Use this URL to cite or link to this record in EThOS:
Title: Bayesian econometric modelling of informed trading, bid-ask spread and volatility
Author: Oduro, Samuel Dua
ISNI:       0000 0004 6349 0107
Awarding Body: University of Kent
Current Institution: University of Kent
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Access from Institution:
Recent developments in global financial markets have increased the need for research aimed at the measurement and possible reduction of liquidity risk. In particular, market crashes have been partly blamed on the sudden withdrawal of liquidity in markets and increases in liquidity risk. To this end, it is important to develop better approaches for inferring or quantifying liquidity risk. Liquidity risk caused by some investors trading on their information advantage (informed trading) has been a subject of market microstructure research in the last few decades. Researchers have employed information-based models that use observed or inferred order flow to investigate this problem. The Probability of Informed Trading (PIN) is a measure which uses inferred order flow to quantify the extent information asymmetry. However, a number of computational issues have been reported to effect the estimation of PIN. Using an alternative methodology, we address the numerical problem associated with the estimation of PIN. Varied evidence of a relationship between volume and bid-ask spread has been documented in the extant literature. In particular, theory suggests that bid-ask spread and volume are jointly driven by a common process as both variables measure an aspect of liquidity. The complex relationship between these variables is time-varying since the informed trading component of order flow changes as trading takes place. Thus, volume and bid-ask spread may provide insight on the time-varying composition of economic agents trading an asset. We exploit the nonlinear relationship between traded volume and bid-ask spread to develop a model that can be used to infer informed and uninformed trading components of volume. The structure of the model and estimation methodology enhances the sequential processing and incorporation of past volume and bid-ask spread as conditioning information. The model is applied to two equities that trade on the New York Stock Exchange. Finally, to increase our understanding on the effects of liquidity risk on volatility, we also examine whether separating volume into informed and uninformed components can provide further insight on the relationship between liquidity risk and volatility.
Supervisor: Griffin, Jim ; Oberoi, Jaideep Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: QA Mathematics (inc Computing science)