q-fin.RM

arXiv:2412.18222v1 Announce Type: cross Abstract: With the development of the financial industry, credit default prediction, as an important task in financial risk management, has received increasing attention. Traditional credit default prediction methods mostly rely on machine learning models, such as decision trees and random forests, but these methods have certain limitations in processing complex data and capturing potential risk patterns. To this end, this paper proposes a deep learning model based on the combination of convolutional neural networks (CNN) and Transformer for credit user default prediction. The model combines the advantages of CNN in local feature extraction with the ability of Transformer in global dependency modeling, effectively improving the accuracy and robustness of credit default prediction. Through experiments on public credit default datasets, the results show that the CNN+Transformer model outperforms traditional machine learning models, such as random forests and XGBoost, in multiple evaluation indicators such as accuracy, AUC, and KS value, demonstrating its powerful ability in complex financial data modeling. Further experimental analysis shows that appropriate optimizer selection and learning rate adjustment play a vital role in improving model performance. In addition, the ablation experiment of the model verifies the advantages of the combination of CNN and Transformer and proves the complementarity of the two in credit default prediction. This study provides a new idea for credit default prediction and provides strong support for risk assessment and intelligent decision-making in the financial field. Future research can further improve the prediction effect and generalization ability by introducing more unstructured data and improving the model architecture.

arXiv:2410.11773v4 Announce Type: replace-cross Abstract: This study is the first to explore the performance of a time-series foundation model for Value-at-Risk (VaR) forecasting. Foundation models, pre-trained on vast and varied datasets, can be used in a zero-shot setting with relatively minimal data or further improved through finetuning. We compare the performance of Google's model, called TimesFM, against conventional parametric and non-parametric models, including GARCH, Generalized Autoregressive Score (GAS), and empirical quantile estimates, using daily returns from the S\&P 100 index and its constituents over 19 years. Our backtesting results indicate that in terms of the actual-over-expected ratio, the fine-tuned TimesFM model consistently outperforms traditional methods. Regarding the quantile score loss function, it achieves performance comparable to the best econometric approach, the GAS model. Overall, the foundation model is either the best or among the top performers in forecasting VaR across the 0.01, 0.025, 0.05, and 0.1 VaR levels. Fine-tuning significantly improves accuracy, indicating that zero-shot use is not optimal for VaR forecasting.

arXiv:2404.17227v2 Announce Type: replace-cross Abstract: In the rapidly evolving cryptocurrency landscape, trust is a critical yet underexplored factor shaping market behaviors and driving user preferences between centralized exchanges (CEXs) and decentralized exchanges (DEXs). Despite its importance, trust remains challenging to measure, limiting the study of its effects on market dynamics. The collapse of FTX, a major CEX, provides a unique natural experiment to examine the measurable impacts of trust and its sudden erosion on the cryptocurrency ecosystem. This pivotal event raised questions about the resilience of centralized trust systems and accelerated shifts toward decentralized alternatives. This research investigates the impacts of the FTX collapse on user trust, focusing on token valuation, trading flows, and sentiment dynamics. Employing causal inference methods, including Regression Discontinuity Design (RDD) and Difference-in-Differences (DID), we reveal significant declines in WETH prices and NetFlow from CEXs to DEXs, signaling a measurable transfer of trust. Additionally, natural language processing methods, including topic modeling and sentiment analysis, uncover the complexities of user responses, highlighting shifts from functional discussions to emotional fragmentation in Binance's community, while Uniswap's sentiment exhibits a gradual upward trend. Despite data limitations and external influences, the findings underscore the intricate interplay between trust, sentiment, and market behavior in the cryptocurrency ecosystem. By bridging blockchain analytics, behavioral finance, and decentralized finance (DeFi), this study contributes to interdisciplinary research, offering a deeper understanding of distributed trust mechanisms and providing critical insights for future investigations into the socio-technical dimensions of trust in digital economies.