# Risk formulation mechanism among top global energy companies under large shocks

## TL;DR

This paper studies how top energy companies manage risks during major global shocks using machine learning and network analysis.

## Contribution

A new method combining machine learning and network analysis is introduced to assess energy market risks under large shocks.

## Key findings

- The proposed method effectively captures energy market risk profiles during major shocks.
- Risk contagion in energy markets shows geographical clustering.
- Company-specific and market factors significantly influence tail risks.

## Abstract

Taking top global energy companies as the epitome, this paper investigates the risk formulation mechanism of the international energy market under the impact of large shocks. We first use the machine learning method in (Liu and Pun, 2022) to calculate the systematic risk level - EMES - for each energy company. Then use network analysis methods to explore the internal risks due to risk comovement among top energy companies. Finally, a dynamic quantile regression model(DNQR) is used to investigate the external risks occasioned by network effects, individual company characteristics, and market environment. Our research finds that the method we use can capture the risk profile of the energy market under different major shocks. Secondly, we find that the risk contagion in the energy market exhibits geographical clustering characteristics, and certain firm-specific factors and market environmental factors of the company have a significant impact on the tail risk of the company. Our research can provide reference and guidance for risk management in the energy market.