Newsletter · · Ashutosh Agarwal

Tariff Refunds Are Flowing and Quietly Inflating Margins - The Squeezed Consumer: Tariffs & Trade-Down - Week of August 11, 2026

Consumer and retail newsletter for the week of August 11, 2026. Liberation Day tariff refunds are flowing back fast, padding this quarter's consumer margin beats at Under Armour, Apple, Amazon, Disney, e.l.f. and Columbia, while operators warn the apparel P&L was broken by markdowns long before tariffs.

The Squeezed Consumer: Tariffs & Trade-Down

Week of August 11, 2026: Tariff Refunds Are Flowing and Quietly Inflating Margins

A few months ago the smart money assumed the tariff refund money was a mirage. The courts would drag it out, the government would appeal forever, and any company banking on getting cash back was kidding itself. So treat the tariff as a permanent cost, we all said, and move on.

Well. About that.

This week we got the receipts, literally. Checks are going out, they're big, and they're landing right in the middle of earnings season. That changes the story in two directions at once. It's a windfall for the companies getting paid back. And it's a trap for anyone reading this quarter's margins at face value, because a chunk of the "beat" you're seeing isn't the business getting better. It's a one-time refund dressed up as profit. Let's get into it.

TL;DR

  • The Supreme Court knocked out the "Liberation Day" tariffs, and the refunds are real and fast. Roughly $100 billion already paid, another $29 billion processed, more than 75% of the $166 billion owed now out the door, "much faster than anyone anticipated." Over 250,000 refund claims were filed.
  • A lot of this quarter's margin beat is refund money, not operating money. Under Armour got a $70M benefit that pushed gross margin up ~6 points; Apple booked $2B+ (adding $0.11 to EPS); Columbia Sportswear got $78M; Amazon $600M+; Disney $100M; e.l.f. Beauty $50M.
  • Columbia's CEO says the refund is a balance-sheet event, not a pricing event. He "ate" 2025's tariffs, can't reprice goods already pre-sold to retailers, and is spending the cash cautiously, some of it going back to the vendors who helped share the pain.
  • The real apparel problem is older than tariffs. A 30-year veteran of Nike and Target laid out why the apparel P&L is broken: 200+ day lead times, ~40% of the assortment ending up on markdown, and a $300B+ annual US markdown bill. Tariffs just exposed it.
  • The China exit shows up in hard port data. China has fallen from ~60% to ~40% of Port of LA imports, but it's a re-route into Southeast Asia, not a return home.
  • The Iran war is a live cost shock. Vessel fuel up 60% at LA, pump gas up 25%, diesel up over a third, and two-thirds of that port's cargo moves by truck.
  • The discounters were silent again. Nobody talked Walmart, Dollar General, TJX or the trade-down trade this week. Noting the gap, not filling it.

What's New

1. The refund money is real, and it's inflating the earnings you're looking at. This is the single biggest shift of the week. On CNBC's Squawk on the Street, trade reporter Megan Casella (a journalist, not an operator) reported that after the Supreme Court struck down the "Liberation Day" duties, the administration has already refunded about $100 billion, with another $29 billion in the pipeline, over 75% of the $166 billion it owed, and "much faster than anyone anticipated." More than 250,000 companies filed claims (Squawk on the Street, Aug 7). Why it matters: this is the exact opposite of what the consensus assumed a few weeks ago, and it means real cash is hitting corporate balance sheets right now.

The catch, and this is the part a PM needs to circle, is where that money lands in the financials. Casella ran the tally: Under Armour flagged a $70M benefit that pushed gross margin up "nearly six percentage points" year-on-year; Apple booked $2 billion+ and added $0.11 to EPS; Amazon more than $600M; Disney $100M; e.l.f. Beauty $50M (Squawk on the Street, Aug 7). As she put it, the refunds are "quietly inflating the earnings picture." Anchor Carl Quintanilla noted that for Under Armour, "nearly half of their operating profit target is going to be tariff-based or refund-based." Translation: strip the refund out and some of these beats look a lot more ordinary. This is a quality-of-earnings flag, not a growth story.

2. Columbia's CEO: the refund is cash, not a pricing lever, and it doesn't come back to you at the register. This is the operator commentary that outranks the rest, because Tim Boyle actually runs the business. Columbia Sportswear got a $78M refund, "we have gotten virtually all of our duties back that we paid under the IEPA system." But he was blunt that it changes almost nothing about how the company prices: its products are "pre-sold to retailers" and bundled in its own stores, so "we can't change the prices on this stuff. So we basically ate the cost of the tariffs in 2025" (Squawk on the Street, Aug 7). Why it matters: for anyone hoping tariff relief flows through to lower shelf prices, or to sustainably fatter retailer margins, Boyle just said no. The refund strengthens the balance sheet, some goes back to the vendors who helped share the tariff hit, and he's spending it cautiously.

He also drew the map of where the real risk sits. Columbia does "a little bit over 40%" of its business outside the US, and footwear and apparel are "virtually impossible to make in the United States", some of his factories employ 100,000 people. The killer line for modeling next year: he's selling to retail partners today for delivery in 2027, and "we know what we can sell it for… but we really, in many ways, don't know what we're going to pay for the merchandise." That "capricious" tariff uncertainty is already making him more cautious on operating expenses. Notably, when asked directly about trade-down, he said the consumer is "more fragile… still quite strong," and pointed out his business is much healthier outside the US than inside it, a direct read on what tariffs are doing to domestic pricing.

3. The apparel P&L was broken before tariffs, tariffs just turned the lights on. The most useful teardown of the week came from Maithili Shenoy, a 30-year operator (she led inbound inventory at Nike through Covid, then ran sourcing across 40+ private-label categories at Target, where she grew nearshoring enough to deliver $150M in tariff mitigation). Her argument, on Supply Chain Now: the industry chased cheap labor for two decades, China, then Vietnam, Indonesia, Bangladesh, now Kenya, and in doing so stretched total lead times from about 180 days to 200+ days, forcing buyers to commit "nine months in advance." Back when sourcing was near-shore, full-price sell-through ran 85-90%. Now roughly 40% of the assortment ends up marked down, and the US apparel industry eats a $300 billion+ annual markdown bill, with 86% of retailers blaming inventory misalignment (Supply Chain Now, Aug 10). Why it matters: this reframes the whole apparel tariff debate. The tariff is a cost shock on top of a model that was already leaking margin through markdowns, so the companies with tight, fast inventory will absorb tariffs far better than the ones drowning in nine-month bets.

4. The China exit, measured at the dock. Gene Seroka, who runs the Port of Los Angeles, gave the cleanest hard number on de-risking I've heard: China has fallen from about 60% of the port's imports at the peak to roughly 40% today, even as total volume kept growing (Inbound Logistics: Supply Chain Reactions, Aug 5). But he was careful: this isn't reshoring. China's latest five-year plan (last November) is explicitly about exporting manufacturing expertise, not just products, building factories across Vietnam, Indonesia, Malaysia, Thailand and Cambodia. Why it matters: the freight and the tariff base are moving to Southeast Asia, but the supply chain is still Chinese-run. The re-route is real; the decoupling is mostly cosmetic.

5. The Iran war is a fuel shock hiding in your landed cost. Same conversation, and it's an operator flagging it: Seroka said vessel fuel is up 60% in LA since the war began, pump gas up 25%, and diesel up "more than a third." Since two-thirds of the port's cargo moves onward by truck, much of it via small operators who "have great difficulty absorbing" higher fuel, those costs get passed to shippers and, eventually, to shelf prices. Fuel is now 25-30% of a vessel voyage (Inbound Logistics: Supply Chain Reactions, Aug 5). Why it matters: it's a second cost line, independent of tariffs, quietly pushing landed costs up right as inventories are already bloated from pre-tariff front-loading.

The Debate

Is apparel facing a temporary tariff bump, or a permanent gross-margin reset?

This week the operators leaned hard toward reset, but with an important twist. Shenoy's whole point is that tariffs aren't even the main event. The margin damage was already baked in by long lead times and a metric obsession with gross-margin rate instead of inventory productivity. Her fix is to stop worshipping gross margin percentage and start measuring GMROI, gross margin dollars earned per dollar of inventory. Her example makes it vivid: a business at 38% gross margin but only 3.2 inventory turns generates $2.16 per dollar of inventory; a business at a lower 30% margin but 6 turns generates $5.10, more than double the cash, from the "worse" margin (Supply Chain Now, Aug 10). The read: the companies that treat tariffs as an excuse to fix a broken model will re-rate; the ones that just wait for tariffs to go away are the reset risk.

The counter-voice this week was Boyle, not arguing the bull case exactly, but showing the middle path. Columbia ate 2025's tariffs, got vendor help, and is now sitting on a refund and a strong balance sheet, treating the whole episode as something to survive rather than a permanent margin cliff. That's the "we're big and well-capitalized enough to absorb it" argument. It works for Columbia. It does not obviously work for a leveraged, markdown-dependent mall retailer.

The trade-down debate, meanwhile, still isn't happening on the podcasts. Nobody made the case that trade-down into discounters is a durable share-gain story, and nobody made the opposite case that it's a late-cycle head-fake. The only consumer read this week came from Boyle calling the shopper "more fragile… still quite strong."

The Names in Play

Guests actually moved theses on specific stocks this week, so here they are, woven, not rattled off.

Shenoy's clear favorites are the velocity operators. Inditex (Zara) is her poster child: FY2024 gross margin of 57.8%, inventory costs of just 7.2% of sales (versus an industry average of 13-15%), ROIC around 20%, and full-price sell-through near 85%, all because it accepts higher unit costs in exchange for speed (Supply Chain Now, Aug 10). Her turnaround story is Abercrombie & Fitch, whose "Always Forward" discipline took operating margin from 1% in FY2021 to 15% in FY2025, revenue from ~$3B to nearly $5B, and free cash flow to about $1B, and which the market re-rated accordingly. Ralph Lauren, she says, is walking the same path now.

Her cautionary tale is Lululemon, which she says "flooded" the market and has "trained consumers" to wait for the "We Made Too Much" clearance rack, a self-inflicted markdown machine. Worth pairing that with a genuinely new demand risk she surfaced: GLP-1 weight-loss drugs are driving apparel returns up over 50% year-on-year (citing a Wall Street Journal piece from the past week), as customers drop several sizes in months and send back whole wardrobes, brutal for anyone with a long return window and a fat private-label mix.

And Columbia Sportswear itself: the bull case is the fortress balance sheet plus innovation-led share gains abroad (its Amaze Puff jacket doubled sales this year; China and Europe are growing). The bear case is Boyle's own admission, he genuinely doesn't know what next year's product will cost him. Next catalyst to watch there: how much of that $78M refund shows up as capital returns versus a war chest for the next tariff round.

Read-Throughs

  • Quality-of-earnings screen, right now. Any consumer name that beat this quarter deserves the question: how much of the beat is a one-time IEEPA refund? Under Armour (half its operating profit target), Apple, Amazon, Disney, e.l.f. and Columbia all named refund dollars this week. The refund is cash, but it's not a run-rate.
  • Southeast Asian contract manufacturers are the structural winners of the China re-route, but remember Seroka's point that the factories are increasingly Chinese-built and Chinese-run. The tariff address changes; the ownership often doesn't.
  • Freight and fuel are a live short-term cost, not just a tariff footnote. With vessel fuel up 60% and diesel up a third, the trucking-heavy last mile out of the ports is getting squeezed, watch small-carrier capacity and any fuel-surcharge language into peak season.
  • Off-price and velocity retailers should widen the gap versus markdown-dependent peers if Shenoy is right that the market rewards inventory productivity over headline margin. The "adapt or get acquired" line was aimed squarely at the laggards.
  • The dollar-store / discounter read-through is still a blank. For the second issue running, the trade-down names went unmentioned. That's a coverage gap to chase, not a signal that nothing's happening there.

What Changed

The big reversal versus where we were two weeks ago: the refund door didn't close, it blew wide open. In the last issue the working assumption was that the courts would grind refunds down and the smart move was to treat tariffs as a permanent, unrecoverable cost. This week the Supreme Court struck down the "Liberation Day" duties and the government is paying back the $166 billion at a pace nobody expected, over 75% already out. That's a genuine change in the facts on the ground, and it flips refund optionality from "closing" to "cash in hand."

Second change: an apparel operator finally spoke on the record about pass-through. For two issues we'd been waiting for a named brand to actually discuss eating versus passing tariffs, Boyle did exactly that, and the answer was "we ate it, and the refund doesn't reset our prices."

Third: we now have hard port numbers on the China re-route (60%→40% at LA) and a quantified fuel shock from the Iran war, both of which had been narrative before and are now measured.