From Numbers to Words: Multi-Modal Bankruptcy Prediction Using the ECL Dataset
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917572749819904 |
|---|---|
| author | Arno, Henri Mulier, Klaas Baeck, Joke Demeester, Thomas |
| author_facet | Arno, Henri Mulier, Klaas Baeck, Joke Demeester, Thomas |
| contents | In this paper, we present ECL, a novel multi-modal dataset containing the textual and numerical data from corporate 10K filings and associated binary bankruptcy labels. Furthermore, we develop and critically evaluate several classical and neural bankruptcy prediction models using this dataset. Our findings suggest that the information contained in each data modality is complementary for bankruptcy prediction. We also see that the binary bankruptcy prediction target does not enable our models to distinguish next year bankruptcy from an unhealthy financial situation resulting in bankruptcy in later years. Finally, we explore the use of LLMs in the context of our task. We show how GPT-based models can be used to extract meaningful summaries from the textual data but zero-shot bankruptcy prediction results are poor. All resources required to access and update the dataset or replicate our experiments are available on github.com/henriarnoUG/ECL. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_12652 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | From Numbers to Words: Multi-Modal Bankruptcy Prediction Using the ECL Dataset Arno, Henri Mulier, Klaas Baeck, Joke Demeester, Thomas Computational Engineering, Finance, and Science Computational Finance In this paper, we present ECL, a novel multi-modal dataset containing the textual and numerical data from corporate 10K filings and associated binary bankruptcy labels. Furthermore, we develop and critically evaluate several classical and neural bankruptcy prediction models using this dataset. Our findings suggest that the information contained in each data modality is complementary for bankruptcy prediction. We also see that the binary bankruptcy prediction target does not enable our models to distinguish next year bankruptcy from an unhealthy financial situation resulting in bankruptcy in later years. Finally, we explore the use of LLMs in the context of our task. We show how GPT-based models can be used to extract meaningful summaries from the textual data but zero-shot bankruptcy prediction results are poor. All resources required to access and update the dataset or replicate our experiments are available on github.com/henriarnoUG/ECL. |
| title | From Numbers to Words: Multi-Modal Bankruptcy Prediction Using the ECL Dataset |
| topic | Computational Engineering, Finance, and Science Computational Finance |
| url | https://arxiv.org/abs/2401.12652 |