Array fires on all cylinders — and the market keeps selling it

A record $2.5B order book and a raised margin outlook meet a 49% ten-week drawdown; the Atlas and AWM bets are the real storyline.

ARRY · Earnings Call · 2026-08-05

A record quarter on a falling knife

Array Technologies reported a Q2 that beat every guidance line it had set: revenue of $342 million, up 53% sequentially and above the $300–320 million guide; adjusted gross margin of 30.8%; adjusted EBITDA of $63 million, up 119%; and a third consecutive record order book of $2.5 billion. “Q2 was a quarter of exceptional momentum across every key metric on the page,” — Kevin Hostetler, CEO · 2026-08-05 CEO Kevin Hostetler opened. The order book was the headline — up 37% year-over-year with over $500 million of new commercial momentum in the quarter, roughly half from Tier 1 customers, several projects above 500 MW, and a 12-month book-to-bill of 1.5x. Yet the tape is screaming the opposite. The recent 90-day series shows a 34.5% decline, with a 49% drawdown in the last ten weeks as the stock sank toward $4.70 from a May 28 peak of $9.26. The full-period series is starker still — down 87% from the January 2021 high of $51.05. The market has spent years repricing Array's story, and this quarter — objectively strong — has not reversed that flow.

The innovation engine as the differentiator

What's genuinely new in this report is the breadth of the product cadence. Products launched since 2023 — OmniTrack, SkyLink, SmarTrack, Hail XP, and APA — now represent roughly 50% of the order book and nearly half of 2026 revenue, versus a third in 2025. COO Neil Manning: “2026 is our largest launch year ever, with 5 significant product introductions,” — Neil Manning, President and COO · 2026-08-05 anchored by the DuraTrack 60-degree variant (a storm-stow option for insurers unwilling to pay for more tracker than a site needs) and the Array Atlas foundation-to-tracker suite, launched just last week. Atlas II is engineered to cut components by 70% versus APA's traditional A-frame, a direct installation efficiency win. Both qualify for the 45X manufacturing credit, and Atlas I opens a >$1 billion traditional-foundation market at roughly $0.03–0.04 of ASP per watt on top of a typical tracker sale. The foundation story matters because it is the first true productization of the APA acquisition. The integration process that brought APA's engineered foundations into the Array tracker platform is explicitly the template Kevin is now applying to the next deal — Affordable Wire Management (AWM), a ~$60 million trailing-twelve-month cable-management business that extends Array into battery storage and even data-center racking applications.

It's disciplined adjacency, not just a roll-up strategy, really expanding our balance of systems offering to a great engineered category that we really understand and with customers that we're already serving domestically.
Kevin Hostetler, CEO · 2026-08-05

The deal prices at roughly 6x EBITDA after step-up benefits and is expected to be high-single-digit accretive to adjusted EPS in year one before synergies — funded entirely from a cash position of $307 million that more than doubled sequentially.

Margin normalcy and the second-half mix

The margin story deserves a closer read. The H1 consolidated adjusted gross margin of 30.8% is flattered by one-time items — a tariff recovery and catch-up 45X benefits that totaled roughly 300 basis points in Q1 and under 50 in Q2. CFO Keith Jennings was explicit about it: “first half margins are to be commended... however, they're not to be fully extrapolated.” — Keith Jennings, CFO · 2026-08-05 Strip those out and the normalized H1 rate is closer to 28.8%; the raised full-year guide of 27–28% is 100 basis points above prior communication. The second-half drag is a step-up in international mix (from ~5% of revenue in H1), higher commodity and logistics costs, and the absence of the one-time Tariff refund tailwind. The offset is productivity — significant enough on the call to keep Keith comfortable with the margin line even as input costs creep higher. The demand commentary connects this to a global theme: nearly every major renewables and grid-equipment reporter in this window is talking about tariff refunds, IEEPA refunds, and 45X. Array is squarely in that camp, but with a twist — it is actively converting the 45X regime into a structural product advantage (the Atlas qualification), not just a P&L tailwind. That is the most differentiated thing on the call.

Cash, leverage, and the PREF decision

The other notable change is cash generation. Free cash flow was $114 million in the quarter, and total available liquidity is now over $640 million with net debt leverage down to 2.1x from 2.7x. That gives Array the capacity — and the conviction — to do AWM in all cash. The Series A preferred flips to cash pay in August, a small ~$12 million cost into 2026 year-end at a 6.25% coupon stepping 50bps annually; Keith said they're comfortable servicing it and will only refinance if a cheaper instrument makes sense. The prior call's framing is consistent — on 2026-05-09 Kevin noted “we've had over $900 million in new orders over the last 2 quarters” — Kevin Hostetler, CEO · 2026-05-09 — and momentum has continued. The definition caveat still holds, too: “we have not made any changes to our order book and how we define that order book” — Kevin Hostetler, Chief Executive Officer · 2026-02-26 — the 37% growth is real volume, not definitional slack. The risk is equally clear: 80% of a $2.5 billion backlog converting over six quarters is a strong signal, but project timing — permitting, interconnection, site readiness — can push revenue below midpoint, and free cash flow conversion is now guided to just 20–25% of EBITDA (half the prior run-rate expectation) because of the Q4-heavy shape pushing collections into 2027. For an investor, the interesting question is whether the market is pricing Array as a one-quarter wonder or as the early beneficiary of a genuine balance-of-systems build. The innovation engine is company-unique; the tariff/45X tailwind is shared; the AWM move into battery and data-center adjacencies is a credible extension of the strategy. On a normalized basis, margins are holding in the high-20s, and the P&L is flexing. The stock's drawdown says the market is not buying it yet — but the order book and the product cadence say the company is executing.