The findings do not establish a general incremental forecasting advantage from transformer-based news sentiment, but highlight the importance of strong benchmarks, temporal alignment, robust inference, and careful evaluation of text-derived predictors.
Abstract
This study examines whether transformer-based financial news sentiment improves equity volatility forecasts beyond conventional market-based predictors. Using data for eleven US stocks, we compare volatility persistence models with specifications incorporating the VIX, AAII investor sentiment, and news sentiment extracted using ModernFinBERT. The forecasting target is annualized volatility constructed from future averages of trailing 252-session Yang-Zhang estimates, evaluated at horizons of 1, 5, and 22 trading days. Models are assessed through walk-forward testing with expanding and rolling estimation windows, fixed initial hyperparameter selection, and statistical comparisons accounting for multiple testing and serial dependence. Results show that a naive persistence benchmark achieves the lowest average QLIKE loss at all horizons. Although HAR-VIX produces the largest number of individual forecasting wins under expanding windows, it does not achieve the lowest mean loss. Adding text-based sentiment improves HAR forecasts in only two of 33 expanding-window comparisons, and neither improvement remains statistically significant under longer HAC bandwidths. A blinded audit of 2,700 news texts further characterizes the agreement and confidence patterns of ModernFinBERT classifications. The findings do not establish a general incremental forecasting advantage from transformer-based news sentiment. They highlight the importance of strong benchmarks, temporal alignment, robust inference, and careful evaluation of text-derived predictors.
Forecasting short-term movements in financial markets remains challenging because market prices are influenced by rapidly changing economic conditions, investor sentiment, and volatility. While machine-learning methods have demonstrated promise for financial forecasting, many forecasting tools remain difficult for non-...
C. Ibebuchi· Proceedings of the 7th Natio...· 0 citations
Stock volatility forecasting serves as a crucial underpinning for derivatives pricing, asset allocation, and risk management. This paper examines volatility proxy variables, empirical applications, as well as forecasting evaluation metrics, and then reviews representative models and empirical evidence across three rese...
Volatility forecasting plays a fundamental role in risk management, asset allocation and derivative pricing. In the Chinese equity markets, where retail participation is high and arbitrage is limited, the volatility of returns is well documented to be driven by investor sentiment. We construct and document an investor-...
Kai-Jie Liu· International Journal of Fin...· 0 citations
The application of cryptocurrency asset volatility forecasting is relevant for real‑world risk‑management procedures and derivatives pricing in the digital‑asset market. This paper assesses out-of-sample forecast performance for three model families: standalone EGARCH, vanilla Transformer, and an EGARCH‑Transformer com...
Ya-Fei Wang, Qiu-Jin Yu, Jiang-Chuan Peng et al.· Advances in Economics, Manag...· 0 citations
Previous studies link investor sentiment to momentum, yet the forecasting content of that link may vary with the sentiment measure and the market environment. We compare the same US momentum outcomes across two published sentiment indices, their available calendar windows, and the winner and loser legs, and then evalua...
Indonesian equity-market volatility reflects both macroeconomic information and investors’ psychological responses to that information. This study examines the effects of geopolitical risk, inflation, the exchange rate, and the policy interest rate on investor sentiment and IDX Composite volatility, and tests whether s...
Naurah Syifa Salsabila, E. Satoto, M. Halim· EKONOMIS Journal of Economic...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.