財務諸表及び経営指標を用いた企業のエネルギーシフト可能性に関する研究

抄録

We are facing a great trend of energy transition from fossil fuel to renewable energy in the world, and companies should play a key role in de-carbonization. In the Energy Transition period, it is important to examine what brings companies to encourage the transition from a financial and business perspective. This paper aims to clarify the characteristics of aggressive companies about energy transition by applying machine learning methods in the Tokyo Stock Exchange; both quantitative and qualitative analysis of the company’s financial/business data were examined. The k-means++, which is a method of clustering, was used for the quantitative analysis, and TF-IDF, which is a method of quantifying the importance of words in documents, was applied to the qualitative analysis. The results showed that companies who worked on energy transition positively (1) earned normal or slightly below financial scores, (2) had larger electricity indirect CO₂ emission than direct CO₂ emission, and (3) were particularly aware of “Strategies” and “Own brand names” in their business plans. It was also shown that newly introduced CO₂ indexes successfully captured characteristics of companies from CO₂ emission perspective.