How order-driven auction markets match buyers and sellers, with a single-price call example that separates price formation from individual fills.
Auction market
An auction market is an order-driven market in which participants' buy and sell orders meet under a common matching rule. Prices emerge from the submitted orders. For a quant modelling execution, the rule matters as much as the price: it determines when an order can trade, which price it receives and whether it receives any shares.
“Auction” does not necessarily mean a room of bidders or one trade at the end of a session. An electronic order book can run as a continuous auction, matching compatible orders as they arrive, or as a call auction, accumulating orders before a batch match. Foucault, Pagano and Röell use this distinction in Market Liquidity, Chapter 1.
The structural comparison is with a dealer market, where customers trade against dealers' quotes. Real venues can combine order matching with designated market makers. “Auction market” therefore describes a mechanism; it does not imply that dealers are absent.
Continuous matching and a call solve different problems
In continuous trading, an incoming marketable order can execute against resting orders at several prices. Its average fill depends on the depth it consumes and the book it encounters. A limit order that cannot trade may wait in the queue.
A single-price call first collects eligible interest, then chooses a common execution price under the auction's rules. This concentrates buying and selling at an event such as an opening auction or closing auction. It exchanges immediacy for an opportunity to match a larger pool together; it does not guarantee a particular price or a full fill.
To develop the continuous side of that comparison, Order Books works through depth, average execution prices and queue priority. The lesson has a public preview; the full lesson requires membership.
A price that lets both sides trade
Consider a synthetic single-price call in one stock. Prices are USD per share, quantities are shares, and the allowed price tick is USD 1. There are only the following limit orders: no market orders, hidden interest, fees, cancellations or incoming orders during the calculation.
Side
Limit price
Quantity
Buy
102
50
Buy
101
60
Sell
100
80
Sell
101
40
A buy limit is a maximum acceptable price; a sell limit is a minimum. At candidate price p, count the shares willing to buy and willing to sell:
D(p)=i:bi≥p∑qi,S(p)=j:aj≤p∑qj.
Here bi and aj are buy and sell limits, with quantities qi and qj. The greatest feasible matched quantity at that price is
V(p)=min{D(p),S(p)}.
Every matched share needs one willing buyer and one willing seller. Adding the two sides would count interest, not trades.
Candidate price
Buy demand D(p)
Sell supply S(p)
Matched shares V(p)
100
110
80
80
101
110
120
110
102
50
120
50
Assume the call chooses a price that maximises matched volume. USD 101 is the unique choice, with 110 shares traded. At lower valid ticks no seller participates; above 102 no buyer participates. The best bid of 102 is not the auction price: raising the common price to 102 excludes the 60-share buyer whose limit is 101.
Synthetic call: buy 50 at USD 102 and 60 at 101; sell 80 at 100 and 40 at 101. At candidate prices 100, 101 and 102, matched volumes are 80, 110 and 50 shares. Price 101 maximises volume. Under better-price-first allocation, sellers fill 80 plus 30, leaving 10 sell shares.
Original synthetic single-price call. Prices are USD per share; quantities are shares. All executions use the chosen price. The allocation panel assumes better-priced limits fill before limits at the auction price.
Choosing the price and allocating the shares are separate steps. If better-priced limits have priority in this example, both buy orders fill, the 80-share seller with limit 100 fills, and 30 of the 40 sell shares at limit 101 fill. Ten sell shares remain unfilled. Every executed share trades at 101, including those offered at 100. A limit specifies an acceptable boundary, rather than a requirement to execute at that exact limit.
There is no need to decide between equal-priced sellers here because only one sell order sits at 101. With several such orders, the venue's allocation rule would determine their individual fills. The volume calculation alone cannot answer that question.
What the simple calculation leaves out
The ASX's auction explanation starts with maximum executable volume, then applies further conditional rules involving surplus, market pressure and a reference price when necessary. Our example stops at the first criterion because its maximum is unique. Other books can have several equally good prices, including ticks at which no order was submitted. A production model must evaluate the venue's eligible prices and apply its actual rules.
Order entry, cancellation cutoffs and price protections also depend on the market and auction. NYSE's auction documentation, for example, distinguishes its markets' timelines and describes imbalance information and auction collars. An indicative price calculated before the match is conditional on the eligible book at that moment. Subsequent permitted order changes can change the outcome.
For an execution backtest, retain the trading phase alongside the price and quantity. A batch print is not evidence that a strategy could have bought the same quantity immediately beforehand. Nor does an auction imbalance identify a buyer-initiated or seller-initiated trade in the continuous-book sense. Model eligibility, price formation and allocation before crediting the strategy with an auction fill.
Sources
Foucault, T., Pagano, M. and Röell, A. (2013), Market Liquidity: Theory, Evidence, and Policy, Chapter 1, §§1.1–1.2.1, pp. 16–23: order-driven and dealer structures, continuous matching and single-price calls.
ASX, Auctions: cumulative quantities, maximum executable volume and the conditional price-selection hierarchy.
NYSE, Auctions: market-specific auction timelines, indicative match information and price collars.
