A Study on the Nonlinear Causality among Dollar Index, Gold and Crude Oil Price

• Main Authors: Liang-Yu Lin, 林亮宇
• Other Authors: Ming-Jui Hsieh
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/v6b965

碩士 === 淡江大學 === 財務金融學系碩士班 === 106 === It was based on three kinds of variables such as U.S Dollar Index (USDX, DXY), XAU/USD, Crude Oil-continuous (NYMEX) in this study, and the time total was 10 years from January 2008 to December 2017, and the daily data total was 2517. Firstly, this study was made the single root verification by the ADF, PP, KPSS, and the study variable to become a stationary sequence I (1), and then using the nonlinear threshold error correction model structure in order to know the long and short term non-linear causal relationship among the variables such as the dollar index, crude oil price and gold price. The methodology was adopted the threshold autoregressive (TAR) model introduced by Enders and Granger (1998), and the MTAR-model is based on the Threshold Autoregressive model (TAR-model) for the conformity test, in order to determine the relationship among those variables such as the dollar index, the gold price and the crude oil price, it was found the long-term equilibrium asymmetry between the crude oil price and the dollar index. Using the Threshold Error Correction Model (TECM) of Enders and Granger (1998) and Enders and Siklos (2001), it was found that the variable of gold price has a short-term and long-term causal relationship with the variable of dollar index, and the dollar index has a long asymmetrical causal relationship with the price of gold and crude oil.