Dealing desk or straight-through: who is on the other side of your trade
A dealing desk broker takes the other side of your trade itself: when you buy, it sells to you from its own book, so your loss is its gain. A straight-through broker passes your order to an outside liquidity provider and earns from a markup or commission instead, so it does not hold the opposite position. The first model creates a conflict of interest; the second removes that conflict but hands you the outside provider's pricing, which can be worse in thin conditions.
A dealing desk broker fills your order from its own book and ends up holding the opposite position, so your gain is its loss and your loss is its gain. A straight-through broker does not fill you itself: it passes your order out to an external liquidity provider, keeps a markup or a commission, and holds no position in the trade at all.
That single difference, who ends up carrying the position, explains almost everything else people argue about regarding the two models.
Why a broker would want to be your counterparty
Internalising orders is not a trick. It solves a real problem.
Most retail orders are small and most of them cancel each other out. At any moment some clients are long and some are short the same instrument. If a broker sent every one of those orders to an outside provider, it would pay a cost on each leg to have an outsider do something the broker could have done by simply matching the two clients against each other. Internalising lets the broker fill both sides instantly, charge a spread, and carry no net position.
That is the core of how a dealing desk actually works, and it is the part most explanations skip. A dealing desk does not usually sit there hoping you lose. It nets opposing client flow and holds only the leftover imbalance. Suppose, as a made-up illustration, that clients are long 600 units of some instrument and short 450 units of the same thing. The desk internally matches 450 against 450 and is left carrying 150 units of risk. Those 150 units are the only part of the book where the broker’s outcome and the client’s outcome genuinely oppose each other.
The desk then chooses what to do with that 150. It can hedge the imbalance with an outside provider, in which case it has converted itself into something close to a pass-through operation for the net amount. Or it can keep the exposure, in which case it profits if that leftover client flow loses money. This decision, made continuously and invisibly, is where the structural conflict lives.
Straight-through routing exists because some brokers do not want that decision on their books at all. By passing every order out, the broker’s revenue becomes a function of volume rather than of client outcomes. More trading means more revenue whether clients win or lose. The conflict of interest disappears, and the broker’s business stops depending on its own risk judgement.
What each model costs you in practice
The conflict in the dealing desk model is obvious, so people assume the other model is free. It is not. It just charges you somewhere less visible.
A pass-through broker buys its pricing from outside. It has to add something on top, either as a wider spread than it received or as an explicit commission. It also inherits whatever the outside provider’s pricing looks like at that moment, including the moments when the provider’s pricing is bad.
Work through an invented case. Imagine a pass-through broker receives a quote from its provider with a spread of 0.2 units and adds a markup of 0.1 units, so you see 0.3 units. Say you trade a position where each unit of spread costs you $4. Your cost on entry is 0.3 × $4 = $1.20. Now imagine a dealing desk quoting a fixed 0.4 units on the same instrument, so your cost is 0.4 × $4 = $1.60. On a calm day the pass-through route is cheaper by $0.40.
Now imagine a sharp move, and the provider’s spread widens from 0.2 to 1.5 units. With the same 0.1 markup you now see 1.6 units, and your cost on entry is 1.6 × $4 = $6.40. The dealing desk, setting its own fixed 0.4, still charges you $1.60. Every number there is invented for the arithmetic, but the shape of the result is the real point: the pass-through model gives you the outside market’s pricing, good and bad, while the dealing desk model gives you the broker’s pricing, which is steadier because the broker is absorbing the variation. The broker charges you for absorbing it on the average day.
The same trade-off shows up in order rejection. A pass-through broker can only fill you at a price an outside provider will honour, so when liquidity thins your order can be requoted or rejected. A dealing desk can fill you at its quoted price because it is the one honouring it. Whether that is a service or a liability depends entirely on whether the desk is quoting fairly.
| Dealing desk | Straight-through | Hybrid | |
|---|---|---|---|
| Who holds the position | The broker | An outside provider | Broker holds the net, routes the rest |
| Main revenue source | Spread, plus any gain on net exposure | Markup or commission on volume | Both |
| Spread behaviour | Fixed or near-fixed, set by the broker | Variable, follows outside liquidity | Varies by instrument and conditions |
| Conflict of interest | Direct, on the net imbalance | Volume incentive only | Direct, but smaller |
| Fills under stress | Broker can honour its own quote | Depends on provider willingness | Mixed |
Most firms are in the third column. A broker that calls itself one pure thing is usually describing its routing for some instruments rather than its whole operation.
The mistake people make reading this
The common error is treating the model as a verdict on honesty. It is not a verdict on anything. It tells you where the risk sits and what the broker’s incentives point at, and that is all.
The specific confusion runs in both directions. People assume a dealing desk must be manipulating prices against them, when a desk with deep internal flow and no appetite for net exposure can be a perfectly ordinary place to trade. And people assume pass-through routing is a safety guarantee, when it says nothing at all about whether your money is protected. Routing is about where your order goes. Whether your balance survives a bad week at the firm is about whether client funds are held apart from the firm’s own money and whether the firm holds enough capital to absorb losses. Those are separate questions with separate answers, and the second one matters more if the firm fails. See what segregated client money means for how that part works.
The second mistake is reading marketing language as a disclosure. The labels used to describe routing are not standardised and not defined by anyone in particular, so a firm can describe its model in whatever terms sound best. What is harder to dress up is behaviour. Watch whether the spread holds steady through a volatile period or widens with it. Watch whether commission appears as a separate line or is folded into the quote. Watch whether orders get requoted when the market moves fast. Those observations tell you about the routing more reliably than the label does.
The third mistake is caring about this at the wrong moment. For someone holding positions for weeks, the difference between the two models is small relative to the cost of being wrong about direction. For someone placing many trades a day, it dominates everything else. Before you spend time on this, work out whether your own trading is sensitive to it. The pre-trade checklist covers the execution details worth confirming first, and the glossary defines the routing terms you will run into.
How to reason about it next time
Ask what the broker earns when you lose. In a pure pass-through operation the answer is nothing beyond the commission it already took. In a dealing desk holding net exposure, the answer is the other side of your position. Then ask what that answer is worth against the thing you are actually exposed to, which is whether the firm holds your money separately and can survive a bad day. A firm that is honest about its model and weak on capital is a worse place to sit than a dealing desk with the opposite profile.
The routing model is one layer of the foundation. It is not the one holding the most weight.
Education only, not financial advice. Trading carries a risk of loss.
Key facts
- A dealing desk broker is the counterparty to your trade and holds the opposite position on its own book.
- A straight-through broker routes your order to an external liquidity provider and does not take the opposite position.
- Dealing desk revenue can come from client losses; straight-through revenue comes from spread markup or per-trade commission.
- Most dealing desks net opposing client orders against each other and only hold the leftover imbalance.
- A straight-through broker can still fail you through poor provider selection, wide markups or weak capital, none of which are conflicts of interest.
- The model a broker uses tells you where the risk sits, not whether the broker is honest.
Frequently asked
Is a dealing desk broker betting against me?
In the narrow sense, yes: if you are long, it is short the same amount unless it has offset that exposure. In practice most dealing desks match opposing client orders first and only carry the net leftover, so the position they hold against you is usually far smaller than the sum of client trades. The conflict is structural and real, but it is a conflict over the residual, not over every trade you place.
Does straight-through routing mean better prices?
Not automatically. Straight-through means your order reaches an outside provider; it says nothing about how good that provider's pricing is or how much markup sits on top. A dealing desk with a deep internal book can show a tighter, steadier quote than a thin external feed, particularly in quiet hours.
How can I tell which model a broker uses?
The behaviour of the quote is the clearest signal. A dealing desk tends to show fixed or near-fixed spreads that hold steady through news, because the broker is setting them. A straight-through route tends to show variable spreads that widen sharply when outside liquidity thins and compress when it returns. Commission charged separately from the spread also points toward pass-through routing.
Which model is safer if the broker gets into trouble?
Neither model protects your balance by itself. What protects it is whether your money sits apart from the firm's own funds and whether the firm holds enough capital to absorb a bad day. A straight-through broker removes the market-risk conflict but can still be thinly capitalised, and a well-capitalised dealing desk can be the sturdier place to hold a balance.
Risk warning: trading is high risk and most retail accounts lose money. Education only, not financial advice.