UBS Just Pegged the Next Five Years at 40,000 Store Closures — and AI Is the New Variable

UBS analysts led by Michael Lasser raised their five-year U.S. store-closure projection to more than 40,000 this week, citing the combination of e-commerce share gains and a new accelerant: AI-aided shopping. The note arrived the same week Moody's downgraded its apparel outlook and the U.S. lost 5,000 net stores year-over-year.

UBS analyst Michael Lasser and his retail team published an update this week that anyone running a brick-and-mortar fleet should read carefully. The bank raised its base-case projection for U.S. store closures to more than 40,000 over the next five years, as Retail Dive reported. The number isn't shocking on its own — UBS has been publishing these closure forecasts for almost a decade — but the variable that pushed the figure up this round is new.

E-commerce is no longer the only force pulling sales out of stores. AI-aided shopping — agentic checkout, conversational commerce, and the LLM-driven product discovery that's now embedded in Google Search, ChatGPT, Perplexity, and the OpenAI-Walmart-Microsoft retail council's tooling — is now siphoning enough volume that the UBS team felt obligated to break it out as its own structural pressure. The bank's modeling now puts e-commerce penetration at roughly 27% of U.S. retail by 2030, up from just over 20% today and 10% in 2019, per the team's note as cited by Shopifreaks.

The data point that should keep buyers and landlords up at night isn't the projection. It's the trailing print. From Q3 2024 to Q3 2025 the U.S. net store count fell by roughly 5,000 locations, as the UBS team flagged. That's the first sustained net-negative store-fleet year since the pandemic recovery, and it broke the multi-year pattern of more openings than closures that landlords had counted on as a baseline. CoStar's parallel forecast, released the same week through StockTitan, said the same thing more politely: closures will stay elevated through 2026 even under a stable-economy assumption.

Department stores and specialty apparel are again the most exposed categories. Saks Global's restructuring is in its final exit-financing phase. Macy's has 14 stores closing this year as part of its 150-location wind-down. Francesca's is liquidating its remaining footprint this month, and Neiman Marcus Last Call is closing the rest of its outlet fleet. The Endcap team has been tracking each individual closure cohort, but UBS's update reframes the cumulative impact: this isn't a category clean-out, it's a structural contraction.

The bull-case carve-outs in the UBS note are the part the trade press is mostly ignoring. The bank named Walmart, Costco, Target, off-pricers (TJX, Ross, Burlington), and dollar stores as beneficiaries of the closure cycle — categories that absorb the orphaned trips when a Macy's or a Saks Off 5th leaves a market. Footwear got an unusual mention as a category that may buck the closure trend, Yahoo Finance noted, because the high-touch fit element of footwear retail is structurally harder to displace with online or AI-driven shopping. That's a reading worth testing — Foot Locker, Famous Footwear, and the DTC running brands that have built out physical fleets are all banking on something close to that thesis.

The UBS note arrived the same week Moody's reaffirmed its negative outlook on the global retail and apparel sector, per WWD's coverage. Moody's expects U.S. apparel and footwear retailers to remain pinched through the first half of 2026 even with the temporary 10% tariff cut on most imported goods, and to absorb most of the residual cost rather than pass it through. That sets up a particularly grim sequence: tariff-compressed margins this year, AI-share leakage compounding next year, and a closure cohort that UBS now thinks reaches 40,000 over five.

The downside scenario in the UBS note is the one that should be cited carefully. If U.S. population growth turns negative — a scenario the bank treats as low-probability but no longer impossible given current immigration policy and flat birth rates — the closure number could rise toward 70,000 over five years. That's a remodeling of the U.S. shopping-center inventory at a scale the country hasn't seen since the 1970s mall consolidation.

For retail real estate, the strategic move is already underway. Primaris REIT in Canada is repurposing former Hudson's Bay anchor boxes into mixed-use formats. RioCan is replacing dead anchor space with pickleball, fitness, and food-hall tenants. The U.S. mall REITs — Simon, Brookfield, Macerich — have been quietly converting Class B and C anchor pads to medical, residential, and entertainment for two years. The UBS forecast is a confirmation that the strategy is no longer optional: the inventory of viable retail tenants for traditional store boxes is shrinking faster than landlords can fill it, and the AI variable just made the curve steeper.

The category to watch for the next quarter is grocery-adjacent and convenience real estate. As DoorDash and Instacart pull more household-essentials volume out of stores (the Q1 prints earlier this week made the trajectory unmistakable), the dark-store and micro-fulfillment footprint becomes the new buildout. UBS's 40,000-closure forecast is in many ways a reshuffling number, not a vacancy number. The boxes that close will not all stay empty. But the retailer who occupied them probably won't be the one who reopens them.