H2O America HTO

Key data

The business

H2O America, the old SJW Group renamed this year, sells a thing nobody shops for: tap water, plus some wastewater service, to about 1.5 million people who have no other pipe to choose. San Jose Water is the big one, then Connecticut Water, then small systems in the Texas Hill Country and Maine. The customer sees a meter and a monthly bill. The regulator sees a rate base, the sum of every pipe, pump and reservoir the company has built, and periodically sets rates that let the utility earn an allowed return on that base. That is the whole machine: build assets, get them into rate base, earn a return, build more. The moat is a legal monopoly plus a century of water rights that no competitor can replicate. What breaks the machine is not competition, it is a regulator who lets the earned return drift below the allowed one, which is exactly what is happening now.

The numbers

Water is seasonal, so read the quarters as a sequence, not one against the next. The summer quarter is the fat one.

The last row is the story. Net income rose 7.7% against Q2 2025, and earnings per share still fell 13%, because diluted shares went from 34.9 million to 42.9 million. Capex converts to rate base for the regulator and dilution for the owner. The June quarter is the proof: more profit and less per share.

Over four years net income compounded about 14% a year, but earnings per share only 9.5%, because the share count kept rising even before this year. Trailing EPS of 2.83 has now slipped below the 2.92 the company earned in 2025. The business is getting bigger and the owner's slice of it, per share, is standing still.

The reason sits in the capital plan. Capex runs about 2.7 times operating cash flow, so the build is funded by debt and equity, not by what the water business throws off. Free cash flow is deeply negative, a yield of about minus 16%.

Debt climbed $0.32B in two years while cash flow barely moved, and the equity raise did the rest. What the market believes is that this treadmill runs indefinitely, which is why the stock sits at 1.4 times book. What the memo asks is whether rate relief lifts the earned return toward the allowed one before the next equity raise resets the treadmill. The print that settles it is realized return on equity, near 6% today against a California allowance close to 9.8%.

Management

The signal here is an owner adding, not selling. Atlas Infrastructure Partners, the infrastructure investor now embedded in the register, bought about $8.0 million of stock in the open market across five purchases in April and May 2026, including two $3.0 million lots on April 10 and $1.8 million on May 29, all near $60. Insider selling was trivial, $0.28 million, plan status not disclosed. The board raised the dividend to $0.44 a quarter from $0.42, and first-half payout ran about 79% of earnings, funded by external capital given the negative free cash flow. The judgment to weigh: management is issuing equity at 1.4 times book, barely above the book value it dilutes, to fund a rate base it has not yet been allowed to fully earn on. That is defensible only if the rate cases land.

How it fails or surprises you

The dilution treadmill. Q2 net income rose 7.7% and diluted EPS fell to $0.62 from $0.71 as shares went 34.9 to 42.9 million. If H2O keeps issuing near book to fund capex, per-share earnings can stall for years while the utility itself grows. Watch the diluted share count and any new equity forward in the next two prints.

Underearning versus allowed. Realized return on equity of roughly 6% sits well under a California allowance near 9.8%. Regulatory lag plus a fresh raise hold it there. If rate decisions keep trailing the capex that feeds rate base, the earned-allowed gap persists and the book multiple stays at its trough. Watch the next California and Connecticut rate orders.

Rate relief and a re-rating (right tail). At 1.4 times book, the low end of a 12-year 1.1 to 2.8 range and below peers near 2 times, any convergence of realized return toward the allowed on the now-larger equity base lifts earnings and the multiple together. An infrastructure owner buying $8 million at $60 is betting on exactly that. Watch for a constructive rate order.

Closing thoughts

The rate case decisions and realized return on equity settle it. Those two convert a utility that is cheap on book into one of two things: a compounder catching up to its allowed return on a base it has already built, or a dilution treadmill that grows the company and not the share. The left tail is perpetual underearning while the share count climbs. The right tail is rate relief off a trough book multiple. My judgment is the right tail is the fatter one, because the multiple already prices disappointment and a real owner is adding, but the near-term per-share math is genuinely negative and could stay negative for a year while the rate cases work through.

The bet is still that H2O America keeps building water systems and regulators let it earn a fair return on what it built. It breaks if the earned return stays stuck near 6% while the shares outstanding keep climbing. The one pair that tells you first is diluted share count against realized return on equity, quarter by quarter. If realized return has not begun turning up toward the allowed line by the time this capex cycle's rate cases are decided, the owner keeps losing to the regulator.