Smart Investor Today

A weekly column about markets, written slowly.

Front pageArchive › Guest column

Guest column · Utilities

Utility capex plans assume a load growth number nobody has stress-tested

Tuesday, February 24 · Owen Marchetti, guest

I spent nine years in regulatory affairs at an investor-owned utility, which means I have written the load forecast section of a rate case and I know precisely how much art is in it. What is happening across the sector right now is not forecasting. It is coordination.

The same number, everywhere

Read six five-year capital plans filed in the last three quarters and you will find annual load growth assumptions clustered in a band about eighty basis points wide, all of them a multiple of what the same companies were assuming three years ago. The plans differ in service territory, climate, industrial mix, and rate structure. They do not differ in this.

Five-year annual load growth assumed in recent filings, by territory type
Territory Assumed Trailing 5-yr actual
Southeast, mixed2.9%0.6%
Mid-Atlantic, urban3.1%0.2%
Mountain West3.4%1.1%
Upper Midwest2.7%−0.3%

The gap between the last two columns is the whole argument. Some of it is genuinely defensible — large interconnection queues are real, and electrification of heat and transport is real. But the trailing column includes years in which those things were already underway.

Why the arithmetic is fragile

A regulated utility earns an allowed return on its rate base. Growing the rate base is therefore the growth story, and capital plans are the mechanism. The load forecast is what makes the capital plan prudent in front of a commission. If the load does not arrive, one of three things happens, and none of them is neutral for the equity.

Either the commission disallows some of the spend, or rates rise enough to depress the demand the spend was justified by, or the plan is cut and the growth story goes with it.

The number I would want disclosed

Signed interconnection agreements with financial commitments, separated from the queue. The queue is a wish list; anyone can join it and much of it is speculative and duplicated across territories. Two utilities I follow now disclose the signed figure. It is between a fifth and a third of the queue. The other four do not disclose it, and I think that is the tell.

Guest disclosure I hold no utility equity, long or short, and have not since leaving the sector. Harriet’s own positions are on the disclosures page. This column is commentary and is not advice.

What would change my mind

Two consecutive years of actual load growth above two per cent in more than half of the territories above, with the growth traceable to signed and energised interconnections rather than to weather. That is a specific test and it resolves within the window of the plans themselves, which is more than most sector arguments can say.

Other guest columns: Priya Rangan on municipal supply.

About guests

Guest columns run a few times a year, from people who have done the job rather than covered it. The standard is described on the about page.

Front page · archive · disclosures