Snapshot
The US installed a record 18.9 GW / 51 GWh of battery energy storage in 2025, up 52% from 37.1 GWh in 2024 (Wood Mackenzie). Globally, BloombergNEF forecasts 92 GW / 247 GWh of new storage in 2025 and 123 GW / 360 GWh in 2026. The five tickers grouped here span three distinct businesses: utility-scale battery system integration (FLNC), residential solar-attached batteries (ENPH, SEDG), AI-software optimization of storage assets (STEM), and solar module manufacturing with no current storage product (FSLR). Of the five, only Fluence (FLNC) is a pure-play storage company; the others derive most or all revenue from solar equipment.
51 GWh
US storage installed in 2025 (record, +52% YoY)
247 GWh
Global storage forecast 2025 est.
$70/kWh
Stationary storage pack price, 2025 (BNEF, down 45% YoY)
152 GW
US utility-scale storage pipeline (project databases, Q4 2025)
500 GWh
US additions forecast 2026-2031 est.
$52.9B
Combined mkt cap, all 5 tickers (Jun 2, 2026)
Battery storage cell prices collapsed 45% in one year to $70/kWh for stationary packs (BNEF, Dec 2025), driven by Chinese LFP oversupply. That makes batteries cheaper to deploy but crushes margins for anyone who resells cells rather than makes them. The US installed 51 GWh in 2025 and has a 152 GW pipeline in interconnection queues — but Wood Mackenzie forecasts a near-term dip (2026-2027) before double-digit growth resumes through decade-end, partly because the Section 25D residential tax credit expired and tariff policy is unstable.
The product & how money is made
Battery energy storage systems (BESS) store electricity in lithium-ion cells (overwhelmingly lithium iron phosphate / LFP chemistry for stationary use) and discharge it later. The product is measured in GWh of storage capacity. A 4-hour / 100 MW system stores 400 MWh. Money is made in four different ways across these five companies:
- System integration (FLNC): Fluence designs and assembles utility-scale BESS (100 MWh to 1+ GWh) by purchasing cells from third-party manufacturers (currently Envision AESC for US-made cells), adding its own battery management software, inverters, and thermal management, then selling turnkey systems to utilities and developers. Revenue is hardware-heavy, recognized on delivery/commissioning. Fluence also sells Mosaic — a software platform for asset optimization — on recurring subscriptions (ARR: $148M at FY2025-end, targeting $180M by FY2026-end).
- Residential battery attach (ENPH, SEDG): Enphase sells IQ Batteries (5 kWh modular units) as add-ons to its microinverter solar systems. SolarEdge sells batteries paired with its power optimizers and inverters. Both target homeowners. Revenue is hardware — boxes shipped through installers. Enphase shipped 706.1 MWh of IQ Batteries in FY2025 (+36% YoY from 521 MWh). SolarEdge recognized 928 MWh of batteries in FY2025.
- Software & services (STEM): Stem’s Athena AI platform optimizes when storage assets charge and discharge to maximize revenue from energy arbitrage, demand-charge reduction, and grid services. Revenue comes from software subscriptions (ARR: $61.1M), edge hardware sales, and battery hardware resale (winding down). Storage assets under management: 1.7 GWh. Separately manages 36.1 GW of solar monitoring.
- Solar modules, no storage (FSLR): First Solar manufactures thin-film CdTe (cadmium telluride) solar panels. It has no battery or storage product line. Its inclusion in this group is a categorization artifact — FSLR is a solar manufacturer, not an energy storage company. It has a 47.9 GW contracted backlog of solar modules.
Sources: FLNC FY2025 earnings release (Nov 2025); ENPH Q1-2026 8-K (Apr 2026); SEDG FY2025 earnings release (Feb 2026); STEM FY2025 earnings release (Feb 2026); FSLR Q1-2026 earnings release (Apr 2026).
Demand
Contracted / already committed
- Fluence backlog: $5.3B as of Sep 30, 2025 (up from $4.5B a year earlier). Management says ~85% of FY2026 revenue guidance midpoint ($3.4B) is covered by this backlog. FY2025 contracted 3.4 GW of energy storage products; pipeline stands at 122 GWh (+52% YoY). contracted
- First Solar backlog: 47.9 GW of solar modules contracted as of Mar 31, 2026. No storage backlog — FSLR does not sell batteries. contracted
- Stem CARR: $67.2M in contracted annual recurring revenue. Backlog of $21.3M in project-based work. contracted
- ENPH / SEDG: Neither company reports a forward storage backlog. Demand is real-time pull from installer channels, heavily influenced by residential solar policy (tax credits, net metering rules, electricity prices).
Forecast demand (not contracted)
- US market: Wood Mackenzie projects 500 GWh of US storage additions from 2026 to 2031. Near-term, 2026 deployments are expected to dip: utility-scale down ~11% YoY, residential down ~2%, before rebounding to double-digit growth by 2028. est.
- Global market: BNEF forecasts 123 GW / 360 GWh of global installations in 2026, growing at ~23% CAGR through 2035 to a cumulative 2 TW / 7.3 TWh installed. MarketsandMarkets sizes the global BESS market at $50.8B in 2025, growing to $106B by 2030 (15.8% CAGR). est.
- Residential (US): The Section 25D federal residential clean energy tax credit expired at end of 2025, removing a key demand driver for ENPH and SEDG battery sales. Enphase’s Q1-2026 IQ Battery shipments fell to 103.1 MWh from 150.1 MWh in Q4-2025 (a quarter that was front-loaded by the expiration rush). The ITC under Section 48 still covers systems with solar, but the standalone residential credit is gone.
- AI / data center demand: Batteries cannot provide 24/7 baseload power (they complement, not replace, gas and nuclear for data centers). AI-driven demand for storage is indirect — grid stability, peak shaving, and renewable firming — not direct procurement by hyperscalers.
Supply
Capacity
- Battery cells: Global lithium-ion cell manufacturing capacity is overwhelmingly Chinese. CATL and BYD together hold 55%+ of global battery cell shipments est.. LFP cell prices have crashed — stationary-storage pack prices fell 45% YoY to $70/kWh (BNEF, Dec 2025). There is no cell shortage, and prices are falling. The US has minimal cell manufacturing (a few plants from LG, Samsung SDI, Envision AESC, Panasonic — almost all focused on EV rather than stationary storage).
- System integration capacity: Fluence has deployed 6.8 GW / 17.8 GWh cumulative through FY2025-end. It sources non-Chinese cells from Envision AESC’s Tennessee plant for US domestic content requirements. Global system integrators also include Tesla (Megapack), BYD, Sungrow, Wärtsilä, and others.
- Residential batteries: Enphase manufactures IQ Batteries at its Texas facility (49.5 MWh shipped from Texas in Q1-2026 alone). SolarEdge sources batteries from third parties. Residential storage supply is not constrained — the bottleneck is demand (end-customer economics, policy).
- Software (STEM): Software scales with near-zero marginal cost. Stem’s 1.7 GWh of storage AUM and 36.1 GW of solar AUM are limited by sales, not capacity to serve.
Bottlenecks
- Grid interconnection: 530 GW of battery storage sits in US interconnection queues (Q4 2025), but interconnection timelines run 3-5 years. This queue is the binding constraint on how fast utility-scale storage gets built — not cell supply, not system integrator capacity.
- Trade policy: US tariffs on Chinese batteries and components create a cost gap between US-sourced and Chinese-sourced cells. Fluence’s pivot to Envision AESC’s Tennessee plant insulates it from tariff exposure but raises costs vs. Chinese competitors. The “FEOC” (Foreign Entity of Concern) rules under the IRA determine which cells qualify for US tax credits; guidance changes create uncertainty for the entire supply chain.
- Tax credit policy: The IRA’s Section 45X manufacturing credits ($35/kWh for US-made battery cells) and ITC for storage projects are key economics drivers. FSLR’s FY2026 guidance assumes $2.1-2.2B of Section 45X credits. Wood Mackenzie’s forecast range includes 52 GW of variance through 2031 depending on trade policy and tax credit stability. est.
Sources: BNEF lithium-ion battery price survey (Dec 2025); Wood Mackenzie US Energy Storage Monitor Q4-2025 and Q1-2026; FLNC FY2025 10-K; ENPH Q1-2026 8-K; FSLR Q1-2026 earnings release.
The gap
| Measure | Demand side | Supply side | Direction |
|---|
| US annual installs (GWh) | 51 GWh (2025 actual) | 530 GW in interconnection queues; cell supply unconstrained | Growing but 2026-27 dip expected |
| Global annual installs | 247 GWh (2025 est.), 360 GWh (2026 est.) | Chinese cell oversupply; >1 TWh/yr mfg capacity est. | Surplus — prices falling |
| Battery cell prices | $70/kWh stationary (2025) | Down 45% YoY | Falling; BNEF expects further decline in 2026 |
| US residential storage | 2.7 GW installed 2025 (+92% YoY, tax-credit pull-forward) | Manufacturing unconstrained | Contracting ~2% in 2026 (25D credit expired) est. |
| Utility-scale pipeline | 152 GW in project databases (US) | Interconnection queue: 530 GW, 3-5 yr wait | Demand enormous; interconnection is the bottleneck |
Cell-level prices are in structural decline from Chinese overcapacity. For system integrators like Fluence, cheaper inputs help margins if contract prices hold, but competition from Chinese integrators (BYD, Sungrow, CATL) compresses the system-level markup. For residential battery sellers (ENPH, SEDG), falling cell costs help unit economics but the demand driver (homeowner payback period) depends on electricity rates and policy, not just hardware cost. Supply surplus at the cell level, demand growth at the system level, with interconnection as the physical bottleneck.
The players
| Ticker | Business | Mkt Cap | FY2025 Rev | Net Inc | Cash | Debt | Storage exposure |
|---|
| FSLR | Solar module mfg (CdTe thin-film) | $33.4B | $5.22B | $1.53B | $2.8B | $0.5B | None. Zero storage product or revenue. 47.9 GW solar backlog. |
| ENPH | Microinverters + residential batteries | $9.5B | $1.47B | ~$174M | $0.9B | $0.6B | 706 MWh batteries shipped FY2025 (+36%). Storage is ~15-20% of revenue est. (not separately disclosed). |
| FLNC | Utility-scale BESS integration + software | $5.1B | $2.26B | ($68M) | $0.7B | $0.4B | Pure-play. 17.8 GWh deployed cumulative, 6.8 GW. $5.3B backlog. 122 GWh pipeline. |
| SEDG | Solar inverters + optimizers + batteries | $4.8B | $1.18B | ($405M) | $0.5B | $0.3B | 928 MWh batteries recognized FY2025. Storage is a minority of revenue (not separately broken out). |
| STEM | AI software for storage optimization | $0.09B | $156M | $138M* | $49M | $356M | 1.7 GWh storage AUM. Software-only model. *Net income includes $220M gain on debt extinguishment. |
Sources: market data Jun 2, 2026 (stockanalysis.com, companiesmarketcap.com); FSLR FY2025 10-K and Q1-2026 8-K; ENPH FY2025 10-K and Q1-2026 8-K; FLNC FY2025 earnings release (Nov 2025); SEDG FY2025 earnings release (Feb 2026); STEM FY2025 earnings release (Feb 2026).
Key distinctions
- FSLR is miscategorized here. It is a solar module manufacturer with $2.8B cash, $9.5B equity, 40%+ gross margins boosted by ~$2.1B/yr of Section 45X manufacturing tax credits, and no energy storage business whatsoever. Its 47.9 GW backlog is entirely solar panels.
- FLNC is the only pure-play storage company. Its $2.3B FY2025 revenue is almost entirely BESS hardware. Margins are thin (13% gross, EBITDA at $19.5M) because the business is project-based system integration with commodity cell inputs. The software/services ARR ($148M) is higher-margin but still small relative to hardware.
- ENPH has a storage business growing inside a microinverter company. Battery shipments grew 36% in FY2025 and the company manufactures in Texas (domestic content eligible). Batteries are an accessory to the solar install, not the primary product.
- SEDG is in recovery. After a $1.81B net loss in FY2024 (inventory write-downs, restructuring), FY2025 narrowed the loss to $405M with gross margins recovering from negative 97% to positive 17%. Battery shipments (928 MWh) were material but the company’s trajectory depends on its core inverter/optimizer business stabilizing.
- STEM is a micro-cap ($90M) with more debt ($356M) than market cap. FY2025’s positive net income ($138M) is entirely from a one-time $220M gain on debt extinguishment. The operating business generates ~$7M adjusted EBITDA on $156M revenue. Its 1.7 GWh of storage AUM is down from 1.8 GWh a year ago.
The price of exposure
| Ticker | Price | Mkt Cap | EV (approx) | Trailing P/E | EV / FY2025 Rev | Net Cash (Debt) |
|---|
| FSLR | $311.01 | $33.4B | $31.1B | 21.9x | 6.0x | +$2.3B net cash |
| ENPH | $72.33 | $9.5B | $9.2B | ~55x | 6.3x | +$0.4B net cash |
| FLNC | $27.91 | $5.1B | $4.8B | N/A (loss) | 2.1x | +$0.3B net cash |
| SEDG | $78.51 | $4.8B | $4.6B | N/A (loss) | 3.9x | +$0.2B net cash |
| STEM | $10.02 | $0.09B | $0.4B | N/A* | 2.6x | ($0.3B) net debt |
*STEM’s FY2025 “profit” was a one-time debt extinguishment gain. On an operating basis, the business is roughly breakeven.
- FSLR trades at 21.9x trailing earnings — but ~$2.1B/yr of that income comes from Section 45X manufacturing tax credits. If those credits disappeared, trailing earnings would drop to roughly $400-600M pre-tax est., making the effective P/E far higher. The credits are law through at least 2032.
- FLNC trades at 2.1x trailing revenue on $2.3B of sales and a $5.3B backlog. Management guides $3.2-3.6B FY2026 revenue with $40-60M adjusted EBITDA. At the EBITDA midpoint, EV/EBITDA is ~96x.
- ENPH at 55x trailing earnings reflects a business that peaked at $2.33B revenue (FY2022) and has been declining since. Q1-2026 revenue ($283M) was down 21% YoY. The 25D credit expiration is a headwind.
- SEDG is priced at $4.8B despite consecutive years of losses ($1.8B + $0.4B). The stock is up 288% in the past year from deeply depressed levels. Gross margins are improving (22% in Q4-2025) but the company still loses money.
- STEM’s $90M market cap sits below its $356M debt. Enterprise value is ~$400M on $156M revenue. FY2026 guidance is $140-190M revenue and $10-15M adjusted EBITDA.
Sources: market data Jun 2, 2026; company filings as cited above. EV calculated as mkt cap minus cash plus debt (simplified).
What to deep-dive next
- FSLR — a deep dive would be a solar module manufacturing analysis, not a battery/storage analysis. Key question: how much of the business depends on Section 45X credits and tariff protection vs. intrinsic cost competitiveness against Chinese thin-film. 47.9 GW backlog and $2.8B cash provide visibility.
- FLNC — the only pure-play worth analyzing for storage specifically. Key variables: can Fluence convert its $5.3B backlog at profitable margins as cell prices drop, or does competition from Tesla Megapack and Chinese integrators compress margins to zero? Software ARR growth trajectory ($148M to $180M guided) is the higher-quality earnings stream.
- ENPH — whether US residential battery demand recovers after 25D expiration and whether the IQ Battery can become a standalone product (grid services, virtual power plants) rather than a solar accessory.
- SEDG — can gross margins reach 25%+ sustainably, and does the balance sheet ($244M net cash) provide enough runway.
- STEM — whether $356M of debt on a $90M market cap and shrinking storage AUM is terminal or whether the software pivot generates enough cash to service it.
Sources & confidence
- FSLR: FY2025 10-K (filed Feb 2026); Q1-2026 earnings release (Apr 30, 2026). High confidence — directly from SEC filings.
- ENPH: Q1-2026 8-K (Apr 28, 2026); FY2025 annual revenue and battery shipments from 10-K summary. High confidence — SEC filings.
- FLNC: FY2025 earnings release (Nov 25, 2025, fiscal year ends Sep 30). Backlog, deployed GWh, pipeline from same release. High confidence.
- SEDG: FY2025 earnings release (Feb 2026). Battery MWh and margin recovery from same. High confidence.
- STEM: FY2025 earnings release (Feb 2026). Net income includes $220M one-time gain; operating metrics from same filing. High confidence on filings; going-concern risk is real given debt/cap structure.
- US market data: Wood Mackenzie US Energy Storage Monitor (Q4-2025 and Q1-2026 editions, via Energy Storage News and ESS News). est.
- Global forecasts: BloombergNEF (via Energy Storage News, May 2026). est.
- Battery prices: BloombergNEF lithium-ion battery price survey (Dec 2025).
- Market caps / prices: stockanalysis.com, companiesmarketcap.com, Jun 2, 2026.