Search for content and authors
 

Causality Link Prediction analysis in OECD stock market indices

Ji Hwan Park ,  Minhyuk Lee ,  Sungyoon Choi ,  Woojin Chang 

Seoul National University (SNU), School of Mat. Sci. Eng., Seoul 151742, Korea, South

Abstract

 Link prediction in multi-relational networks has become an important area in network analysis and is widely used in social network analysis. However, applying this method to financial market is limited. For this reason, we construct networks using the link prediction method. In this research, we analyze the changes of financial networks between OECD countries over five categories of period: pre-crisis, global-crisis, inter-crisis, European-crisis, post-crisis. We apply a mechanism called knowledge dissemination to measure the power of a node by computing the H-index of the node. We analyze explanatory power of the constructed network by observing the change of benchmark minimum spanning tree network. Through this network constructing process, we can obtain the converged stationary network and use the related outcomes for individual institutions’ decision making or predicting market crash.

 

Legal notice
  • Legal notice:
 

Presentation: Oral at Econophysics Colloquium 2017, Symposium A, by Ji Hwan Park
See On-line Journal of Econophysics Colloquium 2017

Submitted: 2017-03-13 09:15
Revised:   2017-04-15 16:44