Computational study of the US stock market evolution: a rank correlation-based network model Academic Article

abstract

• This paper presents a computational study of global characteristics of the US stock market using a network-based model referred to as the market graph. The market graph reflects similarity patterns between stock return fluctuations via linking pairs of stocks that exhibit "coordinated" behavior over a specified period of time. We utilized Spearman rank correlation as a measure of similarity between stocks and considered the evolution of the market graph over the recent decade between 2001-2011. The observed market graph characteristics reveal interesting trends in the stock market over time, as well as allow one to use this model to identify cohesive clusters of stocks in the market. 2013 Springer-Verlag Berlin Heidelberg.

authors

• Butenko, Sergiy

published proceedings

• COMPUTATIONAL MANAGEMENT SCIENCE

author list (cited authors)

• Shirokikh, O., Pastukhov, G., Boginski, V., & Butenko, S.

citation count

• 23

complete list of authors

• Shirokikh, Oleg||Pastukhov, Grigory||Boginski, Vladimir||Butenko, Sergiy

publication date

• June 2013

publisher

• Springer Nature  Publisher

published in

• Computational Management Science  Journal

keywords

• Cohesive Clusters
• Degree Distribution
• K-cores
• Market Graph
• Outliers
• Spearman Rank Correlation

Digital Object Identifier (DOI)

• 10.1007/s10287-012-0160-4

start page

• 81

end page

• 103

volume

• 10

issue

• 2-3

URL

• http://dx.doi.org/10.1007/s10287-012-0160-4