Enphase Is Not Waiting for a Recovery — It Is Betting the Company on Five of Them

A 35% US sell-through collapse and a fourth channel clean-up, offset by a 2027 revenue framework and a $1 billion data-center aspiration for 2030

ENPH · Earnings Call · 2026-10-27

The gap between the quarter and the pitch

Enphase reported $293.3 million of Q3 2026 revenue, down 20% year-over-year, and guided Q4 to $225–255 million, of which $51.2 million is safe harbor — a line item that exists because third-party-owned systems are pre-buying inventory, not because homeowners are buying. Strip the safe harbor out and the core business has been flat at roughly $230 million of global sell-through for three straight quarters. Management did not dress it up: “We are not seeing any meaningful recovery in the U.S. demand.” — Badrinarayanan Kothandaraman, President and Chief Executive Officer · 2026-10-27 U.S. sell-through fell 35% year-over-year, in line with third-party permit data, and the company plans to ship $20 million below sell-through in Q4 to drain a channel it described, again, as “a little above normal for batteries and micro inverters” — Badrinarayanan Kothandaraman, President and Chief Executive Officer · 2026-10-27. That is the fourth consecutive quarter of deliberate channel surgery — overshipped $25 million in Q1, corrected in Q2, under-shipped again now. This is not a cyclical hiccup story. Revenue peaked at $699 million in Q2 2023 and has since halved; the peak-to-trough drawdown on the top line is 52%. What is genuinely new is not the weakness — it is the substitution of a story for a number.

Five growth drivers, and the arithmetic behind them

The single most important sentence on the call was a framework, not a guidance range:

So you take $200 million multiplied by 4 for 2027. And then you add the $200 million of growth drivers, which comprises of — we said that the battery number goes from 450-megawatt hours to 675-megawatt hours. We talked about commercial going from $30 million to $100 million. We did not quantify Propel, but that would be — that could be some upside. We then talked about bidirectional EV charger, adding a modest $15 million to $20 million.
Badrinarayanan Kothandaraman, President and Chief Executive Officer · 2026-10-27

The growth driver framing dominated the prepared remarks to an unusual degree — five of them, laid out like a pitch deck: residential battery shipments up ~50% in 2027 on the Gen 5 stack; a recovering U.S. residential market via Propel; commercial solar and storage going from ~$30 million to over $100 million; bidirectional EV charging at $15–20 million with three U.S. OEMs; and the AI data center play through a solid-state transformer that management says can be a $1 billion business by 2030 — a number Raghu Belur justified by taking an 11–30 gigawatt 2030 SST market and applying "a reasonable market share." Note the composition of 2026: 72% residential solar, 21% residential batteries, 7% everything else. The "everything else" is where the entire equity story now lives.

The things that keep sliding to the right

The pattern worth flagging is not that management is wrong about the terminal value of these markets. It is that the timetables move. On the October 2025 call, the fifth-generation battery was the fix for battery margins: “by moving to non-China sources, which we are going to move by early 2026 when we plan to ramp up our non-China batteries as well as our fifth-generation battery, which will result in a step change in costs.” — Badrinarayanan Kothandaraman, President and Chief Executive Officer · 2025-10-29 By April 2026 it had become "Q4" for Europe. Today it is pilots in Q4, volume shipments in Q1 '27, with management conceding the generation transition itself may take two to three quarters. That is a product that has now absorbed three calendar resets. Propel is the same shape, only better. Installer count went from 40 (February) to 200 (April) to roughly 350 today; states from four to thirteen; originations from 200/week to over 300/week. But the April aspiration was explicit: “my wish is to basically get up to a run rate of 500 originations a week by the end of Q4.” — Badrinarayanan Kothandaraman, President and Chief Executive Officer · 2026-04-28 They are at 300-plus, battery attach has slipped from 84% to ~71% as it moved beyond California, and the funding partner SolSource raised $300 million but is, in management's own words, "actively working to secure additional capital." Propel is a financing-dependent channel dressed as a demand driver. And then the Netherlands, the thesis that has been three calls in the making. Colin Rusch asked the obvious question — why the much-hyped retrofit market isn't materialising — and got the most honest answer of the call: “customers aren't going to react until they see their electricity bill. And they're going to start seeing their electricity bill go up quite a bit in 2027.” — Badrinarayanan Kothandaraman, President and Chief Executive Officer · 2026-10-27 That is a real catalyst (Dutch net metering ends this year, and the company ran 78 homeowner events a quarter to pre-sell it), but it is also a fourth deferral of the same inflection. Back in October 2025 management was already warning that “10% take rate is tough to do because customers have their own way of deciding what is good for them.” — Badrinarayanan Kothandaraman, President and Chief Executive Officer · 2025-10-29 The Net metering keyword, which spiked on the Dutch and French feed-in-tariff resets, has since faded from the company's own keyword set — the exact moment it should be most relevant.

The tape is voting against the new story, not the old one

The stock is down 47% in the last 90 days and 49% from its June 30 high of $49.24, on a $3.3 billion market cap. That de-rating is not really about Q3's miss. It is about the market re-pricing the destination industries. Global theme tape shows AI data center as a 90-day decliner at roughly -5.4%, with 50 negative tickers against 6 positive, and solid state transformer down 7.3% alongside GEV, POWI and ETN. Enphase is formally entering the 800V DC data-center power chain — 800V DC was a fresh global keyword last quarter — just as the complex is de-rating. Its 2030 ambition is being priced in a market that currently discounts the category. The battery story has the mirror-image problem. Management cut U.S. distributor list prices on batteries by 8–9% a couple of months ago, and the Q4 non-GAAP gross margin guide of 40–43% bakes in that price adjustment plus freight. So 50% MWh growth in 2027 translates into meaningfully less than 50% revenue growth, at a margin the company insists normalises to 40%. Meanwhile the tariff-refund tailwind that inflated Q2 — GAAP gross margin of 60% in Q2 versus 43.3% in Q3 — is gone, and reciprocal tariffs still cost 1.3 points.

What to watch

Two things. First, free cash flow: Q3 was negative $11.4 million, and FCF less stock comp was -$18 million in Q2, with the offset being a $108 million IRS refund on 2024 PTCs expected in Q4. Second, the embedded option nobody paid for: a bipartisan extension of the investment and production tax credits post-midterms, which Raghu flagged as plausible. That, not the SST, is what moves 2027 numbers. Everything else — PROPEL, the FEOC regulations advantage in commercial, the Sole Source-structured lease channel — is management asking investors to underwrite execution against a tape that is currently underwriting the opposite.