| The accurate and transparent classification of sentiment in financial texts is a cornerstone of computational finance. This field is currently at a methodological crossroads, dominated by two paradigms: the fine-tuned specialist, represented by domain-adapted models like FinBERT, and the instructed generalist, embodied by modern Large Language Models (LLMs) like Google's Gemini. While performance benchmarks are emerging, a significant research gap exists in the systematic comparison of their performance trade-offs and, crucially, the nature of their explainability. This research conducts a comparative study between a fine-tuned FinBERT model and the Gemini 2.5 Pro LLM on an extended version of the Financial PhraseBank dataset. The analysis is performed along two axes: (1) Classification Performance, evaluated via metrics robust to class imbalance, and (2) Explainability, where FinBERT's predictions are analyzed using SHapley Additive exPlanations (SHAP). Furthermore, this study operationalizes a controlled test of the "overthinking" hypothesis-which suggests that complex reasoning degrades performance on subjective tasks-by comparing a two-step Separated Protocol against a single-step Simultaneous Protocol. The results reveal a nuanced verdict. While FinBERT excels in accuracy (0.85), both Gemini protocols achieve virtually identical performance, providing no evidence of performance degradation under the simultaneous protocol. Qualitative analysis uncovers two distinct reasoning styles: FinBERT's logic is bottom-up and pattern-based, excelling at domain-specific jargon, while Gemini's is top-down and conceptual. Ultimately, this work concludes that the choice between a specialist and a generalist is a strategic trade-off between accuracy, risk sensitivity, implementation cost, and the desired nature of explainability. |
*** Title, author list and abstract as submitted during Camera-Ready version delivery. Small changes that may have occurred during processing by Springer may not appear in this window.