Introduction

Electronic exchanges use deterministic matching policies to decide which resting orders execute when incoming liquidity arrives. Matching rules shape fairness, queue incentives, and market behavior. Most venues rely on a few common models, often with hybrid adjustments by instrument and session type.

Price-Time Priority as the Dominant Baseline

Price-time priority, often called FIFO at each price level, is widely used in equities and many crypto venues.

Core rule:

1. better price executes first
2. for equal price, earlier timestamp executes first

1from dataclasses import dataclass
2from heapq import heappush, heappop
3
4@dataclass
5class Order:
6    side: str
7    price: float
8    qty: int
9    ts: int
10
11buy_book = []
12orders = [
13    Order("B", 100.0, 50, 1),
14    Order("B", 100.0, 40, 2),
15]
16
17for o in orders:
18    heappush(buy_book, (-o.price, o.ts, o))
19
20best = heappop(buy_book)[2]
21print(best.qty, best.ts)

This model rewards queue position and encourages early liquidity posting.

Pro-Rata Matching

Some derivatives venues use pro-rata matching, where incoming quantity is distributed proportionally among resting orders at best price.

1def pro_rata_allocate(resting_sizes, incoming_qty):
2    total = sum(resting_sizes)
3    if total == 0:
4        return [0] * len(resting_sizes)
5
6    raw = [incoming_qty * s / total for s in resting_sizes]
7    alloc = [int(x) for x in raw]
8    remain = incoming_qty - sum(alloc)
9
10    rank = sorted(range(len(raw)), key=lambda i: raw[i] - alloc[i], reverse=True)
11    for i in rank[:remain]:
12        alloc[i] += 1
13
14    return alloc
15
16print(pro_rata_allocate([100, 200, 300], 300))

Benefits include stronger incentive to quote larger displayed size. Tradeoffs include complexity around rounding, minimum fills, and anti-gaming rules.

Auction-Based Matching

Opening and closing auctions match orders in batches instead of continuously. The engine seeks a clearing price that maximizes matched volume under venue rules.

Why exchanges use auctions:

• liquidity concentration at important session times
• reduced continuous queue-racing pressure
• transparent uncrossing event with single clearing price

Many markets combine continuous matching during regular trading with auction mechanisms at open and close.

Hybrid Models in Production Venues

Real exchanges often implement hybrids rather than pure single-rule policies. Common combinations:

• price-time in continuous book plus auction mechanisms by session
• pro-rata with time-priority tie-break overlays
• distinct treatment for hidden and iceberg orders

Instrument-specific rulebooks define these details. Engineering teams should model rule variants explicitly rather than assuming one universal algorithm.

Matching Rules and Market Incentives

Algorithm choice influences participant behavior:

• price-time rewards speed and queue discipline
• pro-rata rewards displayed size
• auctions reward strategic timing into call windows

There is no absolute best model. Venue goals decide the tradeoff among immediacy, fairness criteria, and liquidity quality.

Engine Engineering Requirements

Independent of matching policy, production engines need:

• deterministic replay for audit and surveillance
• strict sequencing under high throughput
• fairness guarantees consistent with published rules
• low-latency operation with predictable jitter bounds

Simple prototype logic is not enough. Exchange engines require careful data structure design, concurrency control, and observability for regulatory and operational trust. Change-management discipline is equally important because small rule adjustments can materially alter queue behavior and participant incentives. Rigorous replay testing helps.

Common Pitfalls

• Assuming all exchanges use pure FIFO without instrument-specific exceptions.
• Ignoring incentive effects of matching policy on participant behavior.
• Implementing pro-rata without robust rounding and tie-break rules.
• Treating hidden-order logic as separate from matching policy.
• Designing engines without deterministic replay and auditability.

Summary

• Price-time priority is the most common baseline algorithm.
• Pro-rata and auction methods are also widely used in specific markets.
• Many exchanges run hybrid rule sets by session and instrument.
• Matching policy directly shapes fairness and liquidity incentives.
• Engine implementation must prioritize determinism, clarity, and operational reliability.