Asymmetric price cycles which look similar to Edgeworth Cycles are appearing in increasingly many retail gasoline markets in the U.S. and worldwide. The cycles can give the appearance of asymmetric prices responses to cost shocks under traditional methodologies. This article shows how to remove the confounding effect of the cycles and test for any true underlying asymmetry in price responses. Designing the correct counterfactual is key. The methodology is demonstrated for one strongly cycling market and some asymmetry to cost shocks is found. Covert collusion is unlikely, but the ability to coordinate cyclical price increases may play a role. Consumers can still reduce xpenditures on gasoline up to 7.7% with simple timing rules of thumb.

This paper examines dynamic pricing behavior in retail gasoline markets for 19 Canadian cities over 574 weeks. I find three distinct retail pricing patterns: 1. cost-based pricing, 2. sticky pricing, and 3. steep, asymmetric retail price cycles that, while seldom documented empirically, resemble those of Maskin & Tirole[1988]. Using a Markov switching regression, I estimate the prevalence of patterns and the structural characteristics of the cycles. Retail price cycles prevail in over 40% of the sample. I show they are more prevalent when there is a greater penetration of small firms. The cycle is accelerated and amplified with very many small firms. In markets with few small firms, sticky pricing is dominant. My findings is consistent with the theory of Edgeworth Cycles.

In this article, I exploit a new station-level, twelve-hourly price dataset to examine the strong retail price cycles in the Toronto gasoline market. The cycles are visually similar to the theoretical Edgeworth Cycles of Maskin & Tirole [1988]: strongly asymmetric, tall, rapid, and highly synchronous across stations. I test a series of predictions made by the theory about how firm behaviors would differentially evolve over the path of a cycle. The evidence is consistent with the existence of Edgeworth Cycles and inconsistent with competing hypotheses. One finding is that smaller firms are more likely than larger firms to initiate rounds of price undercutting but the reverse is true for rounds of price increases. While the cycles are an interesting phenomenon for study in their own right, the evidence has important implications for understanding market power in both cycling and non-cycling gasoline markets.

Retail gasoline prices are known to respond fairly slowly to wholesale price changes. This does not appear to be true for markets with Edgeworth price cycles. Recently, many retail gasoline markets in the midwestern U.S. and in other countries have been shown to exhibit price cycles, in which competition generates rapid cyclical retail price movements. We show that cost changes in cycling markets are passed on 2 to 3 times faster than in markets without cycles. We argue that the constant price movement inherent within the Edgeworth cycle eliminates price frictions and allows firms to pass on cost fluctuations more easily.

We analyze the effect of Wal-Mart's entry into the grocery market using a unique stor-level price panel data set. We use OLS and two IV specifications to estimate the effect of Wal-Mart's entry on competitors' prices of 24 grocery items across several categories. Wal-Mart's price advantage over competitors for these products averages approximately 10%. On average, competitors' response to Wal-Mart's entry is a price reduction of 1-1.2%, mostly due to smaller-scale competitors: the response of the "big three" supermarket chains (Alberson's, Safeway, and Kroger) is less than half that size. We confirm our results using a falsification exercises, in which we test for Wal-Mart's effect on prices of services that it does not provide, such as movie tickets and dry cleaning services.

Motivated by the discovery of apparent Edgeworth Cycles in many retail gasoline markets, this paper extends the Maskin & Tirole [1988] theory Edgeworth Cycles to a wide range of more complicated and realistic settings. Taking a computational approach to search for Markov Perfect Equilibria, I examine models involving duopoly and triopoly, differentiation, capacity constraints, and different sharing rules, discount factors and initial beliefs about price leading behavior. I find Edgeworth Cycles equilibrium in many scenarios outside the homogenous-good Bertrand mold. Cycle characteristics and average markups depend on the scenario.