Use this URL to cite or link to this record in EThOS:
Title: The effects of strategic behaviour on properties of order flow
Author: Palit, Imon Joydipto
Awarding Body: University of Essex
Current Institution: University of Essex
Date of Award: 2012
Availability of Full Text:
Access from EThOS:
This thesis seeks a better understanding of the role that endogenous dynamics plays in financial markets. We investigate the intra-day dynamics of trading in stock markets by means of computer simulations and empirical analysis of high-frequency data. We show that the strategic behaviour of interacting liquidity taking and providing agents plays an important role in explaining the phenomena of volatility clustering, long-memory in order flow and stable- U-shape intra-day volume dynamics that are-observed in pure order driven markets. We first construct a zero-intelligence model to reproduce order flow from naive trader behaviours in a market that uses the double auction market clearing mechanism. We simulate the market by generating random types of event drawn from empirical distributions calibrated against data from the Paris Bourse exchange. We find that whilst the model recovers some realistic features, it does not reproduce volatility clustering. This suggests strategic behaviour is an important feature of markets creating correlated order flow as a result of a subtle interplay of liquidity provision and taking. We then go on to test empirically whether an important part of this correlated order flow occurs due to order splitting. We show that the cause of long-memory at high frequency timescales is due to persistent liquidity taking by single agents. This is in line with the practice of algorithmic trading that strategically splits a large order into smaller pieces to avoid market impact and detection from other market participants. Finally we investigate how the practice of algorithmic order splitting can influence intra-day volume dynamics. We develop a model of intra-day trading volume dynamics and run Monte-Carlo simulations with estimated parameters. We find our model achieves realistic dynamics when compared with observed volume curves. This suggests strategic behaviour from algorithmic trading plays a major part in shaping intra-day trading volume dynamics.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available