cs.AINov 6, 2026

TailDART: Compute-Efficient Foundation Models for Financial Time-Series Forecasting

Authors: Chiang Qin Zhi, Ranjan Satapathy, Erik Cambria, Joyjit Chattoraj

Organizations: Nanyang Technological University Singapore · Agency for Science, Technology and Research Singapore

Abstract

Although time-series foundation models (TSFMs) achieve strong zero-shot transferability across high-information-density physical systems, applying overparameterized TSFMs to quantitative finance is fundamentally constrained by low signal-to-noise ratios, non-stationarity, and prohibitive autoregressive decoding latency. To reconcile representation learning with institutional operational constraints, this study presents TailDART, a dense 2.68 million-parameter foundation model engineered for financial time-series forecasting. By integrating grouped-query attention into a lightweight backbone, the architecture drastically reduces Key-Value cache memory bandwidth overhead, achieving major inference acceleration (up to 48.1%) over prevailing state-of-the-art TSFMs to enable low-latency, high-throughput streaming execution. Furthermore, multi-asset benchmarks across varying temporal resolutions show that a compact model can remain competitive for market forecasting: TailDART matches the predictive fidelity of substantially larger baselines while demonstrating superior error resilience in high-volatility regimes.

Figures & tables

Explore similar work

CardsList
  1. Re(Visiting) Time Series Foundation Models in Finance

    Nov 23, 2025Eghbal Rahimikia, Hao Ni, Weiguan WangTime Series ForecastingTime-Series Foundation Models

  2. EXAONE Forecast for Finance

    Sep 7, 2026Seunghan Lee, Jaehoon Lee, Jun Seo +9Multivariate Time Series ForecastingFinancial Forecasting

  3. KiT: A Foundation Model for Financial Time-Series Forecasting using DiffusionTransformers

    Sep 28, 2026Boyu Zhang, Haorui LiDiffusion TransformerFinancial Forecasting