Background
Background & Problem Analysis
Fundamentals of Perpetual DEX Models
Orderbook Model
In the orderbook model, traders place limit or market orders that are matched via an order-matching engine. Execution is often off-chain matching + on-chain settlement, enabling high-frequency trading similar to centralized exchanges (CEXs). Advantages:
High performance, low latency
Flexible order types Trade-offs:
Requires off-chain infrastructure and trusted operators
Centralized control over matching and ordering
Peer-to-Pool Model
In the peer-to-pool model, traders interact directly with a liquidity pool, bypassing traditional order matching. The protocol itself acts as the counterparty to all trades, with liquidity provided by LPs. Advantages:
Fully on-chain, transparent execution
Open to any liquidity provider, not just professional market makers Trade-offs:
Price discovery often depends on external oracles
Capital efficiency limits due to TVL constraints
Comparison
Orderbook: High performance but more centralized. Peer-to-Pool: Transparent and open, but with performance and dependency limitations.
Structural Issues of the Orderbook Model
Architectural Features
Matching engine usually runs on centralized servers or specialized appchains
Example: dYdX v3 uses a centralized matching engine
Example: Hyperliquid built its own chain for extreme performance
Users sign orders off-chain; the protocol’s matching service determines execution order
Sacrifices decentralization in exchange for availability and consistency (per the CAP theorem)
Market Maker–Dominated Manipulation Risks
Stop-hunting: Market makers push prices to trigger stop-losses/liquidations
Spoofing: Placing fake orders to mislead market participants
Information asymmetry: Market makers have privileged order flow and sequencing visibility
Wash Trading & Inflated Volume
Definition: The same entity matches its own buy and sell orders to generate artificial volume. Causes:
Matching is off-chain and costless
No real asset transfer required
Zero slippage and no gas costs
Only two controlled accounts needed Motivations:
Create a false impression of liquidity
Climb rankings on volume leaderboards
Exploit trade mining rewards Consequences:
Misleading market depth perception
Distorted data affecting governance and valuation
Unfair competition environment
Structural Issues of the Peer-to-Pool Model
Price Discovery Reliance on CEXs
Most protocols use centralized oracles (e.g., Chainlink) for mark price
On-chain trades do not feed back into the mark price
The protocol becomes a price consumer, not a price producer
Limited Liquidation Timeliness
High volatility requires fast liquidations, but constrained by:
Block time
Gas limits and priority fees
Network congestion
Delays lead to LP losses and reduced liquidator incentives
Capital Efficiency Ceiling
All positions are backed by LP pool TVL
High leverage magnifies LP risk
OI (Open Interest) is capped by available liquidity
Centralized Oracle Dependence
Reliance on centralized data sources (CEXs)
Whitelists or rate limits reduce openness
Creates potential “backdoor pricing” authority
LP Naked Exposure Risk
Traders choose position direction; LPs are forced to take the opposite side
No effective hedging, leading to unpredictable returns and possible mass LP exits during extreme moves
On-Chain Infrastructure Limitations
Non-deterministic transaction ordering: Enables frontrunning and unfair execution
Gas limits and block capacity: Complex transactions may fail
Uncertain pending transaction time: Congestion can delay execution for minutes
Non-unique block states: Flash loans and intra-block changes can cause temporary state mismatches affecting trade and liquidation accuracy
Summary
Orderbook Model: Optimized for performance but prone to centralization, manipulation, and wash trading
Peer-to-Pool Model: Transparent but suffers from oracle dependency, delayed liquidations, and limited capital efficiency
Underlying blockchain constraints amplify these issues AZEx’s goal is not to “pick a side” but to design a third path that delivers complete mechanism closure, reliable trading experience, and fairness for both LPs and traders.
Last updated