Enphase Energy, Inc. ENPH

Key data

The business

Enphase sells microinverters, one small box mounted under each solar panel that turns the panel's DC into household AC, plus IQ home batteries and the app that watches the whole system. It sells through solar installers, not to homeowners directly, and the great majority of revenue is US residential rooftop, with Europe a distant second. The moat is real but narrow: installers trained on Enphase gear, a fleet of millions of monitored systems, and US factories that qualify Enphase for federal manufacturing credits rivals importing hardware cannot easily match. What it is not is a utility or a producer of anything scarce; demand rises and falls with electricity prices, interest rates, and whatever the tax code says this year about putting panels on a roof.

The numbers

The last five reported quarters show the whole cycle in miniature: a peak, then an air pocket.

Revenue topped out at $410M in the third quarter of 2025, then rolled over hard. Q1 2026 came in at $283M, down 21% from a year earlier, and gross margin compressed from 48% at the peak to 35%, dropping the company to a GAAP loss for the first time in years. The most recent print, Q2 2026, earned 46 cents adjusted against a 47-cent estimate, missing by a cent. The demand pull-forward into late 2025 emptied the 2026 pipeline.

Zoom out and the pattern repeats at a larger scale.

Revenue nearly halved from the 2023 peak of $2.3B to $1.33B in 2024, recovered 11% in 2025, and the Street models it declining slightly to $1.40B by 2028. The tension sits in the earnings line: consensus has EPS more than doubling to $2.71 by 2028 on revenue that slips 5%. That is a bet on margin, not volume, driven by mix shift toward higher-margin batteries and by domestic-manufacturing tax credits flowing through the cost line. What this memo believes the flat-to-down revenue model understates is the second lever: if US battery attach keeps climbing while microinverter volume merely stabilizes, the earnings recovery arrives faster and heavier than a topline chart suggests. The print that settles it is gross margin: back above 48% means the mix thesis is working; lingering in the low 40s means it is not.

The balance sheet is the reason survival is not the question here.

They repaid roughly $632M of convertible debt in Q1 2026, cutting long-term debt from $1.20B to $573M, while cash rose to $498M. Net debt is half a turn of EBITDA. A company this liquid does not die in a demand trough; it waits.

Management

The capital-allocation record is mixed and worth staring at. Management spent $130M on buybacks in 2025 at an average of $56.58 a share, and the stock now sits at $36, so that cash bought less than it cost. They then halted repurchases in Q1 2026 to retire the convertible, the right call, and extended the buyback authorization to 2029 with $269M left. Stock-based compensation runs near 15% of revenue, high enough that much of the buyback is really dilution offset rather than shrinkage. On insiders, chairman T.J. Rodgers sold about $10.4M across two sales in late 2025 and March 2026, plan status not disclosed, while CEO Kothandaraman put $309K of his own money in on the open market in October 2025 near then-current prices. One founder taking chips off, one operator adding on the dip.

How it fails or surprises you

The credit is gone and demand does not come back. Washington ended the residential solar tax credit after December 2025. Q1 2026 revenue already fell 21% year over year. If US sell-through keeps sliding through the second half of 2026 rather than stabilizing, the "flat revenue" model breaks to the downside and the loss quarters multiply. Watch US revenue in the Q3 and Q4 2026 prints.

The earnings-doubling math is the least-explained fact here. Consensus has EPS going from $1.29 to $2.71 by 2028 on revenue that declines 5%. That requires gross margin recovering toward 48% and manufacturing credits landing as modeled. If margin stays near the 35% of Q1 2026, the multiple that looks like 13x forward is fiction and the real number is far higher.

Battery attach and manufacturing credits compound faster than modeled (right tail). If US storage attach rates keep rising and the domestic 45X credits flow at full rate, margin recovers on stable volume and earnings snap back before 2028. The market pays nothing for this today because it is fixated on the demand hole. First tell: battery MWh shipped rising while gross margin pushes back above 48%.

Closing thoughts

A specific print settles most of this, and it arrives soon. The demand shock is real and precisely dated, the credit expired December 31, so the question is not whether 2026 is weak but whether Q3 and Q4 US sell-through find a floor or keep sliding. A stabilizing US revenue line with gross margin creeping back toward the mid-40s converts this into a cheap recovery; a second leg down in demand with margin stuck near 40% means the forward multiple is a mirage. An ambiguous print, flat revenue and flat margin, means you wait another quarter, and the balance sheet lets you. The left tail is a longer, deeper US demand winter than anyone models; the right tail is battery mix and credits pulling earnings back fast on flat volume. On balance the downside is bounded by $498M of cash and half a turn of net leverage, while the upside needs only margin, not a demand boom, which is why the fatter tail here reads as the upside.

The bet is still that American homeowners keep putting solar and batteries on their roofs, and keep bolting Enphase's boxes under the panels, even without the tax credit. What breaks it is a US market that shrinks and does not recover, and the one pair of numbers that tells you first is US revenue against gross margin: both falling together in the next two prints means the story is wrong, both firming means you were early.