Premium
Forecasting Long‐Run Coal Price in China: A Shifting Trend Time‐Series Approach
Author(s) -
Dong Baomin,
Li Xuefeng,
Lin Boqiang
Publication year - 2010
Publication title -
review of development economics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.531
H-Index - 50
eISSN - 1467-9361
pISSN - 1363-6669
DOI - 10.1111/j.1467-9361.2010.00567.x
Subject(s) - economics , autoregressive integrated moving average , coal , unobservable , econometrics , energy mix , china , investment (military) , price index , macroeconomics , time series , power (physics) , statistics , engineering , electricity generation , physics , mathematics , quantum mechanics , politics , political science , law , waste management
The paper studies the behavior of mid‐ to long‐run real coal price in the Chinese market. The problem is of great importance because the coal takes a 70% share in China's energy mix, and China is the world's second largest carbon emitter. An accurate forecast in coal price is crucial in predicting China's future energy consumption mix as well as the private sector's energy‐type‐related investment decisions. In estimation and forecasting, the shifting trend time‐series model suggested by Robert Pindyck is used to capture the technological progress that is unobservable to the econometrician. It is found that the shifting trend model with continuous and random changes in price level and trend outperforms plain vanilla ARIMA models. It is argued that the model postulated by Pindyck is robust even in a transition economy where energy prices are subject to relatively rigid regulatory control. Out‐of‐sample forecasts are provided.