Explore other Clean Energy & Nuclear stocks
Other companies we cover in this category.
What does it do?
BE is a hypergrowth business that earns money by selling and servicing solid oxide fuel cells that convert fuels like natural gas or hydrogen into electricity without combustion. Customers buy the Bloom Energy Server, a modular power platform, to generate reliable electricity on-site rather than relying on the traditional utility grid. Money flows through three primary channels: the upfront sale of the physical server stacks, multi-year service agreements to maintain those stacks, and occasionally through "Power Purchase Agreements" where Bloom sells the electricity itself. Customers keep paying because Bloom's systems are "always-on," providing higher reliability than a grid prone to blackouts while often delivering lower carbon emissions.
Where does revenue come from?
The vast majority of revenue comes from physical product sales, which accounted for 88% of total revenue last quarter. The remainder is split between long-term service contracts (7%), installation services (5%), and direct electricity sales (1%). While product sales drive the initial surge, the service segment creates a growing tail of recurring revenue as the total base of installed systems expands worldwide.
Revenue Breakdown
Who are its customers?
BE serves a dominant cluster of global technology giants, hyperscalers, and mission-critical industrial facilities including over a dozen AI data center operators. Following the 2.8 gigawatt partnership with Oracle and a multi-gigawatt deal with Brookfield Asset Management, its customer base has shifted toward large-scale infrastructure projects. The company recently surpassed $1.06 billion in quarterly revenue, driven by a 215% surge in product volume. Key historical clients include Equinix, which recently expanded its contract to exceed 100 megawatts, as well as hospitals, semiconductor manufacturers, and retailers like Walmart and The Home Depot who require "always-on" power.
What gives it staying power?
Bloom's staying power comes from the high switching costs and grid-independence its technology provides to data centers. Once a data center is built around a Bloom microgrid, switching to a different power source requires massive infrastructure overhauls and utility wait times that can last five years or more.
Where is it headed?
The company is making a massive strategic bet on becoming the primary power architect for the AI era. Management is doubling manufacturing capacity to 2 gigawatts by the end of 2026 to fulfill its multi-gigawatt backlog. If successful, Bloom will move from being a "backup power" option to the standard primary power source for massive AI "compute factories" globally.
Revenue reached a record $1.07 billion last quarter, a 166% jump that signals a fundamental shift in the company's scale. This acceleration is driven by massive data center orders that have more than doubled product revenue year over year. The business is no longer just growing; it is inflecting into a high-volume utility provider.
Operating cash flow turned positive at $226 million this quarter, a major reversal from the $213 million burn seen a year ago. This confirms that Bloom is finally generating cash from its core business as it scales production. While capital spending remains high to build new factories, the cash quality has improved as customer deposits and product payments now cover the build-out costs.
The balance sheet holds $2.67 billion in cash, providing a massive buffer against its $2.47 billion in long-term debt. While the debt-to-equity ratio of 1.74x is high, the surging cash pile and newfound profitability make the leverage manageable. The company recently simplified its capital structure by removing obsolete share classes, signaling its maturity for institutional investors.
BE has transformed from a loss-making technology bet into a profitable, high-growth infrastructure giant.
Product revenue grew by 215% year over year as hyperscalers validated Bloom as the standard for AI onsite power. This surge in volume drove non-GAAP gross margins to 34.3%, proving that the company can expand its profitability even while rapidly increasing its manufacturing capacity.
A class action lawsuit alleges the company misled investors regarding its reliance on Chinese-sourced scandium. If supply chains for this critical mineral are disrupted or if the litigation reveals deeper vulnerabilities, it could threaten the company's ability to fulfill its massive multi-gigawatt backlog.
The market for on-site electrical equipment is roughly $500 billion today and is on track to exceed $700 billion by 2030 as AI data centers outrun grid capacity. The critical bottleneck for the next decade is "time-to-power," making onsite generation a structural necessity rather than an elective choice. Pricing power is currently high because demand for immediate electricity far exceeds the utility grid's ability to supply it, positioning Bloom as a vital alternative for hyperscalers.
The competitive dynamic is currently a race for capacity rather than a war on price. Barriers to entry are high due to the complex chemical engineering required for solid oxide fuel cells, but the industry is consolidating around players who can manufacture at gigawatt scale. Long-term pricing power will depend on maintaining a lower cost-per-kilowatt than the traditional grid.
Main competitors include GE Vernova and FuelCell Energy. GE Vernova threatens Bloom by leveraging its massive existing utility relationships to bundle power solutions. FuelCell Energy competes head-to-head on technology but lacks Bloom's massive commercial momentum in the data center vertical. The most dangerous threat is the traditional electrical grid itself, which could eventually upgrade capacity and remove the "time-to-power" advantage Bloom currently enjoys.
BE is rapidly gaining share as it transitions from a niche provider to a utility-scale partner. The 2.8 gigawatt Oracle deal is larger than the entire historical footprint of most competitors combined.
The primary protection is the company's proprietary Solid Oxide Fuel Cell (SOFC) technology, which allows for high-efficiency power generation without combustion. This IP is protected by hundreds of patents that prevent rivals from easily replicating the specific material science required for their fuel stacks. The company’s recent 215% surge in product revenue proves that hyperscalers are standardized on this specific technology.
Collectively, the 31.2% gross margins and 24.8% ROE indicate that Bloom is starting to reap the benefits of its technological lead. The transition from GAAP losses to a $182 million operating profit in a single quarter proves that the business model scales effectively. These numbers are consistent with a real moat that is beginning to show its strength through operating leverage.
The Narrow rating reflects Bloom’s reliance on critical minerals like scandium and the risk that cheaper battery storage could eventually compete. While the business is strong, it has not yet proven it can maintain these margins if supply costs for its rare earth inputs spike.
The forward-looking verdict is that the moat is strengthening. The multi-gigawatt long-term agreements with Brookfield and Oracle create a decade of visible, locked-in demand that competitors cannot easily displace.
Delivered 166% revenue growth and surpassed $1B in quarterly revenue for the first time.
Invested in capacity expansion while reaching positive $226M operating cash flow this quarter.
Co-founder CEO has led the company for 25 years with significant equity stake.
Capital Allocation Track Record
K. R. Sridhar has demonstrated exceptional strategic judgment by pivoting Bloom from a niche distributed power player into a critical AI infrastructure partner. His ability to secure multi-gigawatt deals with some of the world's most sophisticated capital allocators, like Oracle and Brookfield, validates his vision for the future of energy. Management’s recent performance shows they can scale production rapidly while finally delivering the GAAP profitability and positive cash flow that investors have demanded for a decade.
The primary governance risk is the heavy dependence on K. R. Sridhar, whose technical vision has been the sole driver of the company since its founding 25 years ago. While he has built a strong bench including a new Chief Commercial Officer to handle the scaling effort, his departure would create significant uncertainty regarding future technical innovation. The board recently modernized the corporate charter to align with Delaware law, which improves protections for officers but also highlights the complexity of managing a founder-led company during a hypergrowth phase.
We expect revenue to grow from $4.1B in FY2026 to $21.0B in FY2031 (~39% CAGR), with EPS growing from $2.65 to $17.00 (~45% CAGR). The rapid expansion of AI data centers is driving massive demand for on-site power generation that bypasses the constrained electrical grid. Manufacturing costs for fuel cell stacks decrease as production volume scales, allowing more revenue to flow to the bottom line. Operating margin expected to reach ~22% by FY2031.
Oracle Project Jupiter scales to full 2.45 GW capacity. Fulfilling this massive contract establishes Bloom as a utility-scale player and triggers similar deals across the data center industry.
Manufacturing costs drop as production volume doubles to 2 GW. Scaling production allows for better materials sourcing and higher automation, widening the gap between revenue and costs.
Expansion into hydrogen-based power for carbon-neutral data centers. Launching hydrogen-ready servers allows Bloom to capture the green-energy mandates of Big Tech firms like Google and Microsoft.
Supply chain disruption for scandium halts production of fuel stacks. Dependence on Chinese-sourced minerals could lead to production delays or massive cost spikes that erode profit margins.
Electrical grid upgrades accelerate, reducing the time-to-power advantage. If utilities significantly shorten the time to connect new data centers, Bloom's primary value proposition of speed disappears.
AI data center build-out slows due to lower ROI. A pullback in hyperscaler capital spending would leave Bloom with massive excess manufacturing capacity and high fixed costs.
Below is our estimate of current and future fair value, with detailed reasoning and assumptions. Fair value is a judgment, not a fact, and other analysts will likely land on different numbers. Use it as one data point in your research, and apply your own discretion in any investing decision.
We use a Normalized P/E approach — calculating value based on average profit over a cycle rather than a single peak year. It fits Bloom Energy because the company has just reached profitability and is experiencing a massive growth surge from AI data centers, which can make a single year's earnings look unsustainably high or low. Normalizing the earnings (averaging them out) helps us find the true, long-term value of the business.
We use a normalized earnings per share (EPS) base of $8.00 and apply a 30x multiple to reach our $240 fair value. Our $8.00 base is the average projected profit from FY2026 to FY2030, which captures the company's transition into a mature, profitable manufacturer. A 30x P/E ratio (investors paying $30 for every $1 of profit) sits conservatively below peers like GE Vernova at 51x and Siemens Energy at 58x. We chose this lower position to account for Bloom’s higher debt levels and the potential risks surrounding its scandium supply chain.
A cross-check using EV/Revenue (total company value vs. yearly sales) produces a fair value of $205 — within 15% of our primary answer, confirming the result. We applied a 15x revenue multiple to the midpoint of management's 2026 sales guidance ($4.05 billion). This multiple is a discount to the current 20x market rate to stay conservative. After subtracting the company's net debt and dividing by 295 million shares, the $205 figure suggests the stock is currently trading very close to its fair value for the current year, while our primary $240 target reflects the average earning power over the next few years.
We're assuming Bloom Energy achieves its manufacturing capacity goal of 2 gigawatts by the end of 2026. This is reasonable because the company is already scaling from its first $1 billion revenue quarter and has secured multi-billion dollar financing through Brookfield to fund the expansion of its production lines.
We're assuming the "time-to-power" advantage remains the primary driver for data center contracts through 2028. Because traditional electric utilities often take 5 to 10 years to upgrade the grid for AI needs, Bloom’s ability to install onsite power in under a year makes it the only viable choice for hyperscalers like Oracle, even at a premium price.
We're assuming the company maintains a gross margin (the profit left after making the product) above 30%. While manufacturing costs typically fall as companies grow, the risk of rising rare mineral prices (scandium) is balanced by the massive scale of the new Oracle and Brookfield contracts, which should allow Bloom to negotiate better volume pricing.
The biggest risk is the ongoing class action lawsuit regarding the company's reliance on Chinese-sourced scandium. This rare mineral is essential for Bloom's fuel cells, and any supply disruption or proof of misleading disclosures could force the price-to-earnings multiple down from 30x to 15x. This would knock roughly $120 off the per-share fair value. Watch for any court rulings or supply chain updates regarding "critical mineral sourcing" in the next two quarters.
Bear case ($176): Scandium supply chain litigation leads to a material halt in manufacturing or a major SEC fine; or A significant deceleration in AI data center orders as hyperscalers digestion period begins in 2027.
Bull case ($350): Oracle Project Jupiter reaches 50% completion ahead of schedule, proving Bloom can scale at gigawatt levels; or Brookfield Asset Management partnership expands beyond $25B, funding a global rollout into the Middle East and Dubai.
Clearthesis wrote this report from 45 sources, including SEC filings, analyst estimates, industry research, and recent news.
How did you like this thesis?
Your feedback helps us make reports better for you
© 2026 Clearthesis.ai · Report generated on August 12, 2026
This is an AI-generated analysis for informational purposes only and does not constitute financial advice. Data and analysis may not reflect recent developments if viewed significantly after the generation date. Always conduct your own due diligence before making any investment decisions.