Both books at once
How the routing decision is really made, ticket by ticket
Read
The A-book versus B-book argument assumes a firm picks one. None of them do. The decision is made per client, per trade, by software — and knowing how it is made tells you more than the label ever could.
AI Mentor here. 🧠 Two lessons, two roads, and a question you are probably holding: so which one is my broker?
The honest answer is both, continuously. Almost every retail firm of any size runs a hybrid, and the interesting part is not the mixture — it is the rule that sorts you into one side of it.
The decision is not about you personally
Nobody in the dealing room looks at your name. The routing is a rule engine, and it runs on every ticket in milliseconds.
What it is really doing is managing the firm's risk, not judging you. It asks: does this order increase or reduce our current net exposure? Is this client's flow historically costly to carry? Is the market about to become expensive to hedge in?
An order that offsets the house position is free money to internalise. An order that piles onto an exposure the firm already dislikes gets pushed out to a liquidity provider. The same client can have one trade internalised and the next one hedged, ten seconds apart.
The mental picture of a trader deciding to "take on" a customer is decades out of date. It is a classification system tuned to keep the firm's net exposure inside limits set by its risk committee.
What actually moves you between books
Firms do not publish their rules. The inputs, though, are not mysterious — they follow from what the risk desk is trying to avoid.
• Profitability over time. A client whose flow has cost the firm money is a client it would rather hedge than carry. Consistent winners tend to get routed out.
• Size. Larger tickets move the net exposure more, so they cross the hedging threshold sooner.
• Style. Very short holding times, trading straight into news releases, or latency-sensitive strategies are all expensive to carry and get treated accordingly.
• What the rest of the book is doing. The same order is internalised when it offsets other clients and hedged when it reinforces them.
If you start winning consistently, you are more likely to be moved to the A-book. That is usually framed as sinister. It is worth noticing it is also the outcome you would want: on the A-book, nothing about your result touches the broker's revenue.
Why you might notice
You are never told which book you are in. You can sometimes feel it.
Internalised fills are typically instant and exactly at the quote, because no external trade needs to happen — the firm simply writes the ticket. Hedged fills involve a real round trip to a liquidity provider, so they can take a few milliseconds longer and can land a fraction away from the shown price in either direction.
So the classic report — *"my fills changed after I had a good month"* — is not paranoia, and it is not proof of anything sinister either. It is what moving from one route to the other feels like.
What this does to the question you should ask
It retires "is this an A-book broker?" as a useful question. Every honest answer is "partly, sometimes, depending".
What replaces it is narrower and answerable: what has the firm committed to in writing, and does what I observe match it? An order execution policy is a published document. It says how the firm decides where orders go, what it treats as best execution, and how it handles slippage. It is dull, it is specific, and it is the only thing you can hold a broker to.
That is where the next two lessons go: the price on your screen, and the fill you actually get.
See you on the desk. — AI Mentor