76% of CFD Traders Lose Money: Is the Game Rigged?

These figures are not marketing scare tactics. They are audited disclosures mandated by the European Securities and Markets Authority under its 2018 product intervention measures. Every quarter, each broker recalculates the figure using a standardized methodology covering the most recent twelve-month period. Quarter after quarter, year after year, the story remains the same: the overwhelming majority of retail CFD traders lose money.

The question worth asking is why. Is it simply that trading is hard and most people are bad at it? Or is there something about the CFD product itself, its structure, its incentives, and its hidden costs, that systematically tilts the odds against the retail participant? The answer, as we will show, is that both factors are at work. But the structural disadvantages of CFDs are not widely understood, and they deserve serious examination.

The ESMA Data: What the Numbers Actually Tell Us

ESMA implemented its product intervention measures for CFDs in August 2018, following extensive analysis of retail trading outcomes across the European Union. The intervention included mandatory leverage caps, negative balance protection, a prohibition on incentive payments (such as deposit bonuses), and the requirement that brokers display standardized risk warnings disclosing their retail loss percentages.

The rationale was straightforward. ESMA determined that the CFD product posed significant investor protection concerns and that existing disclosure requirements were insufficient. In its decision memo, ESMA cited evidence that between 74% and 89% of retail accounts lost money across the industry, with average losses per account ranging from 1,600 euros to 29,000 euros depending on the broker.

Several patterns in the data deserve attention. First, the loss rates are remarkably stable over time. They do not improve when markets are calm or worsen only when markets are volatile. The consistency suggests structural rather than cyclical causes. Second, brokers that cater to more sophisticated clients report lower loss rates. Interactive Brokers, which targets experienced active traders and imposes higher minimum standards, reports 62%. Brokers that invest heavily in marketing to beginners and first-time traders consistently report rates above 75%. Third, the figures have remained largely unchanged since the intervention measures were implemented. Leverage caps and other protections have not materially reduced the retail loss rate, suggesting that the causes run deeper than leverage alone.

The FCA in the United Kingdom adopted equivalent measures and publishes similar data. ASIC in Australia implemented comparable rules in March 2021. The pattern is global. Across every jurisdiction that requires disclosure, the majority of retail CFD traders lose money.

The B-Book Model: Your Loss Is Their Profit

To understand why CFD loss rates are so high, you need to understand how CFD brokers actually make money. The answer, for most brokers, goes far beyond spreads and commissions.

In a B-book model, the broker does not forward your order to an external market or liquidity provider. Instead, the broker takes the other side of your trade internally. When you go long, the broker goes short. When you go short, the broker goes long. Your profit is the broker's loss. Your loss is the broker's profit.

Why would any business voluntarily take risk against its own customers? Because the data is overwhelmingly in the broker's favor. If 76% to 82% of your retail clients lose money, internalizing their flow is not a gamble. It is a near-certainty of profit. The broker collects spreads from every trade (both winning and losing), collects overnight financing fees from every position held overnight, and then captures the net trading losses of the majority of its client base. The combination of fee revenue and client losses makes B-booking one of the most profitable business models in financial services.

Here is a simplified illustration. A broker has 1,000 active retail clients trading CFDs. The average monthly notional volume per client is $200,000. The broker earns an average spread of 1.5 pips per round trip, generating approximately $30 per client per month in spread revenue, totaling $30,000 across the client base. If 78% of clients are net losers with an average monthly loss of $600, client trading losses total $468,000. The 22% of winning clients average $400 in monthly gains, costing the broker $88,000. The broker's net profit from trading alone is $468,000 minus $88,000 plus $30,000 in spreads, totaling $410,000 per month. The client losses are not a side effect of the business model. They are the primary revenue driver.

The Structural Disadvantages: How the Deck Is Stacked

Spread Manipulation and Widening

CFD brokers control their own price feeds. The prices you see on your platform are not exchange prices. They are the broker's proprietary quotes, derived from market data but adjusted at the broker's discretion. During normal market conditions, the broker's prices track the real market closely. But during volatile events, such as economic data releases, central bank decisions, or sudden market moves, the broker can widen spreads dramatically.

A forex pair that normally quotes with a 1-pip spread might widen to 5 or 10 pips during an NFP release or FOMC announcement. Index CFDs can see spread widening of 10 to 20 points. These are precisely the moments when many traders are active and when stop-loss orders cluster near the current price. Wider spreads during volatile periods are presented as a natural market phenomenon, but on a CFD platform where the broker controls pricing and may be on the other side of your trade, the broker has a direct financial incentive to widen spreads as far as clients will tolerate.

Overnight Fees: The Silent Profit Killer

Overnight financing fees on CFDs are charged every night a position is held past the daily rollover time. These fees are calculated on the full notional value of your position, not on your margin deposit. A $50,000 position with 10x leverage means you deposited $5,000, but the overnight fee is calculated on the full $50,000.

Typical annualized overnight rates range from 5% to 8% of notional value. On a $50,000 position at 7% annualized, the daily charge is approximately $9.59. Over a month, that is $287. Over a quarter, $863. Over a year, $3,500. That $3,500 represents 70% of the $5,000 margin deposit. The position must appreciate by 7% just to cover financing costs before any profit is realized.

These rates are set by the broker, not by the market. The broker determines the markup over the benchmark rate, and this markup is rarely competitive because most retail traders do not compare overnight rates when selecting a broker. This is pure, opaque, discretionary pricing on a cost that compounds relentlessly against the trader.

Asymmetric Slippage

Multiple academic studies and regulatory investigations have documented patterns of asymmetric slippage at CFD brokers. Slippage occurs when the executed price differs from the requested price. In a fair execution environment, positive slippage (getting a better price) and negative slippage (getting a worse price) should occur with roughly equal frequency and magnitude. At many CFD brokers, the pattern is skewed. Negative slippage occurs more frequently and in larger amounts than positive slippage.

The FCA took enforcement action against FXCM for precisely this issue. FXCM was found to have systematically retained positive price improvements rather than passing them to clients. When the market moved in the client's favor between the time of order submission and execution, FXCM kept the benefit. When the market moved against the client, the full negative slippage was passed through. This asymmetric treatment meant clients always experienced the downside of execution delays but never the upside.

The Information Asymmetry

B-book brokers know the exact position of every client's stop-loss order, take-profit order, and limit order. They know where order clusters form. They know which clients are profitable and which are losing. They can see the aggregate exposure of their client base in real time. This information asymmetry is inherent to the model and creates opportunities for exploitation that cannot exist on a transparent exchange.

Whether a specific broker acts on this information is a matter of their internal policies and ethics. But the fact that the information exists and that the broker has a financial incentive to use it adversely represents a structural disadvantage for the retail trader that no amount of regulation can fully eliminate.

The Alternative: On-Chain Perpetual Futures

On-chain perpetual futures fundamentally restructure the relationship between the trader and the trading venue. They do not fix CFDs. They replace the entire model with something architecturally different.

There is no B-book. On a decentralized exchange like Hyperliquid, no entity takes the other side of your trade. The protocol matches buyers and sellers on a public order book. The exchange earns a small fee from every trade, regardless of whether the trader wins or loses. Revenue is volume-dependent, not outcome-dependent. The exchange is incentivized to attract traders and facilitate volume, not to profit from trader losses.

Execution is verifiable. Every fill on Hyperliquid is recorded on the blockchain with a timestamp, price, and size. You can independently verify your execution against the order book state at the time of the trade. Asymmetric slippage is not just unlikely. It is cryptographically impossible when the matching engine follows deterministic rules encoded in public smart contracts.

Pricing is transparent. There is no proprietary price feed. The order book is public. The mark price is derived from a documented formula using multiple reference exchanges. Spread manipulation is detectable because every order and cancellation is recorded on-chain. The concept of a broker widening its proprietary spread does not exist in this model.

Overnight costs are market-driven. Instead of broker-set swap rates, on-chain perps use funding rates that are determined by market supply and demand. When the market is balanced, funding costs approach zero. When the market is imbalanced, the funding rate creates an economic incentive for traders to take the other side. Critically, funding can be positive or negative. If you are positioned on the receiving side of funding, your position generates income rather than incurring cost. This is the opposite of CFD overnight fees, which are always a net cost to the trader.

Self-custody eliminates withdrawal risk. On Hyperliquid, your funds sit in a smart contract under your cryptographic control. Withdrawals are permissionless and typically settle in seconds. There is no broker who can delay your withdrawal, freeze your account, or add friction to discourage you from withdrawing profitable balances.

Would Traders Fare Better on Perps?

It would be dishonest to claim that moving from CFDs to on-chain perps will make a losing trader profitable. Trading is inherently difficult. Leverage amplifies both good and bad decisions. Many traders will lose money regardless of the venue.

But the structural headwinds are measurably different. A trader using CFDs faces spread markups, opaque overnight financing, potential asymmetric slippage, possible adverse execution from a conflicted counterparty, and a venue with a financial interest in their losses. A trader using on-chain perps faces transparent trading fees, market-driven funding rates, verifiable execution, and a venue that is architecturally indifferent to whether they win or lose.

These differences do not guarantee better outcomes, but they remove friction that makes success harder than it needs to be. A trading strategy with a genuine positive expected value has a better chance of delivering returns when it does not have to overcome a 3% to 5% annual drag from hidden costs and a counterparty incentivized to work against it.

The 76-82% loss rates published by CFD brokers are not an inevitable feature of leveraged trading. They are the product of a specific market structure with specific incentive misalignments. Change the structure, align the incentives, and the playing field shifts meaningfully toward the trader.

Hyperdash Tip

Before sizing any leveraged trade, calculate your total cost of carry across the expected holding period. On a CFD, this means the spread, overnight financing, and any commission, summed over the number of days you expect to hold. On an on-chain perp via Hyperliquid, this means the taker fee and the projected funding rate. Use Hyperdash to view real-time and historical funding rates, compare them to typical CFD overnight costs, and make an informed decision about which venue gives your strategy the best chance. In most cases, the cost differential is large enough to change the expected value of the trade.

Frequently Asked Questions

Are the ESMA loss disclosure numbers reliable?

Yes. ESMA mandates a standardized calculation methodology that brokers must follow, with the figures covering the most recent twelve-month period and updated quarterly. Brokers are subject to regulatory audit on these disclosures. The consistency of the figures across different brokers, jurisdictions (ESMA, FCA, ASIC), and time periods reinforces their reliability. If anything, the figures may understate actual losses because they measure account-level outcomes and do not capture traders who deposit and lose multiple times within the same account.

If most CFD traders lose, why do people keep trading them?

Several factors explain continued CFD adoption. Brand awareness is high due to massive marketing spending by brokers. Onboarding is easy, with many brokers allowing account creation in minutes. Platforms are well-designed and intuitive. Many traders are unaware of alternatives like on-chain perpetual futures. In some jurisdictions, CFDs provide access to markets that are difficult to trade through other instruments. And many traders underestimate the structural costs and conflicts until they have experienced significant losses firsthand.

Do any CFD brokers operate honestly?

Many regulated CFD brokers operate within the law and manage their conflicts of interest as required by their regulators. The issue is not necessarily dishonesty at the individual broker level but the structural incentive misalignment built into the product. Even a well-managed B-book broker profits when clients lose. Even a broker that never manipulates prices or hunts stops benefits from a system where the majority of retail participants lose. Regulation can mitigate the worst abuses, but it cannot eliminate the fundamental conflict of a venue that earns more when its users lose more.

How can I independently verify execution on an on-chain exchange?

On Hyperliquid, every trade is recorded on the blockchain with a transaction hash, timestamp, execution price, and size. You can look up any trade using the Hyperliquid explorer or through third-party block explorers. By comparing your execution price to the order book state at the block timestamp, you can independently verify that you received a fair fill. This level of transparency does not exist in the CFD model, where execution records are stored solely in the broker's internal systems and are not independently auditable.