A CFD looks like the simplest leverage product there is: pick a market, set a size, go long or short. No exchange, no contract expiry, no delivery. But that simplicity hides the most pointed version of this whole series' question — because with a CFD, the person on the other side of your trade is often the very broker selling it to you, and in one common model, your loss is literally their profit.

This is the fourth piece in our series on how instruments really work.

What a CFD actually is

A CFD — contract for difference — is an agreement between you and a broker to exchange the difference in an asset's price between when you open and close the position. You never own the asset. If you go long oil at 80 and close at 85, the broker pays you the difference times your size. If it drops to 75, you pay them.

The everyday version: it's a private side-bet on a price, settled in cash. You're not buying the barrel — you're agreeing with the bookmaker on what the barrel does.

The mechanics that surprise people

It's leveraged through margin: you post a fraction of the position's value, which magnifies both gains and losses.

It costs you to hold. Keep a CFD open overnight and you pay a financing charge each day (you're effectively borrowing to fund the leveraged position). Over weeks, that quietly adds up.

There's no expiry, but there's a spread on every trade, and your position can be closed automatically if your margin runs out.

What CFDs are good for

• One account, many markets — indices, FX, commodities, shares, all from a single login.
• Easy shorting — going short is as simple as going long.
• Small position sizes and fractional exposure.
• Quick hedging of an existing portfolio.

The other side — your broker, and how it makes money

Here's the part the slick app hides. There are two broad ways a CFD broker handles your trade:

In the A-book model, the broker passes your trade through to the real market or a liquidity provider and hedges it. It earns the spread and commission. Your winning or losing doesn't change its outcome — it's a toll-taker.

In the B-book model, the broker takes the opposite side itself and does not hedge. Now your loss is its gain, and your profit is its cost. Since most retail traders lose over time, internalizing that flow can be very profitable for the broker. Many run a hybrid: B-book the clients who tend to lose, A-book the ones who tend to win.

None of this is necessarily fraud, and a regulated broker can run a B-book legitimately. But you should be clear-eyed about the structural tension: in that model, the firm providing your charts, your platform, your margin terms, and your execution is also the firm that profits when you don't.

What the EU rules changed

To its credit, this is one of the most heavily regulated retail products in Europe. Under EU rules, retail CFD accounts get capped leverage (tighter for volatile assets), mandatory negative-balance protection (you can't lose more than your account), and a required risk warning stating the share of that broker's retail accounts that lose money — a figure that is consistently a large majority. Brokers must show you their own number. It's worth reading.

Where it goes wrong

The conflict of interest in the B-book model is the headline. Beneath it: overnight financing slowly erodes longer-held positions, leverage cuts both ways, and wide or widening spreads in fast markets eat into every trade. The required disclosure exists precisely because the base rate of retail CFD outcomes is poor.

Why they're hard to backtest

A CFD's real cost is broker-specific: the spread you actually pay and the overnight financing you're charged. A backtest using clean mid-prices and ignoring financing will overstate a strategy badly — especially anything held for days, where financing dominates. And you can't assume frictionless fills from a venue that also profits from your losses. Test against realistic spreads and carry, or you're testing a fantasy.

Understand who's on the other side, and the question changes from "is this a good trade?" to "is this a good trade after the spread, the financing, and the fact that my counterparty may be rooting against me?"

Study the past — improve your future. 🥋

Educational content, not investment advice.