Daily Market Note — 2026-08-28

Market commentary on the global ferrous scrap trade. Not trading advice.

The read

For a third consecutive session, imported scrap into Pakistan declined to price in a cost increase that everybody in the market can see coming.

Container freight to Pakistan is publicly expected to rise $400–$500 a box in September. On a twenty-foot container that is $14 to $20 a tonne. It has been public and dated for three days. Sellers cite it. And across those three sessions the bid has not moved a dollar and neither has the offer.

Two sessions of that was an observation. Three is a property of the market. A market that passes a cost increase to the buyer is a market with a buyer able to pay it. This one has now declined three times.

There is a second thing worth saying about today, and it is a credit to the source rather than a complaint about it. The market report states plainly that no fresh Pakistan assessment dated 28 August exists, and that the latest evidence remains a 27 August headline and a detailed report dated 25 August.

That is the right way to publish. It also has a practical consequence: the “$415 assessed” figure circulating today is a Tuesday number. So is the last published index print. When a price is quoted at you this week, ask what date it carries.

Key moves

Imported shredded into Pakistan is unchanged for a third session, and the evidence behind it is ageing rather than refreshing:

• Broad market $412–$415 a tonne CFR — headline dated 27 August
• Assessed shredded around $415 CFR Port Qasim — dated 25 August
• Buyers focused at $410–$412 — unchanged for a third session
• EU and UK-origin offers at $415–$420 — unchanged for a third session
• Last published index print $413.63 on 25 August, now three publisher-sessions old

The much-quoted cargo is the same cargo, for the third time. The 2,000–3,000 tonne UK-origin parcel booked at $414–$415 appears again today. It is a restatement, not a second booking. One print is a clearing reference. Two would be a level. The market still does not have two — and repetition is not corroboration.

Demand is weak and now measured three times identically. Pakistani mill operating rates around 35%, steel sales at 40–45% of normal, high inventories, and monsoon rain still cutting construction. Three identical observations is about as durable as a demand reading gets from public sources.

A statistic returned today that explains the rest of it. Pakistan’s July scrap imports were 408,041 tonnes — up roughly a quarter on June and over 40% on a year earlier. Today the report files that figure alongside mill coverage rather than in a statistics box, and the connection is the point: a heavy July arriving into mills running at 35% is why there is no spot urgency. The import number and the utilisation number are one story seen from two ends.

Domestic prices, unchanged in local currency for a third session: scrap PKR 145,000–147,000 a tonne, billet PKR 216,000–218,000 ex-works, rebar PKR 234,000–236,000 ex-works. In dollars they tick slightly down, because the rupee eased to 277.52: roughly $522–$530, $778–$786 and $843–$850. The rolling margin between billet and rebar works out around $65 a tonne on midpoints — thin, and a fair summary of why mills are not chasing.

On Turkey, the exchange reference is unchanged and the physical evidence behind the argument is getting old. Scrap CFR Turkey, month two, sits around $382.50 — the only ferrous exchange figure published today.

Set that against what actually traded: on 25 August a northern Turkish mill booked 12,000 tonnes of European HMS 80:20 at $370 CFR, and a Marmara mill took a Baltic cargo at $369. The September future is $12.50–$13.50 above the last cargoes anybody booked.

That remains the right comparison — but quote the date with it. No new physical booking printed today, in Turkey or anywhere else, and a three-day-old cargo is fair game for a seller to call stale.

A note on how quickly references come and go here. Three exchange contracts introduced yesterday — rebar FOB Turkey, scrap CFR India and scrap CFR Taiwan — have all disappeared from today’s reporting after a single session. That is the fourth new structure withdrawn inside five days. Anyone building a reference set from daily commentary should record the number, its date and its definition the day it appears.

Freight and shipping

The lane numbers are restated for a third session, which makes them a level rather than a report: twenty-foot freight into Pakistan around $1,450–$1,500 per box, September expected $400–$500 higher toward $1,800–$1,850.

20ft container	Per box	At 28 tonnes	At 25 tonnes
Current	$1,450–$1,500	$51.79–$53.57/t	$58.00–$60.00/t
September (expected)	$1,800–$1,850	$64.29–$66.07/t	$72.00–$74.00/t
Increase	$400–$500	$14.29/t	$20.00/t

The sequencing rule stands: lock September freight before locking September prices. A three-to-five dollar supplier concession does not pay for a fifteen-to-twenty dollar carrier increase.

Two things got clearer today, both usefully.

First, the Far East–Pakistan peak season surcharge now has named dates. The revised charge — $1,000 per twenty-foot and $1,500 per forty or forty-five-foot container — takes effect 31 August for most covered origins, with Vietnam from 1 September and South Korea from 4 September. Anyone with Asian-origin cargo loading around month-end should be repricing now rather than on the day.

Second, and more interesting: port congestion at Karachi and Qasim has been split into the part that can be measured and the part that cannot. Weekly data through 22 August shows median anchorage waiting of about 0.05 days at Port Qasim and 0.16 days at Karachi — roughly seventy minutes and under four hours. Those are not congested ports on the vessel-queue measure.

The commentary draws the right conclusion rather than the convenient one: low anchorage waiting time does not contradict tight equipment or booking conditions, because vessel queuing and container availability are separate constraints.

Both halves of that matter. A congestion surcharge justified by “Karachi delays” can be tested against public anchorage data. But the same data says nothing about container availability, slot cancellations, free days, detention or demurrage — and those are where landside cost actually lands. The Karachi board for 27–28 August lists 14 expected arrivals against 2 expected sailings, which is worth watching: sustained arrivals against thin sailings is how equipment ends up stranded inland even when the anchorage is empty.

Payload remains the cheapest saving on the table. At $1,850 all-in, the difference between loading 25 and 28 tonnes is $7.93 a tonne — wider than the entire current gap between the Pakistani bid and the European offer. It improves without any carrier conceding anything.

Route and energy

Oil reversed on a political headline rather than on inventories. Brent traded around $88.92 a barrel and WTI around $82.42, Brent up about $2.15 (2.5%), after the White House said there are no direct US–Iran negotiations underway — which reduced confidence in a near-term reopening of the strait. The commentary draws the obvious inference: a crude rebound weakens the case for assuming imminent freight relief.

Traffic through the Strait of Hormuz was reported at 10 commodity vessels on Wednesday against 8 on Tuesday, versus a ten-day average of about 15. ⚠️ Note that those are the same figures published yesterday — Wednesday to Wednesday. Today’s report does not advance the series by a day, and this particular count has been revised upward repeatedly. Read it as no new observation rather than as a fourth consecutive improvement.

Iran and Oman are still working on the details of a strait agreement, with no final terms.

One point of care on the most useful number in this story. Yesterday Reuters put Hormuz oil and LNG flows at about one-quarter of pre-war levels, on ship-tracking data. Today the same source is rendered simply as “well below pre-war levels.” A quantified finding became a qualitative one inside twenty-four hours.

The quarter figure is worth holding onto with its date attached, because it is materially more informative than a vessel count — and materially less alarming. Ten transits against a pre-conflict norm of 130–140 a day is about 7%; actual throughput at 25% means the traffic still moving is skewed toward larger or fuller ships. The counts overstate the disruption by roughly three times. That is not an argument for treating the corridor as normal — a quarter of normal flow is a badly impaired waterway. It is an argument for quoting throughput and direction rather than daily hull counts.

Supplier markets

Europe and the Baltic remain where the transacted evidence is, at $369–$370 for HMS 80:20 into Turkey on 25 August data. No newer cargo was reported today from any origin.

US commentary remains soft to sideways rather than bullish, and Canadian domestic ferrous weakened in parts of Ontario and Quebec through August — both restated without a fresh dock or yard number behind them.

And that gap deserves saying out loud, because it is now a month old. Exact current UK and US East Coast dock buying prices were again unavailable from reliable public sources. Where dock data is missing, an approach to a supplier should be framed as a question about their market, never an assertion about it.

Policy

The US measure covering black mass and tungsten waste and scrap is now in force, and today it became specific enough to screen against.

The covered material is identified by tariff line for the first time: black mass under Schedule B codes 8549.13.00.00, 8549.14.00.00 and 8549.19.00.00, and tungsten waste and scrap under 8101.97.00.00. The measure also has an end date — it runs through 27 August 2027 unless adjusted or extended.

The mechanism is worth understanding precisely, because it is not an export control. US sellers of covered material must allocate 100% of monthly sales to US persons, and the material must remain physically in the United States, unless the Bureau of Industry and Security grants an exception or adjustment.

Two practical consequences follow. First, the allocation is measured monthly — it is an accounting obligation on the seller, which is why nothing produced at a port can cure it. Second, there is no licence to apply for. The screening question is not “do you hold an export licence” but “is your material covered, and do you hold a BIS exception or adjustment?”

A twelve-month instrument with a stated expiry is also a different planning object from an open-ended restriction. Anyone describing it as permanent is overstating it. The thing genuinely worth watching is the precedent rather than the scope: an allocation order can be extended to another scrap stream without new legislation.

On Pakistan, policy reporting returned after two quiet sessions, covering the electricity-linked reduced sales-tax treatment for qualifying steel producers, Pakistan’s position among non-OECD countries seeking continued eligibility to receive qualifying EU waste, and the requirement for written eight-digit classification before loading unusual ferrous grades. One caution accompanies the tax point and it is a fair one: the treatment may make imported scrap structurally attractive to mills that qualify, but it does not create demand by itself — qualification is producer-specific and should be confirmed rather than assumed.

Non-ferrous snapshot

Three-month copper around $14,236 a tonne, aluminium around $3,216 and zinc around $3,890, on the latest exchange data. All three fell over two published sessions — copper about $100, zinc about $43, aluminium about $23. Anyone who re-cut a metal-bearing formula upward earlier this week should now be cutting it back down.

But the more useful observation today is about references rather than prices.

A widely circulated day-delayed feed shows copper around $14,253, aluminium $3,225 and zinc $3,893. Those are exactly the same six figures that feed published yesterday, unchanged to the cent — while the exchange prices moved beneath them.

The result is instructive. Two sessions ago that feed sat $83, $14 and $40 BELOW official prices. Today it sits $17, $9 and $3 ABOVE them. Same numbers. Opposite sign. Two days apart.

The lesson generalises well beyond this one feed. A stale reference does not err consistently in a helpful direction — it errs in whichever direction the market last moved, so its sign cannot be known in advance. There is no correction factor to apply and no side of a trade on which it is conservative. Price off a same-day official at the moment of quoting.

And note the trap inside it. Zinc’s $3 gap looks like agreement between two sources. It is not agreement — it is a stale series crossing a moving one. Coincidental proximity on the day of a crossing produces a false pass on exactly the day a reference is least reliable. Check whether a number moved, not whether it matches.

On market structure, copper and zinc are both in backwardation and both widened — copper to about $254 and zinc to about $217 — with zinc’s cash price unchanged while the three-month fell. Aluminium’s contango narrowed to about $3.50. All three metals are falling from the back of the curve while the front holds, which is a tight market selling off rather than a loosening one.

Nickel, lead and tin figures continued to circulate unchanged from the same carried table. Unchanged is not the same as confirmed. Thinly-quoted metals deserve more caution than the majors, not less — and mixed or nickel-bearing lots should be valued on multiple XRF readings and a weighted recoverable-metal calculation, never off a headline.

One general point that today made concrete. Figures published as “three-month” carry a tenor but no price type. Cash, official, closing and screen quotes are different series. A formula naming the metal and the tenor but not the price type — and not the fixing date — is not fully specified.

Glossary

• CFR — Cost & Freight; the quoted price includes ocean freight to the buyer’s port. FOB — the price at the loading port, freight excluded.
• HMS 80:20 — Heavy Melting Scrap, grades 1 and 2 in that ratio. Shredded is processed, size-reduced scrap: cleaner and dearer.
• Billet — semi-finished steel, cast from melted scrap and later rolled into products such as rebar.
• Bid and offer — what a buyer will pay and what a seller asks. The gap is the spread; a deal prints somewhere inside it.
• Assessment — a price a publisher judges to represent the market on a given date. It carries that date, not today’s.
• Month two — the second-nearest futures contract. For a late-August date that is September, not prompt material.
• Cash vs three-month — metal for immediate versus three-month delivery on the exchange. Cash dearer is backwardation; three-month dearer is contango.
• Official price — the exchange’s formally published daily fix. Distinct from the closing price and from electronic-screen quotes.
• Peak season surcharge — a temporary per-container charge added by a carrier on top of base freight.
• Payload — the tonnage actually loaded into a container. Per-container charges divided by a smaller payload produce a higher cost per tonne.
• Anchorage waiting time — how long a ship waits at anchor before a berth. It measures vessel queuing only, not container or equipment availability.
• Free days, detention and demurrage — the days a container may sit before charges begin, and the charges that follow. Landside congestion is paid for here, not in the freight rate.
• Schedule B code — the US export classification number for a commodity. Coverage under a trade measure is determined by the code, not by the description.
• DPAS allocation order — a US instrument directing that covered production be supplied to domestic buyers. It commandeers material rather than licensing its export.
• XRF — X-ray fluorescence, a handheld method of reading the elemental composition of a metal lot.