An autocallable note on an index pays 9% a year if the index is at or above its start on an annual observation date, and then calls. If it ends below 60% of its start after three years, capital is lost one for one. Work three scenarios and say what the coupon is paying for.

The note pays 9% a year when markets are flat to up, and hands over the index's full loss when they fall below 60%. An early call returns Rs 1.09 crore after a year; a late call Rs 1.27 crore after three; a flat market returns capital with no coupon; a 45% crash returns Rs 55 lakh. The coupon is the price the client is paid for taking the index's severe downside and giving up its upside.

Picture a landlord who offers a tenant a rent discount every year the market is calm, on one condition: if the building's value collapses by more than 40%, the tenant pays for the loss. In calm years the deal looks generous. An autocallableA structured note that ends early and pays a fixed coupon if the underlying index is at or above a set level on an observation date. pays a fixed coupon in good and flat outcomes and passes the index's loss to the client in the bad one. Each anniversary asks one question; if the answer is never yes, year 3 has three outcomes.

Run four paths. Index at 106% after one year: called, Rs 1.09 crore, 9% for the year, while the index rose 6%. Index at 94%, 97% and then 103%: called in year 3 at Rs 1.27 crore, about 8.3% a year, while the index rose 3%. Index at 85%, 90% and 80%: never called, above the barrier, Rs 1 crore back and nothing for three years. Index at 88%, 72% and 55%: below the barrier, Rs 55 lakh back. The note beats the index in flat and mildly rising markets, matches its loss in a crash, and can never pay more than Rs 1.27 crore however far the index rises.

Draw the payoff at year 3 against where the index ends. Below 60% the note tracks the index exactly; from 60% to 100% it is flat at Rs 1 crore; at 100% and above it is capped at Rs 1.27 crore. The client has in effect sold the bank insurance against a severe fall in the index, a down-and-in putAn option that only comes into existence if the price falls through a barrier, after which it pays the full fall from the starting level., and given up the index's upside and dividends; the 9% is the premium for both. Selling insurance feels like income in calm years and pays out in the year the client can least afford it.

Two risks sit outside the index. The note is a promise from the issuing bank, so if the bank fails, even the good outcomes are at risk; that is credit risk, and a single issuer should be sized accordingly. And the note is hard to sell before it ends, often only back to the issuer at a price it sets. Say the limitation of the scenarios too: they are illustrations, not probabilities. The fair question for a client is whether she would sell insurance on a 40% crash for 9% a year, and for whom that is a suitable trade at all.