On June 24, the Public Utility Commission of Texas approved ERCOT’s “Batch Zero” — a one-time, systemwide process to evaluate every new large load or expansion of 75 megawatts or more. It replaces the old utility-by-utility interconnection review with a single centralized gate. The reason is in the numbers: ERCOT’s large-load queue reached 226 gigawatts as of late 2025, up from 63 GW a year earlier, and roughly 73% of it is data centers. Texas’s actual 2026 summer peak forecast is about 92 GW. The queue is roughly 2.5 times the size of the entire grid.

The supply side cannot keep pace. ERCOT added about 23 GW of generation across 2024–25, with another ~9 GW due in early 2026 — a rounding error against 226 GW of requests. ERCOT itself cautions that its long-term peak scenarios (up to 150 GW by 2030) are likely inflated by duplicate and speculative filings. That caution is the whole point: when the queue is this detached from physical reality, first-come-first-served stops allocating anything.

Meanwhile, capital is not the problem. JLL’s new Global Credit Intensity Index — built on roughly $9 trillion of bids and loan quotes — hit an all-time high in April, with a near-record field of lenders competing across banks, private credit, agencies, insurers, and family offices, even as an estimated $936 billion in mortgages mature this year. JLL names data centers as a primary driver of that lending surge.

Put the two together and the trade re-underwrites itself. Money is competing to fund data centers; the grid is rationing whether they can ever plug in. Interconnection — not equity, not debt, not tenant demand — is now the gating risk. Batch Zero is Texas pricing that scarcity into the open.

The immediate casualty is speculative power. A site whose value rested on “queue position” just lost its premium, because the queue itself has been declared unreliable. Bankable projects are the ones with firm interconnection agreements, secured generation, or co-location with existing power. Everything underwritten on a request number is now a regulatory coin-flip. Expect a widening basis between sites with real power and sites with only a place in line — and expect lenders, flush as they are, to start pricing that gap.

There’s a second-order loop worth naming. The same buildout straining the grid is helping push energy prices up, and energy just became the leading driver of construction-cost inflation — inputs rose 9.6% year-over-year in the latest read, the fastest since the pandemic, despite the June 8 equipment-tariff cut. So the data-center surge is simultaneously consuming the power, inflating the cost to build everything, and pulling in the lenders — the single force bending three different markets at once. For everyone outside the data-center lane, the lesson generalizes: capital is plentiful and the real constraints are physical — power, energy, entitlement, occupancy. The winners are assets that solve a physical scarcity; the losers are claims on the future that depend on a line that may never move.

Stakeholder lens — Developers: a power letter is not a power commitment; underwrite firm interconnection or co-location, and treat queue position as zero until proven. Lenders & investors: liquidity is not your edge this cycle — collateral quality and physical deliverability are; re-price speculative-power sites accordingly. Owners with firm power or leased core product: your scarcity just got more valuable, and the basis between you and the speculative field is widening.

Still unresolved: whether Batch Zero clears the queue or just reshuffles it. If a meaningful share of the 226 GW is genuine, even centralized triage won’t conjure generation fast enough — and the constraint hardens. And if the Fed actually hikes into this record-liquidity market, today’s lender auction could reprice quickly.

CRE 360 Signal™ — Commercial Real Estate Intelligence

Twelve outside reads we’re tracking this week.