Commentary: There are too many freight cars

The views expressed here are solely those of the author and do not necessarily represent the views of FreightWaves or its affiliates. 

My thesis is that the North American railcar fleet is too large. What makes it too large?

Reasonable people can agree that with almost one-quarter of the available and usable cars in storage, that’s not a sustainable business model.

One of the three largest fleet owners supplying leased cars for Mexico, Canada and the United States last week stated flatly that perhaps 15% to 18% excessive “market supply” is challenging.

This old economist goes a bit further. A best practice evaluation might see no more than 8% to 12% as the desired excess railcar supply.

(Photo credit: Jim Allen/FreightWaves)

Why so low? Two reasons. The first – any of the railroads wants to have enough added rolling stock to accommodate preventative railcar maintenance and periodic heavy car re-building. Statistically – given the vast distances across the North American continental track network – that’s probably in the 5 to 7% of fleet size range.

Secondly, if the long-term business outlook is optimistic about traffic growth, then a railroad needs to have a surge capacity. That adds a cushion of 4 to 5% to the fleet.

Since the railroad freight industry is very cyclical in terms of changing market demand, reasonable people can debate the excess railcar supply over the levels suggested above.   

How did we get to be so railcar “rich”? In the decades since the end of World War II, the business model has migrated from one where the large and the small railroads (Class 1 railways and the short line companies) began owning fewer and fewer freight cars.  The ownership pattern shifted to either independent car leasing firms and/or to rail freight receivers or shippers. I have written previously about this pattern.

(Photo credit: iStock)

There were two reasons for this railcar supply role shift. One was depressed earnings and working capital at most railroads between the mid-1950s and the early 1980s. The second reason was the tendency for many rail customers to want to control their inventory of goods while in transit – as if the railcar was a warehouse on wheels.

There was also a regulatory environment that favored private railcar ownership with “compensatory” per diem rates. This regulatory allowance stimulated a higher market supply investor incentive.

The over-building could have been mitigated if investors had examined several economic signals more carefully.

The first economic signal occurred between 2006 and 2009, as rail freight loadings peaked and then struggled during the recession. 

There followed a recovery period that saw short spurts of growth out towards 2012. These higher growth rates favored intermodal business plus the growth of selected commodities like ethanol and crude oil by rail tank car.

(Photo credit: Jim Allen/FreightWaves)

Big railroad company profits also rose. The operating income, net cash growth and earnings per share of almost every Class 1 railroad improved almost beyond rail executives’ expectations. It was a celebrated financial renaissance period.

There was an accelerated demand for selected car types, particularly: tank cars; small cube sand hoppers; and covered hopper cars for plastics.

(Photo credit: Jim Allen/FreightWaves)

What could go wrong? What went wrong was that there was too little due diligence regarding market risks. Often ignored – or discounted – were alerts about growing global competition and what economists refer to as product substitution competition.  

Instead, there was an exuberance, or excitement, for railcar investment. Why so optimistic? Frankly, the railcar investments appeared sound to outsiders. Many  overlooked the cyclicality of the North American freight industry. 

The first signal that was ignored might have been the relatively long recovery timeline after the recession. It took quite a while for the rail industry to recover to 2006 carload volumes.

(Photo credit: Jim Allen/FreightWaves)

The second economic signal came around 2014-2015. That signal had two different sources. One signal came as the railroads started switching to their so-called precision scheduled railroading (PSR) model. The PSR business model improved productivity by aggressively managing operating ratios (ORs). PSR was income statement-focused as railroads cut the ratio of expenses to revenues.

Perhaps overlooked was the parallel management objective – whereby the PSR railroads slashed unnecessary balance sheet assets. Railways following PSR reduced “railcars on hand,” foreign railcars on their tracks, and locomotives in use.  

My thesis is that railcar builders and railcar leasing companies, as well as investors did not pick-up the PSR asset minimization theme.  

(Photo credit: Jim Allen/FreightWaves)

Too much rolling stock was no longer in the industry game plan. What was good financially for the Class 1 railroads should have been good for the private railcar suppliers. Correct?        

Here is the second post-2009 timeline issue “we” should have strongly considered.  Starting around 2015, multiple indicators suggested that railcar building ought to slow down. The pace of year-over-year volume growth slowed.

You didn’t require a statistics degree to have noticed the growth rate slowing.

 Figure 1

By examining the blue bar in Figure 1, the years 2015 and 2017 should have been considered “carload growth exceptions.” Focusing primarily upon the brief upticks gave a false sense of confidence.

Beyond the Association of American Railroads data like that in Figure 1, there was an implied threat that improved PSR railcar dwell and train velocity numbers should eventually improve railcar loaded cycles per year. If true, that would further threaten future railcar orders.

Too few seemed to have integrated all these economic signs into a more predictive railcar demand model.

Here are some consequences. 

As 2020 began, the North American fleet totaled just over 1,658,000 freight railcars. A hard analysis suggests that might be between 230,000 to nearly 300,000 cars too many.

Some may disagree. Nevertheless, that’s my hypothesis.

Let me put some numbers around that oversupply.

Working off the lower 250,000 excess railcars might represent a net depreciated fleet value of almost $7 billion. That’s using a pro forma $30,000 per railcar unit.  Some will be higher; some lower.

Table 1 breaks down railcars in storage by nine broad railcar types. Be aware that there are by some counts almost 800 specific railcar types.


 Source: The table and the graph are courtesy of the American Association of Railroads

Figure 2 as a graph of same stored units

Want better granularity? Of the 230,000 to 300,000 excessive freight cars, the small cubic capacity hopper car used for the frac-sand market stands out. There are about 48,000 small hopper railcars too many for the current market. Those are part of the approximately 137,000 stored covered hopper units.

Figure 3

REF20 finance conference graph courtesy of Chicago Freight Car Leasing Company- 3/2/20

For a physical assessment, try to visualize that 48,000 cars number. If all of the units were in one location, the railcars would stretch along a single track for about 450 miles.    

That car storage issue is still fluid. As February ended:

  • The number of open hoppers in storage increased by 31,000 month-over-month
  •  The number of gondola railcars going into storage also increased by 23,000 month-over-month 
  • Another 25,000 intermodal railcars went into storage

Other experts like Bascome Majors of Susquehanna Group see the oversupply this way. Railroad freight cars are in the doldrums.

There was not much optimism as the investors and owners gathered to address the railcar supply issue last week. The only exception was for large cubic capacity railcars used for plastic pellets.

Their bottom line as of early March: 

“We see little hope for a broader freight car tightening until North American rail volumes return to consistent growth and idle railcars consistently return to service…”

Lessons learned looking back over the last half decade.

The cars in surplus story is complex.  It varies by car type. One fleet average number doesn’t give you much insight.

Yes, there are some optimistic prospects out there for rail carload traffic. For example, opportunities include both U.S.-Mexico rail trade, as well as global grain exports from the U.S.

How quickly might PSR railroading increase overall car demand and use? That remains unclear. The upside evidence is limited. As an example, the Figure 4 metrics were presented recently by Union Pacific’s (UP) CFO Jennifer Hamann to investors.

Figure 4

Union Pacific scorecard: January 2020 versus January 2019

UP’s trip-plan compliance for merchandise and automotive carloads suggests a relatively poor one-third late final railcar delivery against the proposed scheduled precision-like objective.

Initial conclusions now 20% into year 2020

The comments above do not show an implosion of rail freight markets. There are positives to build upon. 

Fleet productivity might improve by reducing the vast number of railroad car types over time to less than several hundred.

We can extend physical railcar-life by using higher quality parts and improving maintenance. 

Importantly, the rail industry is still generating strong cash flow. We just need a more realistic market outlook to buoy car investor confidence.

Yet, there is a random possibility that a full-blown recession might further exacerbate the railcar supply dilemma.

For better insight, we should next examine the carload markets by commodity segments.


For a second opinion, contact others like David Nahass. He is Senior Vice President at Railroad Financial Corp. in Chicago. David has more than 20 years of experience in risk analysis regarding the railcar market.