1 Company Set to Make a Fortune From the $1.7 Trillion Data Center Build-Out

Southern Company is positioned to benefit from a 75 GW data-center load pipeline and has raised its 2026-2030 capital spending plan to $81 billion to meet demand. The utility completed Vogtle Units 3 and 4 in 2024, giving it the largest nuclear power station in the U.S., while 2025 adjusted EPS rose 6% to $4.30 and 2026 EPS is guided to $4.50-$4.60. It also increased its dividend for 25 consecutive years, including a 2.7% hike in 2026, with the yield at 3.25%.

Analysis

The key second-order effect is that Southern is morphing from a traditional regulated yield vehicle into a quasi-infrastructure bottleneck on the AI buildout. In the Southeast, power availability is becoming the gating factor, so utilities with completed baseload and grid hardening can negotiate from strength: once a hyperscaler commits land and interconnect spend, switching costs become high and cancellation risk falls sharply. That makes the current demand pipeline more valuable than the headline EPS growth rate suggests, because the optionality sits in future rate base expansion and long-dated contracted load, not just near-term earnings.

The market may be underestimating how favorable the mix shift is for cash conversion. Large-load customers typically pull forward transmission, substation, and peaking investments while improving load factors, which should support faster recovery of capex through regulated returns. The hidden winner set extends beyond SO into gas turbine OEMs, transformers, switchgear, and EPC contractors, while the main loser is any utility in data-center-heavy geographies that lacks spare capacity or faces permitting friction.

The contrarian risk is that this story is being priced as if every GW in the pipeline converts cleanly. That is unlikely: interconnection queues, local political resistance, and a single megaproject delay can push revenue recognition out 12-24 months, which matters in a bond-proxy stock trading on dividend and rate sensitivity. Also, if AI capex slows or hyperscalers optimize power usage more aggressively, some of the assumed load growth can get deferred rather than canceled, creating downside to the growth premium without threatening the dividend.

Net: the best asymmetry is to own the utility for the contracted load upside while financing it with a short against a more rate-sensitive peer with less visible load growth. The dividend profile should keep the stock supported on any pullback, but the real upside comes from multiple expansion if investors conclude the pipeline is durable and rate-base accretive, not just a headline demand spike.

Key Decisions for Investors

• Long SO on a 3-6 month horizon into pullbacks of 3-5%: risk/reward favors a 10-15% re-rating if the market starts capitalizing the data-center pipeline as durable contracted load rather than speculative interest.
• Pair trade: long SO / short a Southeastern regulated utility with weaker visible load growth and less nuclear capacity, sized market-neutral; thesis is relative multiple expansion for the utility with the strongest large-load optionality over the next 2-4 quarters.
• Add a tactical long in grid and gas-turbine supply chain names on weakness (e.g., ETN, PH, GEV) for a 6-12 month trade: incremental data-center buildout should force accelerated orders and margin support before SO’s earnings fully reflect the pipeline.
• Use SO upside calls rather than outright equity for event-driven exposure into the next investor day/earnings cycle: limited premium outlay captures the rerating if management confirms load conversion, while capping downside if project timing slips.
• Avoid chasing the stock after sharp outperformance; wait for a rate-driven selloff or broader utility weakness, since the core risk is duration/rates compression that can temporarily swamp the fundamentals.