Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- Within 7 days, start a weekly count of empty availability by box type at your actual load port, measured against the next two weeks of your booking plan; if it falls below plan, pre-book 30 per cent of next month's volume immediately.
- Record the routing on every booking, Cape or Suez, and rebuild the transit assumption per service rather than per trade lane, planning the receiving calendar on the 80th percentile of the last eight voyages instead of the proforma.
- Reconcile surcharges at container level with charge code and voyage, monthly, and dispute any line you cannot attribute within 30 days of the invoice date.
- Recalculate pipeline inventory at the new loop time and report the delta to finance as a one-off working capital step; if it exceeds one month of gross margin, re-phase the order calendar before the next peak.
- For any SKU with under eight weeks of shelf life or a six-week selling cycle, set a written cost ceiling per FEU for buying a faster routing at the time of booking, so the decision is a rule rather than a monthly argument.
A 6 July supply-chain disruption briefing noted that Asia-Europe services still largely route via the Cape of Good Hope, with only selective Suez transits, while Strait of Hormuz traffic remains cautious below normal. Carriers continue to apply war-risk and operational surcharges, and empty-container repositioning stays inefficient across major lanes.
The Analysis
A disruption briefing is a schedule of symptoms, listed in the order the author noticed them. The update dated 6 July says Asia-Europe services still run largely via the Cape of Good Hope with only selective Suez transits, that Strait of Hormuz traffic stays cautious and below normal, that war-risk and operational surcharges remain in force, and that empty-container repositioning is inefficient across the major lanes.
Three of those four items are consequences. The last one, the empty boxes, is the one that moves first and the one almost nobody puts a number on. If you want an early warning on this trade rather than a commentary on it, that is where to look.
Model the routing change as a change in loop time, because that is the only form in which it reaches your invoice. The unit of deep-sea capacity is not a ship. It is a ship-day, and what a fleet delivers is the number of completed loops it can run. Take a loop that used to take seventy days and now takes eighty-four. The same ships deliver one minus seventy over eighty-four, which is about seventeen per cent fewer cycles. Nothing was scrapped, nothing was cancelled, and seventeen per cent of the effective capacity left the market. A carrier has three ways to answer that: add ships, which takes years; skip calls, which breaks the return flow; or raise the price of the slots that remain. Only the third one can be done this quarter, which is why the surcharges are not opportunism. They are an accounting of a physical fact, and they will not disappear while the loop is long.
The empty-container line is the same arithmetic seen from the other end, and it is the part that matters to an exporter. Aggregate box count in the world is fine. What breaks is distribution. A container is a stock item that has to be moved back to where the next load is, and repositioning is a transfer with a lead time, not a phone call. In a normal market the schedule slack absorbs the mismatch: a box arrives late, waits two days, gets loaded anyway. Cape routing removes the slack, because every extra day on the water is a day the box is not available to be loaded somewhere else. So the system loses its shock absorber at exactly the moment the shocks get more frequent, and the failure shows up as a local shortage in a market that has plenty of boxes nationally. That is a queue problem, not a supply problem, and it is why the shortage appears suddenly and in one box type at one port.
Selective Suez transits then do something subtler, and it costs buyers real money. When fifteen per cent of a service's voyages take one route and eighty-five per cent take another, the string is no longer selling one product. It is selling two, at two transit times, under one proforma. The published figure becomes an average of two different services, which means it is accurate for nobody. Your planning calendar needs the spread, not the mean: the difference between the fastest and slowest voyage on the same service. Carriers will give you the proforma. They will not volunteer the spread, and the spread is what determines whether your distribution centre receives a week early or a week late.
The impact sorts by value density and by shelf life rather than by volume. A furniture or ceramics importer on a ninety-day planning horizon barely notices thirteen extra days; the working capital moves, the container still lands before the season, and the surcharge is a line item. A fashion or seasonal goods importer on a six-week cycle is in a different business, because thirteen days is a fifth of its entire planning window and the markdown risk sits on the other side of it. High-value, low-volume electronics sits with fashion on timing and with nobody on cost: the surcharge is trivial against the cargo value, and the schedule variance is the whole exposure.
On the direction axis the asymmetry is sharper than people expect. Asia-based exporters face space and equipment at the same time, which is the worst combination. European exporters into Asia find boxes easier, because empties are accumulating on their side, but they face the same long loop and the same space constraint, so their lever is schedule flexibility rather than equipment.
Timing runs in a sequence that is worth writing down. The first thing you feel is not the invoice. It is the booking confirmation: the moment your forwarder stops confirming a specific vessel and starts confirming a week. Assume two to three weeks after a routing shift becomes general. Equipment tightness at origin follows, and this is where the empty repositioning shows up on your desk, typically four to six weeks in, as a specific box type at a specific port rather than a general shortage. The receiving plan at your distribution centre is the third thing, six to eight weeks out, when the first voyage with an unusual transit time misses its slot. Inventory is last, eight to twelve weeks, when the buffer that was quietly covering the variance runs out and every subsequent delay is a stockout.
Run the arithmetic on one importer, and treat the inputs as my assumptions rather than anything in the briefing. Forty FEU a month, one-way transit going from thirty-two days to forty-five, average cargo value twenty thousand dollars per FEU. Pipeline inventory is monthly volume divided by thirty, multiplied by transit days: forty-two and a half FEU before, sixty FEU after. The difference is seventeen and a half FEU, or about three hundred and forty-six thousand dollars of additional working capital, and you now commit each order thirteen days earlier. At a six per cent carrying cost the cash number is only around twenty-one thousand dollars a year, which is not the point. The point is the thirteen days of forecast. You are buying inventory against a demand signal that is thirteen days staler, and in most categories that forecast error costs more than the surcharge and the carrying cost combined. The visible number on the invoice is rarely the biggest one.
Which brings up what the briefing cannot tell you, and what I would build the whole response around. The empty-container line is the leading indicator, and it is measurable at your own desk, weekly, without any subscription. It is not the rate index, which reprices after the fact and tells you what already happened. It is the answer to a boring question: how many boxes of my type are available at my load port for the next two weeks of my booking plan. When that number falls below your plan, equipment is about to bind; when it rises while rates stay flat, capacity is quietly returning. The rate tells you what you are paying. That count tells you what you are about to be unable to do.
The counterfactual is not a different route, because there is no cheap alternative to a long loop. It is planning on the wrong statistic. A buyer who plans the receiving calendar against the proforma transit time is planning against a blend of two products and will be wrong in both directions, early enough to build stock it does not need and late enough to miss the one that matters. A buyer who plans against the eightieth percentile of the last eight voyages carries slightly more pipeline and stops being surprised. The other avoidable loss is reconciliation: if surcharges are checked at invoice total rather than per container with a charge code and a voyage, you cannot dispute what you cannot attribute, and the argument is lost by default rather than on merits.
↑ Back to the key points