Array Technologies is the #2 single-axis solar tracker manufacturer in the US, trading at $7.58 — down 83% from its 2021 peak and -32.5% in the last week alone (post-Q4 earnings crash, RSI 24). The stock screens cheap at ≈8x NTM P/E. Whether it actually is depends entirely on which version of earnings you believe.

This memo lays out what the business actually is, what the financials show, and where the gaps are between market expectations and primary source evidence. No price targets, no recommendations. Just what the desk needs to know.

Business Overview

What they sell. Single-axis tracker systems that rotate solar panels to follow the sun, increasing energy production 15-25% vs fixed-tilt. Primarily utility-scale projects. Hardware (≈85% of revenue), software (SmarTrack, moving to subscription), and field services.

Two segments:

• Array Legacy Operations (US, 83% of FY2025 revenue): The real business. Albuquerque-based manufacturing. $1.07B revenue in FY2025 at 28% gross margin — though that margin includes $173.6M of Section 45X manufacturing credits. Strip those out and Array Legacy gross margin is ≈12%.
• STI Operations (international, 17% of revenue): The wreckage. Acquired for ≈$651M in 2021. Cumulative impairments: $460.0M ($338.6M goodwill + $91.9M long-lived assets + $29.5M H250 inventory write-down). Now running negative gross profit (-$1.4M in FY2025). Brazil revenue collapsed 76% from 2023 peak ($258M to $61M). The H250 product line is being phased out and replaced with DuraTrack. STI is functionally dead as an independent segment.

APA Solar acquisition (closed Aug 2025, $185.4M): Engineered foundation solutions — ground screws, helicals, c-piles. Contributed $50M in 5 months with 2x book-to-bill and ≈$100M backlog. EBITDA margin runs "a few hundred basis points ahead of core business." APA + new products now comprise nearly half of total order book value. This is the growth engine management is betting on, but 45X credits only apply to A-Frame structural fasteners for utility-scale, not the entire APA product line — some street models overstate this.

Revenue model. 85% recognized over time (percentage-of-completion), 15% at point of sale. Contract sizes range from hundreds of thousands to tens of millions. Two customers accounted for 25.9% of FY2025 revenue (unnamed). Orderbook: record $2.2B, >95% domestic, >2x book-to-bill in Q4 2025, providing 1.5-1.7x revenue coverage on 2026 guidance.

Core patent. US patent on linked-row, rotating gear drive system — the technology that lets one motor drive multiple rows (competitors need one motor per row). Management calls it the source of "greater reliability, lower installation costs, reduced maintenance requirements." This patent expires February 5, 2030 — 3.9 years from now. Zero analyst questions about it across the last four quarterly calls. Zero management commentary addressing post-2030 competitive dynamics. 74 issued US patents total (expiring 2030-2045), 216 pending applications, but the linked-row patent is THE moat.

Financial Profile

Four-year revenue trajectory:

Revenue is volatile — $1.6B to $916M to $1.3B in three years. EBITDA is more stable ($174-189M range) because management adjusts margins with volume. But the big number is free cash flow: it dropped 41% ($135M to $80M) despite 40% revenue growth. Working capital consumed $55M. Capex tripled ($7M to $22M). FCF/EBITDA conversion collapsed from 78% to 43%. Guidance implies similar conversion in 2026.

Capital structure — the real story:

Stockholders' equity is negative $206M and worsening (from -$118M a year ago). Common shareholders are deeply subordinated residual claimants.

The Blackstone preferred — a ticking clock. $493.1M liquidation preference held by BCP Helios (Blackstone), with an additional $90.8M in accrued and unpaid dividends as of 12/31/2025 — effective claim of ≈$584M. No maturity. No conversion rights. Perpetual. Upon any change of control or liquidation, Blackstone receives the GREATER of (i) liquidation preference + accrued dividends, or (ii) 130% of initial investment minus cumulative cash dividends paid — a floor that protects Blackstone's downside at common shareholders' expense. Current dividend: 6.25% PIK (adding ≈$30M/year in additional shares). But on August 11, 2026 — 5.4 months away — dividends become mandatory cash:

• Year 5 (Aug 2026): 6.25% = ≈$30.8M/year cash
• Year 6: 6.75% = ≈$33.3M/year
• Year 7: 7.25% = ≈$35.8M/year
• Year 8: 8.25% = ≈$40.7M/year
• Year 10: 10.25% = ≈$50.5M/year

If they don't pay cash: penalty rate of +200bps. After 6 consecutive quarters of non-payment: Blackstone can force equity issuance at 95% of 30-day VWAP. At $7.58, that's catastrophic dilution.

The math: At 2026 guided FCF (≈$93M midpoint), preferred cash dividends consume 33% of free cash flow in year one, rising to 54% by year 10. Management calls the preferred "reasonably priced long-term capital" and is NOT planning to redeem. The revolver expansion to $370M (Feb 18, 2026 — seven days before reporting a $161M quarterly loss) was almost certainly arranged to fund this transition.

Convertible notes. The 2028 converts ($325M, $23.86 conversion price) are deeply out of the money and will need cash refinancing. The 2031 converts ($345M, $8.12 conversion price with capped calls to $12.74) are 7% out of the money at current prices. Settlement is cash-first (cash up to principal, then shares for excess). If the stock stays below $8.12, the company owes $345M cash at maturity in 2031. Capped calls protect against dilution between $8.12 and $12.74, costing $35.1M.

45X dependency. Section 45X manufacturing credits recognized: $173.6M (FY2025), $137.8M (FY2024), $9.3M (FY2023). These are netted against product costs. Without 45X, Array Legacy gross margin drops from 28% to approximately 12%. FEOC (Foreign Entity of Concern) restrictions under OBBB could disqualify these credits if supply chain contains material assistance from China/Russia/Iran/North Korea. Treasury guidance on material assistance calculation was issued Feb 13, 2026; proposed rule still forthcoming. This is the single largest earnings variable for Array Legacy.

Competitive Position

ARRY is #2. NXT is #1. The gap is widening.

NXT does not mention Array Technologies by name in any of 7 earnings transcripts searched (FY2025 Q1 through FY2026 Q3). Not once. NXT describes market share gains as "flight to quality." NXT is executing the same balance-of-system strategy ARRY is pursuing — foundations (NX Earth Truss, which reduces part count by 10x), eBOS (Bentek acquisition), module frames (Origami), robotic inspection (Onsight), power conversion, fire detection — but with 3-5x the resources and zero balance sheet constraints.

Where ARRY has positioning advantages:

• Tier 1 customer mix improving (50%+ direct, 4 GW from competitor customers in FY2025) — reduces pushout risk
• APA foundation integration launching H2 2026 — first mover on integrated tracker+foundation
• DuraTrack product technically sound — management claims "dollar-per-watt savings that are material"
• Domestic manufacturing (Albuquerque + Ohio) provides tariff insulation and 45X eligibility

Where ARRY is structurally disadvantaged:

• NXT's 45X benefit is structural (≈11% on US revenue from 25+ partner manufacturing facilities) vs ARRY's partially historical amortization benefit that's falling off
• NXT absorbs tariff headwinds better: higher margins (low 30s vs 26-27%) provide more cushion, more diversified supply chain
• NXT can invest in R&D (≈$100M/year, 3x increase) AND buy back stock ($500M program). ARRY's cash goes to debt service and preferred dividends
• Patent cliff (Feb 2030) opens the door for NXT and others to replicate Array's linked-row gear drive advantage

Counterparty confirmation of demand environment:

• FSLR (First Solar): 2 GW/month pent-up demand released post-OBBB. 50.1 GW / $15B backlog. Record 17.5 GW modules sold in FY2025.
• SHLS (Shoals Technologies): Record backlogs, $900M+ quarterly quote volume, safe-harbored projects confirmed through 2030. SHLS explicitly says its revenue follows tracker installations by 1-2 quarters.
• Both independently confirm the solar demand environment is strong. Demand is not the problem. Competitive position within that demand is.

Management & Governance

Executive team. CEO Kevin Hostetler (since March 2022, previously VP at Honeywell UOP) runs an aggressive commercial strategy. Two new hires in January 2025: CFO Keith Jennings (ex-Haynes International, Fortune 500 experience) and CLO Gina Gunning (ex-Vista Equity, transactions background). COO Neil Manning is the most operationally candid of the three — describes international strategy as "intentionally selective."

Tone gap. CEO speaks in superlatives ("thrilled," "phenomenal," "exciting") and forward-looking strategic terms. CFO is measured and precise — explicitly states EBITDA margins will "hold at 15%" near-term before reaching "high mid-teens." COO is sober on operational execution. Truth is closer to the CFO. Revenue growth is real but margin expansion is constrained.

Compensation. CEO total comp: $8,867,007 (FY2024), up 55% from $5,716,945 (FY2023) per DEF 14A Summary Compensation Table — during a year when revenue declined 42%. The increase was driven by $6.99M in stock awards (vs $3.60M in FY2023), primarily supplemental retention RSUs granted September 2024. 79% of CEO comp is equity-linked (RSUs + PSUs). Performance metrics: 50% revenue, 25% adjusted gross margin, 25% adjusted EBITDA. PSU multiplier ranges 0-200% based on 3-year TSR. Current multiplier achievement: 0% on TSR vesting (stock down ≈80% from grant date).

Insider ownership. Director Hill City Capital: 9.4%. Institutional: BlackRock 10.1%, Vanguard 9.1%, GMO 7.3%. CEO holds 201,400 shares (≈$1.5M at current prices) — well below his 5x base salary ownership guideline ($4.375M required). Recent Form 4 filings verified: COO Manning (2/15/2026, 3,749 shares) and CLO Gunning (1/30/2026, 13,386 shares) were both code M (RSU vesting) + code F (tax withholding) — automatic grant transactions, not open-market purchases. Zero code P (open-market buy) transactions by any officer or director. At $7.58, the complete absence of insider buying is a negative signal.

Capital allocation track record. Mixed at best. STI acquisition ($651M) produced $460M in cumulative write-downs — a value-destroying deal that management doesn't adequately address. APA acquisition ($185M) looks better early but is a bet. The 2031 convertible issuance was well-timed (refinanced some 2028s at a discount, extended maturity). The revolver expansion was strategically timed before ugly Q4 reporting. Nine-bank syndicate including three new lenders (HSBC, RBC, BNP Paribas) signals institutional confidence. Management is competent at liability management — less so at M&A.

Factor Profile

Default factors (SPY, AGG, VXX, GLD, SHY, TLT) — 1Y daily:

• Total variance: 3.84%
• Idio variance: 3.37% → 88% idiosyncratic
• Alpha: +49% annualized
• Beta to SPY: 1.06 (p=0.03)

Looks great. Looks like a high-idio stock picker's dream. It's not.

With solar sector factor (TAN added):

• Total variance: 3.84%
• Idio variance: 2.27% → 59% idiosyncratic
• Alpha: -13% annualized
• Beta to TAN: 1.04 (p<0.001)

With TAN + NXT:

• Idio variance: 51%
• Alpha: -38% annualized
• TAN beta: 0.81, NXT beta: 0.41

ARRY fails the 75% idio variance target when properly benchmarked against its sector. The apparent positive alpha on default factors is entirely sector exposure — the solar sector has outperformed broad equity, and ARRY is a leveraged solar play. Company-specific alpha is deeply negative.

What this means: ARRY is a factor bet, not a company bet. Buying ARRY for solar exposure is inefficient — TAN or NXT give you the same sector return without the negative idio alpha. If you don't have edge on the solar sector factor itself, there's no reason to take ARRY's company-specific risk.

Where idio edge could exist: The three non-consensus structural variables — preferred cash drain timing, patent cliff, 45X dependency — are genuinely company-specific. An analyst with a differentiated view on any of these could have idio alpha. But the factor regression says the market's negative view on ARRY company-specific fundamentals (vs the sector) is, so far, justified.

Forward Expectations Gap Analysis

What $7.58 requires. yfinance shows forward P/E of 8.3x, implying NTM consensus EPS of ≈$0.91. But management guides FY2026 adjusted EPS of only $0.65-$0.75 — a 21-40% gap between street consensus and the company's own guidance. Either the street is modeling margin expansion management isn't guiding for, or the consensus estimate includes adjustments management doesn't. This disconnect matters.

The analyst community is split: 52% bullish (13 of 25 at Buy/Strong Buy), 8% bearish, mean target $10.60 (+40%). Deutsche Bank downgraded from Buy to Hold post-Q4 (2/26). Short interest at 17.7% of float with 4.1 days to cover — the market is genuinely divided, not uniformly bearish.

Running a DCF with management's own guided assumptions (not street consensus):

• Revenue: $1.45B growing to $1.7-1.8B by 2028
• EBITDA margin: 15% (management's stated near-term ceiling)
• CapEx: $25-30M/year
• Interest: ≈$27M/year
• Preferred cash: $31-36M/year (escalating)
• Tax rate: 18-22%
• Terminal: 7x EBITDA
• WACC: 12%

At 7x terminal EBITDA, this framework produces equity value well below current price; at 8-9x terminal, it approaches current levels. The exact output is sensitive to terminal multiple and WACC assumptions, but the directional finding is clear: using management's own guided assumptions (not street consensus), the current price already embeds one or more of: (a) margin expansion beyond mid-teens, (b) revenue acceleration above guidance, (c) capital structure improvement not currently discussed by management. The 21-40% gap between street consensus EPS ($0.91) and management guide ($0.65-$0.75) suggests the street is pricing in upside management hasn't committed to.

Key disconnects:

1. 45X stability. Street models assume 45X manufacturing credits continue at $170M+/year. FEOC restrictions and Treasury rulemaking create real risk to continuation. If 45X is impaired, Array Legacy gross margin drops from 28% to ≈12% and the business model breaks.

2. Preferred cash transition. Starting August 2026, ≈$31M/year in cash goes to Blackstone before common shareholders see anything. Escalating annually. The escalating coupon structure ($30.8M in year 5 rising to $50.5M by year 10) creates a compounding cash drain that increases annually — and zero analyst questions addressed this across the last four quarterly earnings calls.

3. H1/H2 skew risk. Q1 2026 guidance: ≈$200M revenue, sub-$11M EBITDA. The full year guide ($200-230M EBITDA) requires $190-220M of EBITDA in H2 — that's 88-96% of full-year earnings in two quarters. Two things have to go right simultaneously: the July 4, 2026 OBBB ITC construction deadline drives orders, AND those orders convert to revenue in H2. If either slips, the full year misses badly.

4. OBBB demand timing. Management explicitly says "I don't see a whole lot of artificial demand acceleration or pull forward" for 2026. The OBBB-driven construction surge is real (confirmed by FSLR and SHLS), but it shows up more in 2027 revenue than 2026. Some bull theses expecting a 2026 demand surge may be a year early.

5. NXT gap as benchmark. NXT trades at ≈24% EBITDA margins and is widening the gap. If ARRY can't close this margin differential — and management is guiding for "high mid-teens" not 20%+ — the competitive discount should persist or widen.

Key Risks

Balance sheet risks (near-term, high probability):

• Preferred cash dividend transition (Aug 2026) — $31M minimum annual cash drain, escalating, with penalty and forced dilution provisions for non-payment
• 2028 convertible maturity ($325M, conversion price $23.86 = deeply OTM) — needs cash refinancing in an elevated rate environment
• Negative equity deepening — $1.15B in senior claims on $1.45B of assets
• FCF conversion at 43% may be insufficient to service both preferred and fund growth

Competitive risks (medium-term):

• NXT executing same BoS strategy with 3-5x resources and zero leverage
• Core linked-row patent expires Feb 2030 — 3.9 years of remaining moat
• Tariff exposure structurally higher than NXT (fewer domestic manufacturing facilities, lower margins to absorb)

Regulatory/policy risks:

• 45X FEOC restrictions (Treasury guidance still evolving) — $174M/year at risk
• OBBB ITC termination for projects not starting construction by July 4, 2026
• Section 301 investigation into Brazil directly impacts STI Operations
• Domestic content threshold increased under OBBB — guidance not final
• AD/CVD rates on SE Asian solar components (up to 3,400%) could raise module costs for ARRY's customers, slowing project economics

Operational risks:

• H1/H2 revenue skew (40:60) creates two weak quarters before inflection — negative sentiment window
• APA integration risk (1 year internal controls window, potential unknown liabilities, earnout cap at $90M)
• STI transition to DuraTrack — unproven in international markets where H250 failed
• Customer concentration (top 2 = 26% of revenue)

Market positioning risks:

• Short interest 17.7% of float (4.1 days to cover) — creates squeeze potential on any positive catalyst but also signals informed bearish conviction
• P/C ratio 1.91 (bearish), ATM IV 75.4% (21st percentile of 52-week range) — options market pricing downside risk but implied vol is compressed vs history
• RSI 24 (oversold) after -32.5% in one week — near-term sentiment extreme, could produce technical bounce independent of fundamentals
• Max pain at $9.00 (19% above current) — options expiry dynamics could create upward pressure near March 20 expiration

Upside risks (things that could go better than priced):

• IEEPA Supreme Court ruling (Feb 2026) could generate tariff refunds — magnitude uncertain
• APA 2x book-to-bill could accelerate beyond expectations — $150-200M revenue contribution by 2027
• STI drag removal is a genuine tailwind — no more negative-margin quarters from Brazil
• SHLS confirms safe-harbored projects through 2030 — demand visibility extends beyond OBBB cliff fear

What to Watch

August 11, 2026: Preferred dividend transition. The single most important near-term date. Does management pay cash, negotiate with Blackstone, or trigger default provisions? The revolver expansion suggests they plan to pay, but at the cost of increased leverage. Watch for any refinancing or amendment discussions on the Q1 or Q2 2026 earnings calls.

H2 2026 revenue inflection. Management guides 60% of revenue in H2. Q3 2026 earnings (expected Nov 2026) is the first data point on whether the OBBB-driven construction wave converts to tracker revenue. If Q3 revenue exceeds $400M, the full-year guide is achievable. If not, the market re-rates downward.

45X FEOC rulemaking. Treasury proposed rule on material assistance from foreign entities of concern. Comment period through March 30, 2026. Final rule could materially impact Array's $174M annual 45X benefit. Watch for the final rule text and management's assessment of supply chain compliance.

2028 convertible refinancing. $325M due December 1, 2028. At $23.86 conversion price, these are cash obligations. Management needs to refinance at acceptable terms. Watch credit spreads and any 8-K filings on new financing. The expanded 9-bank syndicate suggests relationship capacity exists.

APA integrated product launch. H2 2026 timeline for tracker+foundation integrated solution. Commercial success here validates the BoS strategy and wallet share expansion. Failure or delay undermines the M&A thesis.

Patent landscape post-2030. No urgency yet, but by 2028 this becomes a front-of-mind issue. Watch for: new patent filings, SmartTrack IP disclosures, NXT/competitor product development that approaches linked-row performance, any management commentary on post-2030 differentiation.

NXT quarterly comparison. NXT reports on a March fiscal year (Q4 FY2026 likely May 2026). Track the margin gap, backlog gap, and capital allocation gap. If NXT continues widening lead while ARRY treads water, the competitive discount should compress ARRY's multiple further.

Epistemic State: Doorway

Two patterns genuinely fit:

Pattern A — Value trap (60%): NXT gap widening. Balance sheet constraining strategic optionality. Patent cliff approaching. Margins structurally capped at mid-teens. Preferred eating 30%+ of cash flow. Common shareholders are residual claimants on very little. $7.58 is not cheap enough for these risks.

Pattern B — Recovery (40%): STI drag removed. APA integration creating wallet share expansion. Tier 1 customer mix reducing volatility. Record $2.2B orderbook with 1.5-1.7x coverage. OBBB demand wave confirmed by FSLR and SHLS. If H2 2026 materializes and margins hold, there's $0.65-0.75 of EPS with revenue growing double digits.

Catalyst that resolves: Q2-Q3 2026 earnings. If H2 2026 revenue acceleration materializes AND preferred transition goes smoothly AND APA margins hold, Pattern B gains probability. If H2 disappoints, or NXT takes more share, or 45X gets impaired, Pattern A is confirmed. Size for surviving the wrong interpretation.

Sources

All claims sourced from primary documents:

• ARRY 10-K filed 2026-02-25 (line references available in research notes)
• ARRY 8-K filed 2026-02-18 (Amendment No. 5), 8-K filed 2026-02-25 (Q4 results)
• ARRY DEF 14A filed 2025-04-08
• ARRY earnings transcripts: Q1-Q4 2025
• NXT earnings transcripts: FY2025 Q1-Q4, FY2026 Q1-Q3
• FSLR earnings transcripts: Q2 and Q4 2025
• SHLS earnings transcripts: Q2 and Q3 2025
• BX earnings transcripts: Q2 2024 through Q4 2025 (7 transcripts, zero ARRY mentions)
• Cross-corpus transcript searches: "Array Technologies," "APA Solar," "linked-row," "solar tracker"
• yfinance market data (accessed 2026-02-28)
• Factor regressions via iev regress (1Y daily, multiple model specifications)