Nextpower Inc. [NXT] — Company Research
formerly Nextracker Inc.; renamed 2025-11-12
Tier-2 memo · framework: Criteria, 2026-07-29 · as of 2026-07-29 · spot $95.35 (close 2026-07-28)
This document issues no position verdict. It scores Criteria and outputs an analysis. Whether that analysis justifies a position is a question about a particular book, and the book decides.
0. Start here: the company is not called Nextracker any more
The brief names this company "Nextracker Inc." The registrant's current legal name is Nextpower Inc.
SEC formerNames for CIK 0001852131 reads ["Nextracker Inc.", "Nextracker LLC"]. The rename was announced
2025-11-12 with an 8-K carrying Item 5.03 (amendments to articles of incorporation) and the amended and
restated charter and bylaws as exhibits. The FY2026 10-K is filed under "NEXTPOWER" throughout.
The press release states the reason in the company's own words:
"Nextracker Rebrands as Nextpower to Reflect the Company's Position as an Integrated Power Technology Innovator… Announces development of power conversion product line… Hosts Capital Markets Day; announces FY27 outlook and FY30 financial targets" — Exhibit 99.1 to 8-K filed 2025-11-12
"$4.8 billion to $5.6 billion in revenue by FY30, with expectations of one-third of revenue to come from sales of non-tracker products and services"
This is not cosmetic and it is the whole story. A company whose entire revenue base is US utility-scale solar trackers renamed itself, set a five-year target in which a third of revenue is not trackers, and then — in the eight weeks after its fiscal year ended — agreed to buy a battery-storage business and a German steel business. It did that against a US policy deadline that expired 25 days ago.
1. Screen-input audit — one large share-count error, everything else clean
| Input | Screen (NXT_analysis.json) |
Filed | Verdict |
|---|---|---|---|
| Shares | 136,542,423 | 150,274,472 Class A outstanding, 2026-05-11 (10-K cover); no Class B — the Up-C is fully collapsed. Balance sheet 2026-03-31: 149,391,483. Diluted WA FY2026: 152,710,033; basic WA 147,976,256 | WRONG — 9.1% below outstanding, 10.6% below diluted. Market cap understated by $1.31–1.54bn. |
| Net cash | $1,094,976,000 | Cash $1,094.976m, zero debt (revolver undrawn; total liquidity ~$2.0bn) — but there is a Tax Receivable Agreement liability of $372.659m, a real cash obligation to Flex/TPG that the screen ignores | INCOMPLETE. Net cash before TRA is right; net of the TRA it is $722.317m. |
| TTM revenue | $3,559,390,000 to 2026-03-31 | $3,559.390m = FY2026, the full fiscal year ended 2026-03-31 | CORRECT. NXT's fiscal year ends 31 March, so "TTM to 2026-03-31" is a clean audited FY, not a stitched window. No quarter-skip defect. |
| Gross margin | 32.6% | 32.6% ($1,160.095m / $3,559.390m) | CORRECT |
| Operating margin | 19.6% | 19.6% ($697.266m / $3,559.390m) | CORRECT |
| Op margin change | −2.0pp | −2.0pp (21.6% FY2025 → 19.6% FY2026) | CORRECT — and note the sign. Margins are contracting, two years running (23.5% → 21.6% → 19.6%). |
| Demonstrated CAGR | 23.2% | 23.2% = FY2023 $1,902.1m → FY2026 $3,559.4m, 3 years | CORRECT, window now named |
| Entity name | "Nextpower Inc." | Nextpower Inc. | The screen has this right and the brief does not. |
EPS cross-check, as instructed:
585,883 / 147,976,256 basic = $3.959 vs filed basic $3.96 ✓
585,883 / 152,710,033 diluted = $3.836 vs filed diluted $3.84 ✓
585,883 / 136,542,423 (screen)= $4.291 vs filed diluted $3.84 ✗ — off by 11.7%
The filed EPS refutes the screen's share count directly. No dual-class complication remains: Class B was fully surrendered when the Up-C collapsed, and only Class A is outstanding.
Recency. Latest filing is the FY2026 Form 10-K, filed 2026-05-19 for the year ended 2026-03-31 — 71 days old. Q1 FY2027 (quarter ending ~2026-06-27) has not been reported; the prior-year analogue was an 8-K on 2025-07-29 and a 10-Q on 2025-08-01. Q1 FY2027 results are imminent or have just landed and are not in this analysis.
2. The trap: is 23.2% installed demand, or a pull-forward ahead of policy change?
Answer: the growth is real, the deadline is real, and the deadline expired on 4 July 2026 — twenty-five days before this memo. The pull-forward has already happened and the most recent reported quarter already declined year-on-year.
2.1 The policy cliff, dated, from the 10-K
"On July 4, 2025, a U.S. federal budget reconciliation bill known as the One Big Beautiful Bill Act ("OBBBA") was enacted… whereas under the IRA, Section 48E and 45Y credits were available through 2032… the OBBBA substantially reduced this timeframe to require that projects begin construction by July 4, 2026 to utilize a continuity safe harbor that permits solar projects to be placed in service within four calendar years… Alternatively, solar projects that begin construction after July 4, 2026 must be placed in service by December 31, 2027 to qualify for the Section 48E and 45Y credits. Such acceleration in the expiration of these tax credits will reduce the number of projects in future years that would have otherwise qualified for such credits, likely reducing the overall project volume over time." — FY2026 Form 10-K
And the safe harbour was tightened again, mid-year:
"…on July 7, 2025 President Trump issued an Executive Order directing the Secretary of the Treasury to take measures to strictly enforce the termination of the Sections 48E and 45Y credits… Treasury guidance… was issued on August 22, 2025 in the form of IRS Notice 2025-42. Under this guidance the 5% safe harbor was eliminated effective September 2, 2025 and additional requirements were imposed on solar projects for purposes of demonstrating both the start of physical construction and continuous physical construction thereafter."
Plus FEOC (foreign-entity-of-concern) restrictions applying to 45X, 48E and 45Y, with interim Treasury guidance only issued 2026-02-12 (Notice 2026-15) and further guidance still outstanding.
Timeline, so the pull-forward window is unambiguous:
| Date | Event |
|---|---|
| 2025-07-04 | OBBBA enacted. 48E/45Y compressed to a begin-construction-by-2026-07-04 safe harbour. |
| 2025-09-02 | The 5% cost safe harbour is eliminated; physical-construction test imposed. Bar raised. |
| 2026-02-12 | FEOC interim guidance (Notice 2026-15); further guidance still pending. |
| 2026-07-04 | The begin-construction deadline. PASSED — 25 days before this memo. |
| 2027-12-31 | Placed-in-service deadline for anything that began construction after 2026-07-04. |
NXT's fiscal 2026 (Apr 2025 – Mar 2026) and fiscal 2027 Q1 (Apr – Jun 2026) sit exactly inside the safe-harbour rush. Every developer racing to break ground before 2026-07-04 needed trackers ordered, deposited and scheduled. The screen's 23.2% demonstrated CAGR and the >$5.25bn backlog were both built into that window.
2.2 The most recent quarter already declined
| $m | Q1 | Q2 | Q3 | Q4 | FY |
|---|---|---|---|---|---|
| FY2025 (Apr'24–Mar'25) | 719.9 | 635.6 | 679.4 | 924.3 | 2,959.2 |
| FY2026 (Apr'25–Mar'26) | 864.3 | 905.3 | 909.4 | 880.4 | 3,559.4 |
| YoY | +20.1% | +42.4% | +33.9% | −4.7% | +20.3% |
(Q1–Q3 from XBRL Revenues; Q4 derived as FY − 9M and confirmed against the Q4 press release's "$881".)
Q4 FY2026 revenue fell 4.7% year-on-year and 3.2% sequentially. Operating income fell harder:
| $m | FY2026 Q1 | Q2 | Q3 | Q4 | Q4 FY2025 |
|---|---|---|---|---|---|
| Operating income | 186.2 | 181.3 | 176.1 | 153.7 | 195.3 |
Q4 FY2026 operating income −21.3% YoY, and down every single quarter through the year. A 23.2% three-year CAGR is a true description of the past and a poor description of the run-rate.
2.3 The company's own guidance says the same thing
FY2027 outlook, raised at Q4 (8-K 2026-05-12):
| Updated | Previous | FY2026 actual | |
|---|---|---|---|
| Revenue | $3.8 – 4.1bn (mid $3.95bn, +11.0%) | $3.6 – 3.8bn | $3.559bn (+20.3%) |
| GAAP net income | $501 – 559m (mid $530m) | — | $585.9m |
| GAAP diluted EPS | $3.19 – 3.56 | — | $3.84 |
| Adjusted EBITDA | $825 – 900m | $800 – 900m | — |
| Adjusted diluted EPS | $4.21 – 4.59 | — | — |
The company guides revenue +11% and GAAP EPS down 7–17%. That is a raised outlook, and it still contains a decline in reported earnings. FY30 target of $4.8–5.6bn implies 9.9%/yr from FY2026 — less than half the demonstrated 23.2%.
2.4 Where the growth actually came from — geography and volume
"Revenue increased by $600.2 million, or 20%, for our fiscal year 2026… driven by a 13% increase in GW delivered as we delivered approximately 38 GW during fiscal year 2026, compared to 34 GW during fiscal year 2025… Revenue increased approximately $699.1 million, or 34%, in the U.S. while decreasing slightly by $98.9 million or 11% in the Rest of the World." — FY2026 10-K MD&A
| FY2024 | FY2025 | FY2026 | |
|---|---|---|---|
| US revenue | $1,702.6m (68%) | $2,031.6m (69%) | $2,730.7m (77%) |
| Rest of world | $797.2m (32%) | $927.6m (31%) | $828.7m (23%) |
| US YoY | — | +19.3% | +34.4% |
| RoW YoY | — | +16.4% | −10.7% |
Every dollar of FY2026 growth came from the United States, and international revenue shrank. That is the signature of a domestic policy-driven rush, not of broad-based global demand. Volume grew 13% while revenue grew 20% — the extra 7pp is price/mix, of which the 10-K names $365.0m of "point in time revenue… components directly shipped… including software licenses" plus four acquisitions.
Verdict on the trap: the 23.2% demonstrated CAGR is genuine installed demand, but a material and unquantifiable share of the FY2026 portion of it is a safe-harbour pull-forward into a deadline that has now passed. The company's own FY2027 guide (+11%) and its own FY30 target (+9.9%/yr) are the best available statements of what remains.
3. Backlog: duration, coverage — and the number that matters more
3.1 What is disclosed
"At the end of fiscal year 2026, our backlog was over $5 billion and included project-specific purchase orders and VCAs comprising multiple specific projects. We define backlog as executed EPC or VCA contracts or purchase orders with deposits of cash paid or financial equivalents, identified named project sites, product and volume requirements, and ship dates." — FY2026 10-K
The Q4 press release is more precise: "increased backlog to a record level of over $5.25 billion."
| Fiscal year end | Backlog | Revenue that FY | Coverage |
|---|---|---|---|
| FY2025 (2025-03-31) | >$4.5bn | $2,959.2m | 1.52x |
| FY2026 (2026-03-31) | >$5.25bn | $3,559.4m | 1.47x |
Backlog grew +16.7% while revenue grew +20.3% — implied book-to-bill ≈ 1.21x. Coverage against the FY2027 guided revenue of $3.95bn is 1.33x.
Backlog duration is not disclosed. The 10-K gives no conversion schedule, no aging, and no split between purchase orders (near-term) and VCAs ("multiple projects typically to be deployed over multiple years"). This is a real gap and it is stated rather than estimated. The most that can be said from the filing is that coverage of one year's guided revenue is 1.33x and that VCAs are multi-year by the company's own definition.
3.2 The number that matters more — only $410.0m of that backlog is an enforceable ASC 606 obligation
"As of March 31, 2026, Nextpower had $410.0 million of the transaction price allocated to the remaining performance obligations. The Company expects to recognize revenue on approximately 75% of these performance obligations in the next 12 months. The remaining long-term unperformed obligation primarily relates to extended warranty and deposits collected in advance on certain tracker projects." — FY2026 10-K, revenue note
Set the two disclosures side by side:
| Measure | Amount | % of the other |
|---|---|---|
| "Backlog" (company-defined, not a GAAP measure) | >$5,250m | — |
| ASC 606 remaining performance obligations | $410.0m | 7.8% of backlog |
92% of the reported backlog does not meet the ASC 606 definition of a remaining performance obligation, and the residual that does is described as primarily extended warranty and advance deposits — i.e. not even tracker hardware. This is not an accusation of impropriety: company-defined backlog and ASC 606 RPO measure different things, and the company defines its term clearly. It is a statement about how much of the headline number is legally enforceable revenue and how much is a pipeline the company believes in.
The 10-K's own risk factor says the same thing in words:
"Because backlog is inherently uncertain and subject to change, the contracts comprising our backlog may not result in actual revenue in any particular period or at all due to… project cancellations, suspensions, delays, scope reductions, failure to execute anticipated supply agreements under framework or volume commitment arrangements… Cancellation of or adjustments to contracts have occurred in the past and may occur in the future."
This is the AAOI-class finding for NXT. Applied Optoelectronics had a private stocking distributor at 53% of revenue and ~264-day DSO named in no release. Nextpower has a $5.25bn headline number, quoted in every release, of which $410m is the enforceable core — a fact disclosed only in a revenue footnote and never in a press release.
4. Accounting quality — the whole of FY2026's gross-profit growth is a government tax credit
This is the central finding of this memo and it is quantified straight from the MD&A.
4.1 The mechanism
"We have contractually agreed with these suppliers to either share a portion of the economic value of the credit related to our purchases in the form of a vendor rebate or assign their credit directly to us… We account for the 45X Credits shared or assigned to us as a reduction of the purchase price of the parts acquired from the vendor and therefore a reduction of inventory until the control of the part is transferred to the customer, at which point we recognize such amounts as a reduction of cost of sales." — FY2026 10-K
The Section 45X advanced-manufacturing credit is earned by NXT's suppliers on domestically produced torque tubes ($0.87/kg) and structural fasteners ($2.28/kg). Nextpower captures the economics as a credit against cost of sales. It flows through gross profit.
4.2 The size of it
"During fiscal year 2026, we recognized approximately $379.9 million of reduction to cost of sales related to the 45X Credit earned on production of eligible components shipped during the period, compared to $224.9 million recognized in fiscal year 2025." — FY2026 10-K MD&A
| $m | FY2025 | FY2026 | Change |
|---|---|---|---|
| Revenue | 2,959.2 | 3,559.4 | +600.2 |
| GAAP gross profit | 1,008.8 | 1,160.1 | +151.3 |
| 45X credit in cost of sales | 224.9 | 379.9 | +155.0 |
| Gross profit EXCLUDING 45X | 783.9 | 780.2 | −3.7 |
| GAAP gross margin | 34.1% | 32.6% | −1.5pp |
| Gross margin excluding 45X | 26.5% | 21.9% | −4.6pp |
The entire $151.3m of FY2026 gross-profit growth is accounted for by a $155.0m increase in a government tax credit. Excluding 45X, gross profit fell $3.7m — down 0.5% on revenue up 20.3%.
Carried to the operating line:
| $m | FY2026 |
|---|---|
| GAAP operating income | 697.3 |
| less 45X credit in cost of sales | (379.9) |
| Operating income excluding 45X | 317.4 |
| GAAP operating margin | 19.6% |
| Operating margin excluding 45X | 8.9% |
45X supplies 10.7pp of a 19.6% operating margin — 54% of operating income.
4.3 It is legislated to disappear, on a schedule the 10-K prints
"The Section 45X Credit amount is reduced each year by 25% in calendar years 2030, 2031 and 2032. Under current law, there are no Section 45X Credits available for components sold after December 31, 2032." — FY2026 10-K
| Calendar year | 45X available |
|---|---|
| through 2029 | 100% |
| 2030 | 75% |
| 2031 | 50% |
| 2032 | 25% |
| 2033+ | 0% |
Inside a five-year valuation window (FY2027 → FY2031) the credit is intact for the first ~3.5 years and
then steps down. The terminal year, FY2031 (Apr 2030 – Mar 2031), spans nine months at 75% and three
months at 50% — 68.75% of the current credit. This is the derivation behind the terminal margin used in
NXT_Valuation.md, and it is why the implied-path test is run at three terminal margins rather than one.
4.4 Two more items in the same direction
Tariffs are eating the rest.
"…the impact from a $110.7 million increase in tariffs which increased to $130.4 million in fiscal year 2026 from $19.7 million in fiscal year 2025… Gross margin decreased by 150 basis points, to 32.6%… primarily resulting from the increase in tariffs noted above that were not fully included in pricing."
Tariffs are now 3.7% of revenue and, by the company's own words, not fully passed through.
Cash conversion has turned, and the 45X receivable is why.
| $m | FY2024 | FY2025 | FY2026 |
|---|---|---|---|
| Net income | 496.2 | 517.2 | 585.9 |
| Cash from operations | 429.0 | 655.8 | 562.9 |
| CFO / NI | 0.86x | 1.27x | 0.96x |
| Section 45X credit receivable (balance sheet) | — | 215.6 | 352.6 |
"Accounts receivable and contract assets in aggregate increased $64.6 million… coupled with a $267.1 million increase attributable to our Section 45X credit receivable." — FY2026 10-K
A record year of GAAP earnings converted to less cash than the prior year, and the largest single working- capital drag is the receivable for the very credit that produced the earnings. The 45X receivable now stands at $352.6m — 60% of a year's net income sitting on the balance sheet as a claim on the US Treasury, under a statute Congress has already amended once.
4.5 What is clean
- No customer >10% of revenue is tagged for FY2025 or FY2026 (the XBRL carries a
CustomerGrevenue concentration dimension for FY2024 only). The 10-K states: "As of March 31, 2026… our largest customer constituted less than 10% of our total trade accounts receivable and contract assets balances." The risk factor still warns that "a small group of customers have historically accounted for a material portion of our revenue" — the concentration is real but no single name crosses 10% in the two most recent years. - Geographic concentration is disclosed and is the more binding constraint: US 77% of revenue (Brazil was 11% in FY2024, below 10% since).
- Zero debt; revolver undrawn; ~$2.0bn of total liquidity; investment-grade rating achieved in FY2026.
- FY2024's $121.4m cumulative 45X catch-up was adjusted out of the non-GAAP measures as non-recurring. From FY2025, 45X is included in adjusted figures and in management incentive compensation — a change the 10-K discloses explicitly. That is honest labelling; it also means the adjusted numbers no longer strip it.
5. The mechanism — named, specific, evidenced
Nextpower's forward mechanism is the deliberate replacement of a US-tracker revenue base with non-tracker content and non-US geography, executed through named acquisitions with filed prices, because the US tracker policy support is legislated to shrink.
This is unusually well evidenced — every leg has a filing behind it.
5.1 Non-tracker attach, already showing in the mix
"Solar tracker system sales was approximately 88% of total revenue and non-tracker sales was approximately 12% of total revenue, which was up from approximately 8% from fiscal year 2025. The growth in our non-tracker platform solutions sales was higher than our solar tracker system sales, a trend we expect to continue." — FY2026 10-K
Non-tracker revenue: ~$237m (FY2025) → ~$427m (FY2026), +80%. Target: one-third of revenue by FY30.
5.2 Named products, with adoption evidence
| Product | Named where | Evidence of adoption |
|---|---|---|
| NX PowerMerge™ trunk bus connector (eBOS) | Q4 FY26 press release | "record eBOS quarterly bookings, including bookings of over 100 MW"; "in final stages of UL qualification testing" |
| TrueCapture™ (yield software) | Q4 FY26 press release | "record quarterly and annual TrueCapture revenue" |
| NX Horizon-XTR™ (terrain-following tracker) | Q4 FY26 press release | "Exceeded 50 GW of cumulative sales" |
| NX Anchor™ / NX Earth Truss™ (foundations) | Q4 FY26 press release | "now deployed at multi-gigawatt scale" |
| Robotics (bundled) | Q4 FY26 press release | "Booked the first bundled VCA project incorporating robotics" |
| Power conversion systems (PCS) | Rebrand release 2025-11-12; Q4 FY26 release | "first shipments expected in 2026"; agreement to acquire PCS product lines and IP, subject to Spanish FDI approval |
| NX One™ (unified software platform) | Q4 FY26 press release | "Piloted" — earliest stage |
| Cumulative scale | Q4 FY26 press release | "Surpassed 160 GW of cumulative tracker shipments globally"; ">25 GW each in Latam and MEIAT" |
5.3 Named acquisitions, with prices, from 8-Ks
Inside FY2026 (four, $117.2m cash net of cash acquired, per the cash-flow statement): Bentek (2025-05-07), OnSight (2025-05-09), Origami Solar (2025-09-08), Fracsun (2025-11-07).
After the fiscal year end — the two that reveal the strategy:
| Date | Target | Consideration | What it is |
|---|---|---|---|
| 2026-05-28 | Prevalon Energy LLC | up to $365m — ~$150m cash at closing, $50m in Class A stock issued one year after closing (priced on the 60-day VWAP to 2026-05-27), up to $165m contingent cash | Battery energy storage. Entry into BESS and, per the CEO, "facilitate entry into the battery storage and data center verticals" |
| 2026-06-21 | Zimmermann PV-Steel Group GmbH & Co. KG (Germany) | up to €330m — ~€180m cash at closing, €105m in Class A stock at closing, up to €45m contingent cash | German PV steel/mounting. European manufacturing and non-US revenue |
(Both filed under Item 3.02, unregistered sales of equity securities, because part of each consideration is Class A stock issued under the Section 4(a)(2) exemption. Combined announced consideration ≈ $750m, of which roughly $175m is new stock — dilution of ~1.2% not yet in any share count above.)
Read the sequence. The OBBBA passes 4 July 2025. Four months later the company renames itself away from "tracker", sets a FY30 target with a third of revenue outside trackers, and announces a power-conversion product line. Seven months after that — weeks after the fiscal year closes and weeks before the begin-construction deadline — it agrees to buy a battery-storage company and a German steel company. Management is doing exactly what the policy analysis says it should. That is the mechanism, and it is credible precisely because it is expensive and irreversible.
5.4 What the mechanism does not yet do
It does not yet replace the 45X gross-profit contribution. Non-tracker is 12% of revenue; 45X is 54% of operating income. Storage, eBOS and European steel have to carry a margin gap that is currently filled by a tax credit, and none of the acquisitions has reported a quarter under Nextpower ownership.
6. Mention frequency across quarterly earnings releases — required core metric
Source substitution, stated. Alpha Vantage EARNINGS_CALL_TRANSCRIPT returned a rate-limit body for
every requested quarter on 2026-07-29 (shared 25/day free-tier cap, exhausted by parallel runs; the fetcher
correctly declined to cache it). No earnings-call transcript was obtainable. Mention frequency is
computed instead across the quarterly earnings press release furnished as Exhibit 99.1 to each Form 8-K
Item 2.02 — eight consecutive quarters from EDGAR. Primary, dated, management-authored; not a call
transcript, and labelled as such wherever used. Normalised per 10,000 words (releases run 2,917–3,700 words).
Per 10,000 words, Nextpower/Nextracker quarterly earnings releases:
| Term | FY25Q1 | FY25Q2 | FY25Q3 | FY25Q4 | FY26Q1 | FY26Q2 | FY26Q3 | FY26Q4 |
|---|---|---|---|---|---|---|---|---|
| 45X | 40.5 | 34.0 | 32.5 | 35.7 | 13.7 | 6.1 | 5.4 | 5.6 |
| backlog | 0.0 | 0.0 | 5.9 | 5.5 | 3.4 | 12.3 | 10.8 | 11.1 |
| bookings | 0.0 | 0.0 | 0.0 | 8.2 | 0.0 | 18.4 | 13.5 | 16.7 |
| power conversion | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 2.7 | 16.7 |
| eBOS | 0.0 | 0.0 | 0.0 | 8.2 | 10.3 | 12.3 | 5.4 | 13.9 |
| robotic | 0.0 | 0.0 | 0.0 | 0.0 | 20.6 | 0.0 | 2.7 | 11.1 |
| storage | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 2.7 | 2.8 |
| data center | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 2.8 |
| tariff | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 3.1 | 2.7 | 2.8 |
| safe harbor | 0.0 | 0.0 | 0.0 | 2.7 | 3.4 | 0.0 | 0.0 | 0.0 |
| OBBBA | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
Four readings, each load-bearing:
- "45X" collapsed 86%, from 40.5 to 5.6 per 10k words — while the credit grew from $224.9m to $379.9m and became 54% of operating income. Management stopped talking about 45X in its releases at exactly the point it became the majority of the earnings. The FY26Q1 release (July 2025, weeks after OBBBA) is where the drop happens: 35.7 → 13.7. This is the single most striking disclosure-behaviour finding in either name in this memo.
- "OBBBA" appears zero times in eight quarters of earnings releases, while the FY2026 10-K devotes multiple pages to it and names it as materially reducing the future availability of the credits the business depends on. An investor reading only the releases would not learn that the law changed.
- "tariff" first appears in FY26Q2 and never exceeds 3.1 per 10k, while the tariff line went from $19.7m to $130.4m and the 10-K names it as the cause of the 150bp gross-margin decline.
- The pivot vocabulary is loud and new: "power conversion" 0 → 16.7, "robotic" 0 → 11.1, "eBOS" 0 → 13.9, "bookings" 0 → 16.7, "storage" and "data center" first appearing in the last two quarters. The releases are being rewritten around the new story while the old story's economics go unmentioned.
7. Archetype and Quality Criteria
Archetype: COMPOUNDER. Profitable since before the IPO, 19.6% operating margin, positive free cash flow, investment-grade. The screen's classification is correct.
| Test | Value | Basis | Result |
|---|---|---|---|
| ROIC above WACC, with an evidenced mechanism for redeploying capital at that return | Operating income $697.3m on total assets of $4,073.2m of which $489.0m is goodwill and $78.4m is PP&E; the business is working-capital-light and asset-light. Redeployment mechanism is evidenced and named: four FY2026 acquisitions plus Prevalon and Zimmermann, ~$750m of announced post-year-end consideration, into non-tracker adjacencies with a stated FY30 mix target | 10-K; 8-Ks 2026-05-28, 2026-06-21 | PASS — and unusually well evidenced. A compounder that cannot reinvest is a bond; this one is visibly reinvesting. |
| Piotroski F-score / gross profitability (LEVELS apply on a COMPOUNDER) | Gross profitability 1,160.1/4,073.2 = 0.285; ROA rising in absolute terms; leverage zero; current ratio 2.45x; but gross margin, operating margin and CFO/NI all deteriorated YoY | 10-K | MIXED. Levels pass; the changes are negative on three of the classic F-score components. |
| Accruals (retained on both archetypes) | CFO/NI 0.96x (FY2026), down from 1.27x. Driver: the $267.1m 45X-credit-receivable build | 10-K | MARGINAL. Accruals turned positive (earnings above cash) for the first time in two years, and the reason is identifiable and specific. |
| Margin stability | Operating margin 23.5% → 21.6% → 19.6%; gross margin 32.5% → 34.1% → 32.6%; ex-45X gross margin 26.5% → 21.9% | 10-K | DETERIORATING |
Quality Criteria (BINDING): PASS, with the accruals component marginal and the margin trend negative.
The business is genuinely good — 20% growth, 19.6% margins, no debt, real products, evidenced reinvestment.
The concern is not quality; it is that more than half of the reported operating income is a tax credit
Congress has scheduled to expire, which is a valuation question, handled in NXT_Valuation.md.
8. Downside case with a named cause (Downside Criteria, MEASURED — blocks nothing)
Named cause: the 4 July 2026 begin-construction deadline has passed. US utility-scale solar project starts fall in calendar 2027–2028 as the safe-harboured pipeline is worked off and is not replaced, while the Section 45X credit that supplies 54% of operating income begins its legislated 25%/yr step-down in 2030. Nextpower's non-tracker and non-US pivot is real but arrives too slowly and at lower margin.
Every element is filed, dated and quoted in §2 and §4. Nothing here is a scenario invented for the memo.
How it becomes permanent rather than cyclical. Solar tracker demand is not a preference, it is a financing arithmetic: a 30–40% investment tax credit is what makes a utility-scale project clear its hurdle rate. Remove or compress the credit and a fraction of the pipeline never gets built at any price for trackers. NXT cannot price its way out — the 10-K already says the FY2026 tariff increase was "not fully included in pricing," so pass-through is incomplete even today. And the replacement revenue (storage, eBOS, European steel) is bought with cash and stock at ~$750m of announced consideration, against businesses whose margins are unknown and unreported.
Quantification. If FY2028–FY2031 revenue is flat to modestly down (the FY2027 guide of +11% is achieved, then the safe-harboured backlog runs out) and the operating margin reverts toward the ex-45X 8.9% as the credit steps down, terminal EBIT is roughly $350m. At a 15x EV/EBIT exit — below the growth-matched 23.1x because the growth would no longer be there — that is an EV near $5.25bn, an equity value near $5.97bn, and ~$39/share.
- Bear-case price: $50 (−48% from $95.35). (Set above the arithmetic $39 because the balance sheet is clean, there is no debt, and $722m of net cash and a 160 GW installed base put a floor under a franchise that would still be the global tracker leader.)
- Probability: 35%. Higher than a typical bear-case weight, because the triggering event is not a forecast — the deadline already passed — and only the magnitude and timing of the response are unknown.
- This is not a going-concern case. Zero debt, $1,095m of cash, undrawn revolver, ~$2.0bn of liquidity, investment-grade. The impairment is a de-rating of a policy-levered earnings stream, not insolvency.
Logged to the ledger with its cause. Downside Criteria is MEASURED and blocks nothing.
9. What could not be established — stated, not filled in
- Q1 FY2027 results (quarter ended ~2026-06-27). Not filed as of 2026-07-29; prior-year analogue landed 2025-07-29. This is the first quarter that will contain post-deadline order intake and it is the single most valuable missing datapoint.
- Backlog duration. Not disclosed. No conversion schedule, no aging, no PO-vs-VCA split. Coverage is computable (1.33x FY2027 guided revenue); duration is not.
- The revenue contribution of the four FY2026 acquisitions. The 10-K names them and gives $117.2m of cash consideration but no pro-forma revenue disclosure was found, so organic growth cannot be separated from acquired growth. FY2026's 20.3% therefore overstates organic growth by an unquantified amount.
- Street consensus for FY2027/FY2028. Alpha Vantage
EARNINGS_ESTIMATESshares the exhausted 25/day cap. No Street figure is quoted anywhere in this memo. The FY2027 leg of the NTM revenue used in the 12-month target is the company's own guidance; the FY2028 leg is an interpolation between the FY2027 guide and the company's FY30 target, and is labelled as such. - Earnings-call transcripts. Same cap. Mention frequency uses earnings releases instead (§6).
- Prevalon and Zimmermann financials. Neither 8-K discloses target revenue, margin or closing date. Neither has closed. The Zimmermann stock consideration is priced on a 30-day VWAP at closing, so the exact share issuance is unknown.
- Whether the FY2026 backlog was pulled forward. The company does not disclose order dates against the 2026-07-04 deadline, and no filing quantifies safe-harbour-driven ordering. The inference in §2 is drawn from the geography mix, the Q4 YoY decline and the FY2027 guide — it is an inference, and it is labelled as one.
Sources
All figures traced to filings retrieved from SEC EDGAR on 2026-07-29:
- Form 10-K, fiscal year ended 2026-03-31, filed 2026-05-19, accession 0001852131-26-000017
- Form 10-K, fiscal year ended 2025-03-31, filed 2025-05-22, accession 0001852131-25-000021 (FY2025 backlog)
- Form 8-K Item 2.02, filed 2026-05-12, accession 0001852131-26-000014, Exhibit 99.1 (Q4 FY26 results, FY2027 outlook)
- Form 8-K Item 3.02/7.01, filed 2026-05-28, accession 0001852131-26-000020 (Prevalon Energy)
- Form 8-K Item 3.02/7.01, filed 2026-06-22, accession 0001852131-26-000026 (Zimmermann PV-Steel)
- Form 8-K Item 5.03/7.01, filed 2025-11-12, accession 0001852131-25-000069, Exhibit 99.1 (rebrand to Nextpower; FY27 outlook; FY30 targets)
- Quarterly earnings releases, Ex-99.1 to 8-K Item 2.02, FY25Q1 through FY26Q4 (mention frequency, §6)
- data.sec.gov XBRL companyfacts and submissions, CIK 0001852131
- Alpaca Markets daily bars and options snapshots, 2026-07-28 close
- Tier-1 screen: reports/scan_all_v2/NXT_analysis.json, as of 2026-07-28