The minds inside a derivatives account

SEBI’s Department of Economic and Policy Analysis has published a series of studies on individual traders in the equity derivatives segment since January 2023, the latest pair in August 2026. It is unusual for a regulator anywhere to publish this candidly on the losses of the investors it supervises. The August 2026 study of trading behaviour is the most valuable of the lot, because it does not stop at how much individuals lost and starts asking how they traded. In other words, it is less a statistics report than a behavioural report. Almost every headline number lines up with a bias that behavioural finance has documented for decades. The data does not merely tell us that individuals lose money in derivatives. It tells us how they talk themselves into it.

Overconfidence, and the illusion that experience teaches.

Time in the market does not help, at least not here. Among traders with one consecutive year of participation, 91 per cent made losses. The share rises with every additional year, reaching 96.5 per cent at four years. Among traders active through all five years from FY22 to FY26, only 0.5 per cent were profitable in every year. This learning curve slopes the wrong way, because what survives repetition is not skill but conviction. Traders do not learn that they cannot beat the market. They learn that they have not yet beaten it.

The lottery preference.

In the study’s sample, 93 per cent of traders were only options buyers. People overweight small probabilities, and a cheap option has the shape of a lottery ticket: a known small cost, an unknown vivid gain. Prospect theory says people will overpay for that shape, and they do. Options buyers had the highest incidence of loss, about 90 per cent. The sharpest number is their median return on capital employed, minus 114 per cent in FY26. Half of them lost more than the entire capital they had put to work.

Self-attribution.

No dataset observes this directly, but it holds the rest together. Gains are read as skill and losses as bad luck, and that asymmetry is what keeps overconfidence intact through a losing record.

The sunk cost fallacy, and the ego that feeds it.

Only 15.4 per cent of trader-quarters in the study were profitable. The median gain in a winning quarter was ₹4,366. The median loss in a losing quarter was ₹10,525, more than twice as large. Traders sat through losses far larger than the gains they banked, and at the level of the account that habit hardens into sunk cost. Traders with cumulative profits above ₹10 lakh during FY22 to FY24 continued trading at a rate of 88 per cent. Traders with cumulative losses above ₹10 lakh also continued at 88 per cent. Traders whose outcome, either way, was below ₹1 lakh continued at only 53 to 55 per cent. What predicts persistence is not whether you won. It is how much is at stake. Big winners stay because they believe they are good. Big losers stay because leaving now would make the loss permanent and admitted. They want to get back to even. Among the large past losers who continued, more than 95 per cent lost again.

What follows? Disclosure has done some work, but not enough, and the reason sits in the data. Every trader believes the aggregate describes someone else, so telling people that nine in ten lose money is a statistic about strangers. The design principle now should be to make each trader’s own record unavoidable, and to target named biases rather than risk in general.

Four measures are worth considering. Let the account’s own history be the credible witness: a personalised profit and loss statement at every login and each quarter end, showing the net outcome and all costs since account opening, in rupees and as a percentage of the trader’s own capital and portfolio. An appropriateness test at entry in place of a one-time consent screen, with graduated exposure limits for first-time traders. Self-set annual loss limits offered by every broker at onboarding, with a cooling-off period before a limit can be raised. And better framing at the order screen, with the premium shown as a share of capital at risk alongside the share of similar contracts that have expired worthless.

Mental accounting, used deliberately.

One more finding deserves to become the basis of investor education rather than another warning. Traders who lost more than ₹1 crore in derivatives held a median cash equity portfolio of about ₹138. For most of them the derivatives account was not a slice of a portfolio. It was the portfolio. Mental accounting, the habit of keeping money in separate mental jars, is usually filed as a bias. Made explicit, it is also the most workable remedy available. Individuals should be taught to allocate across three buckets: money that must stay safe, money that funds long-term goals through diversified market exposure, and a small aspirational bucket where the odds are poor and the loss is affordable. Futures and options belong in the third, sized in advance as a stated share of total wealth. That converts an open-ended activity into a budgeted one, and it gives a trader a number to check against. Brokers can carry this framing at onboarding, and investor education can make it the vocabulary: not do not trade, but decide first how much of your wealth this bucket is allowed to hold.

None of this should be mandated untested. SEBI now has the data and the broker relationships to run randomised trials of these nudges and see which change behaviour. Regulation has already done a great deal about the product, through expiry frequency, contract size and transaction costs. The frontier now is the account holder, and this study is the map.

Author – Rachana Baid, Dean – NISM