Empty container repositioning

Why do shipping lines sail thousands of empty steel boxes back across an ocean instead of selling them where they pile up?

Stacks of empty containers sit at the edges of import-heavy ports, and every week ships carry thousands back across an ocean holding nothing at all. It looks like obvious waste: fuel, slots and handling charges spent on air inside steel. Sell them where they pile up, buy more where they are wanted, and the problem dissolves.

The industry has had decades and considerable incentive to notice that. So assume it is not a blunder, and ask: what is true about a shipping container that makes hauling it home empty the cheaper option?

Begin with what a container actually is. It is not cargo. It is equipment — a durable, standardised, reusable asset belonging to the line or to a lessor and expected to circulate for a decade or more. Cargo is consumed at its destination; equipment has to come back, or be replaced.

Now write down the one law this system cannot break. At any port, over any period, boxes arriving equal boxes departing plus the change in the stack on the ground. So if a region persistently lands more loaded boxes than it ships out, the surplus has exactly four exits: leave loaded, leave empty, accumulate, or leave the fleet by sale or scrapping. Accumulation cannot go on forever; ground is finite and rented.

Notice what this imbalance is measured in. Not money — boxes. A region can run a large deficit in dollars while its box counts balance, if what it exports is dense and valuable. What produces empties is a mismatch in volume of loaded units, which is why the flows run from manufacturing-export regions toward consumer-import regions and the empties run back.

So why not sell? Here is the step that answers the question. Selling a box in an import-heavy port does not put a box where one is needed. The shortage is at the other end of the lane, and no local transaction can relieve it. The real comparison is therefore not "sell it or store it" but: move this box home, or buy a new one at home and dispose of this one here.

Price both sides. Moving it home costs a slot on the backhaul, handling at both terminals, inland haulage and depot time. But look at which leg that slot sits on: the light direction, with spare capacity precisely because it carries less cargo. The ship is sailing anyway, so the marginal cost of an empty in an otherwise unused slot is low — far below the headhaul rate.

Now price the alternative. Replacing the box means buying a new one near the export end at whatever new-build prices are running. Disposing of the old one means selling into the second-hand market at the import end — a market glutted with used boxes for exactly the reason you have one to sell. That is the crux: the resale value is not an independent number, it is driven down by the same imbalance, while replacement cost is set at the deficit end where boxes are scarce. Repositioning wins whenever the cheap backhaul slot costs less than that gap, and usually it does.

The account predicts things that could turn out otherwise. If box-count imbalance is the driver, empty flows should track lanes with unequal loaded-unit counts and be absent where a large value deficit sits on balanced counts. If move-versus-replace is the arithmetic, then when new-build prices spike repositioning should increase and local sell-off fall — and when backhaul slots are themselves scarce, more boxes should be sold or stored where they lie. The refuting observation would be sustained empty flows running toward the surplus region, or lines routinely selling down at the import end and re-buying at the export end while box prices were high and backhaul capacity cheap.

I will not quote a figure for the share of container movements that run empty, or for fleet size. Published estimates for both vary considerably by source, year and whether inland moves are counted, and a confident number here would be false precision.

Three refinements, the first being the most interesting consequence of the whole arrangement.

The cheap backhaul does not just move empties — it creates trade. Because the line would otherwise carry air, it will accept low-value dense cargo on that leg at a rate barely above the cost of the empty move. That is why import-heavy regions export scrap metal, waste paper, hay and logs in containers: commodities whose value could never support a headhaul rate, riding a leg that was going anyway. An "empty" backhaul is really a spectrum, and every loaded box on it is a repositioning move someone paid part of.

Second, the fixes are about matching, not hauling. A street turn — handing an arriving import box straight to a nearby exporter instead of returning it to the depot and drawing another out — removes two moves and is pure gain. Lessors and pooling arrangements let a box be dropped by one user and taken by another.

Third, a boundary worth drawing. This is not the bullwhip effect and not a forecasting failure. Order amplification is an information problem: each tier reacts to a distorted signal. Here the flows are steady and well understood by everyone involved. What binds is a conservation law on a physical asset that cannot be conjured where it is wanted — better information would not remove a single empty box; only a change in what regions ship to each other would.

Every term the collection defines is gathered in the glossary.