Retail Week, and the 40x Problem Nobody Wants to Name

Before Walmart opens this morning, sit with one number: the stock trades at nearly 40 times forward earnings. Nvidia, the company selling the actual physical infrastructure of the artificial intelligence buildout, trades at roughly 25x. The market has decided that a grocery-and-general-merchandise retailer growing revenue in the mid-single digits deserves a richer multiple than the chipmaker whose products are, by consensus, reshaping the global capital stock. Say that out loud to a client and watch their face. This isn't a Walmart problem specifically. It's what happens to valuation discipline in a market that's forgotten what "expensive" is supposed to mean.

Here's the setup you're walking into. Target reported Wednesday with sales up 5.3% year over year, raised full-year guidance, and beat on both lines. Walmart reports today, one day after Target, and the framing everyone on the sell side has settled on is whether Target is quietly taking market share from Walmart. A Placer.ai foot-traffic read has Target visits up 4.7% year over year in the quarter against just 0.7% for Walmart. That gap is small in absolute terms and enormous in narrative terms, because it inverts the story this sector has told itself for three years: Walmart as the all-weather value winner, Target as the turnaround story that might not turn around. Somewhere in the last two quarters that flipped, and the market hasn't finished repricing it.

What actually moved at Target is worth being precise about, because the desk keeps flattening it into "beat." $994 million of the quarter's operating income came from IEEPA tariff refunds, worth $1.65 of the $4.11 EPS print. That's not a knock on the traffic numbers, which were real — comparable traffic up 3.6%, genuine demand response to Fiddelke's merchandising reset. But the earnings acceleration the market is paying up for is partly a fiscal transfer tied to a tariff regime that is, itself, an input-cost headwind for the entire sector. Walmart doesn't get the same refund cushion in the same proportion relative to its size, and its own CFO commentary today will tell you a great deal about whether "value retailer" still means something when fuel and import costs are structurally higher than they were two years ago.

Structurally, here's the tension the desk needs to be pricing rather than reacting to. July retail sales fell 0.6% month over month, the weakest reading since May 2025, which is the demand backdrop both these companies are reporting into. The Fed held rates steady in July on a 9-3 vote, with three officials dissenting in favor of a hike, a level of internal disagreement that tells you monetary policy has no clean read on this economy either. Retail earnings are supposed to resolve that ambiguity for investors. Instead they're adding a second layer of noise — a tariff-refund accounting effect large enough to swing headline EPS by 40% at one major retailer, landing in the same week the labor market print showed the worst payrolls miss in years. Nobody on this desk should be treating "beat" as a clean signal right now. It isn't one. It's two distortions stacked on top of each other, and untangling them is the actual job this week.

On valuation: multiple compression is usually a slow bleed, priced in over quarters. It can also happen in a single session when a market that's been paying for narrative discovers the narrative was thinner than the multiple implied. Sell-side positioning ahead of today's print is already split — Oppenheimer downgraded the stock this week, Bernstein held its rating but cut its price target from $147 to $142 — which tells you the professional money isn't unified on which story wins even before the print lands. A 40x multiple prices in years of flawless execution. It does not price in a value retailer losing foot traffic to a rival that just admitted a chunk of its own earnings growth was a government check.

Desk positioning: treat today's release as a test of two separate claims, not one. First, whether Walmart's core value proposition — the thing that's supposed to make it recession- and inflation-resistant almost by construction — is actually holding against Target's reinvention, or whether the traffic data is an early tell of real erosion. Second, and more importantly for anyone pricing this sector on a multiyear view, whether the market is willing to keep paying an AI-infrastructure-grade multiple for a company whose growth engine is same-day delivery penetration and advertising revenue, in an economy where the Fed itself can't agree on the direction of travel. If the print disappoints on either axis, don't be surprised if the multiple gives back more than the earnings miss justifies on its own. That's what happens when a valuation stops being about the business and starts being about faith that nothing bad can happen to it.

Flag any client questions on tariff-refund accounting treatment to compliance before responding — this is going to come up, and "one-time benefit" is not a complete answer.