Later peak season for intermodal several years in the making

Later peak season for intermodal several years in the making

Holiday-related freight is moving later in the year, with the rise of e-commerce playing a primary role in the change. Photo credit:

Peak season 2020 was very unusual. Intermodal volumes reached “peak” levels in October and then never really declined. But was 2020 so unique? Peak season has indeed changed. The peak is arriving later in the year. But the change began some years ago.

The Association of American Railroads (AAR) puts out intermodal volume data every week. This allows us to get a very granular picture of the timing of intermodal volume. It allows us to ask the question: What was the biggest week of the year in terms of intermodal volume? The first chart looks at AAR data over the last 15 years, and denotes for each year since 2005 when the biggest week occurred.

For the first 10 years peak season was quite consistent. The peak week generally stayed within a narrow range, taking place in the Week 37 to Week 39 time frame. Although the calendar varies a bit from year to year, these weeks would generally fall into the second half of September or the first week of October. There were only a few outliers. 2009 stands out (Week 46), but of course that year was warped by the onset of the Great Recession and therefore can be set aside for the purpose of this analysis.

What does seem clear is that things changed beginning in 2016. Since then, peak week has generally taken place in the Week 48 to Week 50 time frame. This is early December. The only exception in recent years was 2019, when peak season was particularly anemic. The busiest week in 2019 was Week 2, a unique occurrence during the 15 years covered by this analysis.

There is a potential problem with looking at only the busiest single week. The results might be skewed by one-off occurrences that might have created a brief surge. What if we look at a longer period? For instance, what if we identify the busiest four-week period each year? Does that change the conclusion? The answer, largely, is no. The second chart plots when the end of the busiest four weeks of each year occurred. The message still holds.

The period from 2005 to 2015 was remarkably consistent. With the sole exception of 2005, the busiest four weeks concluded with either Week 39 or Week 40. But the pattern began to break down in 2016 and the peak four weeks have generally occurred later.

Peak season becoming less ‘peaky’

It thus seems clear that holiday-related freight is moving later. Why? Certainly, the rise of e-commerce has played a role in the change. The ease of shopping online and having goods delivered to your door while still getting a good (or great) price is allowing consumers the luxury of avoiding holiday crowds and shopping later. Plus, lean supply chain principles are discouraging early accumulation of inventory as retailers strive to match the buyers’ calendar.

At the same time, it appears that peak season has been getting less “peaky”. The third chart compares the weekly average volume during each year’s four-week peak with the average weekly volume for the year. Generally, the magnitude of the peak has been declining since 2010. The magnificent exception is the year just past. The average weekly volume was adversely affected by the pandemic lockdown in the first half, so the big year-end surge was much larger in comparison. But similar to the Great Recession, this must be regarded as a statistical outlier.

The problem with a later peak is that it leaves precious little room for recovery. When things go wrong or the unexpected occurs, will the system be agile enough to avoid stock-outs and disappointment? Often, it appears, the answer is “no”.

Contact Larry Gross at