Two traders can click “buy” on EUR/USD at the same second and get different prices, different fills and very different outcomes over a year. The reason is rarely the market. It is usually the broker’s execution model: what the firm does with your order after you send it.
The debate is usually framed as A-book versus B-book, with one side painted as honest and the other as predatory. The reality is more nuanced. Both models are legal and widely used, both have real strengths, and each creates different incentives that show up in your spreads, your slippage and the way your orders are handled. This guide explains how retail forex orders are filled, how each model affects trading costs, and how to test what your own broker actually does.
What an execution model actually is
Forex is decentralized. There is no single exchange where all orders meet. Prices are formed across banks, non-bank market makers and electronic trading venues, and a retail broker sits between you and that wider market. An execution model is the broker’s policy for one question: when a client order arrives, does the firm pass it to the market, or take the other side itself?
That single choice defines the two basic categories:
- A-book: the broker routes your order to an external liquidity provider and earns a commission, a markup on the spread, or both. Your trade is hedged in the real market.
- B-book: the broker keeps your order on its own books and acts as your counterparty. If you win, the broker pays you. If you lose, the broker keeps the loss.
The labels are industry slang rather than legal terms, and few brokers fit neatly into one. But the distinction explains most of what traders experience as “broker behavior.”
The B-book: market makers and dealing desks
A B-book broker, often called a market maker or dealing desk (DD) broker, internalizes client orders. It quotes its own bid and ask prices, usually derived from a feed of external prices, and fills your trade from its own inventory. You never trade with the interbank market directly. You trade with the broker.
The model has genuine advantages:
- Fixed or tight spreads. Because the broker controls the quote, it can offer fixed spreads or very stable ones, which suits beginners who dislike variable costs.
- Fast, reliable fills on small orders. Internal matching avoids routing delays, so small market orders often fill instantly.
- Low minimums and micro lots. Taking tiny positions in-house is cheap, which is why many brokers offering $10 accounts and 0.01-lot trades run this model.
- Commission-free pricing. The broker’s revenue comes from the spread and from trading flow, so there is often no separate commission.
The central drawback is a structural conflict of interest. In a pure B-book arrangement the firm profits when clients lose. Most retail traders do lose money over time, and regulators in several jurisdictions require brokers to publish the share of client accounts that lose, which commonly sits in the 70 to 80 percent range. That statistical edge is what makes the B-book profitable, and it creates an incentive to widen spreads at inconvenient moments, to be slow on stop-outs that benefit the client, or to restrict profitable accounts. Many B-book brokers behave honestly, but the incentive exists, and understanding it is the foundation of evaluating any broker.
The A-book, part one: STP brokers
Straight-through processing (STP) brokers send client orders to one or more liquidity providers, such as banks or non-bank market makers, without a dealing desk intervening. The broker aggregates quotes from its providers, adds a markup to the spread or charges a commission, and passes the order along.
The broker’s incentives are much cleaner here. Its revenue depends on trading volume, not on client losses, so it wants you to keep trading for years. That generally aligns its interests with yours, although two caveats matter:
- The prices are still the broker’s. STP brokers typically stream a blended or marked-up price, and the markup is not always visible. Two STP brokers can show very different effective costs on the same pair.
- Execution depends on the providers. If the liquidity provider rejects the order or the price moves before it arrives, you get slippage or a rejection. The broker cannot guarantee a fill it does not control.
STP is sometimes used loosely as a marketing term. A broker can describe itself as STP while still internalizing part of its flow, which is why the label alone proves little.
The A-book, part two: true ECN
An electronic communication network (ECN) account connects you to a pool of competing quotes from banks, institutions and sometimes other traders. Your order is matched against the best available bid or offer in that pool, and the broker charges a transparent commission per lot.
The benefits are visible in the data. Spreads on major pairs can be very thin, sometimes close to zero during busy sessions, and you can see market depth when the platform supports it. Pricing is variable, so spreads widen when liquidity thins out, around news releases and at the daily rollover.
The trade-offs are also real. Commissions add up on high-volume strategies, minimum deposits are usually higher, and large orders can move through several price levels and suffer slippage. ECN is not a promise of better results. It is a promise that the broker is paid by commission and does not trade against you.
Hybrid models: where most brokers actually sit
Most large retail brokers run a hybrid model, sometimes called C-book or risk-managed execution. They use software that scores each client by trading behavior, size and profitability. Orders from clients who are statistically likely to lose are kept in-house, and orders from clients who look consistently profitable are hedged with liquidity providers.
This is economically rational. A broker that hedges everything pays out to providers whenever a client wins, while a broker that internalizes everything carries risk from large, informed traders. Hybrid routing manages both. It also means that your execution model may be decided by your own results. A beginner and a skilled scalper at the same broker can receive different experiences, including different spread behavior, slippage patterns or order limits, without either one knowing.
The models side by side
| B-book (market maker) | A-book STP | A-book ECN | |
|---|---|---|---|
| Counterparty to your trade | The broker | Liquidity provider | Liquidity pool |
| How the broker earns | Spread, client losses | Spread markup | Commission |
| Typical spreads | Fixed or stable, often wider | Variable, markup added | Variable, often tightest |
| Slippage pattern | Controlled by the broker’s quotes | Depends on provider depth | Depends on market depth |
| Main conflict of interest | High | Low | Lowest |
| Typical minimum deposit | Lowest | Low to medium | Medium to high |
| Best suited to | Beginners, small accounts | Most retail traders | Scalpers, high-volume traders |
How execution models shape your spreads
The spread is the gap between the bid and the ask, and it is the most visible trading cost. But the headline spread on a broker’s website is only part of the story, because the total cost per round trip also includes commission and the quality of your fill.
On a B-book account, the spread is a price the broker sets. A 1.2 pip spread on EUR/USD with no commission may look comparable to a 0.1 pip spread plus a commission on an ECN account, and for a casual trader holding positions for days it often is. For a scalper opening dozens of trades a day, the arithmetic changes quickly, because each fraction of a pip repeats on every trade.
On A-book accounts, the spread floats with market liquidity. It is tightest when London and New York overlap and widens sharply when the market goes quiet or news hits. A fixed-spread broker advertises certainty, but that certainty is usually paid for through a wider average spread, and fixed spreads can still widen or be suspended during extreme events.
Two practical rules help:
- Compare all-in cost, not the quoted spread. Add the commission, converted to pips, to the spread, and compare at the times of day you actually trade.
- Look at spreads under stress. Quiet-market spreads tell you little. The first minute after a major data release shows what a broker’s liquidity is really like.
Broker slippage explained
Slippage is the difference between the price you expected when you clicked and the price at which your order actually filled. It is not a fee. It is a consequence of the fact that prices move while your order travels, gets checked and is matched.
Negative slippage fills you at a worse price than requested. Positive slippage fills you at a better one. Both are normal in a fast market, and a broker with an honest, symmetrical policy will show both in roughly comparable amounts over time.
Why slippage happens
- Volatility. During news, prices can jump several pips between ticks, and there may be no liquidity at the price you saw.
- Order size. A large order consumes the best quote and fills the rest at the next levels, a process called walking the book.
- Latency. The delay between your device, the broker’s server and the liquidity source gives the market time to move. A trader in a distant region on a congested connection sees more slippage than one on a low-latency server.
- Order type. Stop and market orders fill at the next available price, which can be worse than the trigger. Limit orders guarantee price but not execution.
How the model changes the pattern
On a B-book, the broker decides how to fill you. A fair one fills at the quoted price or gives symmetrical slippage. A less scrupulous one can apply slippage mostly against the client, or hold orders for a moment (last look) before accepting. The key tell is asymmetry: if negative slippage is common and positive slippage almost never appears, the model is not simply passing through market movement.
On an A-book, slippage mostly reflects real market conditions and provider depth. It can still be large in thin markets, but it is more likely to be symmetrical, and a broker with no incentive to profit from it has little reason to skew it. The honest summary is that A-book slippage is more likely to be a market effect, while B-book slippage can be either a market effect or a policy.
Requotes, last look and rejected orders
Slippage is not the only way execution quality shows up. Three other behaviors are worth knowing:
- Requotes. The broker declines your price and offers a new one, which you must accept or cancel. Requotes are a hallmark of dealing-desk execution and are far rarer on modern STP and ECN setups.
- Last look. Some liquidity providers keep a short window to accept or reject a trade after receiving it. It protects them from stale quotes, but when it is used asymmetrically, rejecting trades that would have been profitable for the client, it works against you. The FX Global Code, the industry’s voluntary code of conduct, asks market participants to be transparent about last look and to use it only for legitimate risk control.
- Delayed or rejected fills. A short, random delay on profitable trades and quick execution on losing ones is a pattern that traders report on some platforms. It is hard to prove from a single trade, but it shows up over a large sample.
None of these automatically means bad faith. Fast markets produce genuine rejections. What matters is the pattern over hundreds of trades, which is why record-keeping beats anecdote.
Conflict of interest and what regulators require
The B-book conflict is not a secret. It is openly acknowledged in regulation, and good brokers manage it rather than deny it.
In the European Union and the United Kingdom, investment firms must follow best execution rules: take all sufficient steps to obtain the best possible result for clients, considering price, costs, speed and likelihood of execution, and publish an execution policy that explains how orders are handled. Firms offering leveraged products must also display the percentage of retail client accounts that lose money, which is why that statistic is easy to find on regulated brokers’ websites. In the United States, forex dealers are registered with the CFTC and are members of the NFA, which imposes capital, reporting and conduct requirements.
Regulation helps, but it does not remove the incentive, and its strength varies sharply by jurisdiction. A broker licensed in a major regulatory regime faces audits, capital requirements and complaint channels. A broker registered in a lightly supervised offshore location may disclose very little about how it executes. When comparing brokers, treat the regulator as part of the execution model: it determines what the broker must tell you and what you can do if something goes wrong.
Common myths worth dropping
- “A-book is always better.” A-book removes the main conflict, but it does not guarantee better prices. A badly connected STP broker with a hidden markup can cost more than a transparent market maker.
- “B-book brokers are scams.” Many operate legitimately under strong regulation. The model has a conflict of interest, not a built-in fraud.
- “No dealing desk means no slippage.” No broker can eliminate slippage, because it comes from the market itself.
- “Zero spread means zero cost.” Zero-spread accounts charge commission or compensate through other fees, and the all-in cost is what counts.
- “My broker’s label tells me how it executes.” Marketing terms like STP, ECN and “no dealing desk” have no single enforced definition. Evidence beats labels.
How to evaluate your own broker
You cannot see a broker’s order routing, but you can measure its results. A short, disciplined review will tell you more than any brochure.
- Read the execution policy. It should say whether orders are internalized, whether the firm may act as principal, and how it handles slippage and last look. Vague language is itself informative.
- Check the regulator and the disclosures. Confirm the license on the regulator’s own register, and look for the published client loss percentage and any order execution statistics.
- Log your fills. For a month, record the requested price, the filled price, the time and the market conditions for each trade. Calculate the average positive and negative slippage. A healthy pattern is roughly symmetrical.
- Compare on a demo and a small live account. Demo accounts often execute better than live ones. Test with real but small money on the same instrument at the same time of day.
- Cross-check against another broker. Open a small account elsewhere and compare prices and fills during a news release. Consistent differences are meaningful.
- Test the exits. Slippage on stops and withdrawals both matter. Request a small withdrawal early to see how the firm behaves when you want your money back.
- Watch for behavior changes after success. If spreads widen, fills worsen or limits appear after a profitable stretch, document it and ask the broker in writing.
Which model suits which trader
There is no universal winner, only a better match for your style:
- New traders with small accounts often do well on a regulated B-book or hybrid broker. Low minimums, predictable pricing and simple platforms matter more than saving a fraction of a pip, as long as the broker is well regulated and transparent.
- Swing and position traders are less sensitive to spreads and slippage, because costs are spread over larger moves. The broker’s regulation, withdrawal record and overnight financing costs usually matter more.
- Scalpers and high-frequency strategies depend on tight, consistent spreads and fast, honest fills. An ECN or well-connected STP account with low latency is typically the better fit, and a hybrid broker that profiles winners is a risk to watch.
- News traders should prioritize execution under stress: depth of liquidity, rejection rates and symmetrical slippage, rather than the quiet-hours spread.
Disclaimer: this article is educational and is not financial advice. Leveraged forex and CFD trading carries a high risk of loss, and most retail accounts lose money.
Conclusion: judge the incentives, then the evidence
A-book and B-book are not good and evil. They are two ways of earning money from retail order flow, and each creates a different set of incentives. A-book and ECN brokers are paid for volume and commission, which aligns them with traders who keep trading. B-book brokers are paid in part by client losses, which creates a conflict that regulation and reputation must keep in check. Hybrid brokers do both, which makes their behavior the hardest to read from the outside.
For your spreads, remember that the quoted number is only one part of the cost, and that all-in cost at the times you trade is the figure that counts. For your slippage, remember that it is a feature of fast markets, so the question is not whether it exists but whether it is symmetrical and explainable. Read the execution policy, verify the license, log your own fills and compare. The best protection against a poor execution model is not a label. It is your own data.
