Market microstructure: ticks, queues and priority

There is the price, and then there is your price. Microstructure is the study of the gap between them — the grid that prices snap to, the queues your orders wait in, and the uncomfortable statistics of who you actually trade against. Most traders ignore it until it quietly eats a strategy that looked profitable on paper. The edge in many short-horizon strategies is smaller than the microstructure costs of running them, which is a polite way of saying the strategy was never real.

None of this requires advanced math. It requires taking seriously three ideas: ticks define where you may stand, queues define when you get served, and adverse selection defines what you get served.

Ticks: the grid prices live on

Prices are not continuous. Every instrument has a tick size — the minimum increment between valid prices. A stock might tick in cents; a major FX pair in fractions of a pip; a small-cap crypto pair in much coarser steps relative to its price. Tick size shapes everything downstream.

• When the tick is large relative to volatility, the spread is often pinned at one tick. Price improvement is impossible — you cannot quote inside the spread — so competition moves entirely into the time dimension. Queues get long, and queue position becomes the whole game.
• When the tick is small relative to volatility, the spread floats across many ticks. Competition happens on price: makers leapfrog each other by a tick to gain priority, queues stay short, and the top of book flickers constantly.
• The same instrument can shift between regimes. A quiet market with a relatively coarse tick behaves like the first case; turn up volatility and it behaves like the second.

This is why an instrument’s “personality” is partly an artifact of its tick. Change the tick size and you change the spread, the queue lengths, and the behavior of everyone quoting it — a fact regulators and exchanges have demonstrated repeatedly by changing tick sizes and watching markets reorganize around them.

Queue position is worth real money

Suppose the best bid is 50.00 and there are 1,000 units queued there. You join with 10 units — you are at the back. For you to fill, sellers must hit the bid for 1,010 units before anyone above you cancels. Now consider the trader at the front of that queue. They fill on the very first sell order. Same price, same instrument, same moment — radically different outcomes.

Here is the asymmetry that makes front-of-queue valuable: if price is about to tick up, the front of the queue gets filled and immediately shows a paper profit. If price is about to tick down, everyone in the queue gets filled — including the back — and everyone shows a paper loss. The front of the queue participates in the good fills and the bad. The back participates almost exclusively in the bad. Being deep in a long queue means you mostly fill when you would rather not have.

Adverse selection: the trade you win and wish you hadn’t

Every passive fill is a small auction you just won, and the winner’s curse applies. The counterparty who crossed the spread to trade with you chose this moment, this price, this direction. Sometimes they are uninformed — rebalancing, taking profit, closing a position for reasons unrelated to the next ten minutes. Sometimes they know something, or their model does. You cannot tell which from a single trade. But in aggregate, the flow that hits resting orders right before the price moves is informed flow, and resting orders systematically end up on the wrong side of it.

A concrete way to see it: track the mid-price 30 seconds after each of your passive fills. If you bought and the mid is reliably lower 30 seconds later, your fills are adversely selected — you earned the spread and lost more than the spread in immediate drift. Plenty of “market making” strategies by retail traders are precisely this: collecting pennies of spread while donating nickels of adverse selection.

Why the same strategy fills differently across venues

Take one strategy — say, join the best bid, take profit one tick higher — and run it on two venues listing the same asset. Results can diverge wildly, for structural reasons:

• Tick regime. A venue with a finer tick has shorter queues and more price competition; your passive order fills faster but earns a thinner spread.
• Queue composition. On a venue dominated by professional makers, you are queuing behind firms that cancel within milliseconds of bad news, leaving you holding the fill. On a venue with slower participants, the queue ahead of you is stickier — and so is your protection.
• Fee structure. Maker-taker fees change behavior. Where makers are paid or charged little, books are thick and queues long. Obsidiate’s tiered schedule — from 0.15% maker at Bronze down to 0.05% at Diamond — means two traders running identical strategies can have different break-even spreads purely from their fee tier.
• Flow mix. A venue whose takers are mostly small, uninformed orders is kind to passive strategies. A venue whose takers are fast arbitrageurs is brutal to them. The strategy didn’t change; the opponents did.

This is why backtests calibrated on one venue’s data routinely fail when deployed elsewhere, and why serious practitioners measure fill quality per venue, per instrument, per time of day — not in aggregate.

Reading the book with microstructure eyes

Once you internalize this, the order book panel stops being decoration. A thick best bid that never fills is a queue you don’t want to join at the back. A spread that suddenly widens is makers stepping away — information may be arriving. A burst of trades chewing through one side while the other refills calmly tells you which side is urgent and which is patient. None of this is a crystal ball; all of it improves the small decisions — where to place, when to cross, when to wait — that compound into the difference between a strategy’s paper results and its real ones.

Key takeaways

• Tick size sets the terms of competition: coarse ticks push competition into queues and time; fine ticks push it into price.
• Queue position has real value — the front of the queue gets the good fills and the bad, while the back mostly gets the bad.
• Adverse selection means your passive fills are biased toward moments when the market is moving against you; measure post-fill drift to see it.
• The same strategy fills differently across venues because tick regimes, fee tiers, queue composition, and flow mix all differ.
• Fee tiers change break-even math directly — a strategy viable at 0.05% maker fees may be dead at 0.15%.
• Fill quality is a measurable statistic, not a vibe. Track it per venue and per instrument before trusting any short-horizon edge.