S09.3 · Retail & Commerce
TJX, Ross and Burlington-led off-price retail is gaining from full-price distress and is comparatively insulated from spec-driven agentic shopping.
Off-price is beating full-price retail, and eMarketer (2026) expects trade-down demand and tariff-driven excess inventory to keep feeding it into 2026. No credible public aggregate market size exists — most published "market size" figures for the category are low quality — but the public anchor does the work: TJX at roughly $56B, Ross at roughly $21B and Burlington at roughly $11B in FY2025 revenue, a roughly $90B combined base with a meaningful independent and regional tail beyond it. What distinguishes the model is where the margin comes from: the buy, not the sell.
The segment is US-concentrated, with TJX carrying a notable European and Canadian presence through TK Maxx. The public players are highly concentrated; beneath them sits a fragmented independent and closeout tail. The chain runs on opportunistic buys of excess or canceled inventory from manufacturers and department stores, sold on to value-seeking consumers — a sourcing model that yields mid-to-high operating margins relative to full-price peers at low capital intensity per store. Buyer expertise, not scale, is the binding asset.
The economics are unusual for physical retail because inventory cost, not real estate or labor, is the primary variable. A format built around ever-changing, opportunistically-sourced merchandise commits far less capital per square foot than a full-price chain planning assortments a year in advance, and it converts distress elsewhere in the retail system — closeouts, canceled orders, overproduction — straight into gross margin. Which is why the category's growth tracks stress upstream so closely: tariff-driven excess inventory and full-price retailer distress are demand drivers for off-price, not headwinds.
Off-price gains structurally from distress in department stores and full-price apparel, so its growth is expansionary by construction rather than a defense of existing revenue. The open question is the buying desk: AI-assisted buying and allocation tools could sharpen the core sourcing skill — reading excess inventory and moving on it quickly — that has so far resisted automation.
Agentic shopping, the threat pressing on other formats, has less to grip here. The category runs on a "treasure hunt" browsing model in which the value proposition is discovery, not a known item at a known spec — a poor match for agents built to purchase against a defined specification and price. That leaves off-price less exposed to spec-driven agentic shopping than categories where the item and price point are known in advance. AI-assisted buying and allocation is expected to mature over a 2-5 year horizon.