Phase A — Understand the business
Lens 1 · Company Overview
Business model (plain terms): Sunrun installs a solar (+ increasingly battery) system on a homeowner's roof, then most often retains ownership and sells the homeowner the electricity under a 20- or 25-year Customer Agreement (lease or PPA), priced below the local utility rate. It is a "solar-as-a-service" subscription business, not an installer. Founded 2007; acquired Vivint Solar in Oct 2020 to become the scale leader.
Two revenue lines:
- Customer agreements and incentives: $1,819.0M (FY25) vs $1,505.2M (FY24), +21%. Within it: Customer agreements $1,708.5M (+23%), Incentives (SREC sales) $110.5M (-5%). This is the recurring subscription spine.
- Energy systems and product sales: $1,138.0M (FY25) vs $532.5M (FY24), +114%. Within it: Energy systems $878.3M (+329%) — driven by a Q3 2025 transaction selling newly-originated Customer Agreements to a third-party investor while Sunrun keeps the servicing relationship; Products (panels/racking to resellers) $259.6M (-21%).
- Total revenue FY2025: $2,956,997K (~$2.96B), +45% YoY.
The "more often the case" tell: Management states most customers buy the service (TPO), not the system. This is the single most important structural fact about RUN — it is why the company can survive the death of the homeowner-facing 25D tax credit (see Lens 10/12).
Main customers: ~1.18M residential households, US-only, >45% of cumulative deployments in California — a geographic concentration that is also a regulatory single-point-of-failure (CA NEM 3.0). customers.csv is empty, so no named-counterparty concentration data in the research layer; the customer base is atomized retail, so concentration risk is geographic/regulatory not counterparty.
Main suppliers: Sunrun is "not a direct importer of modules and batteries," but its suppliers import from China/Vietnam; lithium-ion battery cells are "currently sourced primarily from China". Purchase commitments: $317.6M of batteries by end-Q3 2026 + $1.7B of modules/inverters/batteries across FY2026–FY2029.
Main competitors: Traditional electric utilities (the real competitor — Sunrun prices below retail rates); plus residential-solar peers (Tesla Energy, the now-bankrupt Sunnova, the now-bankrupt/absorbed SunPower).
Contract structure / key payment terms: 20–25yr initial term, auto-renewing annually after, typically at a ~10% discount to then-prevailing utility rates; performance guarantee (refund for underproduction); customers may prepay. Critically, the cash flows are contracted and long-dated but the capital to build them is raised upfront — the structural maturity mismatch that defines the model.
Lens 2 · Supply Chain
Map: upstream cell/module/battery manufacturers (China, Vietnam) → importers/distributors → Sunrun (designer/installer/financier) → homeowner (energy buyer) → tax-equity & debt investors (the real "customer" for the asset's economics).
Named/identified stakeholders along the chain:
- Upstream hardware: Modules, inverters, racking, and lithium-ion battery cells. Sunrun does not name its module OEMs in the ingested filings, but discloses cells are "primarily from China" and that it is actively trying to "identify qualified suppliers outside of China".
supply-chain.md commercial layer is missing (energy KB wiki not yet built), so OEM names are not in the research layer — flagged as a gap.
- Chokepoint #1 — battery cells (China single-source-ish): The explicit China dependency for Li-ion cells is the sharpest physical chokepoint; tariffs flow straight into COGS and "could significantly increase component expenses".
- Sunrun itself: designs, permits, installs (directly or via "energy system partners" and "strategic/retail partners"), interconnects, monitors, maintains, insures.
- Downstream "customer" that actually matters — the capital stack: tax-equity fund investors (monetize ITC + depreciation via HLBV partnership-flip structures) and non-recourse debt lenders + ABS (asset-backed securitization) buyers + third-party asset purchasers (the Q3 2025 system-sale counterparty). This is the binding chokepoint of the whole business: the securitization/tax-equity market is Sunrun's true supplier of oxygen. In FY2025 Sunrun raised $4.11B of non-recourse debt + $2.0B of NCI (tax-equity) contributions + $1.60B of ITC transfer proceeds. If that market closes, growth stops regardless of customer demand.
- Chokepoint #2 — interest rates / advance rates: "elevated interest rates have resulted… in a decrease in our advance rates, reducing the proceeds we receive from certain Funds". Rates are a supply-chain input here as much as silicon is.
Verdict on Lens 2: the physical supply chain (China cells) is a real but manageable tariff risk; the financial supply chain (tax-equity + ABS + advance rates) is the existential one. Names on the physical side are thin because the energy commercial layer isn't built — a research-layer gap to close.
Lens 3 · Competitive Advantages (moats)
Real moats:
- Scale + cost of capital (the dominant moat). Sunrun is the largest residential solar fleet in the US and has "a long track record of attracting low-cost capital from a variety of sources". In a business that is a financing spread, the cheapest cost of capital wins — and scale + a multi-decade securitization track record (investment-grade ABS issuance) is a genuine, compounding advantage over sub-scale installers. The 2025 bankruptcies of Sunnova ($8.9B debt, filed June 9 2025) and SunPower are the moat made visible: when capital tightened, the sub-scale TPO players died and Sunrun absorbed the share.
- Tax-equity origination capacity. Sunrun has "secured expected tax equity capacity to fund about 1,000 MW of new subscriber projects". Few players can structure HLBV partnership flips at this scale — it is a relationship + expertise moat with the bank/insurance tax-equity community.
- Installed-base + upsell optionality. ~1.18M rooftops are a platform for battery retrofits, electrification, and virtual power plant (VPP) / grid-services revenue — Sunrun aggregates distributed batteries and sells grid services to utilities/operators. Storage attachment hit a record 73% in Q1 2026.
- Proprietary siting technology — targets homes with favorable revenue/cost characteristics, "passing through cost savings" to win on price while holding target returns. Modest but real.
Bargaining power:
- Over suppliers: Moderate-to-low. Sunrun is not a direct importer and depends on a China-centric cell supply; tariffs are a price-taker exposure. Scale gives volume leverage but not control of the bottleneck input.
- Over customers: Moderate. Switching costs are high once installed (20–25yr contract, equipment on the roof), but acquisition is fiercely competitive and price-led; customer-acquisition cost is a permanent drag (sales & marketing $709.3M FY25, ~24% of revenue).
- Over capital providers: This is where bargaining power is weakest and most cyclical — when rates rise, advance rates fall and Sunrun takes the terms the tax-equity/ABS market offers.
Honest moat assessment: The moat is real but it is a relative moat (cheapest financier of a commodity service), not an absolute one (no pricing power over the end product, which is electrons priced below the utility). It widens in capital stress (rivals die) and narrows in capital abundance (everyone can fund). It does not protect against the two macro variables that actually move the P&L.
Lens 4 · Segments
Sunrun reports as one reporting unit (it tested goodwill at the enterprise level and wrote it all off — see Lens 10). segments.csv is empty, so the only disaggregation is the revenue-line split (Lens 1) and geography (>45% CA). There is no product-segment EBITDA breakout in the filings — Sunrun runs P&L at the consolidated level. Key trend, by line:
| Revenue line | FY2025 | FY2024 | YoY |
|---|
| Customer agreements | $1,708.5M | $1,388.4M | +23% |
| Incentives (SRECs) | $110.5M | $116.8M | -5% |
| Energy systems | $878.3M | $204.8M | +329% |
| Products | $259.6M | $327.7M | -21% |
| Total | $2,957.0M | $2,037.7M | +45% |
Read: The +45% headline is flattered by the Q3 2025 system-sale-to-investor transaction (Energy systems +329%) — that is a financing/asset-monetization event recognized as revenue, not organic subscription growth. The durable line, Customer agreements, grew a healthier-but-decelerating +23%. Products falling -21% reflects easing supply-chain constraints lowering ASPs. Geography is the real concentration: >45% of cumulative deployments in California means CA NEM 3.0 policy and CA wildfire/utility dynamics are disproportionately load-bearing. Segment-level operating income is n/a — not disclosed (one reporting unit).
Phase B — Measure performance
Lens 5 · Earnings Result (latest print = Q1 2026, reported 2026-05-06)
The headline beat:
- Total revenue $722.2M vs $504.3M, +43% YoY — vs the prior-year quarter, a clean beat.
- Net income attributable to common stockholders $167.6M ($0.71 basic / $0.62 diluted EPS) vs $50.0M ($0.22 / $0.20). The reported EPS "$0.62 vs $0.01 forecast" 6,100% beat figure is real but misleading — see the HLBV caveat below.
- Loss from operations -$43.5M vs -$114.9M — operating loss narrowed sharply (good).
- Interest expense, net -$263.9M vs -$227.4M — still larger than the entire gross profit and the reason the company has a net loss before NCI.
- Net loss (total) -$297.3M vs -$277.2M. The "net income to common" is positive only because net loss attributable to NCI was -$465.0M — i.e. the consolidated tax-equity partners absorbed $465M of loss under HLBV, leaving +$167.6M to common. This is an accounting allocation, not free cash to shareholders.
Operating KPIs (the metrics that actually matter):
- Subscriber Additions: 17,665 vs 23,692 — DOWN 25.4% YoY. Volume is falling.
- Contracted Subscriber Value/sub: $55,464 vs $48,727 — UP 13.8% YoY. Value-per-sub is rising.
- Aggregate Contracted Subscriber Value: $979.8M vs $1,154.4M (-15%) — value-growth did not offset the volume decline this quarter.
- Customers 1,184,634 (+10.3%); Networked Solar Capacity 8,558 MW (+10.8%); Gross Earning Assets ~$21.74B (vs $18.54B).
Guidance/outlook: reaffirmed FY2026 Cash Generation guidance $250M–$450M (ex-equipment-safe-harbor investments) and Aggregate Subscriber Value $4.8B–$5.2B. Storage attachment record 73%. Q1 Cash Generation was -$59M reported (-$31M pro-forma ex-safe-harbor) — management framed it as timing.
Balance-sheet flags: Cash fell from $823.4M (YE25) to $679.6M (Q1'26). Cash paid for interest $210.7M in the quarter. Energy systems net rose to ~$14.06B (VIE-level).
Market reaction: Stock rose ~8% on the Q1 beat. Note the contrast with the prior print: the Q4 2025 report (Feb 26 2026) showed a ~30% decrease in net subscriber value and weak guidance, and the stock fell ~35% the next day. So the tape is whipsawing on subscriber-value disclosure quarter to quarter.
Anomaly vs own history: The pattern is a deliberate strategy pivot — trade volume for value (fewer subs, higher value per sub, record storage attachment). The risk is that if value-per-sub also rolls over (as it did in Q4'25), there is nothing left to offset falling volume.
Lens 6 · Earnings Calls (sentiment trend)
transcripts/ is empty in the research layer, so this lens is ``.
- Tone shift over the last ~4 calls: From a growth-and-volume narrative (2023–24) to a "Cash Generation" + "Net Subscriber Value" + storage-first narrative (2025–26). Management has reframed the entire investor conversation around Cash Generation (a non-GAAP self-funding metric) as the north star — an implicit admission that GAAP and the old "MW deployed" framing don't capture value, and a hedge against the volume deceleration.
- Recurring phrases: "Cash Generation," "storage attachment," "Net/Contracted Subscriber Value," "low-cost capital," "self-funding," "almost entirely subscriptions."
- Things they stopped emphasizing: raw deployment-growth/MW-installed bragging; aggressive customer-add targets.
- The 25D framing (Q1'26 call): CEO Powell explicitly said the December 25D sunset hurt "smaller dealers and some affiliate partners" but that Sunrun's origination is "almost entirely subscriptions," so it is "not seeing similar impacts" — and RBC publicly argued 25D's death pushes customers toward TPO where Sunrun is strong. This is management turning the biggest bear catalyst into a bull narrative.
- Sentiment read: Confident-but-defensive. Confident on relative positioning and storage; defensive on volume and on the constant need to prove self-funding. The reframing to Cash Generation is sophisticated but also a tell — when a company changes its headline metric, scrutinize why.
Lens 7 · Comps
Sunrun is structurally unlike its "peers" — it is a financing/leasing balance sheet (think a specialty finance co with a solar skin), whereas Enphase/SolarEdge are hardware, First Solar is a manufacturer, and Nextracker is utility-scale equipment. Standard EV/EBITDA is nearly meaningless for RUN (no clean EBITDA; the value is in Gross/Net Earning Assets net of the project debt). I report what is sourced and mark the rest honestly.
| Company | Ticker | Market cap | EV/Sales | EV/EBIT | P/E | Div yield | 5-yr avg ROE | Source |
|---|
| Sunrun | RUN | ~$3.0B | n/a | n/m (op. loss) | ~6.5x trailing | 0% | negative | px; mktcap |
| Enphase | ENPH | ~$6.33B | n/a | n/a | fwd ~17x; trailing ~36.7x | 0% | n/a | |
| SolarEdge | SEDG | n/a | n/a | n/a | n/a | n/a | | |
| First Solar | FSLR | n/a | n/a | n/a | n/a | 0% | n/a | FY26 EBITDA guide $2.6–2.8B |
| Nextracker | NXT | n/a | n/a | n/a | n/a | n/a | n/a | |
The right comp frame for RUN is its own book: Gross Earning Assets $21.74B (6% discount) and the implied Net Earning Assets = GEA minus the project debt (~$14.7B total debt incl. non-recourse $13.98B + LOC + converts) ≈ ~$7B of net asset value, against a ~$3.0B market cap. The market is pricing RUN at <0.5x its self-reported net earning assets — either a deep-value signal or the market disbelieving the 6% discount rate / renewal / default assumptions. That gap is the entire RUN debate (see Lens 12/13).
Lens 8 · Stock-Price Catalysts (>5% moves, ~last 5 years)
All ``:
- May 2025 — OBBBA House passage: RUN crashed -40.2% to $6.38 in a day. The single largest catalyst of the period — a tax-policy event, not an earnings event.
- July 4, 2025 — OBBBA signed: 25D residential credit set to die Dec 31 2025.
- Q4 2025 print (Feb 26–27, 2026): -35.1% next day on ~30% net-subscriber-value decline + weak guidance.
- Q1 2026 print (May 6, 2026): +~8% on the revenue beat + record 73% storage attachment + reaffirmed cash-gen guide.
- Dec 2025 — tax-credit-deadline pull-forward + NRG partnership + analyst-target moves: notable up-moves around the 25D deadline demand pull-forward.
- 52-week range $5.38–$22.44 — a ~4x peak-to-trough. What the market reacts to (the pattern): (1) federal tax-policy headlines (the dominant driver — RUN trades as a policy option), (2) subscriber-value disclosure each quarter, (3) interest-rate/macro for the financing model. It reacts far less to raw MW or customer counts. RUN is, functionally, a leveraged call option on residential-solar policy + rates.
Phase C — Judge people & books
Lens 9 · Management
CEO — Mary Powell (CEO since Aug 2021; prior: 21 yrs at Green Mountain Power, the last decade as CEO, where she ran a utility-scale clean-energy + storage transformation).
- Track record: Under Powell, Sunrun added >500K customers, crossed 1M, and pivoted hard to storage-first — batteries installed grew ~800% (≈23K in Q2'21 to ~217K by Q3'25). Named to CNBC Changemakers 2026; Sunrun joined the Fortune 1000 in 2026 on the 45% revenue jump. She is a credible operator who correctly read that batteries (not panels) are the value driver post-NEM-3.0.
- Tenure & skin in the game: ~5 years tenure. Insider ownership is not in the research layer (
insider-transactions.csv absent); a December 2025 insider sale was noted in passing but quantum/holdings are n/a. Founder Lynn Jurich (co-founder, former CEO) stepped back to chair-type roles; this is a professional-manager-led company now, not founder-run.
- Capital allocation: The defining act of this era is the strategic reframe to Cash Generation / self-funding and the willingness to sell originated Customer Agreements to third parties (Q3 2025) to pull cash forward and de-lever the growth — pragmatic, arguably the right move given the rate environment. The flip side: ROE/ROIC on their watch is negative on a GAAP basis (the goodwill writedown alone was $3.1B), and the company has never generated positive operating cash flow at scale (-$421M FY25). Capital allocation is really capital sourcing here — and they have done that well (kept the ABS/tax-equity taps open while rivals went bankrupt).
- Red flags: A December 2025 insider sale into the tax-deadline strength; the repeated non-GAAP metric reframing (Cash Generation, Net Subscriber Value) which can flatter the narrative; HLBV allocation that produces a GAAP "profit" with no cash behind it (not management's invention — it's required — but they lean on the resulting EPS in headlines).
- Archetype: Professional operator/turnaround-and-scale manager, utility pedigree, well-suited to a capital-intensive, policy-exposed, storage-led stage. Not a visionary founder; an executor managing a balance sheet through a policy storm.
Lens 10 · Forensic Red Flags
Forensic equity-analyst lens. Every figure research-layer-sourced from the filings unless marked.
The headline forensic issue: GAAP net income to common is an HLBV artifact, not cash.
- FY2025 net loss (total) was -$1,009.1M, but net loss attributable to NCI/redeemable-NCI was -$1,459.1M, leaving +$449.9M "net income to common" ($1.96 basic EPS). The positive number exists because consolidated tax-equity partners absorb the bulk of losses under the HLBV (hypothetical-liquidation-at-book-value) method. This is the #1 thing a naive screen gets wrong on RUN — a "P/E ~6.5x" is meaningless. Same mechanic in Q1'26 (+$167.6M to common off a -$297.3M total net loss).
- Cash flow diverges hard from "earnings": Operating cash flow was -$421.4M FY2025 (and -$766.2M FY2024). A company reporting ~$450M "net income to common" that burns $421M from operations is the canonical earnings-vs-cash divergence. Cash paid for interest was $741.7M FY2025.
Balance-sheet / leverage flags:
- Total debt ≈ $14.69B: non-recourse $13,708.5M (+$269.5M current) + line of credit $238.3M + convertible senior notes $473.7M. Against total equity $4,274.7M and total stockholders' equity $3,132.5M. The non-recourse debt is project-level (no recourse to Sunrun parent) — a genuine mitigant — but it must still be refinanced/rolled, and the $2.7B recourse-style syndicated Credit Facility (post-Feb 2026 amendment, maturity extended to Feb 2030, now also secured by ITC-sale proceeds) is the near-term liquidity backbone.
- Energy systems, net $16,817.9M is the dominant asset; its carrying value depends on long-dated cash-flow and renewal assumptions. Long-lived assets were tested — no impairment in 2025 (vs the $3.1B goodwill writeoff taken Q4 2024 when market cap fell below book). The fact that the entire goodwill (Vivint legacy) is already gone removes one future landmine but is itself a flag about the 2020 acquisition's value.
- Receivables outrunning revenue: AR rose to $262.6M (from $170.7M), +54%, vs revenue +45% — accounts-receivable change was a -$119.8M operating-cash drag FY2025. Modest but worth watching.
- Prepaid/other current assets swung -$613.7M in operating cash (FY25) — large, tied to safe-harbor equipment prepayments ahead of the tax-credit deadline.
Revenue-recognition nuance: The Q3 2025 sale of newly-originated Customer Agreements to a third-party investor is recognized over time on milestones and drove the Energy-systems +329% — investors should not read that as organic subscription acceleration; it is asset monetization.
SBC: $108.0M FY2025 (modest, ~3.7% of revenue) — not an egregious non-GAAP flatterer by tech standards.
Tax: Pays ~zero cash income tax — NOLs of $720.7M federal + $2.6B indefinite-life federal, $3.5B state, $1.3B foreign. ITC transfers gave a $196.6M income-tax benefit FY2025. ITC indemnification obligations to fund investors exist (insured via policies since 2018).
Internal controls: Unqualified opinion from Ernst & Young LLP on both the FY2025 financials and ICFR (no material weakness). Disclosure controls deemed effective by CEO/CFO.
Regulatory findings (required sub-section):
- SEC Litigation Releases: None. "No LR found for this company" 2021-06-18→2026-06-18.
- SEC AAERs: None.
- 10-K Item 3 (Legal Proceedings): Item 3 points to Note 17, Commitments and Contingencies. Note 17 discloses letters of credit, lease commitments, the $317.6M + $1.7B purchase commitments, warranty accrual (immaterial), and ITC indemnification obligations — but no specific named material litigation is broken out; the company's standard position is that ordinary-course claims are not expected to be materially adverse.
- Non-SEC / web enforcement check: No FTC/DOJ/CFPB enforcement action surfaced. Material legal items from web: (1) a securities class-action investigation opened by plaintiffs' firms (e.g., Schall Law) in March 2026 tied to the Feb 26 2026 Q4 print's ~30% net-subscriber-value drop and guidance — an investigation/announcement, not a filed-and-certified action with a finding; (2) a historical $5.5M TCPA robocall settlement (Do-Not-Call); (3) older 2016–2018-era securities suits re: cancellation-rate disclosure. Residential-solar sales-practice consumer complaints are an industry-wide reputational risk (door-to-door channels) but no current government enforcement against Sunrun specifically was found.
- Summary: No SEC enforcement (LR/AAER) and no material litigation disclosed in 10-K Item 3/Note 17, verified via SEC EDGAR EFTS + 10-K + web as of 2026-06-18. The live watch-item is the March 2026 securities-class-action investigation stemming from the Q4 guidance miss — monitor for an actual filed complaint and lead-plaintiff appointment.
Phase D — Project & stress-test
Lens 11 · Forward Projection
Why EPS is the wrong target for RUN (and what I project instead): GAAP EPS-to-common is an HLBV allocation, not economics. The metric management and the market now use — and the right one to project — is Cash Generation and Aggregate Subscriber Value (volume × value/sub). I anchor on the latest actuals + guidance and label every input.
Anchors:
- FY2025 record Cash Generation $377M (vs negative operating cash flow — Cash Generation is a non-GAAP self-funding metric net of project finance).
- FY2026 guide: Cash Generation $250M–$450M; Aggregate Subscriber Value $4.8B–$5.2B.
- Q1'26 run-rate: Subscriber Additions 17,665 (-25% YoY) but Value/sub $55,464 (+14% YoY); storage attach 73%.
Three-path projection (Cash Generation, ex-safe-harbor), FY2026 → FY2028:
| Path | FY2026 | FY2027 | FY2028 | Key assumptions |
|---|
| Bear | ~$250M | ~$150M | ~$50M | Volume keeps falling ~15%/yr as 25D pull-forward unwinds and CA NEM-3.0 bites; value/sub plateaus; advance rates stay pressured; ABS spreads widen. |
| Base | ~$350M | ~$400M | ~$475M | Volume troughs in 2026 then stabilizes; value/sub +5–8%/yr on storage mix + 48E TPO share gains from dead competitors; rates ease modestly lifting advance rates. |
| Bull | ~$450M | ~$600M | ~$800M | 25D-death drives a TPO share landgrab; storage attach >75%; VPP/grid-services revenue inflects; rate cuts re-open cheap ABS; consolidation gives pricing power. |
Net-asset cross-check (the value the equity is really worth): Net Earning Assets ≈ Gross Earning Assets $21.74B (6% discount) − total debt ~$14.7B ≈ ~$7.0B, vs ~$3.0B market cap. The base case is that the market re-rates toward a fraction of disclosed NAV as the volume trough passes and Cash Generation proves the model self-funds. The bear case is that the 6% discount rate is too low and the "real" NAV (at 8% + lower renewal) is far thinner — the sensitivity table shows GEA dropping to $16.86B at 8% / 80% renewal vs $21.14B base (YE25), i.e. ~$2B of GEA evaporates per 100bps of discount-rate / renewal-haircut.
Per the SKILL's --watchlist rule, I did NOT run forecast.ts create (no Brier forecast logged in unattended breadth mode). If promoted to a tracked call, the natural binary to log is "RUN FY2026 Cash Generation ≥ $350M (midpoint)" rather than an EPS line.
Lens 12 · Bull vs Bear
Bull case. Sunrun is the last scaled financier standing in US residential solar, and the OBBBA just killed its sub-scale competitors and the financing structure (25D) those competitors relied on — while sparing the 48E commercial ITC + depreciation that Sunrun's TPO model monetizes. So as the absolute market shrinks, Sunrun's share of the durable, financeable slice grows. Storage-first (73% attach) turns every rooftop into a higher-value, grid-services-capable asset (VPP optionality is a free call). The company self-funds (record $377M Cash Generation FY25) and trades at <0.5x its own disclosed net earning assets (~$7B NAV vs ~$3B cap). If rates ease, advance rates and ABS spreads improve and the whole spread business re-rates. The contrarian read: the market is pricing RUN as a casualty of OBBBA when it is actually a relative beneficiary.
Bear case (2–3 permanent-impairment risks).
- The financing treadmill breaks. RUN must roll ~$14.7B of debt and raise ~$4B+ of new non-recourse debt + ~$2B tax-equity every year to grow. A securitization-market freeze, a sustained high-rate regime, or a tax-equity-investor pullback (post-OBBBA the tax-equity pool itself is shrinking) would force either a growth halt or dilutive equity at a ~$3B cap — permanent shareholder impairment.
- The NAV is an assumption, not a fact. GEA uses a 6% discount and 90% renewal over an assumed 30-yr relationship. If renewals disappoint, defaults rise, or the right discount rate is 8%+, several billion of "net earning assets" never materialize — and the equity (the thin slice above $14.7B of debt) is the first loss.
- Volume is already rolling over (-25% subs YoY in Q1'26) and CA (>45% of base) is structurally worse under NEM 3.0. If value/sub also plateaus (as in Q4'25), the trade-volume-for-value strategy runs out of room.
Pre-mortem (18 months out, thesis broke): Rates stayed higher-for-longer, ABS spreads widened, advance rates fell further; a key tax-equity partner stepped back as the post-OBBBA credit pool thinned; Cash Generation missed the low end; Sunrun raised dilutive equity or sold assets at distressed marks; the Q4'25-style subscriber-value declines recurred; the securities investigation became a filed action. The stock retested the $5–6 lows.
Are multiples too high? No — on disclosed NAV, RUN is cheap (<0.5x NEA). The bear case is not "overvalued," it's "the NAV and the financing access are fragile." This is a balance-sheet/solvency-cyclicality story, not an expensive-multiple story.
Contrarian view (what the market refuses to see): The market is treating 25D's repeal as uniformly bearish for "solar," lumping RUN with loan/cash players. But 25D's death is asymmetrically bullish for the TPO leader: it removes the cheapest competitor financing and channels demand to leases/PPAs where Sunrun dominates and still gets 48E + depreciation. The contrarian bet is that RUN emerges from the 2026 trough with more share of a smaller but more durable market — if, and only if, the capital markets stay open to it.
Lens 13 · Devil's Advocate (short-seller)
Dismantling the bull case — where this actually breaks:
- It is a finance company pretending to be a growth stock. Strip the HLBV optics and you have a sub-scale-equity, ~$14.7B-levered specialty lender that has never produced positive operating cash flow (-$421M FY25, -$766M FY24, -$821M FY23) and survives only by continuously selling forward its own future cash flows. "Cash Generation" is a constructed metric net of new financing — it can be positive while the GAAP cash account and the equity erode.
- Revenue concentration that bulls underestimate: Not customer concentration — capital-provider and geographic concentration. >45% in California (one PUC's NEM decisions away from a demand shock), and a fundamentally concentrated dependency on a handful of tax-equity sponsors and ABS desks whose appetite is policy- and rate-dependent. Post-OBBBA, the tax-equity pool itself is contracting — RUN is fishing in a shrinking pond.
- The moat is thinner than claimed: "Lowest cost of capital" is a moat only while the capital markets are open and rational. In 2023–25 that moat inverted — RUN's own cost of capital spiked, advance rates fell, and the stock went from >$22 to <$6. A moat that disappears exactly when you need it most is a beta, not a moat.
- Most dangerous competitor bulls underestimate: Not another installer — it's the utilities + regulators (NEM reform that erodes the export-credit value proposition state by state: AZ, NV, HI, IL, and now CA already moved) and the customer's own option to just… buy a battery + panels outright as hardware costs fall (Products ASPs already -21%). And cheap Chinese hardware + a future loan product could re-democratize ownership and undercut TPO.
- Worst capital-allocation / accounting concerns: The $3.1B Vivint goodwill writeoff (2020 deal value destroyed); leaning on HLBV-derived EPS in headlines; the optical "+45% revenue" boosted by a one-off asset sale to an investor; a December 2025 insider sale into deadline strength. None are fraud — EY signs clean — but they are a pattern of narrative management around a structurally cash-negative model.
- What must hold for today's ~$12.72 price: Continuous open access to ~$6B/yr of non-recourse + tax-equity + ITC-transfer capital at advance rates that preserve a positive spread; renewals ≈90% and a defensible ≤6% discount rate; no recurrence of Q4'25-style subscriber-value declines; no dilutive equity raise.
- If growth disappoints 20–30%: Subscriber Additions are already -25% YoY. Push value/sub flat-to-down (the Q4'25 scenario) and Cash Generation misses the $250M low end → the self-funding story breaks → forced dilution at a depressed cap → the equity (the ~$7B NAV slice above ~$14.7B debt, but only if the NAV assumptions hold) compresses violently.
- Single scenario that permanently impairs: A 2008-style securitization/tax-equity freeze coinciding with higher-for-longer rates. RUN can't roll its debt or fund new builds, advance rates collapse, the NAV is marked down, and it must recapitalize at the bottom. Plausibility: low-to-moderate, but it is the whole risk — everything else is noise around the capital-markets-access question.
Lens 14 · Management Questions (ordered by information value)
- Walk through the FY2026 Cash Generation bridge in detail — exactly how much is recurring self-funding vs one-off asset sales / safe-harbor timing / ITC-transfer pull-forward? What is the steady-state, ex-one-off Cash Generation?
- What is your committed, undrawn tax-equity + non-recourse capacity for the next 12 and 24 months, and at what advance rates vs a year ago — and how has the OBBBA-shrunken tax-equity pool changed sponsor appetite and pricing?
- Under what scenario do you raise equity, and at what stock price would you refuse to? How do you think about dilution at the current ~$3B cap?
- Subscriber Additions fell 25% YoY in Q1'26. Where does volume trough, and what is the floor for Aggregate Subscriber Value if value/sub plateaus the way it did in Q4'25?
- Defend the 6% GEA discount rate and 90% renewal assumption against a skeptic who'd use 8% and 75% — what's your evidence on actual realized renewals to date?
- 25D died Dec 31 2025. Quantify the net effect on Sunrun specifically: lost cash/loan volume vs TPO share gains as competitors and loan products exit. Is it net positive or negative to 2026–27 Cash Generation?
-
45% of deployments are in California. What is your concrete plan to de-risk CA NEM 3.0 exposure, and what is the CA cohort's current unit economics vs the rest of the fleet?
- What is the realistic revenue and margin ramp for VPP / grid-services from the ~217K-battery fleet — when does it become a material P&L line, not a talking point?
- China sources "primarily" your battery cells. What's the timeline and cost to diversify, and what's the gross-margin hit per incremental tariff tier?
- The $2.7B syndicated Credit Facility matures Feb 2030 and is now secured by ITC-sale proceeds. What are the covenants, and how much headroom do you have under stress?
- How do you respond to the criticism that HLBV-derived "net income to common" overstates economic earnings — and would you consider leading with cash metrics only?
- What's your capital-allocation priority order today: de-lever, fund growth, buy back the convertible notes, or build VPP capability?
- What is current insider ownership, and how should investors read the December 2025 insider sale?
- What did you learn from the Vivint acquisition that wrote off $3.1B of goodwill, and what does it imply about future M&A in a consolidating market?
- If the securitization market froze for two quarters, what is your survival playbook, and how long can the business run without new external capital?