Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
Standard next-token fine-tuning often fails on financial OHLCV data because individual price increments have minimal signal-to-noise ratio, whereas higher-order temporal dynamics like volatility clustering retain exploitable structure. To address this, the authors introduce Volatility-Clustering Adaptation (VCA), a fine-tuning framework that regularizes the cross-entropy objective with a differentiable penalty matching the autocorrelation of squared returns across multi-step autoregressive rollouts to realized targets. Across three diverse asset benchmarks, VCA consistently outperforms pre-trained base models and standard fine-tuning under primary evaluation protocols, driven largely by a marked reduction in variance error.
Standard next-token adaptation fails on financial time series unless fine-tuning objectives explicitly penalize rollouts for distorting higher-order temporal structures like volatility clustering.
Time-series foundation models are increasingly adapted to new domains through fine-tuning on target data, under the implicit assumption that more target data yields better forecasts. We show that this assumption can fail in financial forecasting, where individual price changes are difficult to predict, but large moves tend to cluster, creating alternating calm and turbulent periods. Using financial foundation models trained on price bars of open, high, low, close, and volume, we argue that adapting to financial domains requires training signals beyond next-token prediction. We introduce Volatility-Clustering Adaptation (VCA), which augments next-token cross-entropy with a differentiable penalty on the autocorrelation of squared returns, the standard statistical signature of volatility clustering. This additional objective provides a multi-step training signal by matching the resulting dependence structure of autoregressive rollouts to those of the realized future. Across three asset sets and two evaluation conventions, VCA improves adaptation over the pre-trained model, with the strongest gains under the primary evaluation (\textsc{fore}), driven primarily by reduced variance error. Overall, our results suggest that effective financial adaptation requires objectives that capture domain-specific temporal structure beyond token-level prediction.