A quoted price is a summary of something more detailed. The underlying structure is an order book, and its mechanics explain several things that look arbitrary from outside.
The structure
An order book holds all resting limit orders for a security, sorted by price.
On one side, buyers with the prices they are willing to pay. On the other, sellers with the prices they will accept.
The highest buy price and the lowest sell price form the spread, and the quoted price you see is derived from these.
Below the best prices sit further orders at worse prices, forming the depth of the book.
Price-time priority
The standard matching rule on most venues.
Better prices execute first. Among orders at the same price, the one entered earliest executes first.
Which means position in the queue has value, and it is why participants care about latency — arriving earlier at the same price means executing sooner.
Some venues use different rules, including pro-rata allocation where orders at a price share fills proportionally, which changes the incentives considerably.
What happens when a market order arrives
It consumes resting orders starting from the best price and working outward until filled.
A small order fills at the best price. A large one walks the book, filling at progressively worse prices.
The average price achieved is therefore worse than the quoted price, and the difference is market impact.
Which is why quoted spreads understate the true cost of trading in size, and why institutional execution is a specialist discipline.
Hidden and iceberg orders
Participants wishing to trade size without revealing it can use orders that display only part of their quantity.
As the displayed portion fills, more is revealed from the hidden reserve.
Which reduces information leakage and generally costs queue priority, since hidden quantity typically ranks behind displayed quantity at the same price.
That trade-off — visibility against priority — is one of the genuine design tensions in market structure.
Order types proliferate for a reason
Beyond market and limit orders, venues offer a large number of conditional types.
Stop orders that activate when a price is reached. Orders that execute only at the closing auction. Orders that cancel if not immediately filled. Orders that will not remove liquidity from the book.
Each exists because some participant wanted specific behaviour, and the proliferation has attracted criticism on the grounds that complexity favours sophisticated participants.
The counterargument is that each type addresses a genuine need and that simple types remain available.
The stop order problem
Worth a specific warning because it catches retail participants regularly.
A stop order becomes a market order when triggered, which means it executes at whatever price is available.
During a rapid move with thin depth, that can be far from the trigger price.
Stop-limit orders avoid this by becoming a limit order instead, at the cost of possibly not executing at all.
Neither is strictly better. The choice depends on whether the greater risk is a bad price or no execution.
Auctions
Continuous trading is interrupted by auctions at the open and close, and sometimes after volatility.
In an auction, orders accumulate without executing, then a single price is calculated that maximises the volume tradeable, and everything executes at that price.
Which concentrates liquidity at a single moment and produces a more robust price than continuous trading in thin conditions.
Closing auctions have grown substantially in importance, since index funds and many benchmarks reference the closing price, and a large share of daily volume now occurs in them.
Why a retail participant might care
Mostly for two practical conclusions.
Use limit orders where the security is thinly traded, because the book may be thin and a market order will walk it.
And avoid trading in the first and last minutes of the session, when the book is reassembling or concentrating and prices are least representative.
Neither requires any deeper engagement with market structure, and both follow directly from how the book works.
Latency and colocation
Where queue position turns into an infrastructure question.
Because priority depends on arrival time, the physical distance between a participant's systems and the matching engine matters at microsecond scale.
Venues sell space in their own data centres, which equalises access for those who pay and creates an advantage over those who do not.
Whether this is a problem depends on your view of what the resulting speed competition contributes, and the argument has run for years without resolution.
Some venues have introduced deliberate delays applied uniformly, which removes the advantage of raw speed while preserving price competition.
Tick sizes
The minimum price increment, which is set by rule rather than by the market.
A larger tick widens the minimum spread, which benefits liquidity providers and costs takers.
A smaller tick allows finer pricing and encourages queue-jumping by trivial amounts, which discourages resting orders.
Regulators have adjusted tick sizes in both directions in different markets, and the evidence suggests the optimal size varies with the security's price and liquidity rather than being uniform.