Close a valve too quickly and the water behind it does not simply stop. The moving column still carries its momentum, and that momentum has to go somewhere. It converts into a pressure wave that races back up the pipe, reflects off the far end, and returns. Engineers call it water hammer. It is the bang you hear in old plumbing, and in industrial systems it is powerful enough to split a pipe.

Order flow behaves in a strikingly similar way, and the resemblance is useful enough to be worth formalising. What follows is a model, not an accusation. Its value is that it converts a vague worry about broker conduct into quantities you can actually measure.

The pipe, the valve and the fluid

A broker sits between client orders and the external liquidity pool. When the broker passes an order through to that pool, the valve is open and flow moves freely. This is conventionally described as A-book execution. When the broker absorbs the order onto its own balance sheet instead, the valve is shut and the flow stops at the broker. This is B-book, or internalisation.

Internalisation is not inherently wrong. A broker that nets a buyer against a seller is performing a real service and sparing both of them a spread. The problem is not the closed valve. The problem is how fast it closes.

Borrowing the Joukowsky equation

Hydraulics gives us a clean expression for the pressure spike caused by a sudden stop. The Joukowsky equation says the surge is the product of three things: the density of the fluid, the speed at which a pressure wave travels through it, and the change in flow velocity. Written compactly, the pressure change equals density times wave speed times the change in velocity.

Each of those three terms has a market counterpart.

• Density becomes the concentration of directional flow. A book where buyers and sellers roughly offset is a thin fluid. A book where almost every client is leaning the same way is a dense one.
• Wave speed becomes the rate at which a price shock propagates across venues. Faster, more interconnected markets carry the wave further and sooner.
• Change in velocity becomes the abruptness of the change in net flow. Not the size of the flow, the suddenness of its arrival or withdrawal.

Multiply them together and you get the dislocation that reaches the client, which shows up in their account as slippage, a requote, or a gap through a stop.

The critical closure time

Hydraulics also tells us when a closure counts as sudden. A pipe has a characteristic period: twice its length divided by the wave speed, which is the round trip a pressure wave needs to travel to the far end and back. Shut the valve more slowly than that period and the returning wave relieves the pressure as it builds. Shut it faster and the full surge lands.

The market equivalent of pipe length is the distance between the broker book and genuinely deep liquidity. A broker with a short path to a deep pool has a short characteristic period and very little room to manoeuvre. A broker sitting several intermediaries away from real depth has a long one, which sounds safer but actually means a larger surge when the closure is abrupt.

This is why the model is worth the trouble. It predicts that the damaging cases are not the largest internalised books. They are the books where a hedging policy changes quickly, under pressure, during exactly the conditions that already have flow pointing one way.

The ram, and why it matters

There is a device called a hydraulic ram that exists purely to exploit this effect. It deliberately slams a valve shut so the resulting surge lifts a fraction of the water far above its source, with no external power at all. The shock is not a side effect. It is the engine.

That is the uncomfortable half of the analogy. A structure that converts the shock of stopped flow into lift is not a malfunction. It is a design. Whether any given broker is running a ram or simply has slow plumbing is an empirical question, and the model tells you where to look: at the asymmetry of slippage, not its average.

What hydraulics would prescribe

Plumbers solved this problem a long time ago, and every one of their fixes has a market analogue.

• The surge tank. A reservoir that accepts displaced volume during a spike. In a trading book this is a partial hedge held permanently, sized so the book never has to move from fully internalised to fully hedged in one step.
• The air chamber. A compressible cushion that absorbs small shocks continuously. The equivalent is an inventory tolerance band, a range of net exposure the book is content to carry without reacting.
• The slow-closing valve. Rather than switching hedging on or off, vary the hedge ratio continuously as a function of net exposure. The closure then takes longer than the characteristic period by construction.
• The relief valve. A hard limit that vents flow to the external market once exposure passes a threshold, regardless of what the model thinks about the flow.

Where the analogy ends

No analogy is evidence. Fluids are conserved and order flow is not; a market participant can simply stop trading, which no volume of water can do. Wave speed in a pipe is a material constant, while in markets it varies with the time of day and the state of the system. Treat the equation as a way of organising measurement, not as a law.

What it does give you is a short list of things to instrument: how concentrated your flow is, how quickly your hedge ratio changes, and how far you sit from real depth. At AsquareC we build those measurements into FinAI because the difference between a well-damped book and a ram is not visible in an average. It is visible in the tails.

This article is general market commentary from AsquareC FinTech FZCO. It is not investment advice and does not take account of your circumstances.