A local non-parametric model for trade sign inference

Adam Blazejewski and Richard Coggins
Additional contact information
Adam Blazejewski: University of Sydney
Richard Coggins: University of Sydney

Finance from University Library of Munich, Germany

Abstract: We investigate a regularity in market order submission strategies for twelve stocks with large market capitalization on the Australian Stock Exchange. The regularity is evidenced by a predictable relationship between the trade sign (trade initiator), size of the trade, and the contents of the limit order book before the trade. We demonstrate this predictability by developing an empirical inference model to classify trades into buyer-initiated and seller-initiated. The model employs a local non-parametric method, k-nearest-neighbor, which in the past was used successfully for chaotic time series prediction. The k-nearest- neighbor with three predictor variables achieves an average out-of- sample classification accuracy of 71.40%, compared to 63.32% for the linear logistic regression with seven predictor variables. The result suggests that a non-linear approach may produce a more parsimonious trade sign inference model with a higher out-of-sample classification accuracy. Furthermore, for most of our stocks the observed regularity in market order submissions seems to have a memory of at least 30 trading days.

Keywords: Order submission; Trade classification; K-nearest-neighbor; Non-linear; Memory (search for similar items in EconPapers)
JEL-codes: G (search for similar items in EconPapers)
Pages: 17 pages
Date: 2004-08-30
New Economics Papers: this item is included in nep-cmp
Note: Type of Document - pdf; pages: 17
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
https://econwpa.ub.uni-muenchen.de/econ-wp/fin/papers/0408/0408009.pdf (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:wpa:wuwpfi:0408009

Access Statistics for this paper

More papers in Finance from University Library of Munich, Germany
Bibliographic data for series maintained by EconWPA ( this e-mail address is bad, please contact ).