Southern Company is no longer a post-construction cleanup story. The latest quarter recasts the Atlanta holding company as the Southeast's contracted power supplier for data-center and large industrial load, while Georgia Power and Alabama Power keep retail base rates frozen into the late decade. That pairing is the entire investment debate. Growth arrives as signed megawatts. Cash recovery still waits on construction, interconnection, and the next rate-plan window. Adjusted earnings of $1.13 for the second quarter cleared the company's own estimate by thirteen cents, but the print is not the thesis. The thesis is whether a contracted book now above seventeen gigawatts converts into allowed return without reopening the political bargain that keeps household rates still.
The earnings bridge still leans on construction accounting and a non-repeat tax item at Georgia Power, not on a return from the new OpenAI or Alabama contracts. Those contracts do not show up in this year's allowed return. Weather-normal retail sales rose in the first half at the fastest June-to-date pace in nearly two decades, and commercial sales in the quarter were led by large-load ramps. System data-center load now exceeds 1.2 gigawatts. That is the visible load, not the contracted backlog. Parent equity issuance in the first half approached $2.6 billion as the company cut the remaining equity need through the end of the decade and sourced more at-the-market forwards. Dilution is the price of keeping leverage inside the credit box while property additions outrun operating cash.
Management now projects full-year adjusted earnings near the top of a range that still tops out at $4.60. That guide implies a second half that barely matches last year's back half, even after a first-half beat. The OpenAI agreement near Savannah adds 3.2 gigawatts of contracted load on a twenty-five year tenor, with service phased from 2028 and a gigawatt of flexible demand response. Alabama Power added about three gigawatts across three projects in the same quarter. The question for the next several years is not whether the Southeast wants the power. The question is whether frozen retail rates, incremental-cost contracts, and another round of generation requests for proposals can carry that book without a second Vogtle-style political rupture.