A Bayesian inference for time series via copula-based Markov chain models. Issue 11 (1st November 2020)

• Record Type: Journal Article
• Authors: Sun, Li-Hsien; Lee, Chang-Shang; Emura, Takeshi
• Is Part Of: Communications in statistics. Volume 49:Issue 11(2020)
• Page Start: 2897
• Page End: 2913
• Publication Date: 2020-11-01
• Subjects: Clayton copula -- Nonstandardized Student's t-distribution -- Bayesian inference -- Markov chain Monte Carlo -- Metropolis-Hastings algorithm; 62M10 -- 62H12 -- 62H20 -- 60J20; Mathematical statistics -- Periodicals; Mathematical statistics -- Data processing -- Periodicals; Digital computer simulation -- Periodicals; 519.5
• DOI: 10.1080/03610918.2018.1529241
• Languages: English
• ISSNs: 0361-0918

Abstract

Abstract: This paper studies the nonstandardized Student's t-distribution for fitting serially correlated observations where serial dependence is described by the copula-based Markov chain. Due to the computational difficulty of obtaining maximum likelihood estimates, alternatively, we develop Bayesian inference using the empirical Bayes method through the resampling procedure. We provide the simulations to examine the performance and also analyze the stock price data in empirical studies for illustration.