Stock price network autoregressive model with application to stock market turbulence
- PDF / 1,053,842 Bytes
- 15 Pages / 595.276 x 785.197 pts Page_size
- 4 Downloads / 203 Views
THE EUROPEAN PHYSICAL JOURNAL B
Regular Article
Stock price network autoregressive model with application to stock market turbulence Arash Sioofy Khoojine a and Dong Han School of mathematics and statistics, Shanghai Jiao Tong University, Shanghai, P.R. China Received 29 August 2019 / Received in final form 2 February 2020 Published online 13 July 2020 c EDP Sciences / Societ`
a Italiana di Fisica / Springer-Verlag GmbH Germany, part of Springer Nature, 2020 Abstract. In this article, the authors develop a Stock Price Network Autoregressive Model (SPNAR) to probe the behavior of the log-return based network of the Chinese stock market. We consider 105 companies of Shanghai and Shenzhen stock market, CSI300, during the steep sell-off in 2015–2016. This model is based on three effects of previous time effect, market effect, and independent noise effect. The results show that the accuracy and performance of this model are more than some time series models like Autoregressive (AR), Moving Average (MA), Autoregressive Moving Average (ARMA), and Vector Autoregressive (VAR) models. Furthermore, the parameter estimation in SPNAR model is more convenient and feasible than time series models as mentioned earlier. Moreover, In this article, the characteristics of three various periods, preturbulence, turbulence, and post-turbulence are analyzed, and findings show there is a significant difference between turbulence period with other periods in topological structure and the behavior of the networks.
1 Introduction 1.1 Framework of financial networks Network analysis is used to define the attribute or behaviors of complex networks. Complex network methods have become essential tools for depicting and determining complex systems in various branches of science [1,2]. In recent decades, there have been some research that conduct to model the stock market applying complex networks. Utilizing complex networks to analyze the financial markets is because the particular stocks in a market are generally correlated, either for the market trend or the periodicity of the specific market segments [3]. In modeling the stock market using network analysis, different stocks are represented as different nodes, and the correlation between them is supposed as edges. Several kinds of research has been conducted to explore the nature of financial markets, the most influential papers among them are Onnela et al. [4], Tumminello et al. [5], Heimo et al. [6] and Gan and Djauhari [7] for New York stock exchange, Huang et al. [8], Qiao et al. [9] and Long et al. [10] for Chinese stock market, Vizgunov et al. [11] for the Russian stock market, Radhakrishnan et al. [12] for Thailand SET index, Birch et al. [13] for German DAX30 Index. In addition, different linear and non-linear methods of defining the correlation between nodes have been constructed to shape the financial networks such as a
e-mail: [email protected]
threshold method, and information-theoretic based distances [14,15]; however, various techniques are deployed for visualization of
Data Loading...
