Goodman Group: our top ASX stock pick for AI play in 2026

The year 2026 marks a genuine turning point for Goodman Group.

While the final weeks of 2025 were dominated by the announcement of a $3.9 billion data centre partnership with the Canada Pension Plan Investment Board (CPPIB), the market reaction was surprisingly muted. That response may suggests that the market thinks Goodman lacks the growth required for its valuation.

The CPPIB transaction matters partly because of its size. But more importantly for the share price, it validates Goodman’s growth plans. It shows that some of the world’s most powerful institutional investors are willing to back its data centre strategy at scale.

The share price reaction implies the market has yet to fully appreciate how powerful that setup becomes in a falling interest rate environment, of the sort that may emerge through 2026.

Non-commoditised land

At the heart of the bull case is Goodman’s roughly 5.0 gigawatt global “power bank”. In today’s race to build AI data centres, land is no longer the binding constraint. Electricity is.

Secured access to high-voltage power in major capital cities has become scarce and highly valuable. This type of land allows AI data centres to run faster and more efficiently – which is important to Microsoft, Amazon, Google and Meta.

Goodman accumulated much of its land and power pipeline years ago, before data centres were fashionable, particularly across Frankfurt, Paris and Amsterdam. In many countries, electricity supply is tightly regulated, and new permitting approvals can take years as governments worry that data centres can take electricity away from crucial infrastructure like hospitals.

This means new entrants cannot simply throw money at the problem. Power queues cannot be bought; they must be endured. That gives Goodman a durable barrier to entry. In certain corridors, it operates with something approaching monopoly characteristics. With Microsoft, Google, Amazon and Meta all chasing AI capacity, Goodman has leverage that few global property groups can replicate.

A shift in capital strategy

The way Goodman monetises its developments is also changing.

Historically, the group focused on developing buildings such as logistics and ecommerce warehouses, then selling down stakes to pension funds. That model worked well, helping Goodman build a substantial funds management platform and crystallise performance fees along the way.

More recently, Goodman has been maintaining ownership of its buildings more. The reason is simple: development margins have expanded dramatically.

JP Morgan estimates yield on cost for recent data centre developments at around 9.2 per cent. By comparison, global cap rates are closer to 5.1 per cent. That spread implies embedded value creation of roughly 45 to 50 per cent, even before allowing for any further yield compression.

The move up the value chain reinforces this advantage. Rather than delivering basic powered shells, Goodman is increasingly providing fully fitted data centres, installing cooling and electrical infrastructure itself. Net operating income rises faster than capital cost, preserving attractive yields on cost even as construction prices fluctuate.

By retaining these assets through completion, Goodman captures the full valuation uplift as projects transition from development to stabilised income.

Why lower rates matter more than investors think

 Goodman enters this phase with a conservative balance sheet. Look-through gearing sits around 17 to 18 per cent, which creates a convex relationship between asset values and net tangible assets.

The mechanics are straightforward. Asset values are determined by capitalising income. When market yields compress, gross asset values rise, while the debt stack remains largely unchanged. The uplift therefore flows almost entirely through to equity.

On current estimates, a 100 basis point compression in cap rates across Goodman’s look-through portfolio would add roughly $20 billion to gross assets. Because debt does not increase, that uplift is effectively mirrored in net assets. On a per-security basis, underlying NTA could approach a doubling under such a scenario.

This is not a forecast, but it does highlight why Goodman is far more sensitive to changes in the cost of capital than traditional developers or REITs.

The valuation disconnect

Despite this setup, Goodman continues to trade on EV/EBITDA multiples that resemble those of mature, lower-growth peers. That sits uneasily with a platform offering high-return development, embedded optionality and a growing annuity-style management income stream.

That income behaves more like a subscription business than a conventional rent roll. As the roughly US$110 billion data centre pipeline is built out, the fee base expands without a commensurate increase in overhead.

The CPPIB partnership confirms that global institutions increasingly view Goodman as a preferred gateway to AI infrastructure. As execution risk continues to fall and interest rates move lower, it becomes harder to justify valuing the group using traditional REIT frameworks.

For investors still anchored to yesterday’s models, 2026 may be the year Goodman finally forces a rethink.

Another way to play the theme

For those who would prefer to invest in Goodman’s hyperscaler tenants, the ETFS Magnificent 7+ ETF (ASX: HUGE) and ETFS US Technology ETF (ASX:WWWW) both take large positions on Microsoft, Google, Amazon and Meta.