How Do Standard Time Series Forecasting Benchmarks Translate to Finance? A Broad Study of Risk Adjusted Performance
Year: 2026
Type: article
Abstract: Forecasting financial asset returns is challenging due to low signal‐to‐noise ratios, nonstationarity and heavy‐tailed distributions. We benchmark 16 deep learning architectures for time series forecasting, including Transformers, State Space Models (SSMs), xLSTMs, and recurrent hybrids, on a 15‐year daily futures dataset using an end‐to‐end volatility‐targeted Sharpe ratio optimization task. Our results challenge prevailing assumptions in the literature. Several widely used “state‐of‐the‐art” architectures, including generic Transformers and SSMs, underperformed a simple AR(1) benchmark in a realistic financial forecasting task. The strongest results came from comparatively simple hybrid approaches using recurrent structures: a Variable Selection Network combined with an LSTM achieved the highest Sharpe ratio, while xLSTM showed the greatest robustness to transaction costs. These findings suggest that success on standard time‐series benchmarks does not reliably transfer to finance, and that architectures tailored to the unique properties of asset returns may remain more effective than generic attention‐based approaches for financial forecasting.
Source: Journal of Forecasting
DOI: 10.1002/for.70233
Institution: University of Oxford
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: hybrid
Estimated APC paid: $3,400
OpenAlex ID: W7133307728
