Retail Gasoline Price Cycles: Further Evidence from Guelph, Ontario Using Bi-Hourly, Station-Specific Retail Price Data
A short note to test a theoretical prediction not tested in previous research published by one of your Ride On writers
While one of your Ride On writers was going through his electronic files to do research for a new series on retail gasoline pricing, he came across a note he wrote in 2013 using data collected for his PhD thesis in late 2005. Since this note was not published anywhere, we decided to start this series by publishing it here. It is a lot more technical than our usual writings, but we hope you find it valuable nonetheless, particularly given all of the talk in the media lately about rising gasoline prices.
1. INTRODUCTION
In several cities in North America, Australia and Europe, it has been documented that retail gasoline prices appear to move in cycles where they rise by large amounts in one or two days, followed by several days of small consecutive price decreases. Some people cite these cycles as evidence of anti-competitive behaviour, while others argue they are an indication of intense competition. Therefore, a clear understanding of these cycling dynamics is important for both economists and policymakers.
The Edgeworth cycle theory, which will be reviewed in Section 2, is often used to explain these cycles. This theory (and extensions of it) has been tested by authors using various data sets. For example, Atkinson (2009) tests several predictions of the Edgeworth cycle theory using bi-hourly retail prices gasoline, which were observed at 27 gasoline stations in Guelph, Ontario for 103 days in late-2005. The author also uses wholesale (rack) prices obtained for London, Ontario to approximate marginal costs. However, there remains one key prediction of the basic theory that was not tested by Atkinson (2009), so the purpose of this note is to test it using these same data.
To summarize the results of this note, it is found the Guelph data appear to be consistent with a variant of a prediction of the basic Edgeworth cycle theory, which states average daily price decreases will be non-decreasing in size when prices are below an optimal peak. However, prices seem to fall by greater amounts when they are above this peak; this result is expected from the theory, but it has not been found in previous studies of retail gasoline price cycles.
This note proceeds as follows. First, Section 2 reviews the relevant theoretical and empirical literature. The data and empirical strategy are then introduced in Section 3. Section 4 presents the results of the empirical analysis, and Section 5 concludes.
Keep reading with a 7-day free trial
Subscribe to Ride On: The Drive for Better Transportation Systems to keep reading this post and get 7 days of free access to the full post archives.