LLM-generated news sentiment features did not improve ETF price-forecasting model performance in this project¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly

Subclaims¶

q2 i? Across all regression models tested, models given sentiment inputs underperformed their price-only counterparts. → Freyaa Chawla 2026

Evidence¶

Freyaa Chawla 2026¶

Freyaa Chawla, Ahan Chawla, Rishi Singh, Joe Germino, Grigorii Khvatskii. (2026). Human-AI Co-Mentorship in Project-Based Learning: A Case Study in Financial Forecasting. arXiv preprint. https://arxiv.org/abs/2605.05144 doi:10.48550/arxiv.2605.05144

q2 · i? · design · r2

Comparative evaluation of regression-plus-classifier combinations reported in the results section (Table 2 reports MSE and MAE for each combination). The article states that in all cases sentiment-including models "underperformed when compared to their counterparts that had no access to the sentiment information." No effect size is printed.

"Additionally, it should be noted that in all cases, the models that included the sentiment as one of their inputs underperformed when compared to their counterparts that had no access to the sentiment information."

Discussion¶

Related Claims¶

• SVM-based regressors outperformed other regressors in the ETF price-delta forecasting benchmark — related
• Sentiment scores of LLM-generated replies differ significantly from human replies under both classifiers — related
• Counterfactual fine-tuning reduces sentiment bias in LLM-generated forum replies — related