, sized

Contract value answers the question of how much exposure passed through the account. Turnover answers a narrower question, which is how much business activity the year contained for the purpose of deciding whether accounts must be audited. A single index option lot carries a contract value in lakhs while contributing a difference measured in thousands.

Across a year of ordinary retail activity the gap runs to a factor of fifty or more. A trader using the wrong figure is not slightly over the threshold. They are over it by an order of magnitude, and they respond by commissioning an audit they never needed, or by concluding that the presumptive scheme is unavailable when it was available all along.

The absolute sum, and why a flat year still has turnover

The method for futures is unambiguous. Take each settled position, take the profit or loss, discard the sign, and add. The signs are discarded before the addition, not after, which is the whole of the mechanism and the part that surprises people.

The consequence is that turnover measures activity, not outcome. A year in which four positions produced two gains and two equal losses nets to nothing and still generates a turnover figure equal to the sum of all four. There is no such thing as a year with meaningful trading and no turnover.

This is not an accident of drafting. Audit applicability is meant to scale with the volume of transactions a set of books has to record, because that is what an audit examines. Netting would make an active account look dormant.

The premium question, and the edition that changed the answer

Options introduce a component futures do not have. When an option is sold, premium is received. The question is whether that premium forms part of turnover in addition to the difference on the position when it is closed or expires.

The answer has moved. The eighth edition of the Guidance Note on Tax Audit, which applied from assessment year 2022-23, removed premium on sale from the computation. The tenth edition, revised in 2025 and applicable from assessment year 2026-27, restored it.

The restoration carries a proviso, and the proviso is where the practical work lives. Where a broker's profit and loss statement has already absorbed the premium into the per-trade result it reports, adding premium again on top counts the same rupees twice. Whether you add it therefore depends on the format of the statement in front of you rather than on a formula that holds universally.

That makes the broker statement a document to be read rather than a number to be copied. Two traders with identical positions at two brokers can correctly arrive at the same turnover by different routes, and can incorrectly arrive at different ones by assuming their statement is formatted like everyone else's.

One year, two editions, two different obligations

Consider a derivatives book with differences summing to nine lakh forty thousand in absolute terms and premium received on written options of sixty-two lakh across the year, on a statement that reports premium separately rather than netting it.

The two figures differ by a factor of seven and both are defensible, for different assessment years. That is the entire point. A number is not correct in isolation; it is correct under a stated edition for a stated year, and any worked example that does not name its basis cannot be checked.

Push the premium figure to two crore and the same book crosses the 44AB(a) threshold under the tenth edition while remaining far below it under the eighth. The obligation changed without a single trade changing.

What the figure decides, and in what order

Turnover is not reported for its own sake. It is an input to two separate tests, and running them in the wrong order produces a confident answer to a question nobody asked.

The third row deserves attention because it inverts the usual intuition. An inflated turnover figure does not merely trigger an audit. If the presumptive scheme is then used, the deemed income is six percent of that inflated figure, and the tax is real even though the profit was imaginary.

The speculative book is computed separately, and to a different end

Intraday equity is a separate business under section 43(5), so it carries its own turnover figure. The arithmetic is the same in principle: each squared-off position produces a difference, the sign is discarded, and the differences are summed. There is no premium component because there is no premium.

What differs is what the figure is used for. Speculative turnover feeds the audit test for that business, but it can never feed a presumptive computation, because section 44AD(6) excludes speculative business from the scheme entirely. Computing intraday turnover in order to decide whether to declare six percent on it is work spent on a question that has no answer.

The two figures are also not added together into a single number. They are two businesses reported within one return, each with its own computation, and merging them obscures both. A trader whose derivatives turnover is comfortably inside every threshold can still have an audit question on the speculative side, or the reverse.

Rebuilding the figure from the primary records, in order

The computation has to be reproducible from contract notes, because the contract notes are what survive a query. A turnover figure that exists only as a number copied from a dashboard is an assertion. The order below produces one that can be defended.

One. Separate the books first. Split every transaction into speculative and non-speculative before any arithmetic. Doing this last means doing it twice.

Two. Reduce each settled position to a single difference. Not each leg and not each order. A position opened across three orders and closed across two produces one difference, not five.

Three. Discard signs, then sum. In that order. Summing first and taking the absolute value at the end yields the net result, which is a different and much smaller number.

Four. Establish how your broker presents option premium. Read the statement rather than assume. If premium is reported separately from the trade result, it is added under the tenth edition basis. If the trade result already absorbs it, it is not added again.

Five. Keep the cost stack out. Brokerage, Securities Transaction Tax, goods and services tax, stamp duty and exchange charges are expenses in the computation of income. They reduce taxable profit and they do not touch turnover.

Six. State the basis on the working. Write down which guidance note edition the figure was computed under and which assessment year it is for. A figure without its basis cannot be checked by anyone, including you, a year later.

Where this goes wrong in practice

Copying the broker's headline. Broker statements are built for reporting activity, not for section 44AB. Whatever figure appears next to the word turnover is a starting point that has to be rebuilt.

Netting before taking absolute values. Summing the year first and taking the absolute value of the total produces the net result, not turnover. The order of the two operations is the method.

Using a worked example from the wrong edition. The most common current error, and it will persist for years because the older articles are well established and do not say which edition they were written under.

Adding premium a broker has already netted. The mirror of the previous error, and it inflates turnover just as badly, with the same downstream effect on the deemed income if the presumptive scheme is used.

Merging the speculative and non-speculative books. Intraday equity and derivatives are separate businesses with separate computations. Adding their turnover into a single figure obscures both tests.

What the number is actually for

Turnover exists to answer an administrative question: how much activity did this business contain, and is that enough to warrant the cost of an audit. It is not a performance measure, not a size measure, and not something a trader should feel anything about.

It is also entirely knowable during the year rather than after it. Every component comes from contract notes that arrive daily. A trader who maintains the computation as they go knows in October whether an audit is coming, which is when the answer is still useful. Reconstructing it in July from a statement that was never built for the purpose is how a manageable obligation becomes an expensive surprise.