A-Book and B-Book are industry shorthand for what a forex or CFD broker does with the market risk created by a client’s trade. They are not official regulatory classifications, and they do not tell you by themselves whether a broker is honest, fast or well regulated. An A-Book broker offsets client exposure externally. A B-Book broker retains it. Most large OTC brokers use some form of hybrid risk management in between.
| Model | What happens to broker market risk? | Where broker economics mainly come from | Main retail issue |
| A-Book / fully hedged | Client exposure is offset with an external liquidity provider, one-to-one or in aggregate | Spread/commission/funding less hedging and liquidity costs | Execution quality and external liquidity |
| B-Book / unhedged | Broker keeps the exposure on its own book | Fees plus gains/losses from customer trading performance | Direct conflict if customer losses become broker gains |
| Hybrid / partial hedge | Opposing client flow is netted and some residual exposure is hedged according to limits | Fees plus risk-management outcome | How the broker decides what to internalise or hedge |
The Short Answer: A-Book and B-Book Describe Risk Management, Not the Trading Screen
The labels are easy to misunderstand because the customer experience can look identical. A trader opens EUR/USD in MetaTrader, sees a fill, watches floating P&L and later closes the position. Nothing on that screen necessarily reveals whether the broker hedged the exposure with a bank, offset it against another customer’s trade, retained it internally or changed its hedge later.
European regulators use more precise language. ESMA has described three possible CFD counterparty models: a firm that hedges all client orders one-to-one or in aggregate; a firm that acts as counterparty without hedging; and a firm that partially hedges based on limits, volumes or selected subsets of flow. Those are effectively the economic models retail traders call A-Book, B-Book and hybrid.
The same broker can use more than one model. It can hedge gold aggressively during a major data release, internalise small EUR/USD flows in normal markets, and automatically hedge a large client order that pushes net exposure beyond a risk limit.
Figure 1. A-Book, B-Book and hybrid are shorthand for how the broker manages market exposure after accepting an OTC client trade.
The First Misconception: Your A-Book Trade Does Not Necessarily Go Straight to a Bank
A retail CFD or rolling-spot forex trade is normally an OTC contract with the broker. The broker may then enter a separate hedge with a liquidity provider. Those are two legal transactions, not one customer order travelling unchanged through a pipe.
IC Markets Global’s 2026 order-execution policy provides a useful real-world illustration. It says the company acts as principal at all times and is the sole contractual counterparty to client trades, even though it may transmit orders for execution to third-party liquidity providers. It also says its prices are built from multiple liquidity and data providers.
That structure is compatible with externally hedged execution. The client is still facing the broker; the broker is separately managing its exposure in the wholesale market. This is why ‘STP’, ‘ECN’, ‘no dealing desk’ and ‘A-Book’ should not be treated as synonyms for an agency brokerage model.
The retail implication is important in a dispute. Your legal claim is generally against the broker that filled your OTC contract, not against the bank or non-bank liquidity provider the broker may have used behind the scenes.
How A-Book Execution Works
In a fully hedged model, the broker seeks to neutralise the market exposure created by client positions. If a client buys one standard lot of EUR/USD, the broker can sell an equivalent amount externally. If the euro rises and the client makes money, the broker’s external hedge should make a corresponding gain that funds the client liability. If the euro falls, the broker gains on the customer contract but loses on the hedge.
The hedge does not have to be one-for-one. ESMA explicitly recognises aggregated hedging. A broker can wait until multiple customer positions are combined and then hedge only the resulting net exposure, provided its risk controls allow it.
This makes A-Book revenue comparatively fee-like. The broker tries to earn the spread, commission or financing charge while passing most directional market risk to liquidity providers. But it introduces other risks: hedge slippage, liquidity-provider rejection, basis differences, latency, financing costs and the possibility that the retail fill and hedge fill occur at different prices.
A-booking therefore does not guarantee better fills. During a fast market, a broker that must immediately source external liquidity can face thin depth or rapidly moving quotes. A broker internalising flow may be able to fill a small order from its own risk book without waiting for an external hedge. The quality question is the price and execution outcome, not the marketing label.
How B-Book Execution Works
In an unhedged B-Book model, the broker accepts the client’s OTC position and keeps the resulting market risk. Economically, the broker is short the client’s profit and long the client’s loss.
If a customer buys EUR/USD and later loses $500, an unhedged broker’s position gains roughly the opposite amount before fees and costs. If the customer wins $500, the broker must pay that gain from its own resources. That is the conflict regulators focus on: the firm can benefit financially when its customer loses.
The conflict does not automatically make the model illegal. Market makers across financial markets act as principal. The regulatory question is whether conflicts are identified and controlled, prices are fair, orders receive best execution where required, customer interests are not manipulated and the firm has enough capital and risk controls to survive adverse client performance.
ESMA has stated that where a CFD provider is the counterparty and does not hedge client orders, customer loss can correlate directly with firm profit. The FCA likewise describes conflicts as inherent in CFD business models and focuses supervision on whether providers generate poor outcomes by encouraging excessive frequency or size.
Original Model: Why the Two Books Produce Very Different P&L
A simplified one-lot EUR/USD example makes the distinction concrete. Assume one standard lot has a $10 pip value, the broker captures a 1.0-pip round-trip customer spread and an external hedge costs 0.4 pip. Ignore financing, slippage, commissions, netting and operating expenses.
Under the fully hedged case, a 50-pip customer win or loss is largely offset by the broker’s external hedge. The illustrative gross economics are therefore about $6 either way: $10 of customer spread less $4 of hedge cost.
Under the unhedged case, the same 50-pip move dominates the spread. A client losing 50 pips produces about $510 of gross broker P&L; a client winning 50 pips produces about negative $490. The expected economics may still be attractive across a large diversified customer base, but the variance and conflict are completely different.
Figure 2. Illustrative one-lot EUR/USD economics. The purpose is to isolate directional risk; real broker books include netting, financing, commissions, slippage and hedging at portfolio level.
Hybrid Execution Is Closer to How Large OTC Dealers Actually Operate
The A-versus-B debate often assumes every trade must be placed into one of two boxes. Real dealer risk management is usually more dynamic.
Suppose customers are collectively long 100 standard lots of EUR/USD while other customers are short 80 lots. The broker has 180 lots of gross customer flow but only 20 lots of net directional exposure. Hedging all 180 lots externally would create unnecessary turnover because most of the customer risk offsets internally.
If the broker nets the opposing positions and hedges only the remaining 20 lots, just 11.1% of gross client flow needs an external hedge in this simplified example. At an illustrative external hedge cost of 0.4 pip per lot and $10 per pip, gross hedging cost falls from $720 on 180 lots to $80 on 20 lots – an $640 difference before considering the timing and quality of fills.
Figure 3. Netting 100 long lots against 80 short lots leaves 20 lots of residual market exposure. Internalisation reduces external hedge turnover without requiring the broker to bet against one specific client.
IG describes a version of this architecture in its annual reporting. It says it is the counterparty to OTC client trades, centralises market risk from customer activity, offsets opposing positions and externally hedges residual exposure when it approaches Board-approved limits. Because of its scale, IG says the vast majority of trades naturally offset against other customer flow.
That is not a pure A-Book or pure B-Book. It is an internalisation and residual-hedging model. The broker earns transaction revenue while trying to keep net market risk inside predefined tolerances.
Internalisation Is Not Unique to Retail Forex
The concept has a broader market-structure precedent. The BIS 2025 Triennial Survey says spot and most FX derivatives are OTC and that dealers match more than 80% of customer trades within their own internal liquidity pools.
Wholesale bank internalisation is not identical to a retail CFD B-Book. Banks may hold inventory only briefly, operate across institutional flow and hedge through interdealer venues. But the underlying economic insight is the same: a dealer with diverse two-way customer flow can offset clients against each other and avoid paying the spread and market impact of immediately hedging every trade externally.
The BIS also reported average OTC FX turnover of about $9.6 trillion per day in April 2025. At that scale, internal matching is a core piece of market plumbing, not an exotic retail-broker trick.
What Plus500’s Accounts Reveal About Retained Customer Risk
Public-company disclosures can show the economics more clearly than broker marketing. Plus500 reports a metric called Customer Trading Performance – the gains or losses arising from customers’ trading positions.
In Q1 2026, Plus500 reported Customer Income of $270.6 million and Customer Trading Performance of negative $38.9 million. Trading income was therefore $231.7 million, with a further $10.4 million of interest income taking total revenue to $242.1 million.
In other words, customer trading performance reduced that quarter’s trading income by about 14.4% relative to Customer Income. The company says this component is expected to be broadly neutral over time and describes real-time exposure monitoring and risk limits. It has also said it is testing more holistic automated hedging.
Figure 4. Plus500 Q1 2026 disclosures illustrate how customer trading performance can increase or reduce an OTC broker’s revenue. This does not label the entire business a pure B-Book.
This is an important distinction. A broker can retain some customer exposure without making a deliberate directional prediction about each trader. The book can be managed statistically across customers, instruments and limits. But the more client P&L is retained rather than hedged, the more broker earnings can move with aggregate customer performance.
Why B-Book Is Not Automatically ‘The Broker Manipulates Your Trade’
Retail forums often jump from ‘the broker is my counterparty’ to ‘the broker controls whether I win.’ Those are not the same statement.
A principal broker has a conflict, but regulated execution rules still matter. FCA rules treat dealing on own account with clients as execution of client orders and require firms to take sufficient steps to obtain the best possible result. For OTC products, the broker must check the fairness of the proposed price using market data and comparable products where possible.
Execution policies should also disclose when orders are executed outside a trading venue and the counterparty risk that creates. Regulators can examine connected-party venues, pricing, slippage, rejections and whether the firm has commercial incentives that interfere with best execution.
The practical danger is therefore not the existence of a B-Book by itself. It is a broker using its control of pricing, execution or account terms to exploit the conflict – for example through asymmetric slippage, artificial delays, unfair re-quotes or withdrawal abuse.
Why A-Book Is Not Automatically Better
A fully hedged broker has less incentive to benefit from a customer’s directional loss, but that does not make every trade cheap or high quality.
The broker may still mark up liquidity-provider prices, charge commissions or earn financing spreads. Its hedge provider may reject an order. Large orders may receive partial fills. A volatile market can create slippage between the customer order and the hedge. A broker using one weak liquidity source can provide worse execution than an internalising broker with a deep, diversified book.
IC Markets’ execution policy, for example, explicitly discusses slippage, partial fills and the fact that market orders can execute away from the displayed top-of-book price when liquidity is insufficient.
The relevant metric is therefore total execution quality: spread plus commission, fill rate, speed, positive and negative slippage, rejection rate, financing cost and the behaviour of stops during volatility.
The Marketing Labels That Often Confuse the Issue
| Marketing term | What it may suggest | What it does NOT prove |
| STP | Orders/exposure may be passed to external liquidity providers | That the broker is legally acting as agent rather than principal |
| ECN | Access to aggregated or electronic liquidity | That the retail customer is a direct participant in an institutional ECN |
| No Dealing Desk | Little or no manual dealer intervention in routine execution | That the broker never internalises or holds risk |
| Market maker | Broker quotes prices and may act as principal | That every customer loss is deliberately retained |
| A-Book | Exposure is generally hedged externally | That each customer order is copied one-for-one |
| B-Book | Exposure is retained internally | That prices are necessarily manipulated or withdrawals are unsafe |
What Should a Retail Trader Actually Check?
The first document to read is the order-execution policy. Look for phrases such as ‘act as principal’, ‘sole counterparty’, ‘execution venue’, ‘liquidity provider’, ‘hedge’, ‘aggregate’, ‘internalise’ and ‘market risk’. Those words usually tell you more than an FAQ claiming ‘ECN execution.’
Second, check execution statistics if the broker publishes them. A broker that can show fill rates, rejection rates and price-improvement data gives the trader something measurable to evaluate.
Third, look at conflict-of-interest disclosures. A broker that retains risk should explain how it prevents staff, pricing systems or account-management incentives from benefiting unfairly when customers lose.
Fourth, separate execution risk from insolvency risk. Whether a trade is A-Booked does not determine whether client money is segregated, whether the legal entity is well capitalised or whether a compensation scheme applies if the broker fails.
| Question | Why it matters |
| Who is the contractual counterparty? | Determines who owes your trading P&L and who bears the legal obligation to execute. |
| Does the broker hedge every trade, aggregate exposure or use limits? | Reveals whether the model is fully hedged, internalised or hybrid. |
| Where do prices come from? | Multiple independent price sources can reduce dependence on a single liquidity provider. |
| How is slippage handled? | Symmetric positive and negative slippage is more informative than an ‘instant execution’ slogan. |
| What are the rejection and fill policies? | Execution quality can deteriorate even in an A-Book model when liquidity is thin. |
| Is client money segregated? | Execution model and client-asset protection are separate issues. |
| What regulator supervises the entity? | Best-execution and conflict rules depend on the actual contracting entity. |
What Would Prove the ‘B-Book Is Bad, A-Book Is Good’ Thesis Wrong?
It is already too simplistic. A well-regulated internalising broker can offer tight prices, instant small-order fills and robust capital because diverse customer flow offsets naturally. A poorly designed A-Book broker can offer wider effective costs, unstable liquidity and frequent rejections because it depends on weak external counterparties.
The thesis would be decisively disproved by execution data showing that an internalising broker consistently produces equal or better total outcomes after spreads, commissions, slippage and rejection rates. Conversely, a broker claiming to be A-Book would lose credibility if its execution policy shows it retains substantial exposure or if its fills systematically fail to track the external market.
The real dividing line is therefore not whether the broker ever internalises risk. It is whether the broker manages that risk transparently, remains financially resilient and gives the customer a fair execution outcome despite the conflict created by principal dealing.
Bottom Line
A-Book means the broker offsets client market risk externally. B-Book means it retains that exposure. Hybrid means it nets, internalises and hedges dynamically. Those labels describe the broker’s risk book, not a guaranteed level of honesty or execution quality.
The most important misconception is that A-Book turns a retail OTC trade into a direct bank-market order. It often does not. The broker can remain your sole contractual counterparty and execute a separate hedge with its own liquidity provider.
The second misconception is that B-Book automatically means manipulation. It does create a direct economic conflict when client losses become broker gains, which is why regulators focus heavily on fair pricing, best execution and conflicts management. But internalisation itself is a normal part of FX market structure; the BIS says major dealers internally match more than 80% of customer trades.
For retail traders, the useful question is not ‘A-Book or B-Book?’ in isolation. It is: who is my counterparty, how is the price formed, how is my order filled, what happens when liquidity disappears, how much risk does the broker retain, and what protections apply if the broker itself fails?
Methodology
Research is current through October 8, 2026 and prioritises regulator guidance, broker order-execution policies, public-company filings and BIS market-structure research. The article uses A-Book, B-Book and hybrid as industry shorthand while mapping them to ESMA’s formal fully hedged, unhedged and partially hedged counterparty models.
The one-lot EUR/USD P&L illustration assumes a $10 pip value, a 1.0-pip round-trip customer spread and 0.4-pip external hedge cost. The internalisation illustration assumes 100 customer lots long and 80 lots short, leaving 20 net lots. Gross client flow is therefore 180 lots and the residual hedge is 11.1% of gross flow; the implied external hedge-cost comparison is 180 × 0.4 × $10 = $720 versus 20 × 0.4 × $10 = $80.
The Plus500 bridge uses the company’s Q1 2026 disclosures: Customer Income of $270.6 million, Customer Trading Performance of negative $38.9 million, trading income of $231.7 million, interest income of $10.4 million and total revenue of $242.1 million. It is used to illustrate sensitivity of broker revenue to customer P&L, not to classify every Plus500 trade as unhedged.
Sources
1. ESMA — Q&A on CFDs and Other Speculative Products — Link. Primary regulatory description of fully hedged, unhedged and hybrid CFD counterparty models and the associated conflicts.
2. FCA Handbook — COBS 11.2A Best Execution — Link. Current UK rules treating principal dealing with clients as execution and requiring fair pricing/best execution.
3. FCA — Contract for Differences — Link. Current supervisory focus on conflicts inherent in retail CFD business models.
4. FCA — CFD Price and Value Multi-Firm Review — Link. 2025 review of CFD pricing, charges and matched client positions.
5. FCA — Conduct Risk in FX Markets — Link. Conduct-risk discussion covering mark-ups, principal dealing and execution controls.
6. BIS — FX Trade Execution Landscape, 2025 Triennial Survey — Link. Current evidence that dealers internally match more than 80% of customer FX trades.
7. BIS — Global FX Markets When Hedging Takes Centre Stage — Link. 2025 market-structure analysis of internalisation and dealer risk management.
8. BIS — 2025 Triennial Central Bank Survey — Link. Final 2026 release of the 2025 global FX survey.
9. BIS — OTC FX Turnover in April 2025 — Link. Global OTC FX turnover of approximately $9.6 trillion per day in April 2025.
10. IG Group — 2025 Annual Report — Link. Primary description of IG’s internalisation, client-flow offsetting and residual external hedging model.
11. IG — Best Execution — Link. Current execution disclosures, pricing sources and order-quality information.
12. IG — Order Execution Policy — Link. Current UK order-execution framework and execution-venue disclosures.
13. IC Markets Global — Order Execution Policy v1.4 — Link. Primary disclosure that IC acts as principal/sole counterparty while using third-party liquidity providers and multiple price feeds.
14. Plus500 — FY2025 Annual Report — Link. Risk-management framework, customer trading performance and hedging disclosures.
15. Plus500 — Q1 2026 Trading Update — Link. Q1 2026 Customer Income, Customer Trading Performance and revenue figures used in the analytical bridge.
16. Plus500 — Investor Relations — Link. Current company reporting archive and business overview.
17. FCA — Best Execution Review — Link. Regulatory findings on execution monitoring, connected parties and conflicts.
18. ESMA — MiFID CFD Q&A via FCA Handbook — Link. Detailed guidance on connected execution venues, hedging arrangements and conflicts.
Johan Shamshad is a financial markets writer at Dave Finances covering cryptocurrencies, trading platforms, brokers, fintech, financial regulation, and developments across global markets. He previously worked at Gulf News, adding newsroom experience to his coverage of fast-moving financial and digital-asset markets.
His work focuses on identifying market-moving events, company developments, regulatory changes, product launches, and shifts in trading and financial infrastructure.
Johan contributes news and analysis designed to help readers understand not only what happened, but why a development matters and how it may affect the wider financial landscape.