Dynamic clustering of energy markets: An extended hidden Markov approach
作者：
Highlights：
• We analyze cycles in energy prices using an extension of the hidden Markov framework to panel data.
• U.S. energy markets are heterogeneous with two distinct clusters.
• Electricity markets are modeled by five states.
• Oil and gas markets are in just two states 96.6% of the time.
• We find little synchronization of states.
摘要
•We analyze cycles in energy prices using an extension of the hidden Markov framework to panel data.•U.S. energy markets are heterogeneous with two distinct clusters.•Electricity markets are modeled by five states.•Oil and gas markets are in just two states 96.6% of the time.•We find little synchronization of states.
论文关键词：Hidden Markov models (HMMs),Clustering,Time series,Energy markets
论文评审过程：Available online 13 June 2014.
论文官网地址：https://doi.org/10.1016/j.eswa.2014.05.030
