IG lost 22 percent in a day because its clients started winning

IG Group shares fell 22 percent on Friday 2 October, and at one point in the session the drop reached 27 percent. The trigger was a third-quarter trading update in which IG said it expected revenue of about 240 million pounds, down 14 percent on the same quarter a year earlier, and cut its full-year growth guidance from 10 to 15 percent down to a mid-single-digit range. Volume was not the problem. The problem was a single number most retail traders never think about: revenue retention. IG kept roughly 70 percent of the revenue its over-the-counter clients generated, against an average of about 80 percent since it changed how it runs its market-making book in the second half of 2025. Put plainly, more of IG's clients made money, and IG paid for it.

What revenue retention actually measures

A CFD broker that internalises client flow earns money in two ways. It earns spread and commission on every trade, and it earns or loses the net result of client positions it chooses to hold rather than hedge. Revenue retention is the share of gross client trading revenue the broker keeps after paying out to the clients who win. At 80 percent, a broker is keeping four pounds of every five its clients generate. At 70 percent it keeps seven of every ten. On a quarter of IG's size, that ten point gap is the difference between a guidance upgrade and a guidance cut.

Retention moves with market conditions, and it moves most when markets trend cleanly in one direction for long enough that clients riding the move can close out in profit. The third quarter delivered several of those trends, from a sharp rally in Bitcoin to sustained moves in rates and currencies after the Federal Reserve resumed hiking. Clients who were positioned with those trends won, and the broker on the other side of their positions lost a portion of what it would otherwise have kept.

IG's update also put a number on the size of the miss. Net trading revenue for the quarter was expected at about 210 million pounds, down from 249.5 million a year earlier, according to Bloomberg's reporting. That is roughly 40 million pounds of revenue that went back to clients rather than staying on IG's income statement, inside a single quarter.

The one percent who move the whole book

The timing of IG's update is striking because, four days earlier, Finance Magnates published data from iSAM Securities' Radar risk platform showing how concentrated retail winnings really are. Over the past year, the top 1 percent of winning retail accounts captured 66.5 percent of all client profits, and the top 5 percent captured 85.5 percent. The same data showed that 1 percent of clients can account for 30 percent of a broker's total drawdown, which is the money a broker pays out to profitable traders.

Read together, those two data points explain the quarter better than any analyst note. A retail broker's risk is not spread evenly across its client base. It sits in a small group of skilled or simply well-positioned traders whose results swing the entire book. When that group has a strong quarter, retention falls, and a broker that had guided investors to expect stable retention has to explain why the most predictable part of its model suddenly was not.

This is the uncomfortable truth behind the business model, and it is one the industry rarely says aloud. A CFD broker that internalises risk does well when most clients lose and badly when its best clients win. Regulators already require brokers to publish the share of retail accounts that lose money. What the market learned on Friday is that the share of clients who win, and how much they win, matters just as much to the share price.

• Q3 revenue expected at about 240 million pounds, down 14 percent year on year
• Net trading revenue about 210 million pounds, against 249.5 million a year earlier
• OTC revenue retention about 70 percent, against an 80 percent average
• Full-year growth guidance cut from 10 to 15 percent to mid-single digits
• Top 1 percent of winning retail accounts took 66.5 percent of client profits, per iSAM Radar

Why the whole sector sold off with it

IG did not fall alone. CMC Markets dropped about 9 percent on the day, Plus500 slid and later issued a reassuring trading update of its own, and XTB was off around 3 percent. Investors were not reacting to any specific weakness at those firms. They were repricing the whole category, because if retention can fall ten points at the most established listed broker in London, it can fall anywhere.

That is the part of the story with the longest tail. Listed retail brokers have spent two years selling investors on more predictable earnings, through better risk management, more hedging of toxic flow, and diversification into share dealing and investment accounts. IG's own diversification is real: it has been building stock trading, crypto and wealth products for exactly this reason. But one quarter showed that the core CFD book can still overwhelm everything else, and investors will now demand a larger discount for that uncertainty.

For private brokers in Asia and the Gulf the lesson is quieter but sharper. Many of them run more of their flow internally than IG does, with less diversification and thinner capital buffers. A trending quarter that costs a FTSE 100 company 40 million pounds can cost a smaller firm far more in relative terms, and those firms do not publish trading updates. Their clients usually learn about it through slower withdrawals.

A CFD broker that internalises risk does well when most clients lose and badly when its best clients win.

The split between winners and the rest keeps widening

IG's quarter fits a pattern we have tracked all year. When we compared NAGA's return to profit with iFOREX's collapsing EBITDA, the dividing line was not market share but how deliberately each firm managed its risk and revenue mix. The brokers that diversified early and priced their risk honestly held up. The ones that relied on a stable loss rate among clients found that stability was a market condition, not a business model.

The same logic explains why large brokers keep moving into adjacent businesses. CMC Markets is launching a proprietary trading offer on 1 October, a move we examined in our analysis of CMC's prop launch. Prop challenges earn fee revenue that does not depend on whether a funded trader wins or loses in the same way a CFD book does. After Friday, expect more listed brokers to explain to investors exactly how much of their income is exposed to client performance, and how much is not.

What smart brokers do differently after a quarter like this

The firms that come through these quarters well have three things in common. They identify their consistently profitable clients early and route that flow to liquidity providers instead of holding it, which caps the damage when those clients have a good run. They publish retention as a range, not a target, so investors and partners are not surprised when markets trend. And they build revenue lines, such as subscription research, interest on client cash and share dealing, that do not depend on clients losing.

None of that is new thinking, and IG already does much of it. What Friday showed is that doing most of it is not enough when a quarter delivers the kind of clean trends that suit active traders. The brokers watching this most closely are the ones about to raise money or list, because they now know exactly which question the first serious investor will ask.

The headline will say IG had a bad quarter. The more useful reading is that the market just put a price on a risk the retail brokerage industry has always carried and rarely discussed in public: a few hundred very good traders can decide a listed company's year. Brokers that understand their own top 1 percent, and manage that flow deliberately, will keep their valuations. Those that treat every client the same will keep getting surprised by the clients who are not.