Predicting equity market price impact with performance weighted ensembles of random forests
Predicting equity market price impact with performance weighted ensembles of random forests
For many players in financial markets, the price impact of their trading activity represents a large proportion of their transaction costs. This paper proposes a novel machine learning method for predicting the price impact of order book events. Specifically, we introduce a prediction system based on performance weighted ensembles of random forests. The system’s performance is benchmarked using ensembles of other popular regression algorithms including: liner regression, neural networks and support vector regression using depth-of-book data from the BATS Chi-X exchange. The results show that recency-weighted ensembles of random forests produce over 15% greater prediction accuracy on out-of-sample data, for 5 out of 6 timeframes studied, compared with all benchmarks
1-8
Booth, Ash
e23d78c8-4b8c-421c-962f-b875136b8e25
Gerding, Enrico
d9e92ee5-1a8c-4467-a689-8363e7743362
McGroarty, Frank
693a5396-8e01-4d68-8973-d74184c03072
2014
Booth, Ash
e23d78c8-4b8c-421c-962f-b875136b8e25
Gerding, Enrico
d9e92ee5-1a8c-4467-a689-8363e7743362
McGroarty, Frank
693a5396-8e01-4d68-8973-d74184c03072
Booth, Ash, Gerding, Enrico and McGroarty, Frank
(2014)
Predicting equity market price impact with performance weighted ensembles of random forests.
Computational Intelligence for Financial Engineering and Economics (CIFEr), London, United Kingdom.
27 - 28 Mar 2014.
.
Record type:
Conference or Workshop Item
(Paper)
Abstract
For many players in financial markets, the price impact of their trading activity represents a large proportion of their transaction costs. This paper proposes a novel machine learning method for predicting the price impact of order book events. Specifically, we introduce a prediction system based on performance weighted ensembles of random forests. The system’s performance is benchmarked using ensembles of other popular regression algorithms including: liner regression, neural networks and support vector regression using depth-of-book data from the BATS Chi-X exchange. The results show that recency-weighted ensembles of random forests produce over 15% greater prediction accuracy on out-of-sample data, for 5 out of 6 timeframes studied, compared with all benchmarks
Text
ash_booth_IEECISE_2013.pdf
- Author's Original
More information
Published date: 2014
Venue - Dates:
Computational Intelligence for Financial Engineering and Economics (CIFEr), London, United Kingdom, 2014-03-27 - 2014-03-28
Organisations:
Centre for Digital, Interactive & Data Driven Marketing, Agents, Interactions & Complexity
Identifiers
Local EPrints ID: 363655
URI: http://eprints.soton.ac.uk/id/eprint/363655
PURE UUID: 52b0d5e0-d3c7-4454-87a9-d41bfcc3a326
Catalogue record
Date deposited: 28 Mar 2014 14:56
Last modified: 15 Mar 2024 03:23
Export record
Contributors
Author:
Ash Booth
Author:
Enrico Gerding
Author:
Frank McGroarty
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics