RefineBridge: Generative Bridge Models Improve Financial Forecasting by Foundation Models
This work addresses financial forecasting for practitioners by enhancing foundation models, though it is incremental as it builds on existing methods.
The paper tackled the challenge of financial time series forecasting with transformer-based foundation models by proposing RefineBridge, a refinement module using a Schrödinger Bridge generative framework, which improved state-of-the-art models across multiple benchmarks.
Financial time series forecasting is particularly challenging for transformer-based time series foundation models (TSFMs) due to non-stationarity, heavy-tailed distributions, and high-frequency noise present in data. Low-rank adaptation (LoRA) has become a popular parameter-efficient method for adapting pre-trained TSFMs to downstream data domains. However, it still underperforms in financial data, as it preserves the network architecture and training objective of TSFMs rather than complementing the foundation model. To further enhance TSFMs, we propose a novel refinement module, RefineBridge, built upon a tractable Schrödinger Bridge (SB) generative framework. Given the forecasts of TSFM as generative prior and the observed ground truths as targets, RefineBridge learns context-conditioned stochastic transport maps to improve TSFM predictions, iteratively approaching the ground-truth target from even a low-quality prior. Simulations on multiple financial benchmarks demonstrate that RefineBridge consistently improves the performance of state-of-the-art TSFMs across different prediction horizons.