Thermos — tariff exposure and the forward-buy window
On April 6, 2026 the basis for Section 232 metal tariffs changed. It stopped applying to the metal
content of a product and started applying to the entire customs value. For a company
whose products are substantially stainless steel, that is a step-change in landed cost — and every
price increase it forces opens a window in which retailers buy ahead at the old price.
Read this first. Thermos L.L.C. is privately held — a subsidiary of Taiyo Nippon Sanso,
under Nippon Sanso Holdings (TSE: 4091). No public financials exist for it. Every dollar figure
on this page is a model, driven by the assumptions you set below — not Thermos data. The tariff
rules are real and sourced at the foot of the page. The numbers are yours to set.
What changed on April 6, 2026
The rate did not move. The basis did — and for a derivative product the basis is most of the story.
Before
—
Section 232 applied only to the declared metal content of the product
→
From April 6, 2026
—
Section 232 applies to the full customs value, regardless of metal content
Derivative articles more than 15% steel, aluminum or copper by value pay the full
derivative rate on the whole entered value.
The old split of customs value between metal and non-metal content was eliminated.
Steel, aluminum and copper duties do not stack with each other — the highest applicable
rate governs.
Section 232 does stack with Section 301 China duties where both apply, subject to the
April 2025 anti-stacking order.
Model the landed-cost shock
Set the assumptions on the left. Everything on the right is computed from them — nothing is pre-baked.
232 duty — old basis (metal only)—
232 duty — new basis (full value)—
301 duty (stacks, full value)—
Increase per unit—
Price increase to hold margin—
Annual cost increase—
"Increase per unit" is the change in Section 232 duty alone — the basis change, holding the rate
constant. Section 301 is shown separately because it was already being paid on full value. The price
increase shown is the pass-through required to hold absolute gross margin per unit.
Duty increase per unit
—
from the basis change alone, same 232 rate
Increase, as % of duty
—
what the same rate now costs on the new basis
Annual cost increase
—
at the unit volume you set
Forward-buy exposure
—
one quarter of volume bought ahead at the old price
The commitment ledger is what makes this actionable
Knowing the tariff cost is arithmetic. Recovering it depends on whether you can produce what was
agreed when the deduction lands months later — sealed, timestamped, and unchanged since.
Why a tariff number alone does not recover the money
The cost is arithmetic. The recovery depends on four things that sit upstream of any software, and
that is where middle-market CPG most often loses the value.
The deal is agreed before it is recorded. Commitments get struck on calls and
deal sheets, then re-keyed at least once before reaching a system. The ledger moves the record
to the moment of agreement — the only moment it is certainly right.
Deduction coding is inconsistent. The same reason gets coded three ways by
three people, and matching on inconsistent codes produces confident, wrong answers. Thirty-five
years of knowing how deals are actually struck is what makes the coding right; no tool supplies
that judgment.
Most teams have no dedicated trade analyst. Platforms in this category quietly
assume one. The engagement supplies the analyst; the ledger is what lets the work survive after
the engagement ends.
The variance has to be explained upward. For a subsidiary reporting through a
foreign parent, a trade variance travels up a chain that was not in the room. Documentation that
proves the number is worth more there, not less.
Why this becomes a trade-spend problem, not a customs problem
Customs handles the duty. The margin damage happens afterward, in the commercial system.
The increase has to be announced. A cost shock of this size cannot be absorbed, so a
price increase goes to the trade with an effective date.
The announcement opens the window. Retailers get 30–90 days to buy at the old price.
For seasonal drinkware, that window can collide with the back-to-school build — the largest buy of the
year — and the loading is enormous.
Shipments stop telling the truth. One quarter is flattered by the buy-in, the next is
starved. Neither reflects a consumer.
Promotions then run on forward-bought inventory. The allowance gets calculated against
a price basis that was never agreed — old cost, new price, or the full loaded quantity rather than the
promoted quantity.
The deduction lands 60–120 days later, when the people who made the deal have moved on
and the commitment was a conversation, an email, or a deal sheet.
Naomi — grounded
The Section 232 basis change is a matter of public record and applies to any importer whose product is
substantially steel: the duty now attaches to the entire customs value rather than the metal content, and
it stacks with Section 301 on Chinese-origin goods.
Naomi — refused
Two things on this page cannot be asserted, and are not.
1 — Whether Thermos's core products are on the derivative list. Vacuum flasks classify
under HTS 9617. Commerce has added hundreds of HTS codes to the Section 232
derivative inclusion list through the petition process, but I could not confirm from public sources that
9617 is among them. It must be checked against the current CBP list before any of this is treated as a
live exposure. That check is itself worth doing — the answer is the difference between
zero and the figure above.
2 — Every dollar on this page. Thermos is private. There are no public unit volumes,
landed costs, or steel-content percentages. The figures here are whatever you set them to. They are a
model of the mechanism, not a measurement of the company.