An Edgeworth price cycle is cyclical pattern in prices characterized by an initial jump, which is then followed by a slower decline back towards the initial level. The term was introduced by Maskin and Tirole (1988)[1] in a theoretical setting featuring two firms bidding sequentially and where the winner captures the full market.
A price cycle has the following phases:
It can be debated whether Edgeworth Cycles should be thought of as tacit collusion because it is a Markov Perfect equilibrium, but Maskin and Tirole write: "Thus our model can be viewed as a theory of tacit collusion." (p. 592).[1]
Edgeworth cycles have been reported in gasoline markets in many countries.[2] Because the cycles tend to occur frequently, weekly average prices found in government reports will generally mask the cycling. Wang (2012)[3] emphasizes the role of price commitment in facilitating price cycles: without price commitment, the dynamic game becomes one of simultaneous move and here, the cycles are no longer a Markov Perfect equilibrium but rely on, e.g., supergame arguments.
Edgeworth cycles are distinguished from both sticky pricing and cost-based pricing. Sticky prices are typically found in markets with less aggressive price competition, so there are fewer or no cycles. Purely cost-based pricing occurs when retailers mark up from wholesale costs, so costs follow wholesale variations closely.
There is a separate literature, which has explored conditions under which price cycles like the ones observed gasoline markets and found that consumer search models can rationalize cycling under various conditions.[4][5][6] Here, the intuition is that there is a small subset of consumers that are not informed about prices and therefore will buy from a firm regardless of the price charged. Once prices get low enough, a firm may find it optimal to charge a high price and exploit this small loyal segment rather than trying to win the whole market.