Testing for the dimension of the nonlinear state space model using particle filter

Research Project

Project/Area Number	15K03394
Research Category	Grant-in-Aid for Scientific Research (C)
Allocation Type	Multi-year Fund
Section	一般
Research Field	Economic statistics
Research Institution	Yokohama National University
Principal Investigator	Kobayashi Masahito 横浜国立大学, 大学院国際社会科学研究院, 教授 (60170354)
Project Period (FY)	2015-04-01 – 2018-03-31
Project Status	Completed (Fiscal Year 2017)
Budget Amount *help	¥2,990,000 (Direct Cost: ¥2,300,000、Indirect Cost: ¥690,000) Fiscal Year 2017: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000) Fiscal Year 2016: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000) Fiscal Year 2015: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Keywords	copula / stochastic volatility / state space model / particle filter / maximum likelihood / 粒子フィルター / 最尤推定 / 非対称性 / Stochastic volatility / ラグランジュ乗数検定 / 分布の退化の検定 / Stochastic Volatility / 非線形カルマンフィルター / Lagrange 乗数検定
Outline of Final Research Achievements	In this project I have obtained the Lagrange multiplier test for the null hypothesis that bivariate time series has a single volatility process in common against the alternative hypothesis that the two series have different volatility process. This test can be used to check that two financial returns has a common factor of volatility changed and hence that the cause of financial turmoil can be identified. I also developed a new asymmetric copula which is more flexible than the popular Joe-Clayton copula and obtained a model where asymmetric dependence of bivariate financial time series changes over time in the fame work of the state-space model. and estimated it using particle filter method. These models were applied to actual financial data.