The Margin the Tariff Ate

During Liberation Day week, a Saudi television network asked me to come on air and explain what the new tariffs would do to American industry. I said no. Partly because my business partner serves in the National Guard and we keep our firm a long way from geopolitics, and partly because the honest answer at that moment was that nobody knew, and honest uncertainty makes terrible television.

I could give the technical answer, though. I have spent years inside plants and supply chains where tariff exposure is a line on a costed bill of materials. And the technical answer, then and now, is that the tariff story in the lower middle market is really a margin accounting story, and most of the P&Ls I read are getting it wrong in both directions at once.

What Actually Happened, in Order

Walk the sequence, because the sequence is the whole analysis. Through 2025, importers paid a stack of duties layered on top of each other. The reciprocal tariffs and the trafficking tariffs on China, Canada, and Mexico were imposed under IEEPA, the emergency economic powers statute, and they sat on top of the older Section 301 duties on Chinese goods and the Section 232 duties on steel, aluminum, and a growing list of other categories. A single imported component could carry 3 separate tariff layers, each with its own legal basis, plus the underlying duty rate in the tariff schedule.

Then this past February the Supreme Court held that IEEPA never authorized those tariffs. The administration terminated them within days, and customs stopped collecting the IEEPA layers entirely. Roughly $142B had been collected under the struck-down orders in 2025, and the refund question went back down to the Court of International Trade, which has the power to order the money returned. A substantial portion of it is expected to flow back to the importers of record who paid it.

Here is what did not go away, and this is the part sellers and buyers both misread. Section 232 duties survived untouched, and they now reach steel, aluminum, copper, vehicles, trucks, timber, and semiconductors, with the metals duties sitting at punishing rates. Section 301 on China survived. The de minimis exemption for small parcels was killed again within days under a different authority. The average effective tariff rate on US imports, even after the biggest judicial reversal in trade history, still sits near its highest level since the 1940s. The regime narrowed, hardened, and got more legally durable.

Reading a 2025 P&L in 2026

Now put on a buyer's glasses, or a seller's, and read a manufacturer's trailing twelve months. It is a geological record of all of this, and normalizing it takes real work.

The cost side first. Quarters from 2025 carry IEEPA duty expense that, as of February, no longer exists. A shop that imported $6M of components under a 20% combined IEEPA layer was carrying roughly $1.2M of annualized cost that vanished by court order. If you value that business on unadjusted trailing earnings, you are underpaying the seller for a cost structure the law already repealed. This is the rare normalization argument that runs in the seller's favor, and I make it for them when their own advisors miss it, because paying a fair price on true earnings is cheaper than winning an argument and losing a deal.

The revenue side cuts the other way. Most operators pushed price through 2025, and the surcharges that stuck are now sitting in margin. Some of that margin is real and durable. Some of it evaporates at the first contract renegotiation, now that the customer's procurement team reads the same court decisions I do. Separating those two is the actual diligence work, customer by customer, and a tariff surcharge that was never broken out on the invoice is the hardest kind to defend.

And then there is the refund. If the business imported meaningfully in 2025, it may be owed real money, and I have yet to meet a lower middle market owner who has quantified the claim. It requires matching entry summaries to the struck-down orders, confirming the company was the importer of record and not its freight forwarder or supplier, and pursuing the claim through the process the courts are still shaping. On a $6M importer that money is potentially 7 figures. In a sale, it should be negotiated explicitly, priced or excluded, rather than left as a silent gift to whoever notices it first. When I find one, I put it on the table. The seller earned it. It arrived under his ownership.

Somewhere in the trailing twelve months of almost every American manufacturer sits a cost that no longer exists, next to a refund nobody has claimed.

Paul W. Swaney III

The Questions That Sort Operators

Tariff exposure has also become the cheapest management test I know. I ask an owner what his blended duty cost was last year as a percentage of material spend, and whether his HTS classifications have been reviewed by anyone other than his customs broker since the rates changed. The classification of a part, and the true country of origin under substantial transformation rules, can move a duty rate by double digits. Misclassification runs both ways too. Some companies have been overpaying for years, which is more found money, and some have exposure that a buyer's quality of earnings will eventually surface, which is better raised by the seller first.

The strong operators answer from memory. They know their exposure by supplier and by part family, they dual-sourced or re-quoted the worst of it during 2025, and they can tell me which customers absorbed the surcharge and which pushed back. The weak ones tell me their broker handles it. Both answers price the business, just in different directions.

The overlooked tools sit one layer deeper. Duty drawback returns 99% of duties paid on imported inputs that are later exported in finished goods, and most small manufacturers who qualify have never filed. Foreign trade zones and bonded warehouses defer duty until goods enter commerce, which matters when rates are moving between order and delivery. None of this is exotic. It is paperwork with a payback, and whether anyone in the building has done it tells me how much margin is still lying on the shop floor.

Underwriting Fog

Nobody can tell you the tariff schedule 2 years out. The struck-down layers may return under slower, more durable statutes. Section 232 keeps adding product categories. Anyone who claims certainty about the end state is selling something.

What a buyer can do is refuse to average the fog into a multiple. As a fundless sponsor, I underwrite one company at a time, so I can afford to build the tariff bridge line by line, entry by entry, for the single business in front of me, and price what I can prove. No committee needs a clean sensitivity table by Thursday. The margin the tariff ate is usually findable, one way or the other. So is the margin it left behind. The work is telling them apart before you sign, and that work does not average

I am the founder of Swaney Group Capital, a fundless sponsor focused in the lower middle market. LeverUp® is published weekly. More if I have something else to say.