On Thursday morning Walmart posted a set of numbers that looked almost too good on paper. Revenue for the second quarter of fiscal 2027 rose 5.9 percent to 187.9 billion dollars. Adjusted earnings came in at 81 cents a share against the 74 cents analysts had modelled. Operating income grew 28.8 percent. The gross profit rate improved by 96 basis points. The company raised its full-year outlook. And then the stock fell about six percent in premarket trading.
The easy response is to call that a mood swing, and the easy response misses what actually happened. The market did something quite precise on Thursday: it largely set aside the profit line and priced the one metric that is genuinely about the customer. Because a substantial part of this quarter’s profit jump did not come from retailing at all. It came from U.S. Customs and Border Protection, in the form of a 2.9 billion dollar refund that Walmart is receiving because the Supreme Court struck down the legal basis for an entire round of tariffs. Anyone trying to understand how a 29 percent profit increase produces a six percent selloff has to keep those two cash flows apart.
Two growth rates for the same profit — 28.8 versus 17.4 percent
The release carries two figures for operating income growth side by side, and the distance between them is the real news. As reported, operating income grew 28.8 percent. Adjusted and in constant currency, it grew 17.4 percent. More than eleven percentage points separate the two, and that gap holds essentially two things: currency effects and the tariff refund.
The 96 basis point improvement in the gross profit rate tells the same story from another angle. For a retailer running a gross margin in the 24 to 25 percent range, most of a percentage point is a very large move. Read as operating performance, it would imply that Walmart suddenly buys far better or sells far dearer. Either would be a minor sensation for a company whose entire model rests on a promise of low prices, and neither happened. What happened is that duties booked as cost of goods in earlier quarters flowed back in this one and pushed the same line in the opposite direction.
That is accurate accounting and still misleading if you extrapolate a margin from it. A refund is not income from operations. It is the correction of a payment that should never have been made. It appears once. It does not recur. And it says precisely nothing about whether Walmart will sell more detergent next quarter.
Where the 2.9 billion comes from
On February 20, 2026, the Supreme Court held in Learning Resources, Inc. v. Trump that the International Emergency Economic Powers Act — an emergency statute dating from 1977 — does not give the president authority to impose tariffs of indefinite scope on his own. The ruling covered the so-called reciprocal tariffs as well as the levies described as fentanyl duties, which together applied to goods from very nearly every country of origin, the European Union included.
The scale of the decision is not in the opinion, it is in the till. Roughly 165 billion dollars in duties were collected under that authority, spread across more than 53 million customs entries filed by over 330,000 importers. The Court of International Trade ordered Customs to repay it. The agency has built a dedicated electronic process for the job, whose first phase went live on April 20, 2026. It does not pay out by itself: an importer has to come forward, produce its importer-of-record numbers, file declarations and register for electronic disbursement. Skip any of that and you get nothing, however right you were.
Walmart did not skip it. According to Chief Financial Officer John David Rainey, the company was eligible for roughly 2.9 billion dollars, with just under 100 million still outstanding at the time of the print. The trajectory of that estimate is worth noting: in May the same company put the figure at about 2.4 billion. The difference is half a year of process work, and it shows how much the actual recovery depends on how thoroughly an importer reconstructs its own customs history.
Why a refund is not a profit — and why Walmart says so itself
The most revealing statement of the day is not in the earnings release. It is in how the company handled its own guidance. Walmart kept the expected tariff recoveries out of it. Rainey had explained the reasoning in the spring: the company felt it best to provide guidance reflecting expectations for the underlying business, excluding any recovery of tariffs paid. Translated, management itself does not count this money as the business.
Consistently, it is not meant to stay in the business either. Rainey said the single best return available on a dollar of capital right now is to invest in the customer and in price, and that the effect of the refunds should show up as lower prices in the third quarter. That completes the map of where the money travels: in as a one-off through second-quarter gross margin, out again in the third quarter as price investment. The margin that looks so impressive this week is a way station, not a destination.
For valuation that has an awkward consequence. Applying a multiple to this quarter’s earnings per share means applying a retailer’s price-to-earnings ratio to a customs disbursement. That is exactly what the market declined to do on Thursday. And because the raised guidance — net sales growth of 4.0 to 5.0 percent in constant currency, up from 3.5 to 4.5, and adjusted earnings of 2.80 to 2.87 dollars a share, up from 2.75 to 2.85 — explicitly excludes the refund, what remained as the basis for valuation was precisely what the business produces without a court ruling.
2.6 percent — the only number that is about the customer
And that number came in soft. U.S. comparable sales grew 2.6 percent against a consensus of 3.7 percent. It is not a disaster, but it is the one metric at a retailer that no special item can substitute for: it measures whether the same store sold more than it did a year ago.
Part of the shortfall has an identifiable cause. Around 80 basis points came out of the health and wellness business, where price caps on certain medicines took effect. Strip that out and the core business grew about 3.4 percent, and the growth came from transactions — from more shopping trips, not higher prices. Qualitatively that is the better version, because traffic growth is harder to buy than ticket growth and tends to last longer. At Sam’s Club, the warehouse arm, U.S. net sales rose 8.8 percent to 25.7 billion dollars.
The tension still stands, though. The company visibly earns more while selling more slowly across the floor than expected. For a stock that has spent recent years being valued like a quality compounder rather than a grocer, the second half of that sentence is the expensive half.
The real margin lever is advertising, not customs
It would be wrong to conclude from all this that nothing in the margin improvement is real. The real part simply sits somewhere else. Global advertising revenue grew 38 percent in the quarter. Membership fee income rose 17 percent globally. E-commerce grew 23 percent worldwide and 24 percent in the United States. Inventory rose 6.7 percent, barely faster than sales, which is respectable discipline in a quarter with moving procurement costs.
Advertising, memberships and marketplace fees have a property retailing lacks: they scale with almost no incremental cost of goods. An extra advertising dollar on a product page that exists anyway falls nearly whole to operating income. An extra dollar of canned-goods revenue, at a gross margin of a quarter, delivers a quarter. When a retailer shifts the composition of its earnings toward those revenue types, the group margin rises without a single item on a shelf getting more expensive.
That is the structural change capable of justifying a higher multiple, and it is exactly the change the tariff refund obscured this week. An advertising business compounding at 38 percent is a recurring lever. A court decision is not. The fact that both landed on the same line in the same quarter makes the numbers harder to read, not better.
Who actually collects the refund — and who gets nothing
Here an accounting question turns into a competitive one. The Court of International Trade has made clear that court-ordered refunds on entries that have already finally liquidated go only to importers that filed suit themselves. Roughly 3,700 such cases are pending. For context: more than 330,000 importers paid the duties. For the overwhelming majority there is currently no confirmed equivalent mechanism; they are dependent on unresolved appeals and on whether a class is certified.
That produces a redistribution that has nothing to do with trade policy any more. Two importers may have taken the same container out of the same port at the same duty rate. The one with a legal department that filed on time in the spring of 2026 and protected its protests gets its money back with interest. The one without may get nothing. The tariff was identical for both; the refund is not. The process ends up functioning as a subsidy for scale and legal sophistication, and Walmart’s 2.9 billion dollars is the single most visible line item in that redistribution.
A second asymmetry matters more to suppliers than to shareholders and still touches both: the refund goes to the importer of record, the party that filed the entry. The customs process makes no attempt to determine who bore the duty economically. If a vendor granted a price concession in 2025 to make the tariff bearable for its American buyer, the vendor paid and the buyer is repaid. Disputes between contracting parties over who ultimately keeps the tariff money are now a live area of commercial litigation in their own right.
What it means for investors outside the United States
For anyone holding international consumer names, this story has two hooks. The first is a reading instruction. Every company with meaningful U.S. import volume will be able to show a comparable one-off in this year’s accounts, and in each case it is worth checking whether management excludes it from guidance the way Walmart did, or whether it quietly disappears into an improved margin. The second variant is the dangerous one, because it creates a trend line that cannot be repeated next year.
The second hook is exposure itself. Consumer-facing importers with heavy U.S. sales — Nike and Lululemon in apparel, Whirlpool and Stanley Black & Decker in hardlines, and on the European side names such as Adidas, LVMH or Moncler — carried the 2025 tariff round in their cost structures, partly as direct duty through their own U.S. subsidiaries, partly as concessions to trade partners. The decisive question for each is who appeared formally as the importer. A company importing through its own U.S. entity sits in the refund chain; a company shipping to an American wholesaler has to renegotiate a contract. For domestic peers such as Target, Costco, Kroger and Dollar General, the useful comparison is different: their next prints will show whether soft traffic is a Walmart problem or an American consumer problem.
The case against this reading
One objection deserves serious weight: refunded money is real money. A company with 2.9 billion extra dollars in the till that spends it on price is buying market share in an environment where many competitors have no such ammunition — the 330,000 importers without a lawsuit above all. The one-off leaves the income statement, but its effect on competitive position can stay. Economically that is the opposite of a bookkeeping artefact.
A second objection concerns the weak comparable. A 2.6 percent print with an 80 basis point drag from regulated drug pricing is not the same animal as a 2.6 percent print from a tired consumer. Price caps on medicines reduce revenue without anyone having stayed home; on the contrary, the customer pays less for the same product. Look at the core business at 3.4 percent with rising traffic and you see a healthy retailer.
And finally the customs process itself is unfinished. Appeals are running, the disbursement mechanics are only partly built, and timelines in this matter have slipped more than once. Treating the refund as a fixed quantity for other companies means counting on money that has been awarded in court but has not landed everywhere.
The third quarter is the actual test
The next set of numbers will either confirm this reading or break it, and at a very specific point. If the gross margin gives back visible ground in the third quarter while comparable sales accelerate, Walmart will have done what it said: converted the refund into price and bought traffic with it. That is the best of the available outcomes, even though it will not feel good at the earnings-per-share line at first. If the margin stays high and traffic stays weak, the price pledge was rhetoric. If both fall, the American consumer is the problem, not the bookkeeping.
The transferable lesson holds either way. In a year when a court decision pushes 165 billion dollars back from the state to companies, the profit line temporarily becomes a worse indicator of business performance than same-store revenue. The market demonstrated that in real time on Thursday: 28.8 percent more operating income was answered with a six percent decline, because one number — 2.6 instead of 3.7 percent — was the only one no government agency had wired in.
Try TradingView Free for 30 Days
Plus get a $15 discount on your first subscription through this link.
