Date of Award
2013
Document Type
Master's Thesis
Degree Name
Master of Science in Applied Natural Resource Economics (MS)
College, School or Department Name
School of Business and Economics
Advisor
Mark C Roberts
Abstract
This thesis analyzes the long-term trend behavior and cyclical components of real base metal prices. A decomposition approach is used to separate the real prices of base metals into a time trend and cyclical components. In this regard, linear and quadratic detrending methods along with Hodrick-Prescott and Baxter-King filters are applied to the base metal prices. Linear and quadratic trendlines are good estimates of the trend component and easy to interpret if the coefficients of estimated trends are significant and the coefficient of determination is relatively high. The Hodrick-Prescott and Baxter-King filters are good fits to the price series, but they affect the cyclical component. Loss of data points and altering the moments of the cyclical component are the disadvantages of filtering methods. On the other hand, the linear detrending methods are weak in removing unit roots in the series. This study shows that the choice of detrending method affects the cyclical component of base metal prices, and consequently the identification of cycles depends on the detrending method.
Recommended Citation
Modirroosta, Mahmoodreza, "A DECOMPOSITION ANALYSIS OF BASE METAL PRICES: COMPARING THE EFFECT OF DETRENDING METHODS ON TREND IDENTIFICATION AND CYCLICAL COMPONENTS", Master's Thesis, Michigan Technological University, 2013.