The Difference Between A-Book and B-Book Execution Models

The mechanics of how a broker routes your order, the latency, the price sourcing, and the fill confirmation are covered in How Brokers Route Orders and Why It Matters for You. This article addresses the deeper question behind those mechanics: what commercial incentives govern the broker’s behaviour, and what do those incentives mean for your outcomes over hundreds of trades?

The A-book and B-book distinction is the most important structural feature of retail CFD brokerage, and the one least honestly explained in industry marketing. Understanding it requires setting aside the language brokers use to describe their models and examining the financial logic instead.

The Commercial Logic of Each Model

In an A-book model, the broker earns from the spread or commission charged on each transaction. A client who trades 500 times per year generates 500 rounds of spread revenue for the broker, regardless of whether those trades win or lose. The broker is indifferent to your performance because it does not participate in the outcome. Its revenue function is: volume × spread margin. Your success or failure is commercially irrelevant to it.

In a B-book model, the broker takes the other side of your trade. Your profit is the broker’s loss. Your loss is the broker’s profit. The broker’s revenue function is: aggregate client losses, aggregate hedging costs. In a purely B-book operation with no hedging, the broker and its clients have precisely opposite interests on every single trade.

The distinction is structural, not subtle. It defines how a B-book broker is incentivised to behave, irrespective of whether it actually exploits that incentive or manages it ethically.

What the Numbers Look Like for the Broker

To understand why the B-book model generates such consistent broker revenue, it helps to see the arithmetic.

Retail CFD brokers typically publish, either in their own risk disclosures or through regulatory requirements, the percentage of retail client accounts that lose money. Across the major FCA-regulated CFD providers, this figure consistently sits between 70 and 80 per cent of retail accounts. The regulatory requirement to publish this figure exists precisely because regulators recognise the inherent conflict in the B-book model and want clients to understand the base rate of retail trading outcomes.

If a B-book broker has 10,000 retail accounts with an average balance of £3,000, the total client equity is £30 million. If 75 per cent of those accounts lose money at an average annual rate of 15 per cent of their balance, the aggregate annual loss across the losing accounts is approximately £3.375 million. Under a pure B-book model with no hedging, that £3.375 million is the broker’s gross revenue from client losses, against which hedging costs and operating expenses are set.

Compare this with the A-book model on the same client base. At an average spread of 1 pip on EUR/USD, 10 trades per account per month, and a position size averaging £5 per pip, the annual spread revenue across all accounts is approximately £6 million. The A-book model generates more gross revenue in this scenario, but the B-book model’s revenue is entirely uncorrelated with the broker’s own costs: it rises when clients lose and falls when clients win.

That inverse relationship between client outcomes and broker revenue is the commercial structure that creates the incentive for a B-book broker to minimise client profitability.

How the Conflict Manifests: Three Scenarios

Understanding the B-book conflict of interest requires distinguishing between how it manifests in practice across a spectrum of broker behaviour.

Scenario 1: The well-run B-book. A reputable B-book broker recognises that its long-term commercial interest is better served by building a sustainable client base than by maximising short-term extraction from losing clients. Its execution policy prohibits fill manipulation. Its internal pricing model closely tracks the external market. Price improvement is passed to clients consistently. The conflict of interest exists structurally but is managed to near-neutrality. The broker earns from natural attrition in retail client capital, not by systematically manipulating fills.

Scenario 2: The marginal manipulation. A B-book broker allows its execution infrastructure to introduce small, systematic differences between the external market price and the internal fill price. On an individual trade, the difference might be 0.8 pips, imperceptible to any single client and easily attributable to spread variation or market movement. Across 200,000 trades per year at an average of £5 per pip, a systematic 0.8-pip execution disadvantage generates £800,000 of additional broker revenue. No individual client has grounds for complaint. The aggregate effect on the client base is substantial.

Scenario 3: Regulatory enforcement. The FCA has taken enforcement action against retail CFD brokers specifically for manipulating fill prices on B-booked trades. In documented cases, brokers used software to introduce artificial delays in the execution of winning client orders, ensuring that the fill on a winning trade was fractionally worse than the market price at submission, while fills on losing trades were processed at or near the quoted price. The asymmetry was systematic, the per-trade difference was small, and the cumulative revenue extracted from clients was material. The pattern only became visible through statistical analysis of large datasets comparing fill prices against contemporaneous external market prices.

What 1 Pip of Systematic Disadvantage Actually Costs

The second scenario above describes something difficult to detect from the client side. But the mathematics of systematic execution disadvantage is worth making explicit.

A trader executes 400 round-trip trades per year on EUR/USD at £10 per pip. In a neutral execution environment, slippage is symmetrical: approximately half the fills are slightly better than quoted, half are slightly worse, and the aggregate slippage across all trades is close to zero.

In an environment where the broker systematically produces fills that are 1 pip worse than the market price at submission, retaining price improvement and adding a small adverse bias to fills, the trader experiences an additional cost of £10 per trade. Across 400 trades per year, that is £4,000 of additional cost that appears nowhere on a trade summary, cannot be attributed to any single transaction, and is entirely invisible unless the trader is comparing fill prices against external market timestamps.

Over three years, the cumulative extraction from a single account is £ 12,000. The broker’s revenue from this account would have been £12,000 higher than the legitimate spread revenue alone. The client’s account performance would have been £12,000 worse than the strategy’s genuine edge would have produced.

The pattern is documented. It describes the mechanism behind the fill manipulation that FCA enforcement cases have confirmed.

The Hybrid and How Client Profiling Works

Most retail brokers operate a hybrid model that B-books most small client trades and A-books others. The routing decision is made by the broker’s risk management system based on criteria that are supposed to be disclosed in the execution policy, but are often described only vaguely.

The most commercially relevant routing criterion is client profitability. Clients identified as consistently profitable, sometimes called “toxic flow” in internal risk systems, a term that reveals the commercial framing, present a problem for a B-book operation. If the client wins consistently, the broker loses consistently on those B-booked positions. The rational response is to A-book that client’s orders, passing the exposure to the external market and earning only the spread margin rather than bearing the position risk.

The consequence for the consistently profitable retail trader is counterintuitive. The better you trade, the more likely it is that your orders are routed A-book, where execution quality is determined by the external market. Your analytical edge can fully express itself. The worse you trade, the more likely your orders remain B-booked, where the broker benefits from your losses and has a structural incentive, however managed, to maximise them.

This means that the execution model a client experiences can change as their account history develops, without any notification. A new account is typically B-booked by default. A profitable account over time may find its routing profile changing. The execution policy should disclose the criteria for assigning and updating this routing profile. Most do not describe it specifically.

What to Ask Before Opening an Account

Asking a broker direct questions about its execution model before depositing capital reveals more about the platform than any spread comparison or marketing claim.

What percentage of retail client orders are routed externally versus filled internally? A broker comfortable with its model will give a substantive answer. One that deflects to general statements about best execution is telling you something.

Does client profitability affect routing decisions? The honest answer from any hybrid broker is yes. What matters is whether the broker openly describes the criteria.

Is price improvement passed to clients under both routing paths? The execution policy should explicitly state this. If it does not, the default is typically for the broker to retain it for B-booked orders.

Can you provide execution statistics showing the distribution of slippage outcomes across a sample of recent fills? Few brokers publish this proactively. Requesting it before opening an account is a reasonable due diligence step, and the response, or its absence, is informative.

The Bottom Line

The A-book model generates revenue from trade volume with no interest in client outcomes. The B-book model generates revenue from client losses with a structural incentive to limit client success. Most retail brokers operate a hybrid that routes based on criteria that reflect these commercial interests.

The difference is not abstract. Across a trading career of hundreds of trades, the systematic execution disadvantage in a poorly run B-book environment compounds into thousands of pounds in additional cost that never appears as a visible fee. Evaluating a broker’s execution model through its policy documentation, direct questions, and empirical fill quality testing is the diligence that reduces exposure to that cost.

Darren Clarke is a Senior Trader at Lunaro Financial Services. He has spent 40 years on trading desks ranging from institutional inter-bank FX to retail-focused fintechs and brokerages in the City of London.

Disclaimer:

This material is a marketing communication and is provided for general information and educational purposes only. It does not take into account your personal circumstances, objectives or needs. Any opinions are those of the author at the time of writing and may change without notice. Nothing in this material constitutes (or should be construed as) financial, investment, legal, regulatory or tax advice, or a recommendation to engage in any investment activity. You should not rely on this material when making investment or trading decisions.