<![CDATA[The movement of energy sector stock returns is dynamic and influenced by external factors that cause changes in market conditions over time. These characteristics indicate the existence of regime shifts that cannot be optimally modeled using conventional linear time series. Therefore, this study aims to model energy sector stock returns in Indonesia using the Non-Homogeneous Markov Switching Autoregressive (NHMS-AR) approach with a focus on the transition probability between regimes influenced by exogenous variables. The data used are monthly logarithmic returns of PT Perusahaan Gas Negara Tbk (PGAS), PT Adaro Energy Indonesia Tbk (ADRO), and PT Medco Energi Internasional Tbk (MEDC) for the period September 2008 to December 2024. The exogenous variables used include the Climate Risk Index, Geopolitical Risk (GPR), and Global Economic Policy Uncertainty (GEPU). The NHMS-AR model is implemented with the assumption that exogenous variables influence the probability of regime shifts, while the autoregressive structure in each regime is homogeneous. The results show that the dynamics of energy sector stock returns can be represented by two hidden regimes: a stable regime and a volatile regime. Transition probability estimates indicate that CRI and GEPU increase the probability of switching between regimes, whereas GPR exhibits an asymmetric effect, depending on the direction of the transition. Furthermore, the volatile regime has a higher degree of persistence than the stable regime. The main contribution of this study lies in the application of NHMS-AR with non-homogeneous transition probabilities to Indonesian energy sector stocks, as well as in presenting empirical evidence on the role of external factors in shaping regime switching dynamics in emerging markets.]]>
Copyrights © 2025
