Information-based models in centralised and decentralised exchanges: spoofing and private order flow

Thesis

In this thesis we focus on information-based models in both traditional exchanges (limit order books) and decentralised ones.

We propose a dynamic model of the limit order book to derive conditions to test if a trading algorithm learns to manipulate the order book. Our results show that as a market maker becomes more tolerant to bearing inventory risk, the learning algorithm will find optimal strategies that manipulate the book more frequently. Manipulation helps to revert inventory to an optimal level and to execute round-trip trades with limit orders at a higher probability than was otherwise likely to occur. Spoofing is a special case of quote-based manipulation where the market maker prefers that the manipulative limit orders are not filled. We use high-frequency data to check our conditions and show that algorithms will learn to manipulate Nasdaq's limit order book. Finally, we extend our model in several directions and we see manipulation can still arise in all of them. In particular, when two market makers use learning algorithms to trade, their algorithms can learn to coordinate their manipulation.

We study how the design of blockchains shapes the interaction between traders in decentralised exchanges (DEXs) and participants in the blockchain security protocol. On blockchains such as Ethereum, traders can route their transactions through either a public memory pool, where transactions are visible and subject to attacks, or private memory pools, where transactions are hidden but can only be executed by specific builders. These features create a fundamental trade-off between execution certainty and protection from attacks. We develop two complementary models to analyse this trade-off. The first model studies the effect of competition among builders (MEV-Boost auction). We show that DEX liquidity depth governs equilibrium fragmentation: when liquidity is high, traders concentrate in one private pool, whereas scarce liquidity leads to balanced order flow across pools. The second model examines the role of builder composition-specifically, the share of builders able to execute private orders. Here, traders’ behaviour and equilibrium fragmentation depend on this composition, with higher shares of public-only builders inducing a shift toward public execution. Together, the models explain how liquidity conditions and builder structure jointly order flow fragmentation.

Authors: García Arenas, G

• DOI: 10.5287/ora-ao9zqpdky
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English