Drewry WCI Snaps Three-Week Slide, Rising 1% to $4,297 on Transpacific Strength

The Analysis

The number is 4,297, and the sign in front of it changed. Drewry's World Container Index rose 1% to US$4,297 per FEU in the week to 6 August. That is the first gain after three straight weekly declines. Transpacific did the lifting: Shanghai to New York jumped 4% to US$7,893, and Shanghai to Los Angeles climbed 3% to US$5,894, which the report puts at 133% above a year ago. Carriers landed their August general rate increases, and congestion in central and southern China trimmed the capacity that was available. Those are the prints. The interesting part is not the 1%. The interesting part is why the sign flipped in the same week a GRI took effect.

A market that rises because a rate increase was scheduled to take effect is not the same animal as a market that rises because cargo got scarce. Both show up as a green number. Only one of them lasts past the next print. And a 1% composite move, sitting on top of a 4% lane and a 3% lane, tells me the average is not where the action is. So before anyone reads first gain as the turn, I take the number apart: what the 1% is made of, what the 133% is measured against, which lanes are pulling and which are dragging, and where the line sits that makes this a real turn rather than a date on a calendar. Same drill as always. Numbers first.

One up week after three down is the most over-read number in this market, and I have watched people lock a quarter on it. I do not. So let us find out whether this is a turn or a date, and price it accordingly.

A green print after three red ones is the most over-read number in this market. One up week is not a turn; it is a pause in the decline. Base effect does the rest: after three weekly falls the base is lower, so the same demand reads as a plus. Calling a trend from a single up week is the standard way to get burned in this trade, and the burn is usually felt about three prints too late. So the first thing I do with a first gain is refuse to be excited by it, and go and look at what is underneath.

Then read the shape, because a composite is a weighted average and averages drag. The index is 4,297. Shanghai to New York is 7,893. Shanghai to Los Angeles is 5,894. The composite is pulled down by the other lanes in the basket, the ones priced below transpacific. So a transpacific shipper feels 4% and 3%, not 1%. The same sentence, the index rose 1%, lands on you as a 4% lane, or as a number averaged into nothing. You pay a lane, not the sentence. The index tells you why the conversation is happening; it does not tell you what to accept. This gap is exactly why I keep saying the headline is the wrong place to make a decision.

Now handle the 133%. It is a year-on-year figure, and year on year is a different instrument from week on week. A 133% gain says the base a year ago was low; read it as distance from a low starting point, not as momentum. The composite rose 1% in a week; the lane is 133% above a year ago; both are true and they measure different things. Using a big year-on-year number to conclude the market is on fire is comparing two rulers with different scales. Note the caliber as well: the Chinese-language figure gives New York up 106% year on year and Los Angeles up 133%, two different gains on the same transpacific trade, because the base and the commodity differ. Establish the caliber before you quote the number, every single time.

Here is the practical split between those two measures, because people mix them and then make the wrong call. The year-on-year number is a health check on the base: it tells you how far the lane has travelled, and a 133% move says it travelled a long way from a low starting point. The week-on-week number is a timing signal: it tells you whether anything is happening right now. You budget off the year-on-year and you act off the week-on-week. If you act off the year-on-year, you will lock at the top because the number looks scary. If you budget off the week-on-week, you will under-provision because one calm week is not a year. Two instruments, two jobs, and the mistake is always borrowing the dramatic one for the calm job.

Separate the two pushes. Carriers landing their August general rate increases is a scheduled increase; it has a calendar, and it shows up in the print by design. Congestion in central and southern China trimming capacity is a genuine supply cut; but congestion is the kind of constraint that clears and gives the capacity back. A scheduled lever and an empty berth do not behave the same way. The part of this 1% that came from a GRI can be handed back once the increase is absorbed or rolled back; the capacity lost to congestion returns when the ports clear. Neither push sits on the demand side. That is the whole question: does this 1% deserve half your confidence or all of it.

So one print needs one confirmation. The first gain after three declines should make you wait for next Thursday. If the next print holds 4,297 or moves up again, treat this as a turn. If next week hands the gain back, this was not a turn, it was a GRI artefact. The market is very good at using a dated increase to rush people into locking early. I want two prints at or above 4,297 before I call it a turn, and I am happy to miss the first week of a real move rather than pay for the last week of a fake one.

So let me be explicit about what a real turn would look like, so we are not arguing about feelings next week. A real turn is more than one lane up one week. I want to see the gain spread beyond the two transpacific lanes, because a turn led by two lanes and a dated GRI is a narrow turn. I want to see the composite hold or build rather than hand the 1% straight back. And I want to see the cause shift from a scheduled increase to something structural, a capacity cut that does not clear on its own, or demand that shows up in the load factors. A GRI that lands and sticks is real money, but it is a top-up, not a trend. A GRI that lands and is absorbed is a non-event. Next Thursday will not answer all of that, but two or three Thursdays will, and waiting for them costs less than being wrong.

Translate it into cost. A transpacific shipper is not looking at 1% this week; it is looking at 4% and 3%. New York 7,893, Los Angeles 5,894, two coasts, two lanes, 1,999 dollars apart on the same box. That is a port and inventory decision, not a rate-direction decision. The year-on-year side bites harder. The report puts the LA lane 133% above a year ago, which implies the same lane was about 2,530 a year back, working back from 5,894 divided by 2.33, on the assumption that the reported 133% is what it says. From 2,530 to 5,894 is a very concrete addition to landed cost. Freight is a small share of unit cost and a large share of the variance, and this year that sentence is doing a lot of work.

Peak and timing matter here too. August is when the build for the fourth quarter is running, and space and equipment tighten. A GRI landing in August is no accident; it is timed to the tightest part of peak. When peak is tight, the scarce things are space and empty boxes, not the printed rate. You will find your sailing closed on a Tuesday before you find the rate moved. So the order this week is space and time first, price second, and the GRI is the least important of the three if your cargo cannot get a slot.

Who eats a GRI is its own calculation, and it is not symmetric. On a CIF or DDP term the seller carries the ocean freight, so a 1% index move and a landed GRI show up on the seller's side of the invoice before they ever reach the buyer. On an FOB term the buyer carries it from the load port, so the same GRI lands on the importer. The question of whether this rise matters is therefore not the same question for both parties at the table, and the answer flips with the Incoterm you signed.

Before anyone argues about the rate, settle which side of the box the freight sits on, because a rise that is a rounding error for one party is a margin line for the other, and the number in the index does not know the difference. That is why the first call this week is not to the carrier, it is to your own commercial team, to fix which Incoterm you are quoting on and therefore who is holding the exposure.

Let me do the arithmetic with the assumptions written down, because a number you cannot decompose is a number you cannot use. Assume you ship Asia to the US West Coast at the 5,894 print. Assume 20 FEU a month, so 240 a year. Assume the reported 133% year-on-year gain applies to this lane, which puts the same lane at about 2,530 dollars a year ago, working back from 5,894 divided by 2.33. One box is about 3,364 dollars more than a year ago. Twenty boxes a month is 67,280 dollars a month, about 807,000 over the year. That is the magnitude the year-on-year line adds to your budget.

Now assume the August GRI lands on this lane at 500 dollars a box, an assumption used only to test the billing rule; 20 boxes a month is 10,000, or 120,000 a year, if it sticks. Put the two together: the 807,000 from the year-on-year move is money already inside your cost base; the 120,000 from the GRI is still hanging, and whether it holds depends on the booking-date-versus-sailing-date rule. Knowing which of the two is a fact and which is provisional beats arguing about whether the rate rise has an impact.

The split between the 807,000 and the 120,000 is where I see budgets go wrong. The 807,000 is sunk: the lane is 133% above last year and your invoice has been paying it for months, so the live question is not whether to accept it but whether to renegotiate the base or re-source around it. The 120,000 is provisional: it depends on whether the GRI sticks, whether the billing rule catches your booking, and whether the congestion that trimmed capacity clears. One number you manage by changing the plan; the other you manage by managing the timing. Teams that treat both as the same kind of number either panic over the sunk one or ignore the live one, and both mistakes cost money. So when the GRI print lands, do not ask whether the rate is up or down. Ask which of your two numbers just moved, and by how much.

The actions this week are specific. Watch the next two Thursday prints; do not call a turn off one week, and only treat this as a turn once two prints sit at or above 4,297. Pin down the August GRI billing rule in writing, whether it bites on the booking date or the sailing date, because the sign flipped on a scheduled increase and not on demand. Price transpacific by coast, New York 7,893 against Los Angeles 5,894, a 1,999 dollar gap per box, with both numbers in hand before you commit. And put names on it: procurement pulls the Thursday print and puts out a one-page lane sheet by Friday, logistics confirms empty-box availability and booking cut-off with the forwarder the same day, and finance has the year-on-year 807,000 and the GRI 120,000 in the budget by month end.

The contract and spot mix needs a wider gap this time. A gain held up by a GRI and a gain held up by congestion unwind differently: a GRI can be absorbed or rolled back and the whole move handed back, while congestion clears and the capacity returns. So I would not, on a rise that may be a GRI artefact, fix the whole quarter's volume at a high level. Fix the must-ship peak volume, and leave enough on spot that a pullback is worth something to you. The carrier will lean on this rise to sell you more fixed volume, because a market that looks like it is turning is a good time to sell a long contract. Your job is to work out whether the rise is scheduled or scarce, then decide how much to fix, and the answer to that question is not the same as the answer the seller wants.

Alternatives, each with a price. There is already a 1,999 dollar gap between the two transpacific coasts, New York 7,893 and Los Angeles 5,894; if you can switch the receiving port, you can trade that gap for freight, and you pay for it in inland distance, transit days and a distribution footprint that may not exist on the other coast. You can change mode or routing for a better transit-to-cost trade, and you pay in days. You can change the timing, pulling the shipment forward before peak tightens further or pushing it back if you can hold inventory, and you pay in inventory carrying cost, which is out of fashion in a high-rate year and may still be cheaper than a panic spot booking. There is no free route, only a choice about which bill you would rather pay.

And one thing the composite hides that a transpacific importer cannot afford to miss: the two coasts are not one market. New York at 7,893 and Los Angeles at 5,894 are the same ocean and two different businesses, separated by 1,999 dollars a box. The East Coast number carries an inland, a canal and a rail story that the West Coast number does not, so a shift between coasts changes freight and lead time together. When someone says the index rose 1%, ask which coast, because one of them rose 4% and the other 3%, and the composite is neither. Then ask about equipment on the coast you actually use, because in August the box is often harder to find than the rate is to negotiate. A rate you can argue with. An empty box you cannot.

The pitfalls sit in a few places. The GRI billing rule is the one that bites silently: it is quoted per box but applied by a rule, and the rule holds which cargo, which booking date, which sailing date, which commodity; a booking made the day before the effective date can still be charged if the box sails after it. Congestion at the Chinese load ports bites too: a box that would clear in two days can sit for five, demurrage and detention start after free time, the index never mentions them, and you should get free time in writing and decide who pays the extension. The fuel mechanism works underneath all of it: a falling BAF offsets a rising base, and tracking only the base makes you panic at the wrong moment. And keep the calibers straight, because the 133% is year on year and the 1% is week on week, and using one to stand in for the other is using two rulers as one.

So here is the position. The first gain in four weeks, up 1% to 4,297, with the transpacific leading, New York at 7,893 and Los Angeles at 5,894, and the report attributing it to August GRIs landing and congestion in central and southern China. One push is a date and the other is a berth that clears, and neither is demand, so give this rise half your confidence. The next two Thursday prints decide the other half: two prints at or above 4,297 and you treat it as a turn, taking the fixed quote at a 3% to 5% narrow gap and staying on spot above about 8%; one print that hands back the 1% and it was a GRI read, and you do not fix a quarter for it.

The lock window is not when your palms are sweating. It is the Friday after your lane posts its second print above the line. Rates in this trade start rising on a dated announcement and start falling on a space you were not watching, so put the date and the hedge on your side of the table and leave the rest to next Thursday. Cross your line, and sign that afternoon.

↑ Back to the key points

— By Vivian Zhao