Information frictions cause variable monetary policy lags

A new SNB working paper by Diego M. Hager and Samuel Reynard introduces a microfounded information mechanism explaining variable monetary policy transmission lags.

Firms, even when rational and without exogenous price change costs, exhibit endogenous price stickiness.

This arises because they form two distinct expectations: a forecast of future demand and a nowcast of the current unobserved state, as information arrives at different frequencies.

The model, a continuous-time optimal stopping problem, shows firms receive continuous noisy signals and discrete precise signals.

An endogenous inaction region emerges, causing rational delays in price adjustments until discrete signals provide sufficient actionable information.

This framework reproduces observed Swiss price-adjustment patterns and variable transmission lags, providing a rational-agent microfoundation for lag heterogeneity.

Macroeconomic implications reveal interest rate shocks accelerate medium-term inflationary pressure transmission, while elevated prediction uncertainty dampens it.

Empirical validation with Swiss data supports these findings, suggesting rational firms have limited insight into broader economic interrelations.

The paper's central theoretical contribution maps price-adjustment decisions under informational frictions into a real-options framework, providing a rational-agent foundation for price stickiness.

Rigidities emerge intrinsically from the sequential interaction of forecasting and nowcasting, not from exogenous menu costs or information-acquisition costs.

Firms optimally maintain a stable hazard rate of price changes, preferring fixed-interval repricing, but unexpected high-frequency information triggers unscheduled revisions.

This approach builds on canonical partial-equilibrium frameworks and connects to noisy- and sticky-information literature as well as models of Knightian uncertainty.

It distinguishes itself by abstracting from specific demand forms or utility assumptions, demonstrating that macrolevel rigidities arise solely from dual-frequency expectation formation.