CenterPoint's $5 Billion 'Savings' Scheme: Data Centers Shift Fixed Costs, Not Bills

CenterPoint Energy announced on August 11, 2026, that up to 14 gigawatts of new large-load projects, including data centers, could save Texas electric customers more than $5 billion over the next decade by having those high-demand users pay a larger share of infrastructure charges.^[1] The pitch is simple: more customers spread fixed costs thinner, so residential bills go down. It is not true.

Here is the mechanism CenterPoint is not naming. Under rate-of-return regulation, the utility files for a revenue requirement, and the Public Utility Commission of Texas divides it among customer classes and meter volumes. When CenterPoint adds 14 GW of new load, it does not erase fixed costs; it reallocates the recovery. The data-center customer pays more because the utility charges them a higher demand rate, and fixed charges rise for new connections. But the utility's total revenue requirement stays set by its authorized return on rate base. Adding megawatt-hungry tenants increases CenterPoint's capital investment, which expands rate base, which grows the revenue requirement, which goes back into rates for everyone. The residential customer does not save; the residential customer gets the new construction costs folded into the next general rate case, offset by a slightly lower per-megawatt demand charge because more megawatts are now in the pool. The net effect: cost shifting, not cost reduction.

The other half of the trap: stranded-asset risk. CenterPoint is claiming it will build out grid infrastructure in advance of firm demand commitments from these data-center operators. If a single large customer signs a 10-year contract and then backs out, or if Texas tax policy shifts, or if another utility gets the next tranche of ERCOT headroom, CenterPoint still owns the plant. Under traditional rate-of-return regulation, CenterPoint recovers those investments regardless, because the commission sets rates to ensure the authorized return. That means ratepayers absorb the risk of overbuild, while the utility keeps the 9, 10% return on a larger base. The risk matrix is asymmetric by design.

The White House Ratepayer Protection Pledge, which CenterPoint cites in the release, calls for cost-allocation transparency and protection against cost-shifting.^[1] CenterPoint is doing the opposite. It is proposing to accelerate capex outside a contested general rate case, using long-term contracts with affluent customers to justify upfront investment that ratepayers will finance for decades. This is a docket play, not a press release: CenterPoint will file for recovery of these costs in Texas Railroad Commission proceedings, either through a new rate case, a grid-modernization rider, or a transmission cost pass-through. No formal rate case means no discovery, no intervenor questioning of forecast accuracy, and no earnings test to claw back overearnings if demand materializes faster than promised.

The alternative is clear. Texas should require that any capex associated with large-load customer additions be recovered through a general rate case with a historic test year, so the commission can examine whether the utility over-forecast demand, inflated the cost of upgrades, or double-counted projects already in the base-case plan. If CenterPoint's 14 GW thesis is sound, the company can wait for actual signed interconnection agreements and then file for recovery of proven, audited costs. The cost-shifting claim dissolves when the utility must prove it.