Use this URL to cite or link to this record in EThOS:
Title: Limit order book dynamics and market impact estimation
Author: Malik, M. A. A.
ISNI:       0000 0004 2738 2651
Awarding Body: University of Essex
Current Institution: University of Essex
Date of Award: 2011
Availability of Full Text:
Access from EThOS:
This thesis focuses on two closely related areas of liquidity and market impact. The historic limit order book of Stock Exchange Trading Sys- tem (SETS) operated by the London Stock Exchange (LSE) is rebuilt using this framework for empirical analysis of information contained in the limit order book. The concept of Notional Volume Weighted Average Price (NVWAP) is introduced to construct liquidity supply and demand curves based on real time bid and ask schedules of the full length of the limit or- der book. This unique approach is used to determine how the order book behaves and I find consistent wave-like patterns between up- ward and downward price trends. Regression coefficients of the slope of the curves for each market event are estimated using an exponen- tial model. Four statistics are defined to identify bullish and bearish trends without prior knowledge of the market price. Detailed analy- sis shows that these statistics correctly identify market conditions for 88% to 97% of the observations. The intraday patterns of regression coefficients are revealed using a nonparametric kernel regression model. These intraday patterns are not found to be consistent between stocks over time. A resampled and deseasonalised set of estimated regression coefficients is analysed for temporal dependence using a multivariate vector autoregression (VAR) model. Inferences drawn from marginal probabilities regarding Cranger-causality do not show significant impact of slope coefficients on the opposite side of the limit order book implying that each side of the market is simultaneously rather than sequentially influenced by prevailing market conditions. The VWAP concept is extended to estimate the average shape of the limit order book and average market impact. The average market impact estimates are found to be superior than the order imbalance based approach. A time-of-day market impact for a given aggregate volume is estimated using a multivariate kernel regression model with monotonicity constraint. The estimated market impact shows stock- specific and wave-like impact that is asymmetric for buyers and sellers.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available