Paper: arXiv 2410.11773

Authors: Anubha Goel, Puneet Pasricha, Juho Kanniainen

Abstract

This study is the first to analyze the performance of a time-series foundation AI model for Value-at-Risk (VaR), which essentially forecasts the left-tail quantiles of returns. Foundation models, pre-trained on diverse datasets, can be applied in a zero-shot setting with minimal data or further improved through finetuning. We compare Google’s TimesFM model to conventional parametric and non-parametric models, including GARCH and Generalized Autoregressive Score (GAS), using 19 years of daily returns from the SP 100 index and its constituents. Backtesting with over 8.5 years of out-of-sample data shows that the fine-tuned foundation model consistently outperforms traditional methods in actual-over-expected ratios. For the quantile score loss function, it performs comparably to the best econometric model, GAS. Overall, the foundation model ranks as the best or among the top performers across the 0.01, 0.025, 0.05, and 0.1 quantile forecasting. Fine-tuning significantly improves accuracy, showing that zero-shot use is not optimal for VaR.

Complexity vs Empirical Score

  • Math Complexity: 3.5/10
  • Empirical Rigor: 8.5/10
  • Quadrant: Street Traders — practical and empirical, lighter on theory

Why this score: This paper presents a highly novel application of pre-trained time-series foundation models (Google’s TimesFM) to Value-at-Risk forecasting. The empirical framework is rigorous, utilizing 19 years of daily data across S&P 100 constituents with extensive out-of-sample backtesting, and the authors provide open-source code for full reproducibility.

Research Flowchart

  flowchart TD
    A[Research Goal: Evaluate TimesFM for VaR Quantile Forecasting] --> B[Data Input: 19 Years Daily Returns of S&P 100 & Constituents]
    B --> C[Model Evaluation Setup: Compare Foundation Model vs GARCH/GAS]
    C --> D[Training Approaches: Zero-Shot vs Fine-Tuned TimesFM]
    D --> E[Backtesting: 8.5+ Years Out-of-Sample Evaluation at 0.01 to 0.1 Quantiles]
    E --> F[Key Outcome: Fine-Tuned TimesFM Outperforms Traditional Models]
    E --> G[Key Outcome: Fine-Tuning Superior to Zero-Shot Strategy]