A market maker quotes a stock with a 5 paise spread, earning 2.5 paise per filled share plus a 0.2 paise exchange rebate, but 40% of fills are followed by a 5 paise adverse move. What does it net per share, and at what share of toxic fills does it break even?

Chakrika nets 0.7 paise a share, about Rs 10,500 a day, and breaks even when 54% of fills are toxic. Half the spread plus the rebate earns 2.7 paise; adverse selection costs 40% x 5 = 2.0 paise. Each extra 10 points of toxic flow costs 0.5 paise, so a modest shift in who trades with Chakrika wipes out the business. Widening the spread helps only if the toxic share does not rise with it.

A money changer at an airport buys dollars at one price and sells at a slightly higher one, and earns the gap on every customer. Most customers are travellers who simply need currency. A few know the rate is about to move. The market maker earns half the spread on every fill but loses to the informed minority, and its profit is the gap between the two, not the spread itself. The loss to informed traders is called adverse selectionThe cost a market maker bears because the counterparties most eager to trade with it are often those who know the price is about to move against it.: the fills Chakrika gets most readily are disproportionately the ones it should not want.

Income per share is 2.5 paise of half-spread plus 0.2 paise of rebate, 2.7 paise. Adverse selection costs 40% of fills times 5 paise, 2.0 paise averaged across all fills, so net is 0.7 paise a share. On 1,500,000 shares a day that is about Rs 10,500, before technology, exchange fees and the capital the inventory ties up. The rebate is 0.2 paise of a 0.7 paise margin: without it, the business would keep only about 0.5 paise.

Net per share is 2.7 minus 5 times the toxic share. It reaches zero when the toxic share is 2.7 / 5 = 54%, only 14 points above today's 40%. Each 10 points of toxic share costs 0.5 paise a share, Rs 7,500 a day at this volume. A new fast trader in the stock, or a large fund starting to work an order through it, can move the toxic share by that much in a week, which is why market makers watch their post-fill price moves more closely than their fill volume.

Partly, and less than it looks. A 6 paise spread raises the half-spread to 3.0 paise. But a wider quote mainly drives away the uninformed traders, who have a choice, while informed traders still trade because their 5 paise edge covers the extra cost, so the toxic share rises. If it rose from 40% to 50%, net would be 3.0 + 0.2 - 2.5 = 0.7 paise, no better than before, on fewer shares. The levers that work attack adverse selection directly: faster quote updates so stale prices are picked off less often, skewing quotes away from signals of informed flow, and separating flow by source where the venue allows it.