What Is a CFD? Leverage, Margin and Risk Explained

A contract for difference is a bet on a price you never have to own. You and a broker agree to settle the gap between an asset's price when you open the trade and its price when you close it - in cash, both ways. If the price rises, the broker pays you the difference. If it falls, you pay the broker. No shares change hands, no barrels of oil arrive at your door. That single design choice - settle the difference, skip the ownership - is what makes a CFD cheap to enter, flexible across markets, and far more dangerous than its low entry cost suggests.

I edit broker reviews and the education library here, which means I read the risk disclosures most traders skip. So let me start where those disclosures end up: in 2016 the UK's Financial Conduct Authority looked at a sample of CFD client accounts and found that 82% of them lost money, with an average loss of £2,200 per account. European regulators found much the same. Keep that number in your back pocket for the rest of this piece. Everything CFDs do well - the leverage, the global reach, the ability to go short in a click - is the same mechanism that produces it.

How a CFD actually settles

Picture two parties, a buyer and a seller, and one number between them: the change in an asset's price. The contract pays out that change, nothing more. Gold goes from 2,400 to 2,420 and you were long? You collect twenty points times your size. It dropped instead? You hand the twenty points over. There is no expiry date forcing the trade to end, no time decay eating at it, and no underlying asset sitting in an account. A CFD simply mirrors whatever it tracks for as long as you hold it.

CFDs were developed in Britain in 1974 as a way to gain leveraged exposure to gold, but the modern version is younger. It is widely credited to Brian Keelan and Jon Wood of UBS Warburg, who built it as an equity swap traded on margin during a Trafalgar House deal in the early 1990s. The first users were hedge funds and institutional desks, and the appeal was specific: they could get exposure to London Stock Exchange shares on a small margin, and because no physical shares moved, they sidestepped stamp duty. The instrument was a tax-and-capital efficiency tool long before it became a retail product.

That retail turn came in the late 1990s, when UK firms - GNI first, then IG Markets and CMC Markets - wrapped the contract in live online platforms and marketed two features hard: leverage, and tax-free status under UK spread-betting rules. The product spread to Australia, Canada, Singapore, South Africa and across Europe. It did not spread everywhere, and where it stopped tells you a lot.

Where you cannot trade them, and why that matters

The United States bans CFDs outright for retail traders; the SEC and CFTC will not let them list on regulated exchanges or trade on any platform reaching US clients, on grounds of risk. Belgium bans them for over-the-counter instruments. Hong Kong treats a securities-based CFD as a futures contract that must be exchange-traded, which effectively shuts the door - though it carves out a separate regime for leveraged spot FX.

A ban is not bureaucratic noise. When a market as large and litigious as the US decides retail investors should not touch a product, that is a verdict on the product's risk profile, written by people with full access to the loss data. Read the map of prohibitions as a second opinion on the 82% figure. The places that let you trade CFDs are not safer; they are simply more permissive.

Leverage and margin: the same lever cuts both ways

Here is the mechanic everyone is sold and few are taught. You put down a fraction of a position's value - say 10% margin - and control the whole thing. Deposit 10,000 to control 100,000. A standard illustration: a CFD on a stock that moves from 10 to 11, on a 100,000 position held with 10,000 of margin, returns roughly 1,000. A clean 10% gain on the move becomes a 100% gain on your cash. Now run the same move the other way and the arithmetic is merciless: an identical small drop wipes out the deposit just as fast.

Leverage does not change the odds of a price move. It changes the size of your stake relative to your wallet - and it does so symmetrically. The thing that makes a small favourable tick feel like a windfall makes a small unfavourable tick a margin call. If prices move against an open position, the broker can demand more cash to keep it alive, and in a fast market that demand arrives with little notice. Miss it and the position is liquidated at a loss you are still liable for. You can lose more than you deposited. That sentence is the whole risk warning, and it is true.

This is why the Australian regulator's own trader-information site says, in plain words, that trading CFDs is riskier than betting on horses or going to a casino. It is also why most regulators now cap leverage. After the 2016 ESMA warning, Cyprus's CySEC limited retail leverage to 50:1 and banned bonus incentives; the UK's FCA followed with restrictions landing at 30:1; Australia's ASIC set the same 30:1 ceiling in 2021. The caps exist because the uncapped version reliably emptied retail accounts.

CFDs versus futures: a quick scorecard

The CFD most resembles a futures contract, and the differences are exactly the ones that decide which tool fits a job. I keep a comparison table for the education library; here is the short version.

The practical reading: CFDs win on access and flexibility for small accounts, and on transparent pricing that simply tracks the underlying. Futures win on maturity and counterparty safety, which is why professionals reach for them on indices and rates. Note the last row. A CFD has no expiry, so a broker writing a CFD over a futures contract has to "roll" your position to the next period as the old future's liquidity dries up - a quiet cost that accrues whether or not the market moves your way.

The costs nobody advertises

"Zero commission" is the headline on a lot of CFD platforms, and it is technically true while being financially misleading. The cost did not vanish; it moved. Brokers earn from the bid-ask spread - the gap between what you pay to buy and receive to sell - which is often a touch wider than the raw market. You pay it on the way in and the way out, every single trade.

Then there is the holding cost. Leverage is borrowed money, and borrowed money has a price: a daily financing charge, the "swap," accrues on overnight positions. Hold a leveraged trade through a flat, sideways market and the swap quietly erodes your break-even while the price goes nowhere. Spreads also widen during news and thin-liquidity sessions, which is precisely when a stop-loss might trigger - so the very moment you most need a clean exit is when execution is worst. Add these up before you size a trade, not after.

A stop-loss is a tool, not a guarantee

The standard advice is to cap your downside with a stop-loss order, and you should use them. But understand what a stop actually does. When the market reaches your stop level, the order becomes a market order - it fills at the next available price, not the price you set. In a calm market that distinction is trivial. In a gap - a weekend, a central-bank shock, a sudden geopolitical headline - the price can leap straight past your stop, and you fill far below it. The protection you bought evaporates exactly when you reach for it.

There is a real tension here worth naming. Set stops too tight and ordinary market noise knocks you out of perfectly good trades; set them too loose, or skip them, and a single bad gap takes the account. There is no setting that escapes the trade-off, only settings that suit your size and your stomach. A guaranteed stop (offered by some brokers for a fee) removes the gap risk but adds a cost. Pick deliberately, and never assume a plain stop will hold at your number.

Counterparty risk: the danger behind the danger

Because a CFD is over-the-counter, your contract is with the broker, not a clearing house. If the broker fails, the contract can be worth little or nothing - even if your market call was dead right. You can pick the correct direction and still lose, because the party owing you the difference cannot pay. Regulated CFD providers must segregate client funds to protect balances in a default, but the collapse of MF Global is a standing reminder that protections can break.

This is why, on every broker review we publish, regulation and the ability to withdraw come before spreads and well before bonuses. A tight spread on an offshore-only licence is a trap with good marketing. Before you weigh a single feature, ask two questions: who regulates this broker, and can clients actually get their money out? If the answers are weak, nothing else on the page matters.

The regional angle: a market drifting east

The centre of gravity in retail FX and CFD trading appears to be shifting toward Asia - emerging-market participation in India, Southeast Asia and beyond is where much of the new account growth now sits, and our own coverage skews that way for a reason. One industry estimate put Asia at around two-thirds of FX-and-CFD web traffic in early 2025, though that figure comes from a single market-research source we have not independently verified, so treat the precise share as directional rather than gospel.

The shift matters less as a statistic and more as a regulatory fact. A trader in a Tier-2 or Tier-3 market often faces the choice the leverage caps were designed to remove: a tightly regulated local or near-shore broker at 30:1, or an offshore entity dangling 500:1 and a welcome bonus. The honest framing is that strict regulation pushes some flow toward the unregulated edge, where the 82% loss rate runs without a safety net and no regulator will recover your funds. The eastward drift in volume is real; so is the gap in protection that comes with it.