Brookfield Looks Fair at 1.1x Book-AI Infrastructure Could Reprice It Higher

Brookfield looks fair value today after its 2025 rally

Brookfield looks about fair value today, and the stock's 21% return in 2025 narrowed the gap between price and value. This is no longer a deep-discount setup; it is a valuation-reset setup. That makes the next few quarters more important than another pass on Brookfield's long-running quality story.

The valuation debate now depends on the right mental model

Investors anchored to traditional asset-manager multiples can miss how much physical infrastructure sits inside the broader Brookfield ecosystem. Even though Brookfield still describes itself as an asset-light investment manager, that framing does not tell the whole story.

The more useful question is not whether Brookfield becomes a tech stock. It is whether the market starts assigning scarcity value to parts of the business that touch AI power, data centers, and associated infrastructure in the same way it does for more obvious infrastructure assets.

Brookfield is moving from AI narrative to infrastructure deployment

The repricing path is straightforward: if investors begin to view parts of Brookfield's portfolio as AI bottleneck assets, those holdings could command scarcity value rather than standard real-estate multiples. Brookfield does not need to become a chip stock for that to matter. It only needs the market to recognize that the assets involved sit where compute, power, and customer demand must physically connect.

The market may still be underestimating the scale of the buildout

Industry estimates point to roughly $6.7 trillion investment by 2030 to match compute demand, with AI workloads accounting for about 70% of the expansion. In that context, data centers start to look less like ordinary industrial real estate and more like critical infrastructure with durable operating importance.

That research also suggests occupancy could rise from 85% in 2023 to above 95% by late 2026. If a property class gets that tight, landlord economics can improve in a recognizable way: stronger rent growth, better pricing power, and longer tenant dependence. Markets often delay that realization until the evidence becomes hard to ignore.

India shows the operating logic already at work

Brookfield's India footprint helps make the case more concrete. Its Digital Connexion venture has about 160 megawatts of data-centre capacity in India, of which 60 megawatts is operational and fully leased. That matters because it shows demand already exists in at least part of the market; the cash-flow case is not purely theoretical.

Brookfield says AI inferencing demand in India is expected to grow, while India currently has about 1.5 gigawatts of installed data-centre capacity, mostly for non-AI uses. The venture is also a partnership between Brookfield, Digital Realty, and Reliance Industries, which suggests Brookfield is accessing operating expertise and anchor demand rather than simply assembling land parcels.

The assets that could drive a rerating

What could be revalued is not "AI" as a slogan, but control over the scarce links in the chain:

• power availability and renewable supply
• data-centre capacity that is preleased or close to full utilization
• connectivity and land needed to bring compute customers online

That combination matters because scarcity value tends to accrue to the choke points of a buildout, not just to the most visible technology brands.

Brookfield still needs operating proof to justify a higher multiple

The upside case now depends on proof, not prose. Bears are not wrong to demand it: Brookfield already trades near fair value after a 21% return in 2025, and a "smart money" narrative can fade quickly if new AI projects remain theoretical. The opportunity exists because management appears to be moving from theme to deployment, and earnings commentary can either validate that shift or leave it hanging.

What would make the market pay up

A rerating does not require full disclosure on every asset. It requires enough visible traction to make investors treat Brookfield as more than a static premium asset manager.

Look for these confirmation signals:

• new AI-infrastructure partnerships or fund deployments that move from announcement to execution
• evidence that data-centre and power assets are being leased or contracted as demand rises
• management commentary that connects AI demand to operating results, not just long-term vision

What would break the thesis

The bear case is straightforward: if these assets fail to convert into contracted, power-constrained, AI-useful income streams, investors will revert to the old multiple.

Watch for:

• delays in permitting, power connection, or construction
• softer preleasing or occupancy trends if industry capacity comes online faster than demand
• repeated narrative pushes without measurable deployment or earnings contribution

Brookfield does not need to become a trophy AI name for this thesis to work. It only needs one earnings cycle in which the conversation shifts from "fair value" to "fair value with visible AI-infrastructure optionality."