Chassis have always been an Achilles’ heel of container or intermodal transport.
Chassis utilization is now about 90%, a high figure. And a chassis is essential to move a container. So people are holding on to chassis so they can reuse them, say for a reverse load. But that means the dwell time for the chassis is higher than it should be.
Some of the holding is due to the shortages; it’s too hard to get another chassis, so I’ll hold onto the one I just got, and even pay the dwell fee to have it for my outbound load.
If containerized cargo goes down from its current heights, the chassis situation will improve. But that would mean a reduction in cargo, and probably a recession; certainly decreased demands. For demand at this level, we definitely need a larger buffer of chassis, so there is some liquidity in the system.
One interesting point mentioned is the pressure ‘gray pools’, which hold chassis from multiple vendors near a large port or logistics hub, are seeing. We’ll find out if the cooperative approach can hold up under this stress.
Essentially, the pools provide a single shared inventory to a number of users. If a user, or a group of them, holds chassis on their own without sending them back, they are separating from the coalition, and they’re probably doing it because they see it as more favorable than returning the units. The separating group sees that they can do better by separating rather than remaining in the pool.
It’s a classic example of a breakdown in a cooperative game from operations research. Inventory pools have been studied for quite a while, by me and many others. The success of the cooperative scheme requires a ‘fair’ allocation of the benefits. If an individual participant, or some group of participants, are not seeing a better allocation of the gains than they would get separating, they will stay apart. This definition of fairness is called the ‘core’ of the cooperative game for the inventory pool. Under some fairly generous assumptions, we find there is always a core set of allocations, in which every group does better with the allocation than it does separately.
And a core allocation can be computed (there may be many of them), which will be fair. However, it’s virtually impossible to define a core allocation by using a pricing scheme for the use of the chassis. It’s almost never fair. There have to be subsidies beyond the price to make the groups stay in the pool.
That’s what is happening when the pools need to badger firms to return chassis, or when they charge dwell fees.
This would make a good project for an operations researcher, to study the rewards of using a pool over time, and examine how disruptions would affect the reward schemes that are in use now.
To find a fair allocation for a given pool is a good task, but once found, it’s probably not readily explainable to the participants. So they would be skeptical of their rewards, and might still split off. Hard computational results don’t always get the job done; the adopted solution must be ‘explainable’ or ‘interpretable’. There’s a growing body of literature on interpretable results, but not much on interpretable results for cooperative inventory games like this.
John Kingston Friday, September 16, 2022
IANA panel: Intermodal chassis squeeze easing, but it’s far from over – FreightWaves