AbstractThis study explores the dynamic relationship between economic news sentiment and the US stock market using a non‐linear empirical framework. The analysis focuses on both sectoral indices and the aggregate stock market index from November 2011 to November 2021. Using causality tests and a rolling window detrended cross‐correlation coefficient, the study reveals several key findings. First, the causal effect of investor sentiment on sectoral returns varies over time, with each sector responding differently. Second, while no evidence of dependence exists for time scales less than 2 months, a positive relationship emerges for time scales greater than 6 months, except for the utilities sector, which is found to be negative. Third, the study shows that the relationships between all pairs of variables are time‐dependent. Finally, economic news sentiment might have a varying impact on market inefficiency over different periods, making it challenging to predict market behaviour based on sentiment data.