The side supplying the money fell instead
The largest capital supplier to AI data centers is Blackstone. It led a $7.5B debt financing for CoreWeave and bought AirTrunk, the Asia-Pacific's largest data center operator, for $16B. Yet its stock moved opposite the AI boom: +83% in 2023, then -8% in 2025 and -19% in 2026. Over the same span NVIDIA, the chip inside those data centers, cooled from +239% to +5% but stayed positive throughout. The AI infrastructure the money flows into rose; the side supplying the money fell.
The five consumer and finance nodes this wiki now tracks (Blackstone, Brookfield, Visa, Walmart, Coupang) pose one question. AI capex propagated layer by layer through the tech chain, but does it spread to the real consumption and capital markets outside that chain? The returns answer clearly: it does not.
Capital runs on the interest-rate clock
Blackstone and Brookfield are capital deeply exposed to AI infrastructure. Brookfield ties the power layer to AI through a 10.5GW renewable framework with Microsoft. Even so, in 2026 Brookfield fell -11% and Blackstone -19%, side by side. An asset manager's stock moves on interest rates and the fundraising-and-realization cycle, not on how exposed its portfolio is to AI. Even capital that funds AI trades on the cost-of-capital clock.
Line up the whole capital layer, banks, asset managers and Berkshire, and the same clock repeats. What lifted JPMorgan, the largest arranger of data-center debt, was the deposit-margin spread rather than AI, and BlackRock, which supplies capital directly to AI data centers, actually lagged the banks. The name-by-name verification continues in Capital that funds the boom runs on the rate clock.
Consumption runs on the consumption clock
The consumer side is the same. Visa's revenue is payment fees, so it runs as a proxy for consumer spending. It put generative AI into VisaNet fraud detection and earmarked $10B in five-year technology spending, but its 2026 stock traced a mild consumption curve at +3.7%. Coupang, though it runs on AWS, fell -21% in 2026, taking Korea's consumption slowdown directly. Being an AI-adopting demand node does not change the clock the stock runs on. The consumer nodes added this round point to the same clock. Nike, a leading discretionary-consumer name outside the chain, fell -30% in 2026 even as semiconductors rose, pressed by weak Greater China sales and inventory correction. E-mart took Korea's consumption slowdown directly and ended -6%.
Adoption is cost-cutting, not a stock driver
Walmart shows the boundary well. It is a leading AI-adopting retailer, running search and workflows on Azure and Google Gemini. But its +74% in 2024 came from retail-media advertising and e-commerce margins, and it cooled to +0.8% in 2026. AI adoption was a tool to cut costs, not demand that lifts the stock. To the seller (NVIDIA) AI is revenue; to the user (Walmart) AI is a cost. So even the same AI points the two camps' stocks in different directions. Starbucks stands in the same place. It runs Deep Brew AI on Microsoft Azure, but that touchpoint is one line on an Azure bill, and its +25% rebound in 2026 was a turnaround hope made by bean prices and Chinese consumption. Further, even companies that depend entirely on the cloud do not run on the AI clock. Netflix, Airbnb and Disney run their services on AWS without data centers of their own, yet their stocks move on subscription, travel and content demand. Depending on AI infrastructure and running on the AI cycle are two different things.
The defensives rose instead, and only consumer tech was the exception
Widen the consumer set and the boundary sharpens. Coca-Cola, with virtually no AI exposure, rose +22% in 2026, the year AI computing cooled, the classic rotation into defensives when AI pauses. Costco (+11%) and McDonald's moved mildly on the consumption and real-estate clock. Even Mastercard, an AI-adopting payment network, earmarked $7B over five years for fraud detection yet was -5% in 2026: adoption did not lift the stock. Adding the purest staples leaves the picture unchanged. P&G rose +7% in 2026 and Nestlé +20%, holding up or gaining as AI cooled. The off-chain control set on this map now passes twelve names, thickening the observation that pure defensives like Coca-Cola, P&G and Nestlé are where money rotates when AI rests. The nodes added this round draw the same boundary. PepsiCo stayed a single-digit defensive alongside Coca-Cola (+3% in 2026), while luxury name LVMH fell at the opposite, discretionary extreme on China's slowdown (-20% in 2025, -8% in 2026).
The one exception is consumer tech, Shopify. It ran +124% in 2023 and +51% in 2025 before giving back -26% in 2026, driven by a growth-stock multiple sitting atop its GMV consumption clock rather than by AI. Pure defensives (Coca-Cola) rise when AI cools, while growth consumer names (Shopify) swing with the growth-and-rate multiple. Neither runs on the AI clock.
Verdict
The AI cycle is contained within the tech value chain. Inside the chain it propagated from demand through chips, memory, substrates and power, but the capital markets and real consumption outside the chain run on their own clocks: capital on rates, consumption on the economy. The sharpest scene is Blackstone, which funds AI data centers, falling amid the boom. There is a scene on the other side too. Uber posted +149% in 2023, powered by its turn to profitability and passing the rate peak. That was a large gain from outside the chain, a reminder not to read every big move as an AI benefit. Netflix's +83% in 2024 and Toyota's +63% in 2023 are the same kind: the first from a subscriber recovery, the second from hybrid demand and the yen, both large gains from outside the AI chain. Being exposed to AI and running on the AI clock are different things, and knowing that boundary shows how far the phrase "AI beneficiary" actually holds. How off-chain consumption fans out into staples, discretionary and luxury continues in Consumption does not run on one clock, and the full arrangement of clocks in Many clocks.