Alphabet this week raised $85 billion in equity — exceeding its initial $80 billion target after investor demand surpassed expectations. The oversubscription is a data point worth sitting with. In a raise of that magnitude, oversubscription doesn't happen by accident. It signals that institutional investors are comfortable with the thesis, the execution capacity, and the projected returns on AI infrastructure at scale.
But the size of the raise isn't the story. It's what the money is for — and why having it is only half the problem.
The Construction Gap Nobody Talks About
A JPMorgan analysis published last month found that more than 60% of data center capacity planned for completion in 2027 is not yet under construction. A further 7% is already delayed. In aggregate: two-thirds of announced capacity for 2027 exists today only as plans, permits, and press releases.
The reasons are structural, not cyclical. Supply chain bottlenecks have extended lead times for critical components — high-voltage transformers now carry wait times of 18 months or longer in some markets. Permitting fights are slowing or blocking projects in high-demand markets. And grid interconnection queues — the formal process of connecting a new facility to electricity supply — are running years long in markets like Northern Virginia, the UK, and parts of the EU.
"Capital is not the constraint. The speed at which you can physically connect, permit, and build is."
Every hyperscaler is facing this simultaneously. Microsoft, Amazon, Meta, and Google are all deploying capital at a rate the infrastructure supply chain — and the grid — cannot absorb without significant delays. The companies that navigate this execution bottleneck fastest will have a structural compute advantage that compounds over years, not quarters.
Google's Actual Competitive Edge
What distinguishes Google's position isn't the magnitude of the capital raise — its peers are committing comparable amounts. It's the $4.75 billion acquisition of Intersect Power, completed earlier this year, and the operational strategy that acquisition enables.
Intersect Power is a wind and solar developer that made a deliberate pivot in recent years: rather than building renewable projects for grid sale, it began building energy projects specifically designed to co-locate with and power data centers. This week, Google and Intersect announced a new facility in the Texas Panhandle that will generate its own power on-site — a behind-the-meter configuration that sidesteps much of the grid interconnection queue entirely.
| Strategy | Grid Dependency | Interconnection Risk | Timeline Advantage |
|---|---|---|---|
| Standard grid-connected data center | High — full grid supply | High — full interconnection queue | Baseline |
| Partial behind-the-meter generation | Moderate — grid backup | Moderate — reduced interconnection size | 6–18 months faster |
| Full behind-the-meter (Google/Intersect model) | Low — self-generated | Low — minimal grid tie | Potentially 2+ years faster |
Google has also built load-shifting capability: the ability to migrate computing workloads geographically to follow available power supply. Where electricity is abundant and inexpensive, compute runs at higher intensity. Where supply is constrained — a grid event, a maintenance window, a peak demand period — compute migrates. Power variability becomes an operational parameter rather than a construction or reliability liability.
Analysts and power industry experts cited in recent market reporting assess that these two capabilities — owned generation and flexible compute load — give Google measurably faster grid connection timelines relative to competitors still dependent on utility supply agreements.
What This Means for the Capital Stack
For investors and advisors working in data center finance, the Alphabet raise and Google's power strategy together illustrate a shift in how the sector should be underwritten.
The traditional data center financing model — secure a lease with a creditworthy tenant, finance construction against contracted revenue, refinance on stabilisation — remains sound for established assets. But for new development at hyperscale, the risk profile has changed. Construction timeline risk, power availability risk, and permitting risk have become the dominant variables. Projects that have solved the power problem — through co-located generation, secured grid capacity, or power purchase agreements with genuine delivery certainty — are materially lower risk than those that haven't.
The implication for capital structure: projects with secured power deserve lower cost of capital and higher leverage than power-uncertain equivalents. Lenders and equity investors that can differentiate on this dimension will make better decisions than those treating all hyperscale development as equivalent.
"The winners in AI infrastructure won't be the ones who raised the most. They'll be the ones who figured out how to build — and power — the facilities faster than everyone else."
The Broader Investment Thesis
Investor enthusiasm for data center and AI infrastructure exposure is clearly intact — the oversubscribed Alphabet raise is the latest evidence. But the investment questions are becoming more granular. It's no longer sufficient to ask whether capital will be deployed into AI infrastructure. The more interesting questions are: which operators have demonstrated the execution capability to convert capital into commissioned facilities on schedule? Which projects have solved the power constraint? And which positions in the capital stack — equity, mezzanine, construction debt, stabilised refinancing — offer the most attractive risk-adjusted returns given where the real risks are concentrated?
Google's answer to the power problem is one approach. It won't be the only one. But it sets a useful benchmark for what genuine execution capability looks like in a sector where the gap between announced plans and physical data centers is, by JPMorgan's count, larger than most market participants appreciate.
Financing Your Data Center Project?
OAKRG advises on data center project finance, construction debt, hyperscale equity raises, and energy-linked infrastructure capital. Tell us your project stage, power strategy, and capital requirement.
Speak with an Advisor