News narratives on supply shocks lift inflation risk premia
An ECB working paper shows that inflation risk premia in the euro area and the US are inversely related to media narratives attributing inflation to demand versus supply drivers. Using text from 2005 to 2024, the study finds supply narratives consistently elevate premia.
Commodities dominate narrative channels
The researchers extract causal inflation narratives from the Financial Times and Wall Street Journal between 2005 and 2024 using natural language processing.
The central metric, NetDemand, measures the volume difference between articles attributing inflation to demand triggers versus supply shocks.
Across two-, five-, and ten-year maturities, inflation risk premia fall as demand narratives dominate and rise when supply narratives prevail.
This inverse relationship holds after controlling for composite PMI and the VIX.
Breakdown by narrative type shows that supply-side drivers, particularly commodity and energy shocks, exert the strongest upward pressure on premia in both currency areas.
Robust across models and volatility
Asset pricing theory indicates that risk aversion and inflation volatility amplify premia dynamics.
Interacting NetDemand with variance risk premia in the US and option-implied inflation volatility in the euro area significantly strengthens the estimated effects at short and medium horizons.
Furthermore, the news-based measure retains explanatory power when tested against macroeconomic shock decompositions, survey forecasts from professional economists, and GPT-based sentiment trackers.
High frequency beats black boxes
Targeted causal extraction outperforms noisy large language models by isolating genuine economic narratives at daily frequency.
Yet media metrics remain prone to reporting bias during acute commodity shocks.
Policymakers gain a useful real-time sentiment tracker, though it cannot replace structural model identification.
Source: Inflation narratives and risk premia
IN: