Title: An analysis of market power in Iran's electricity market with machine learning

Authors: Naser M. Rostamnia

Addresses: Department of Economics, Kharazmi University, Tehran and Karaj, Iran

Abstract: The Iranian electricity market was reformed over the last three decades primarily to promote competition and improve its production efficiency. This paper provides an analysis of competition in the Iranian electricity market. Although other works have provided similar assessments, none has provided a thorough probe over a long period. This paper analyses the Herfindahl Hirschman Index (HHI) of the market for the last decade which has not been done. Also, the paper forecasts the index in the market for the next year to project its direction. Long Short-Term Memory (LSTM) was implemented to forecast the indices in an efficient computational time. Grid search is used to select the optimal model for forecasting, and interactions analyses provide insights into the parameter options that lead to significantly improved accuracies. The results show that the market was unconcentrated from 2012 to 2021. Also, the forecasts show that the market will remain unconcentrated for the next year. Furthermore, the analysis shows that the entrance of new powerplants into the market could reduce the concentration in the market.

Keywords: market power analysis; Herfindahl Hirschman index; long short-term memory algorithm; hyperparameter optimisation; grid search.

DOI: 10.1504/IJGEI.2023.132009

International Journal of Global Energy Issues, 2023 Vol.45 No.4/5, pp.489 - 502

Received: 09 Nov 2021
Accepted: 28 Mar 2022
Published online: 06 Jul 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article