Tag
This paper examines how sentiment arcs in ECB and Fed press conferences predict policy rate changes and inflation expectations, showing that the sequencing of sentiment carries significant policy signals.
Central bankers are discussing a future where AI better understands monetary policy than humans, potentially requiring the Federal Reserve to adjust communication to avoid exploitation by trading agents.
This paper introduces LabelFusion-TS, which fuses a fine-tuned RoBERTa encoder, a prompted LLM, and time-series transformers over market data to classify Federal Reserve communication as hawkish, dovish, or neutral. The fused system achieves 70.2% weighted F1, outperforming a zero-shot LLM and showing early evidence that market time series help financial text classification.