Multi-Modal Deep Learning for Credit Rating Prediction Using Text and Numerical Data Streams

Paper: arXiv 2304.10740

Authors: Mahsa Tavakoli, Rohitash Chandra, Fengrui Tian, Cristián Bravo

Abstract

Knowing which factors are significant in credit rating assignment leads to better decision-making. However, the focus of the literature thus far has been mostly on structured data, and fewer studies have addressed unstructured or multi-modal datasets. In this paper, we present an analysis of the most effective architectures for the fusion of deep learning models for the prediction of company credit rating classes, by using structured and unstructured datasets of different types. In these models, we tested different combinations of fusion strategies with different deep learning models, including CNN, LSTM, GRU, and BERT. We studied data fusion strategies in terms of level (including early and intermediate fusion) and techniques (including concatenation and cross-attention). Our results show that a CNN-based multi-modal model with two fusion strategies outperformed other multi-modal techniques. In addition, by comparing simple architectures with more complex ones, we found that more sophisticated deep learning models do not necessarily produce the highest performance; however, if attention-based models are producing the best results, cross-attention is necessary as a fusion strategy. Finally, our comparison of rating agencies on short-, medium-, and long-term performance shows that Moody’s credit ratings outperform those of other agencies like Standard & Poor’s and Fitch Ratings.

Complexity vs Empirical Score

• Math Complexity: 3.5/10
• Empirical Rigor: 6.5/10
• Quadrant: Street Traders — practical and empirical, lighter on theory

Why this score: The paper provides a practical and thorough comparative analysis of multi-modal deep learning architectures for credit rating prediction without relying on heavy mathematical proofs. Its empirical focus on data fusion techniques (early vs. intermediate, concatenation vs. cross-attention) offers useful applied insights, though novelty is somewhat limited by relying on standard neural network building blocks.

Research Flowchart

  flowchart TD
    A[Research Goal: Optimize Multi-Modal Deep Learning for Credit Rating Prediction] --> B[Data Collection: Structured Numerical & Unstructured Text Streams]
    B --> C[Model Exploration: CNN, LSTM, GRU, BERT]
    C --> D[Data Fusion Strategies: Early vs. Intermediate Fusion via Concatenation & Cross-Attention]
    D --> E[Agency Comparison: Evaluate Moody's, S&P, and Fitch Across Time Horizons]
    E --> F[Key Outcome 1: CNN-based Multi-Modal Model Outperforms Other Architectures]
    E --> G[Key Outcome 2: Cross-Attention Essential for Attention Models; Simpler Models Often Win]
    E --> H[Key Outcome 3: Moody's Ratings Predictability Outperforms S&P and Fitch]