近年來ESG大量出現在各個金融、商業管理的文章中，ESG分別是環境保護（Environment）、社會責任（Social）以及公司治理（Governance）的縮寫。2004 年，UN Global Compact所提出的Who Cares Wins報告首次提出將 ESG 納入投資決策，他們認為，有心經營ESG的企業，才能奪得競爭優勢。疫情和全球氣候的變遷，企業開始思考如何在保護地球的同時實現營收增長。另一方面，網路社群的快速發展，企業的一舉一動都被人所熟知和放大檢視，任何不恰當的行動都會影響到投資人以及客戶的觀感，進而降低聲譽。因此，永續發展和重視企業社會責任已經成為當代公司的主要核心競爭能力。如今，大型企業紛紛開始揭露ESG的相關資訊。實驗證明，透過資訊透明公開化，公司能減輕風險。

近年來，疫情使得醫療負擔日趨嚴重，人力成本、維持設備運作的成本使永續經營方式受到考驗，本研究以美國上市的醫療照護公司為例，探討公司財報內的風險揭露程度對於ESG之影響。藉由BERT模型找出企業風險的四種文字指標，再用迴歸分析對比文字指標和傳統財務指標，實驗證明，添加了文字指標衡量風險揭露，對於永續經營是有所貢獻的。

In recent years, ESG ratings have been mentioned in various financial and business management articles. ESG is the abbreviation of environmental protection, social responsibility, and corporate governance. In 2004, the Who Cares Wins report released by UN Global Compact was the first of its kind to propose incorporating ESG into investment decisions. The report believed that only companies with the heart to attain ESG ratings can the companies gain a competitive advantage over peers. Amid the crisis of global pandemic and climate change, companies are starting to think about ways to grow revenue while protecting the planet. On the other hand, with the rapid development of online communities, every move of any company can be made transparent and magnified for public inspection. Any inappropriate action will influence the perception of investors and customers, thereby reducing company reputation. Therefore, sustainable development and emphasis on corporate social responsibility have become the two main core competitiveness for contemporary companies to possess. Today, large corporations are beginning to publicize ESG-related information. Studies have shown that corporations can mitigate risks by making information transparent and open.

The worsening global pandemic has made the medical care industry overloaded with increasing burden. The cost of labor and the expenses to maintain equipment operation have put the sustainable operation method to the test. This study exemplifies on an US healthcare company listed on the stock market to explore the impact of risk disclosure in its financial statement on ESG. The BERT model is used to find four textual indicators of enterprise risk. Then regression analysis is used to compare textual indicators and traditional financial indicators. Experiments show that adding textual indicators to measuring risk exposure contributes to sustainable management.

論文審定書 i
誌謝 ii
摘要 iii
Abstract iv
TABLE OF CONTENTS vi
LIST OF FIGURES viii
LIST OF TABLES ix
CHAPTER 1 Introduction 1
1.1 Research background and motivation 1
1.2 Objective of the research 2
1.3 Organization of the research 2
CHAPTER 2 Literature Review 4
2.1 ESG Criteria 4
2.2 Text mining in finance 6
2.3 Word representations by BERT 7
CHAPTER 3 Proposed research approach 11
3.1 Data collection 12
3.2 Data preprocessing 14
3.2.1 Text cleaning 15
3.2.2 Stemming 17
3.2.3 Lemmatization 18
3.2.4 Label risk categories 18
3.3 Bert word embeddings 19
3.4 ESG measurement model 21
3.4.1 Textual and financial indicators on risk 21
3.4.2 Financial control variables 22
3.4.3 Regression Analysis on panel data 24
CHAPTER 4 ESG model development 26
4.1 10-k Data Processing 26
4.2 Bert expansion on textual risk indicators 32
4.3 ESG Model results 35
4.3.1 Primary results 36
4.3.2 Advanced results with Control variables 38
4.3.3 Managerial implications 39
CHAPTER 5 Conclusions 41
5.1 Concluding remarks 41
5.2 Research limitations 42
References 43

1. Brundtland, G.H., Our common future—Call for action. Environmental Conservation, 1987. 14(4): p. 291-294.
2. Elkington, J., Accounting for the triple bottom line. Measuring business excellence, 1998.
3. Compact, U.G., Who cares wins: Connecting financial markets to a changing world. New York, 2004.
4. Amel-Zadeh, A. and G. Serafeim, Why and how investors use ESG information: Evidence from a global survey. Financial Analysts Journal, 2018. 74(3): p. 87-103.
5. Fatemi, A., M. Glaum, and S. Kaiser, ESG performance and firm value: The moderating role of disclosure. Global Finance Journal, 2018. 38: p. 45-64.
6. Kang, T., D.-H. Park, and I. Han, Beyond the numbers: The effect of 10-K tone on firms’ performance predictions using text analytics. Telematics and Informatics, 2018. 35(2): p. 370-381.
7. Mushtaq, R., et al., Do financial performance indicators predict 10-K text sentiments? An application of artificial intelligence. Research in International Business and Finance, 2022: p. 101679.
8. Pröllochs, N. and S. Feuerriegel, Business analytics for strategic management: Identifying and assessing corporate challenges via topic modeling. Information & Management, 2020. 57(1): p. 103070.
9. Shirata, C.Y., et al., Extracting key phrases as predictors of corporate bankruptcy: Empirical analysis of annual reports by text mining. Journal of emerging technologies in accounting, 2011. 8(1): p. 31-44.
10. Song, Y., H. Wang, and M. Zhu, Sustainable strategy for corporate governance based on the sentiment analysis of financial reports with CSR. Financial Innovation, 2018. 4(1): p. 1-14.
11. Devlin, J., et al., Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805, 2018.
12. Mikolov, T., et al., Distributed representations of words and phrases and their compositionality. Advances in neural information processing systems, 2013. 26.
13. Pennington, J., R. Socher, and C.D. Manning. Glove: Global vectors for word representation. in Proceedings of the 2014 conference on empirical methods in natural language processing (EMNLP). 2014.
14. Wiki. Wiki. Available from: https://en.wikipedia.org/wiki/Main_Page.
15. BooksCorpus. BooksCorpus. Available from: https://www.english-corpora.org/googlebooks/.
16. Devlin., J. BERT. Available from: https://github.com/google-research/bert.
17. Yang, R., et al., Corporate risk disclosure and audit fee: A text mining approach. European Accounting Review, 2018. 27(3): p. 583-594.
18. SEC. SEC Website. Available from: https://www.sec.gov/.
19. Chen, S.-H., X.T.K. Do, and F.-J. Hsu, ESG and Firm’s Risk-Taking: International Study.
20. Younas, Z.I. and A. Zafar, Corporate risk taking and sustainability: A case of listed firms from USA and Germany. Journal of Global Responsibility, 2019.
21. Friede, G., T. Busch, and A. Bassen, ESG and financial performance: aggregated evidence from more than 2000 empirical studies. Journal of Sustainable Finance & Investment, 2015. 5(4): p. 210-233.
22. Global, S.P. S&P Global Website. Available from: https://www.spglobal.com/esg/solutions/data-intelligence-esg-scores.
23. NLTK. Natural Language Toolkit. Available from: https://nltk.org/.
24. Buallay, A., Is sustainability reporting (ESG) associated with performance? Evidence from the European banking sector. Management of Environmental Quality: An International Journal, 2018.