CenterPoint’s $4 Billion Growth Spurt Meets a Cooling AI-Power Tape

The Houston utility just raised capex and extended guidance, but the stock is 6.8% off its June peak while the data-center power trade wobbles.

CNP · Earnings Call · 2026-10-27

What Changed: Houston’s Queue Became a Capital Plan

CenterPoint’s third quarter was not a routine utility print. Management raised 2026 guidance to $1.90–$1.92, introduced 2027 at $2.07–$2.09, and extended mid-to-high-end 7–9% EPS growth through 2030. More importantly, it converted the transmission study promised last quarter into a $4B increase in the 10-year capital plan, now $70.7B, plus $1.5B accelerated into the first five years. The CEO framed it as customer demand already here: “approximately 14 gigawatts of customer demand we submitted as eligible load for ERCOT's Batch Zero process” — Jason Wells, Chair and Chief Executive Officer · 2026-10-27. This is the first quarter where Batch One outranks Batch Zero in the company’s own keyword set — a subtle shift from the current ERCOT process to the next, larger queue. The prior call set this up. Jason said in July: “we intend to provide a more comprehensive transmission study update the second half of this year.” — Jason Wells, Chair and CEO · 2026-07-28 They delivered. What is new is not the existence of Houston growth; it is that the growth is now being capitalized at scale. Greater Houston demand, rate base CAGR above 14% through 2030, and a $10B-plus opportunity set through 2035 are the new spine of the story. Results support the tone: total revenue rose 11% year over year to $2.2B, and operating margin reached 24.7%, up 3.4 points year over year.

The Load Is Real—But the Froth Is Being Separated

The details matter. CenterPoint submitted 14 GW of eligible load into Batch Zero, remains confident in 10 GW of baseload, and now expects about 2 GW to remain as studied load. Some studied-load customers withdrew during ERCOT’s audit; management says the process is working and that Batch One is already trending above Batch Zero. That is a more mature message than the generic large-load enthusiasm of prior quarters. The keyword large load customer actually lost momentum in the latest quarter, while study load, a top theme in Q2, fell out of the top ranks. The language has moved from speculative queue chatter to firm baseload, brownfield transmission, and trackers. That is healthy for credit but less exciting for a tape that had been paying for any AI-power adjacency. The global price tape shows the AI data center theme still up over 360 days but down about 5.6% over both 90 and 30 days, with data-center capacity and cost-of-power baskets also lower. CNP itself is 6.8% below its June peak and roughly flat over 90 days. In other words, the customer queue is accelerating while the market’s willingness to fund the resulting capex is cooling. That is the central tension of this report.

The Bill Comes Due: Equity, Rate Base, and a Conservative EPS Bridge

The financing ask is the part analysts would not let go. CenterPoint increased remaining 10-year equity financing by $1.6B to $4.5B, including $500–$700M of common equity in 2027. On the $5.7B of plan increases announced this year, incremental equity is only about 25% — below the 47% rule of thumb — but it is still dilution. Chris Foster framed it carefully:

We are increasing our remaining 10-year equity financing guide by $1.6 billion to $4.5 billion, well below our previously communicated 47% funding guide.
Christopher Foster, Chief Financial Officer · 2026-10-27

The stock’s drawdown suggests investors heard the second half of that sentence more than the first. The analyst question was blunt: why 14% rate base growth but only 7–9% EPS growth? Jason Wells conceded the gap is mostly conservatism after equity dilution: “It's definitely more of the latter. Really, again, we're building a plan that we can consistently deliver.” — Jason Wells, Chair and Chief Executive Officer · 2026-10-27 That is a notable admission. A 14% rate base CAGR normally implies more earnings leverage; management is deliberately sandbagging, or regulatory lag and equity are eating more than the headline suggests. The fundamentals show why financing matters: latest-quarter capex of $1.4B was up 21% year over year, while operating cash flow was $778M. That implies capex is running at roughly 1.8x operating cash flow. Non-current debt has reached $22.9B, up 11% year over year, and interest coverage sits near 2.0x. Credit is not yet the problem. Adjusted FFO/debt was 13.7%; the Ohio LDC sale adds roughly 50bp, and the AMT refund another 60bp, leaving stable outlooks across agencies. The funding plan also has levers not in guidance: a temporary-generation sublease or sale, demand charges from Batch One, and portfolio optimization. On the prior call, Foster reminded investors that “we have not yet pulled in the benefit from the demand charges from these large customer loads.” — Christopher A. Foster, CFO · 2026-07-28 That remains true. If those levers land, the $1.6B equity number can shrink.

Bottom Line

CenterPoint is one of the cleaner regulated-utility ways to own the U.S. data-center and advanced-manufacturing buildout: existing capacity, a constructive Texas tracker framework, and a customer queue that is still growing. But the market is no longer paying indiscriminately for AI-power capex. The stock’s 6.8% drawdown from June, despite a guidance raise and a bigger plan, says investors want proof on three things: Batch One conversion, equity funding, and the EPS bridge from 14% rate base growth. Management’s answer is that regulatory lag is being reduced through capital tracker mechanisms, and that “The path to affordable rates is through growth.” — Jason Wells, Chair and Chief Executive Officer · 2026-10-27 The next real test is the first-half 2027 transmission update and the Batch One submission — the point where the queue either becomes another $4B of steel or the growth story starts to look priced.