Google will invest €13bn in four Finnish data centre sites over two years

Google plans to invest €13 billion in data-centre infrastructure across four Finnish sites—Hamina, Kajaani, Muhos and Vaala—during 2027-28, marking its largest single investment in Europe. The expansion is accompanied by a 22-year power purchase agreement with Fortum, supporting the long-term electricity needs of the facilities. The investment strengthens Finland's digital infrastructure and provides a material boost to regional data-centre and clean-power demand.

Analysis

For GOOG, the relevant signal is not the regional asset build itself but the commitment to lock in low-carbon, long-duration power for AI capacity. European inference workloads are likely to carry lower utilization initially than U.S. frontier-training clusters, making near-term returns on invested capital dependent on enterprise AI adoption rather than capacity scarcity. The spend reinforces a multi-year capex step-up that favors the power-and-cooling supply chain—VRT, ETN and Schneider Electric (SU.PA)—but does not independently change Alphabet earnings until utilization and Cloud revenue convert, likely over 2028-30.

FORTUM gains more from contract duration and credit quality than from volume upside: a long-dated hyperscaler offtake can lower cash-flow volatility and support renewable/nuclear investment underwriting, but the valuation impact depends on disclosed pricing, generation source, collateral terms and incremental capex. A second-order risk is Finnish wholesale-power compression if additional clean generation is built faster than data-center load ramps; conversely, persistent grid congestion could raise delivered-power costs and delay facility commissioning. The more investable European implication may be tightening competition for firm, dispatchable low-carbon power, benefiting nuclear-exposed utilities such as Fortum and Vattenfall rather than intermittent-power developers.

Consensus may overread this as an immediate AI-equipment order catalyst. Most construction and electrical procurement will be phased, competitively tendered and partially sourced locally; the nearer catalyst is disclosure of contracted capacity, power pricing and grid-connection milestones. The thesis is weakened if Google Cloud growth fails to reaccelerate enough to absorb elevated depreciation, or if European permitting/grid bottlenecks push utilization beyond the planned commissioning window.

Key Decisions for Investors

• Maintain GOOG as a core long, but do not add solely on this announcement; add on any post-earnings capex-driven pullback if Google Cloud growth remains above 25% and operating-margin guidance holds. Upside is multiple support from credible AI capacity leadership; risk is depreciation growth outpacing Cloud monetization over the next 12-24 months.
• Use a 6-12 month relative-value expression: long VRT or ETN versus short a broad European utilities basket (EXV1/utility proxy), sized modestly. Data-center power density raises cooling and electrical-content intensity, while regulated utility upside is capped unless contract economics are disclosed; exit if hyperscaler capex guidance decelerates materially.
• Place FORTUM/FRTTF on watch rather than initiate. Upgrade only if management discloses incremental contracted load, attractive inflation-linked pricing, and limited new balance-sheet capex; a disclosed low-price PPA or material grid-investment requirement would be a reason to avoid.
• Monitor Finnish day-ahead power spreads and grid-connection approvals over the next 3-9 months. Sustained narrowing between Finnish zones or delayed connections would undermine the scarcity-premium thesis; confirmed firm-capacity shortages would support a tactical long FORTUM relative to Nordic renewable developers.