Academic journal article Journal of Money, Credit & Banking

Linking Individual and Aggregate Price Changes

Academic journal article Journal of Money, Credit & Banking

Linking Individual and Aggregate Price Changes

Article excerpt

Especially in countries having adopted some definition of price stability as the main focus of their monetary policy, policymakers are seeking to possess advance knowledge of forthcoming price changes. Business analysts trying to project real returns on investment in financial assets are also keen to learn about the evolution of inflation. As price aggregates tend to exhibit a substantial degree of inertia, price dynamics at short horizons is of particular interest. (1) Despite its central importance for policy and business, however, characterizing the nature of short-term variation in aggregate price changes has been a daunting task for macroeconomists. While a vast amount of evidence has been amassed to address the issue, mostly cast in univariate time-series or Phillips curve-type models and drawing on national or sector level data, recently documented empirical regularities question the merit of traditional approaches. (2)

This paper takes a step toward examining inflation dynamics from a hitherto unexplored perspective. It starts out from direct microeconomic evidence documenting that store level price sequences exhibit periods of inaction followed by intermittent, discrete adjustments and that lumpiness is coupled with heterogeneity in the timing of price changes, especially across different stores. (3) This description of pricing behavior points decisively to the existence of fixed costs in re-pricing. The consequences of fixed costs in microeconomic price setting are formally examined in (S,s) pricing models, originally devised to provide microeconomic foundation for business cycle analysis. (4) Are, however, these models, through an explicit aggregation of intermittent and heterogeneous individual pricing actions able to carry implications for a better understanding of aggregate price dynamics? By setting up an empirical model of microeconomic price setting rooted in (S,s) pricing theory and applying the model to a unique, highly disaggregated panel of retail prices, this paper provides evidence for an affirmative answer.

Specifically, the paper develops a microeconomic model of price setting that conforms to basic features of retail pricing policies such as lumpiness and staggering, and allows for persistence in pricing shocks. The central object in the model is the price deviation, the latent, store-, product- and time-specific, potentially non-zero log difference between the actual and the target price. Two-sided (S,s) theory suggests that the actual pricing action is sparked off when the price deviation gets sufficiently eroded to surpass one of the non-adjustment boundaries. As a result, the net number of stores adjusting price in a particular direction is determined by the relative bunching of price deviations at the two ends of the price deviation density. The reasoning implies that the shape of the cross-sectional density of price deviations is likely to contain information of aggregate price changes.

The proposed model is applied to the study of inflation in a particular sector of the economy, processed meat products. (5) While the data analysis based on a unique sample of store level prices has limitations in its size and scope, the underlying model is potentially useful for studying any component of the aggregate price index reflecting market-based prices. The main findings of the empirical analysis are as follows. First, the shape of the cross-sectional density of estimated price deviations conveys extra information over simple linear models of inflation. Asymmetry in the density particularly matters. Second, the presence of store- and product-specificity in pricing shocks alters the size, but not the direction of changes in inflation. Finally, the empirical framework based on two-sided (S,s) decision rules is superior in explaining inflation dynamics, relative to naive ones, in which the microeconomic price deviation is measured relative to the cross-sectional mean of prices. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.