The index is at 22,000, the one-month future at 22,120, the rate 6.8%, the dividend yield 1.2% and round-trip costs 0.08%. Is a new cash-futures arbitrage worth putting on, and what return does the basis lock in?

The basis locks in 0.55% for the month, 7.75% annualised with dividends, which after 0.08% of costs is 6.79%: fair value, not an arbitrage. The fair future is 22,103, so only 17.3 of the 120 points exceed carry, and the round trip costs 17.6 points. Against the fund's 6.3% liquid-fund alternative the trade adds about 0.5% a year, which is thin for a position that ties up margin and capital for a month. Put it on only if nothing better is on offer.

Buying a flat today and renting it out for a month costs you the interest on the price less the rent. An index basket is the same: holding it costs the financing rate and earns the dividends. The fair future is spot plus net carry: 22,000 times (1 plus (6.8% less 1.2%) over 12), which is 22,103. The quoted 22,120 is 17.3 points above that. Everything the arbitrage can earn beyond the financing rate lives in those 17.3 points, and the basisThe difference between the futures price and the spot price of the index. Positive basis means the future trades above spot. of 120 is not the prize, most of it is the cost of carrying the basket.

Count every rupee the position earns from entry to expiry. The basis converges to zero, paying 120 points; the basket pays about 22 points of dividends; the round trip costs 17.6 points. Net, the trade earns 124.4 points on 22,000, which is 0.565% for the month and 6.79% a year, almost exactly the 6.8% financing rate. That is the implied repo rate of the future after costs: a lender's return, with a lender's upside.

Only marginally, and the honest answer says so. The trade beats the fund's cash alternative by about 0.5% a year, which on Rs 100 crore is about Rs 0.04 lakh for the month, before the margin it ties up and the risk that a stock in the basket cannot be bought at the close. An arbitrage fund lives on these pickups, so a manager would not refuse it, but the sizing would be modest and the better use of the capital is to wait for a day when the basis blows out to 160 or 180 points on heavy futures buying. Two things sharpen the view: dividends are an estimate until the companies declare them, and a 0.08% cost assumes orderly execution. Both move the net return by more than the pickup.