Stories that (are) Move(d by) Markets: A Causal Exploration of Market Shocks and Semantic Shifts across Different Partisan Groups

Abstract

Macroeconomic fluctuations and the narratives that shape them form a mutually reinforcing cycle: public discourse can spur behavioural changes leading to economic shifts, which then result in changes in the stories that propagate. We show that shifts in semantic embedding space can be causally linked to financial market shocks -- deviations from the expected market behaviour. Furthermore, we show how partisanship can influence the predictive power of text for market fluctuations and shape reactions to those same shocks. We also provide some evidence that text-based signals are particularly salient during unexpected events such as COVID-19, highlighting the value of language data as an exogenous variable in economic forecasting. Our findings underscore the bidirectional relationship between news outlets and market shocks, offering a novel empirical approach to studying their effect on each other.

Figures (6)

• Figure 1: The proportion of significant pairs for when {$y=\text{text}$, $x=\text{econ}$} and {$y=\text{econ}$, $x=\text{text}$}. Significant proportions are marked with a star.
• Figure 2: Significant proportion for different market shocks for each political groups across all lags. An * denotes a significant effect inline with \ref{['sec:stat_test']} and \ref{['sec:p_hack']}, and the shade of blue represents the size of the group effect, with darker colours pertaining to larger effects.
• Figure 3: Deviation in MSE improvement of the enhanced model over the base model, $\Delta_{v} - \overline{\Delta}$, for each test window. Different orientations are indicated with the following colors: left, right, center, and all.
• Figure 4: The proportion of significant pairs when {$y=\text{text}$, $x=\text{econ}$} and {$y=\text{econ}$, $x=\text{text}$} for all news outlets. Comparing the Non-linear and Linear.
• Figure 5: Proportion of significant pairs where {$y=\text{text}$, $x=\text{econ}$} that have a confounding variable by orientation. The category of the confounding variables are denoted by the colour of the bar section.