FinForesight: A Diagnostic Backtesting Benchmark for Deep Research Systems in Financial Forecasting

Anticipating financial outcomes requires synthesizing evolving evidence and reasoning under uncertainty, making deep research systems promising tools for forecasting. Yet evaluating forecasting ability is challenging because historical backtests may reward memorized outcomes or information retrieved after the forecast date. Live forecasting avoids direct access to realized outcomes, but feedback arrives only after events resolve. To obtain timely feedback and assess historical scores, we introduce FinForesight, a diagnostic backtesting benchmark for financial deep research systems. Across the Chinese, U.S., and Hong Kong markets, the benchmark evaluates four forecasting tasks: valuation, return ranking, event direction, and future financial fundamentals. Two capability diagnostic tasks assess counterfactual reasoning and financial calculation, scored separately from forecasting performance. Our harness-compatible framework restricts retrieved evidence to information available at the forecast date. We compare pretraining-era, resolved post-cutoff, and live evaluations to test whether controlled backtests approximate future performance. Pretraining-era scores exceed post-cutoff scores in most temporal comparisons, and stronger retrieval controls narrow the gap between post-cutoff backtests and live results. Forecasting attribution and robustness analyses reveal how scores respond to retrieval policy and how predictions respond to prompt rewrites and numerical perturbations. The code and benchmark data are publicly available.

1. 23 Oct 2026

   Sealed Before the Open: A Verifiable Standard for Measuring AI Forecasts of Financial Markets

   A transparency standard with cryptographic timestamping allows independent verification of AI forecasts before market opens, making backtests tamper-proof and comparable across models.

   SSRN

   3fanfare
2. 15 May 2024

   TKAN: Temporal Kolmogorov-Arnold Networks

   The article presents Temporal Kolomogorov-Arnold Networks (TKANs), a new neural network design that merges the benefits of Recurrent Neural Networks and Long Short-Term Memory for improved multistep time series forecasting.

   SSRN

   195cites
3. 4 Oct 2023

   Deep learning models for price forecasting of financial time series: A review of recent advancements: 2020–2022

   The article explores deep learning models for financial forecasting, discussing their structures, uses, pros and cons, and potential future research areas.

   arXivIn Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery

   188cites
4. 21 Feb 2024

   Robust agents learn causal world models

   The research suggests that intelligent agents must learn an approximate causal model to generalize to new domains, impacting fields like transfer learning and causal inference.

   Machine learningIn International Conference on Learning Representations

   108cites
5. 2 Feb 2026

   Directional-shift Dirichlet ARMA models for compositional time series with structural break intervention

   The article introduces a new Bayesian model that analyzes compositional time series data, effectively handling structural breaks and enhancing forecasting accuracy during these changes.

   arXivIn International Journal of Forecasting

   4cites
6. 12 Nov 2025

   Blameocracy: Causal Rhetoric in Politics

   U.S. Causal Rhetoric: Growing blame/credit language in congressional tweets changed donation patterns, fueled protests and polarization, and shifted public trust.

   arXiv

   1cites