Drewry: Just 34 Blank Sailings Across East-West Trades in Weeks 42-46, 95% Still Sail

The Analysis

The announcement says 34 of 713 scheduled sailings on the main east west trades are blanked between week 42 and week 46, covering 12 October through 15 November. That is 4.77%, rounded to 5%. The inverse is 679 sailings, or 95.23%, still expected to operate. The aggregate looks calm. The distribution is not.

Half of the cancellations sit on Asia to North Europe and the Mediterranean, 32% on the eastbound transpacific, and 18% on the transatlantic. The network plan ranges from 99% of sailings for Gemini to 89% for Premier Alliance. Ocean Alliance is at 98%, MSC 97%, and independent carriers 93%. A 10 percentage point gap separates the top and bottom plans. That is larger than the entire market's rounded 5% blank share.

We should calculate before calling 95% reassuring. The published percentages imply roughly 17 blank sailings on Asia to Europe and the Mediterranean, about 11 on the eastbound transpacific, and about 6 on the transatlantic. Those are rounded allocations, not exact lane cancellation rates, because the scheduled voyage denominator for each trade is not supplied. The useful question is not whether most ships sail. It is whether the one carrying the purchase order still sails, and what the importer or exporter pays in time when it does not.

Start with the base. Thirty four cancellations divided by 713 scheduled voyages equals 4.77%. Subtract 34 from 713 and 679 remain. Divide 679 by 713 and the planned operating share is 95.23%. The summary rounds those to 5% and 95%. That is fair at market level. It is not precise enough for a shipment plan. A single cancelled weekly departure is either 0% or 100% of the available departure for cargo tied to that week. Portfolio percentages become binary at the container door.

The date window is 35 days, from 12 October through 15 November, spanning five numbered weeks. The simple average is 142.6 scheduled voyages per week and 6.8 blank sailings per week. This average is arithmetic, not a timetable. The source does not say that cancellations are evenly spread across weeks. Seven blanks in one week and none in another can produce the same five week total as a smooth 6.8. Cargo does not travel on an average Tuesday. It travels on a named vessel with a cutoff.

The trade allocation has the same denominator problem. Half of 34 is 17. Thirty two percent of 34 is 10.88, which reads as roughly 11 voyages. Eighteen percent is 6.12, roughly 6. The rounded implied counts add back to 34. Yet we cannot say Asia to Europe has a higher cancellation rate than the transpacific, because we are not given each trade's scheduled sailing count. We know where the cancelled voyages sit. We do not know the blank percentage within each lane. Count and rate are not interchangeable.

The alliance comparison is the sharper signal. Gemini plans to run 99%, Ocean Alliance 98%, MSC 97%, independent carriers 93%, and Premier Alliance 89%. Convert the operating shares to planned non operation shares and the ladder is 1%, 2%, 3%, 7% and 11%. On a per 100 planned sailing basis, Premier's missing share is 11 times Gemini's 1%. The operating difference sounds like 10 points. The non operating ratio sounds like elevenfold. Both describe the same published plan, but the second exposes why a small aggregate can hide a large network choice.

Do not turn those percentages into voyage counts by alliance. The source does not provide how many of the 713 scheduled voyages belong to each network. If Gemini had one denominator and Premier another, multiplying every percentage by 713 would be fiction. The comparison is valid as a rate within each published network plan. It is not valid as a claim that Premier cancelled a particular number of the 34. This number has a problem whenever a slide shows five percentages without their five bases. Ask for the service list, not another pie chart.

Why now? The tracker covers the post Golden Week planning window and measures capacity decisions already announced for weeks 42 through 46. Carriers use blank sailings to remove a departure from the plan. The supplied facts do not state each carrier's commercial motive, so we should not invent one load factor or one rate target. What we can say is that capacity pressure is being handled unevenly. If all networks faced and managed the same conditions in the same way, planned operation would not range from 89% to 99%.

The causal chain for cargo is straightforward. A sailing is removed. Booked cargo needs another departure, another service or another arrangement. The risk then moves from network capacity into shipment timing and possibly into commercial terms. The source quantifies the first link, planned operation, but not the later links. It gives no roll delay, transit time, space guarantee or freight rate. Those missing numbers are not permission to fill the cells with zero. They are the questions the booking owner must close before cargo reaches cutoff.

Who is most exposed depends on concentration. An importer or exporter using several networks across many weekly departures may experience something near the aggregate 95% plan. A company with all critical cargo on one Premier service is not diversified merely because 679 market sailings still operate. A company on Gemini is not risk free merely because the network plan says 99%. One percent can still contain its named voyage. Market breadth protects only a shipper able and willing to switch across that breadth.

Lane concentration adds another layer. Roughly 17 of the 34 blanks are allocated to Asia to North Europe and the Mediterranean. That is half the cancellation count. It does not prove a 50% chance of cancellation on those lanes. It does tell cargo owners there to demand more service level detail than the global 5% headline provides. The transpacific's roughly 11 and transatlantic's roughly 6 need the same treatment. Every numerator must be paired with the local schedule denominator before a probability is quoted.

Timing transmission begins before the cancelled vessel would have sailed. A blank sailing normally becomes an operational problem when the booking must be reassigned, not on the original departure date. The source does not give announcement lead times. Therefore the usable control point is booking confirmation and every subsequent schedule update. Waiting until cutoff to discover the voyage is absent converts a planning percentage into a physical queue. The window ends on 15 November, but the decision window for each shipment closes earlier at its own production and terminal milestones.

Here is the point the 95% headline misses. A sailing being expected to operate is not the same as a container being expected to depart on time. The tracker counts blanked sailings. It does not report late departures, port omissions, rolled bookings or connection failure in the supplied material. A voyage can remain in the 95% operating bucket while a specific container misses its intended movement. Therefore 95% is a capacity plan metric, not a 95% on time shipment promise. That distinction is the operational heart of the story.

The inverse is also true. A blank sailing does not prove every booked container suffers the full interval to the next weekly ship. Some cargo may be reassigned to another service. The supplied facts do not state recovery options or delay. The correct entry is unknown until the carrier confirms the alternative. Writing seven days automatically for every blank would be as careless as writing zero days. We can calculate a scenario, but the assumption must be visible.

Assume a company has 100 otherwise comparable shipments distributed in the same proportions as a network's planned operating share, and assume one shipment uses one planned sailing opportunity. Under that simplified planning model, 99 shipment opportunities remain on Gemini and 89 on Premier. The difference is 10 per 100. Ocean leaves 98, MSC 97 and independents 93. These are not forecasts of actual boxes moved. They translate the published percentages into an exposure scale management can read.

Now attach time, explicitly as a scenario. Assume each affected service is weekly and a blanked departure has no same week substitute. One blank then creates a seven day departure gap. Five affected shipments would create 35 shipment days of departure delay; 11 would create 77 shipment days. This calculation does not claim that the market's rounded 5% or Premier's 11% will hit a particular shipper. It shows why a few missed departures can outweigh a large number of normal ones when customer promises are measured in days.

Change the recovery assumption and the result changes. If every affected shipment receives an alternative three days later, five affected shipments create 15 shipment days rather than 35. If half recover in three days and half in seven, the average for those affected lies at five days under an equal split. The tracker supplies none of these recovery intervals. Procurement should therefore ask for the named recovery sailing and date, not accept the phrase next available vessel as a number.

There is a second calculation around concentration. Suppose 20 urgent shipments are all assigned to a network planning 89% operation, and use 11% only as a portfolio planning factor. The expected exposure scale is 2.2 shipment opportunities, not a promise that exactly two will fail. On a 99% network, the same arithmetic is 0.2. The difference is 2 shipment opportunities across 20. Small decimals should not be rounded into certainty. Their use is to size contingency, not predict the identity of a cancelled voyage.

The aggregate comparison can mislead in the other direction too. Thirty four blanks across 713 sounds limited. If an importer has flexible cargo and access to multiple operating services, paying heavily for a blanket emergency plan may be wasteful. The 679 expected operations are meaningful capacity. The action is selective protection: identify critical orders, concentrated lanes and lower operating share networks. Do not treat every container as an emergency merely to look prudent. An undifferentiated buffer is just expensive fog.

The contrary case deserves equal space. Announced blanks may not translate into severe cargo disruption if carriers provide workable alternatives and the remaining services have space. Under that condition, the 95% operating plan is genuinely reassuring and switching carriers could add cost without improving delivery. The source gives no utilisation or recovery data, so we cannot declare disruption. We can only mark where proof is needed. A confirmed alternative with an acceptable date is stronger evidence than the alliance ranking alone.

Another contrary case is deterioration after the tracker snapshot. The 34 cancellations are the count in the stated week 42 to 46 view. Plans can matter only as of their observation point; the supplied material does not promise the count will never change. A procurement decision should not freeze the tracker as permanent truth. Recheck the named sailing. The market total is a screening signal. The live booking record is the control document. If those conflict, the booking record wins for the cargo while the discrepancy is escalated.

The trigger line should use the published network spread. Treat 97% planned operation as the monitoring threshold because it is the stated MSC level and the midpoint between the high pair at 98% to 99% and the lower figures of 93% and 89%. Cargo on a network at 97% or below needs a named backup before cutoff. Cargo on 98% or 99% still needs confirmation, but a blanket backup is not automatic. This 97% threshold is an internal control derived from the supplied ladder, not a claim that 97% is universally safe.

Criticality needs a second trigger. If missing the planned departure would breach a customer commitment before the next assumed weekly opportunity, do not rely on the market's 95%. Split the booking across confirmed services where the shipment can be divided, or move the cutoff decision earlier. No cost figure for splitting is supplied, so quote it before approval. The threshold is temporal: zero spare days between the planned departure and the customer promise. When the buffer reaches zero, aggregate reliability has no commercial value.

Price discipline still matters, although no rate is reported. A blank sailing can tighten space on remaining departures, but the source supplies no evidence of a specific rate increase. Do not invent a surcharge or tell finance that 5% fewer sailings means 5% higher freight. Capacity and price are linked through market conditions, not by that arithmetic identity. Obtain an all in written quote tied to the confirmed operating vessel. Compare price only after the sailing status, equipment, route and included charges match.

The point to lock is therefore conditional, not a fabricated dollar level. For critical cargo on a network at 97% planned operation or below, lock the confirmed space and its written rate as soon as a named operating sailing and acceptable recovery option are provided. For networks at 98% or 99%, lock when the named sailing is confirmed and the shipment has no spare weekly cycle. If the carrier cannot name the vessel or recovery date, a cheap quote is not lockable capacity. It is a number waiting for a ship.

Responsibility should follow the data layers. Procurement owns the network comparison and the written rate. Logistics owns vessel status, cutoff and backup sailing. Sales or customer service owns the promised delivery date and available buffer. Finance owns any premium approved for protection. The factory owns cargo readiness. A network percentage should not be passed around until everyone assumes someone else checked the vessel. Give each critical booking one named owner from confirmation through loading.

Build the control sheet with the smallest useful fields: trade, network, service, vessel, planned departure, published network operation share, current sailing status, cutoff, backup vessel, backup departure, customer promise and spare days. Mark 97% or below amber at booking. Mark a cancelled or unnamed voyage red. Mark zero spare days red even on a 99% network. This uses the tracker as an input without pretending it is the shipment record.

The regional distribution should guide review frequency, not produce fake probabilities. Asia to North Europe and the Mediterranean receive daily review for critical cargo because they contain half the 34 cancellations. Eastbound transpacific receives the same review where its roughly 11 implied blanks intersect a critical order. Transatlantic receives review where its roughly 6 intersect one. The count tells where to look. The named schedule tells what to do.

Do not compare alliances without matching the service product. A 99% network plan may not serve the required origin, destination, date or equipment. An 89% network may offer the only commercially acceptable routing. The source measures sailing execution plans, not route suitability or transit time. Rank feasible services first, then apply the reliability lens. Otherwise procurement may select a highly operating service that cannot carry the cargo where it needs to go, which is a very reliable way to fail.

Diversification also needs a denominator. Splitting ten critical containers five and five across two services reduces concentration only if the services do not share the same cancelled leg or operational dependency. The supplied data is at network level and does not map shared ports or vessels. Ask whether the backups are genuinely separate. Two booking references on the same removed sailing are not diversification. They are duplicate paperwork with matching disappointment.

The five week window should be managed as rolling exposure. At the start of each internal review, remove sailings already completed, update changed voyages and count critical containers on each remaining service. Do not keep dividing a shrinking company portfolio by the original market total of 713. The 713 describes the tracker universe. The company's denominator is its own remaining booking opportunities. The smaller denominator is often the more dangerous one.

After each planned departure, record four outcomes: operated and loaded, operated but cargo rolled, blanked with confirmed recovery, or blanked without acceptable recovery. This creates the distinction the public headline cannot provide. An operated vessel with a rolled container belongs in the 95% market bucket but not in the shipper's success bucket. A blanked sailing with a rapid alternative belongs in the 5% market bucket but may still meet the customer date. Outcome quality is not identical to sailing count.

The rate negotiation should use facts without exaggeration. The buyer can say the overall plan removes only 34 of 713 voyages, yet the chosen network's operation share and the lane concentration create a specific exposure. That is stronger than claiming a capacity crisis the data does not show. Ask the carrier to bind the quote to a named sailing, state the remedy if it is blanked and identify the recovery date. The request is precise because the evidence is precise.

Watch language in carrier replies. Expected to operate is not confirmed loaded space. Next available sailing is not a date. Priority is not a measurable remedy unless the booking terms define it. The source's 95% is already an expectation. Repeating another expectation adds no control. Convert each phrase into vessel, date, cutoff, allocation and written charge. If it cannot be converted, leave it unresolved on the sheet.

There is no basis here for changing inventory across every lane. A seven day buffer may be useful in the explicit weekly no substitute scenario, but it is not a factual delay from the tracker. Inventory decisions require demand, carrying cost and actual recovery time, none of which is supplied. Use the scenario to test whether a customer promise survives one missed weekly departure. If it does not, protect that order. Do not turn one arithmetic example into a network wide stock policy.

The decision sequence can stay short. Verify that the service is feasible. Check the network operating share and lane concentration. Confirm the named sailing. Measure spare days to the customer promise. Obtain a written backup and all in rate. Lock when the reliability threshold or zero buffer trigger is met. These are not ordinal slogans; they are gates. A booking moves forward only when the previous uncertainty is closed or consciously priced.

For the remaining week 42 to 46 window, the most exposed file is not necessarily on the network with the lowest percentage. It is the critical order with one feasible sailing, no spare week and no named recovery. A Premier booking may have two workable alternatives. A Gemini booking may have none. The network ladder screens risk; shipment geometry decides it. This is why the fifth question, what the headline missed, matters more than reciting 95% again.

The market level conclusion is narrow. Capacity removal is limited in aggregate, concentrated by trade and sharply different by network plan. The shipment level conclusion is stricter. An expected operating sailing still needs a named vessel, live confirmation, space and a date that protects the customer promise. No contradiction exists between those statements. One describes 713 voyages. The other describes the container that pays the invoice.

Set the lock point now. Critical cargo on 97% or lower network plans should have confirmed backup space and a written rate locked as soon as the carrier names an operating sailing; any booking with zero spare days to the customer promise gets the same treatment regardless of alliance percentage. Recheck through 15 November. Ninety five percent is an excellent comfort blanket, provided the missing five percent belongs to somebody else. When the plan falls through the 97% line, that is when space and price should already be locked.

↑ Back to the key points

— By Vivian Zhao