The main objective of this doctoral research is to develop models that detect sequences of Limit Order Book (LOB) events that have an exceptionally large market impact and could, from this perspective, be related to market manipulation. This objective is both academically and practically important. Academically, in addition to contributing to research on market surveillance, it allows for the analysis of market dynamics in situations where LOB events form strong cascades, which is crucial for better understanding price formation mechanisms and microstructure noise. These topics will be addressed as part of this research. From the perspective of practitioners, this research can yield entirely new methods for the preliminary identification of illegal trading maneuvers in financial markets. Moreover, the methods could be used to design optimal order splitting strategies, a crucial question for many institutional traders and exchanges to maintain stable stock prices.

2025

2027