Research on Structural Change of China's A-share Market

doi:10.13766/j.bhsk.1008-2204.2016.0228

Abstract

Abstract:

In recent months, China's A share showed a very large volatility with more than 1 000 shares falling by their 10 percent daily limit or increasing by their daily 10 percent limit usually on the same day. Is it possible to find stock market structural change point in advance based on historical information? For each component, a certain RuLSIF model is built. Taking the six dimension matrix as the sample, the non parametric density function ratio of the time series sample is estimated to be the structural change point of A stock market. Empirical results show that since the establishment of A shares, October 16 2014 has been the market structure mutation points. That is, beginning from that day, A-share market structure has undergone fundamental changes, to be very different from the market structure before. The model based on the sample data obtained from the sample forecast and the actual trend is consistent, and then the model can provide reference for the adjustment of the investment strategy in the actual investment.

Key words: China's Share Market, A-share Market, volatility, RuLSIF model, six dimension matrix, non parametric density function ratio, structural change point

CLC Number:

F830.9

FANG Zhaoben, WANG Chuanhao. Research on Structural Change of China's A-share Market[J]. JOURNAL OF BEIJING UNIVERSITY OF AERONAUTICS AND A, 2018, 31(3): 55-60.

References

[1] HSU D A, MILLER R B, WICHERN D W. On the stable Paretian behavior of stock-market prices[J]. Journal of the American Statistical Association, 1974, 69(345):108-113.
[2] BOOTH N B, SMITH A F M. A bayesian approach to retrospective identification of change-points[J]. Journal of Econometrics, 1982, 19(1):7-22.
[3] WORSLEY K J. Confidence regions and tests for a change-point in a sequence of exponential family random variables[J]. Biometrika, 1986, 73(1):91-104.
[4] MIAO B Q. Inference in a model with at most one slope-change point[J]. Journal of multivariate analysis, 1988, 27(2):375-391.
[5] 缪柏其, 赵林城, 谭智平. 关于变点个数及位置的检测和估计[J]. 应用数学学报, 2003, 26(1):26-39.
[6] RAMANAYAKE A. Tests for a change point in the shape parameter of gamma random variables[J]. Communications in Statistics-Theory and Methods, 2005, 33(4):821-833.
[7] 谭常春, 缪柏其. 至多一个变点的Г分布的统计推断[J]. 中国科学技术大学学报, 2005, 35(1):51-58.
[8] 雷鸣, 谭常春, 缪柏其. 运用生存分析与变点理论对上证指数的研究[J]. 中国管理科学, 2007, 15(5):1-8.
[9] 杨继平, 陈晓暄, 张春会. 中国沪深股市结构性波动的政策性影响因素[J]. 中国管理科学, 2012, 20(6):43-51.
[10] LIU S, YAMADA M, COLLIER N, et al. Change-point detection in time-series data by relative density-ratio estimation[J]. Neural Networks, 2013, 43:72-83.
[11] KAWAHARA Y, SUGIYAMA M. Sequential change-point detection based on direct density-ratio estimation[J]. Statistical Analysis and Data Mining:The ASA Data Science Journal, 2012, 5(2):114-127.
[12] ALI S M, SILVEY S D. A general class of coefficients of divergence of one distribution from another[J]. Journal of the Royal Statistical Society:Series B (Methodological), 1966,28(1):131-142.