Built by Grok 4.6 High

Reports Tuesday, October 27, 2026. Prepared September 24, 2026. This is the Q3 2026 print (quarter ending September 30). It has not been reported.

ENPH | Earnings Preview

Q3 2026 has not printed. The last reported baseline is Q2 2026 (July 28, after the close): revenue $291.9 million (−19.6% YoY vs $363.2 million), non-GAAP gross margin 46.8% (−180 bps vs 48.6%), non-GAAP diluted EPS $0.46 (−33.3% vs $0.69). Safe harbor was $84.3 million, 28.9% of the quarter. Ex-safe-harbor revenue was about $207.6 million, −35.7% YoY. That ex-safe-harbor figure is arithmetic, not a separate fundamental. The reaction session was July 29, −3.4% (Daloopa adjusted close, $36.32 on July 28, which was still pre-news, to $35.07). A 0.7% revenue beat and a one-cent non-GAAP EPS miss did not hold the opening gap.

Growth trajectory — still decelerating, and the Street is underwriting the next step down. Year-over-year revenue went +7.8% in Q3 2025, −10.3% in Q4 2025, −20.6% in Q1 2026, and −19.6% in Q2 2026. The headline only improved 90 bps from Q1 to Q2. Core revenue got worse, from about −17.8% to −35.7%, because safe harbor went from $34.5 million to $84.3 million. U.S. revenue has worsened for six comparable quarters, to −16.4% on $226.8 million. Strip safe harbor and the U.S. decline in Q2 was about −38%. International is the only geography whose decline is narrowing: −51.5% in Q4 2025, −47.2% in Q1 2026, −29.2% in Q2 on $65.0 million. Management's "Europe +35%" was sequential, not year over year. Non-GAAP EPS growth has been negative and getting slightly worse: −24.5%, −30.9%, −33.3%. Non-GAAP margin compression narrowed from −500 bps in Q1 to −180 bps in Q2. The Q3 guide does not extend that improvement.

The quarter they guided is not a recovery of the year. Midpoint $305 million is −25.7% versus Q3 2025 revenue of $410.4 million. Even the high end, $320 million, is −22% versus that same quarter. FMP revenue of $305.1 million sits on the midpoint. The street's $300.4–$311.1 million band sits inside $290–$320 million. There is no cushion and no embedded beat. Safe harbor of $75 million is already in the range, so a Q3 2025-style blowout — the only range beat in seven quarters, and it was safe harbor that had not been given a dollar — should not be extrapolated. Core at the midpoint is about $230 million, still about −32% YoY.

Watch items

• Guidance. Enphase guides one quarter only. No EPS and no full year. The bar is revenue $290 million–$320 million, batteries 130–150 MWh, non-GAAP gross margin 44%–47%, non-GAAP opex $76 million–$80 million. Over 70% of the midpoint was booked on July 28, thinner than the 85% and 90% cited on the prior two guides. The last two prints landed inside the range and about 1% under the midpoint. This range is not a sandbag.
• Tariffs, and the margin range that refused to rise. Reciprocal-tariff drag in the guide stepped from about 3 points to about 2 points. The 44–47% non-GAAP range did not move up. The midpoint, 45.5%, is −370 bps versus Q3 2025 49.2%. Price cuts are the reason they gave. Ignore Q2 GAAP margin of 60.0%. That print included $45.4 million of IEEPA tariff refunds. The Q3 GAAP range is 42–45%. The August 6 polysilicon price-floor order is effective December 4. It does not touch this quarter's cost of goods. It is a 2027 system-price risk, and Enphase does not sell modules.
• The cost promise that is not in the guide. The January 23, 2026 restructuring letter targeted non-GAAP opex of $70–$75 million a quarter starting this quarter. The live guide is $76–$80 million, a $1 million step up after Q2 non-GAAP opex of $79.8 million slipped through a $79 million ceiling. A print inside $76–$80 million makes the quarterly guide and misses the letter.
• What the call will try to sell, and what FY2027 does not underwrite. Fifth-generation batteries are a Q4 U.S. ship. Propel was still about 200 originations a week in early July, with 500 a week an exit-rate hope. Netherlands storage is a Q4 story, not this quarter's Europe assumption (flat). IQ SST is Texas modules now, November demos, 2027 pilots. FMP FY2026E revenue of $1,184.6 million is −19.6% versus FY2025 $1,473.0 million. FY2027E of $1,238.0 million is only +4.5% versus that estimate. The street believes the $305 million. It is not capitalizing the long-cycle story.

Hold. The microinverter franchise is intact, and they usually land inside a revenue range they set. The trajectory that matters — core demand, U.S. dollars ex safe harbor, non-GAAP EPS — is still down, and the recent beat has shrunk to noise. There is no internal model. Paying for a beat of a midpoint they have missed the last two times is not the trade. The bar is core near $230 million, batteries whose year-over-year decline actually narrows, non-GAAP margin that stays inside 44–47% without another refund, and a Q4 comment that does not need a fresh safe-harbor step-up to show core growth.

No inverter, installer, or home-battery peer has a confirmed date before October 27. SolarEdge and Sunrun are early November on third-party calendars. Tesla's date conflicts (October 21 versus October 28) and is not used.

The operative outlook is the July 28 release and call. Daloopa tags it to 2026Q2, the quarter in which it was issued. There is no 2026Q3 fundamentals column. Nothing filed since July 28 has revised the ranges. Classification: in line on revenue, not conservative and not a stretch versus the Street. Batteries and non-GAAP gross margin are slightly conservative versus the last print. Non-GAAP opex is a miss versus the January letter.

Consensus, FMP only, as of September 24. Not Visible Alpha and not Bloomberg. Thirteen analysts on Q3 revenue, twelve on Q3 EPS. The earnings-endpoint EPS of $0.48 is a different field. It is not a low of the $0.375–$0.600 analyst range, and it is not averaged with $0.514.

Q3 revenue band $300.4–$311.1 million sits inside the guide. Endpoint EPS is $0.48, only two cents above the last print of $0.46. EBITDA average $77.5 million is street-only. Q4 is not a management guide and not a fifth-generation step-up. The Q4 growth rate is −12.0% versus Q4 2025. There is no annual guide. H1 actual revenue is $282.9 million plus $291.9 million. Adding the Q3 and Q4 FMP averages is close to, not equal to, the annual panel. FY2027 is consensus on consensus: a stall, not a product cycle. G5, the prepaid lease, the Dutch retrofit, and IQ SST are not in that +4.5%.

Revenue versus the company's own range, last seven scored quarters. Columns are the quarter being guided. The guide was issued on the prior call. One above-range beat in seven, and it was safe harbor that was not in the guide. The last two prints were under the midpoint.

Internal is n/a on every row. No model was available. Where a guide midpoint exists, the difference versus consensus is essentially zero on revenue. Elsewhere the year-over-year compare is versus the reported quarter, because there is no guide. Columns on the history tables are periods. Rows are metrics. Q3 2026 has not been reported. The right-hand column is the guide or FMP, not an actual.

(a) Current quarter — Q3 2026e

Reported path into the print. YoY is versus the year-ago quarter, never the prior quarter. Safe-harbor history in this series starts in Q1 2025. Core is revenue minus safe harbor.

Revenue growth re-accelerated through early 2025 and has been decelerating since Q3 2025. The Q3 midpoint extends the decline. Battery growth went from +125% to −40% and got slightly worse in Q2. The guide is the first time management is asking the market to believe that rate of decline narrows, to about −28% at 140 MWh. Microinverter megawatts flipped back to +7.4%. That flip moved with U.S. safe-harbor dollars, and microinverter channel inventory was still slightly elevated, which is why Q3 is another undership quarter. Margin compression narrowed in Q2 and the guide re-widens it.

(b) FQ+1 — Q4 2026e

No Q4 revenue, margin, battery, opex, or EPS range. The only numeric Q4 figure is safe harbor. The CEO was asked whether Q4 revenue should be higher than Q3 and answered with the Q3 bridge instead.

(c) FY+1 — FY2027, and the year still in progress

Enphase does not guide the year. There is no old annual revenue range to revise. FY2027 in this table is the next fiscal year. FY2026 is shown because it is the year the quarterly guides add up to, and it is also not guided.

Five-year revenue, so the quarterly path has a base. FY2025's +10.7% did not stick. It was a U.S. year (+27.2%) while international was already −28.2%.

The words have warmed. The year-over-year math has not. Coverage of the midpoint has gotten thinner: about 90% into Q1, about 85% into Q2, over 70% into this quarter on July 28. The deliberate undershipment has been walked down, from about $45 million into Q4 2025, to about $25 million into Q2, to about $15 million into Q3 against sell-through they put near $245 million. That is a less defensive channel stance. It is not a claim that U.S. residential demand has recovered. Management said U.S. sell-through was still down 34% year over year in Q2. June permits, on the third-party data they cited, were still about 30% below the prior year.

Tone across the last four calls. Q3 2025 was a record quarter — revenue $410.4 million, batteries 195.0 MWh — and an explicit cliff. They then guided Q4 down and put Q1 2026 directionally around $250 million because the Section 25D credit ended December 31, 2025. February raised that Q1 range to $270 million–$300 million and still called Q1 the low for underlying demand, with a January letter that promised non-GAAP opex of $70–$75 million from this quarter. April said sell-through was roughly 10–15% below the prior view, so Q2 became a safe-harbor quarter with a $25 million undershipment. July said core bottoms in Q2. The core comment on that call was sequential. The year-over-year core rate at this midpoint is still about −32%. Batteries are the beat, and the January opex target is not in the guide. Propel originations were still about 200 a week, the same rate as April. The margin range did not rise when the tariff assumption improved.

What the guide assumes. More than 70% of $305 million was booked. Safe harbor of $75 million is the 5% method, not the physical-work-test book. The Q2 10-Q states year-to-date safe-harbor agreements of about $1,081 million: $202 million under the 5% method and $878.6 million under the physical-work test, with no physical-work-test revenue recognized and recognition described as likely beginning in 2028. Europe is assumed flat despite summer, with the battery breakout deferred to Q4 and the Netherlands. Reciprocal tariffs of about 2 points are in both gross-margin ranges, and further battery price cuts can still be absorbed inside 44–47%. PTC transfer sales are already in the margin guide. New products — fifth-generation battery, bidirectional charger, IQ Vault, IQ SST — are outside this quarter's revenue.

What would break it. Core near Q2's $208 million would falsify "Q2 was the trough" one quarter after they said it. February's "Q1 was the low" already failed that test once safe harbor is removed. Revenue below $290 million would break seven straight prints that landed inside or above the range: six inside, and Q3 2025 above the high end. None of those seven missed the low end. Non-GAAP margin below 44% cannot be blamed on the old 3-point tariff. Opex above $80 million would be a second miss. A Q4 comment that walks the 5% safe-harbor total down, or that pulls physical-work-test revenue into 2026, rewrites the bridge.

What changed after July 28. No 8-K, release, or conference revised the ranges. Daloopa has no 2026Q3 document.

• The September 16 rate hike, to 3.75–4.00% from a hold at 3.50–3.75% on July 29, landed after the guide. Rates were already the stated reason distributors are cautious. The undershipment assumption is less conservative than it looked on July 28. It does not, by itself, revise $290–$320 million.
• September 8: IQ SST 4 kW modules are being built in Arlington, Texas. Full-rack demonstrations are expected in November, after this call. Customer pilots remain 2027. Not revenue.
• September 21: IQ Meter Collar approvals reached 80 utilities serving more than 38 million customers, including the three large California investor-owned utilities and SMUD, up from 71 utilities on August 10. Useful for whole-home backup attach. Late for a quarter that ends September 30. Read it as Q4 and 2027, not as evidence the 130–150 MWh range is already through the high end.
• Operating management is unchanged. The August 10 board decision kept a director and moved a committee chair. It is not a CEO or CFO change.
• IRS Notice 2026-15 was already out before the Q2 call. Effective-control guidance is still the open item. It is unlikely to change units already in the Q3 channel. It can change the October commentary on 2027 TPO demand.

The practical setup, with the quarter ending six days after this note: safe harbor at $75 million is the highest-confidence line, because Q2 landed on the dollar they named. The open third of the midpoint is whether they shipped it and whether they stuck to the $15 million undershipment. Batteries and non-GAAP margin are where price cuts can still move the result inside the range. The bullish sentences are 2027 and 2028. The quarter they guided is still a mid-20s year-over-year revenue decline.

Consensus is pinned to the $305 million midpoint, a deeper year-over-year decline than Q2. The catalysts management will spend the call on are mostly not in that number, and FY2027E at +4.5% does not underwrite them. Peers have not reported a comparable Q3.

Listen in this order. Split safe harbor from core before calling $305 million a beat. Battery megawatt-hours versus 130–150, and whether any of those megawatt-hours are safe-harbor storage. U.S. dollars ex safe harbor, which were about −38% in Q2. Originations per week versus "a bit above 200," not the state count. Whether the Q4 G5 date holds. Europe on a year-over-year basis, not the sequential +35%. Non-GAAP margin inside 44–47%, knowing 2 points are still tariff. SST and the bidirectional charger are milestone items only.

Window: after the July 28 print through September 24. Nothing in the window revises the Q3 guide. No named data-center customer, purchase order, acquisition, or CEO or CFO change. The trajectory is execution on adjacencies, not a reported turn in U.S. residential demand.

Also in the window, and too small to move a $290–$320 million quarter: Enphase Care passed 7,500 U.S. enrollments on September 1, and the homeowner AI assistant went live on August 20. Neither disclosed revenue. The August 4 U.S. manufacturing note is positioning for domestic-content and FEOC compliance, not a new factory and not Q3 revenue.

Mixed versus the street, and the recent beat has shrunk to noise. Over the last 12 reported quarters Enphase beat FMP non-GAAP EPS in 5 of 12 and revenue in 6 of 12. The last 4 are 2 of 4 on EPS and 4 of 4 on revenue, but three of those revenue beats are under 1.5% and the EPS line has gone from +45% to a miss. Beat magnitude is deteriorating. Q3 2026 is not on the grid.

Scored actuals are net revenue and non-GAAP diluted EPS. On every quarter in this window, FMP's actual equals those lines. GAAP would turn recent prints into large misses and is the wrong score. Q2 GAAP diluted EPS was $0.27 against a non-GAAP $0.46. A beat is any positive surprise. Q1 2026 EPS was a $0.0011 miss that rounds to $0.47 on both sides and is flagged in-line. Versus the company's own revenue range the pattern is different: 11 of 12 inside, one above (Q3 2025). The last two prints landed under the midpoint.

The last-four EPS average of +18.5% is Q3 2025 at +45.2% and Q4 2025 at +31.5%, averaged with −0.2% and −2.4%. Take out Q3 2025 and the revenue average of the other three recent prints is +0.8%, not +5.9%. The last two prints are −1.3% on EPS and +0.5% on revenue. Q2 non-GAAP EPS of $0.46 missed $0.4711 because the bar had already been cut, not because earnings held up. Non-GAAP operating income was $56.7 million, −42.5% versus $98.6 million.

Management explains the variance against its own guide, not against FMP. The street miss is usually where inside the range they land, how much safe harbor is in the quarter, and how much margin the credit, tariffs, and mix produce. Q3 2025 cleared the high end because about $70 million of safe harbor was pulled forward and was not in the guide. This quarter's $75 million is already in the guide, and the street's revenue average is the midpoint. A repeat of $0.46 misses both the $0.48 endpoint and the $0.514 average. Nothing on the last two calls said that step-up is in the Q3 guide.