Skip to main content
eScholarship
Open Access Publications from the University of California

UC Berkeley

UC Berkeley Electronic Theses and Dissertations bannerUC Berkeley

Three Essays on Retail Price Dynamics

Abstract

This dissertation consists of three essays on the dynamics of retail prices.

The first chapter uses a novel data set of weekly-sampled store-level retail prices for narrowly defined goods observed across 12 cities in Mexico to study the relative magnitude of aggregation biases in estimates of convergence to the Law of One Price (LOP). I find that temporal aggregation can severely bias estimates of persistence in relative prices. Both panel estimations of higher-order autoregressive processes and Monte Carlo experiments suggest that using quarterly aggregated data (from weekly-sampled data) can overestimate the half-life of deviations from the Law of One Price (LOP) by a factor of 4. I do not find evidence that pooling across goods with heterogeneous dynamics biases persistence estimates. The analysis also suggests that intercity prices converge rapidly to the LOP in an absolute sense (the median half-life is estimated at 3 weeks) and the existence of only a weak association between price gaps across cities and physical distance.

The second chapter studies patterns of retail price adjustment at the store level using a unique scanner data set of weekly retail prices, quantities sold and wholesale costs for a cross-section of retailers in Chile. In line with evidence reported for the U.S. (Eichenbaum, Jaimovich and Rebelo, 2010; Klenow and Malin, 2010), posted prices tend to revolve around more persistent reference prices. The implied duration of reference prices is estimated at 2-3 quarters versus 3-4 weeks in the case of posted prices. I find strong evidence that reference prices respond to retailer-level shocks. Comovement in the reference price of a given barcode across retailers is found to be significantly larger for stores belonging to the same retail chain than for stores that belong to different retail chains. Furthermore, most of the variation in the frequency of reference price adjustment is explained by "chain effects". Evidence on the synchronization of price changes suggests that price changes tend to be staggered across stores belonging to different retail chains but synchronized within chains.

The third chapter uses a scanner dataset including weekly prices and costs from a large retailer in Chile to study the relationship between price rigidities and intra-national deviations from the law of one price (LOP). I find that, controlling for transportation costs (proxied by distance), more flexible prices are associated with a larger volatility of deviations from the LOP. The effect is econominally non-negligible and holds for both retail- and wholesale-level prices. The distance equivalent of a 0.01 change in the frequency of retail (wholesale) price change is estimated at 370 (294) kilometers.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View