Overview

This concluding chapter connects a calibrated Federal Reserve model of endogenous planning with Bank of Italy evidence from Italian firms with at least 50 employees, 2017Q3–2024Q4, without treating them as the same measurement. Together they suggest that persistent or unusual shocks can make inflation sensitivity state dependent, while signal weighting, market position, uncertainty, margins, liquidity, and information shape how cost surprises reach prices and expectations. In the 2022–23 Italian episode, estimated input-cost channels account for meaningful portions of expected and realized price growth but not the whole episode. The model matches a correlation near 0.43 with 0.44 in its baseline yet explains only about 14% of measured uncertainty volatility. The result is a conditional framework, not a universal forecast or complete theory.

Learning outcomes

The model and firm survey connect shocks to inflation through decisions, but they measure different objects and cannot be substituted quantitatively.
In 2022–23, estimated input-cost channels accounted for meaningful portions of expected and realized price growth, but not the whole episode.
Communication can reduce expectation spillovers, while model scope, survey identification, and changing shock volatility keep the overall account partial.

Keywords

FirmsInflationPlanningPass-throughCommunicationUncertainty

What Firms’ Planning Can—and Cannot—Explain About Inflation

The subject is the price-setting firm confronting a shock whose meaning changes with time, information, and financial capacity. The final question is what follows when the value of looking ahead changes at the same moment that firms are deciding whether to pass a cost through to prices. The evidence supports a coherent but bounded judgment: inflation can become more sensitive and persistent when firms’ planning and signal-processing responses adapt to the state, but neither a calibrated model nor a firm-level survey supplies a universal forecast or a complete account of inflation.

The common causal account

The strongest connection between the two papers is not that they measure the same variable. It is that both place firms’ decisions between an economic disturbance and the aggregate inflation outcome. The Federal Reserve model asks how far a price-setting firm evaluates future contingencies before setting a reset price. The BIS study observes how an unexpected input-cost movement is followed by changes in firms’ own prices, their inflation expectations, and their broader economic outlook. One supplies a mechanism for state-dependent planning; the other shows that the response to a cost surprise differs by direction, regime, and firm condition.

The combined logic begins with the state of the firm and the economy, not with a fixed pass-through coefficient. A shock has a size, a sign, and a persistence. It also arrives against a background of longer-run beliefs. In the model, a large or persistent disturbance can make the consequences of a short planning window more important. The firm is then more willing to pay the incremental cognitive cost of evaluating another period. If the selected horizon lengthens, inflation’s loading on current demand and supply disturbances rises, while the relative weight of the continuation-value reference falls.

That chain explains how a private decision can create endogenous inflation uncertainty. The underlying shock variance need not change. A state-dependent horizon can change the sensitivity of prices to the same innovations, so the conditional variance of next-period inflation moves with the state. Forecast revisions and uncertainty can rise together because the disturbance both changes the outlook and changes the value of looking further ahead. A fixed horizon cannot generate that second movement in sensitivity, even though finite fixed planning can still produce predictable forecast errors.

The empirical evidence adds an important qualification to this mechanism. An input-cost surprise does not have one inevitable route into prices. In the Italian data, the baseline standardized relationship is about 0.299 percentage points of expected own-price growth and 0.187 points of realized own-price growth for a one-point input-price surprise, with the realized outcome measured over the following year. The response is incomplete and delayed. A firm can revise its expected price immediately but wait, adjust partially, or absorb some of the cost before the realized price changes.

The direction of the surprise matters. Positive input-price forecast errors have estimated coefficients of about 0.362 for expected own-price growth and 0.228 for realized own-price growth, compared with about 0.185 and 0.102 for negative errors. Cost increases therefore pass through much more strongly, while the accurate claim about declines is that their own-price effects are smaller, not that all firms never reduce prices. For aggregate inflation expectations, the asymmetry is sharper: positive cost surprises raise short-term expectations, whereas negative surprises do not produce a comparable statistically significant revision. This creates a route through which an upward shock can leave more of a near-term inflation imprint than a downward movement of similar size.

The two forms of asymmetry should remain separate. The model’s clean analytical asymmetry comes from a moving trend-inflation reference point: with positive trend beliefs and stationary demand, a negative disturbance can be farther from the reference than an equally sized positive one. The BIS asymmetry is estimated from positive and negative input-price surprises and is interpreted through several possible mechanisms, including state-dependent adjustment, market power, strategic complementarities, and financial constraints. The sources support comparison, not the claim that one mechanism has been proven to generate both patterns.

Published by DeepMox

FAQ

Are the planning model and the Italian pass-through study measuring the same thing?

No. The model selects a planning horizon in a calibrated partial-equilibrium setting, while the Italian study estimates how input-price surprises relate to firms’ prices and expectations. They are conceptually complementary but not quantitatively interchangeable.

What did input costs explain during Italy’s 2022–23 inflation episode?

They explained a meaningful portion of firms’ price growth, not the whole episode. The study attributes more than 2 percentage points of peak expected sales-price growth to input-price expectations and roughly 2 of about 6 percentage points of peak realized growth to the input-cost channel.

Can this framework forecast the next inflation episode?

No. It organizes conditional channels—shock persistence, learned beliefs, signal weighting, and firm constraints—rather than producing a point forecast. The sources do not identify universal planning quarters or a complete theory of inflation.

Previous13 minItalian Firms’ Input-Cost Pass-Through

All chaptersBack to Topic