Fervo Energy ($FRVO) & Ormat Technologies ($ORA): Who Gets Paid for Geothermal Growth?

Ormat Is the More Underwritable Business. Neither Is a Demonstrated Bargain.

Prioritize ORA for the September 8 event. Keep FRVO in a speculative, evidence-gated research lane.

At $105.45, ORA offers roughly an 8% annualized return in our funded-growth base scenario through year-end 2028. That is below the 12% research hurdle used here. An unchanged thesis becomes more interesting around $85–$90; alternatively, better cash generation or stronger, credibly financed earnings could justify today's price.

FRVO has the more direct enhanced-geothermal growth exposure, but its current valuation cannot be supported simply by multiplying contracted megawatts by a plant multiple. A no-credit, no-new-equity reverse model needs approximately 1.64 GW operating by year-end 2030 to earn a 12% annualized return from $18.16—and would require substantial additional borrowing. That is a financing stress test, not a forecast.

Market data: [1] FRVO close [2] ORA close Shares and accounts: [3] [4] Calculations and scenarios: Sections 5, 9–11.

Evidence confidence: high for filed historical figures; medium for reproduced call commentary and public estimate snapshots. Underwriting status: preliminary comparative initiation. ORA has a conditional valuation framework; FRVO does not yet have a bankable project-by-project equity valuation. This is not a short recommendation, a pair trade, or authorization to place orders.

The Scarce Asset Is Deliverable Power With Attractive Returns

The useful investment distinction is not simply “new geothermal versus old geothermal.” FRVO is trying to industrialize subsurface development. ORA combines operating assets, manufacturing and development. Both must turn underground heat into dependable net electricity delivered to a paying customer.

For FRVO, faster drilling can lower the cost of each successful well, but pumps, surface plants, transmission, makeup drilling and finance still consume cash. A technically better well does not automatically mean a better equity return. For ORA, existing sites and operating experience can support incremental projects and contract repricing, but weak legacy margins or unusually profitable storage quarters can obscure that improvement.

Where a Genuine Variant View Could Emerge

FRVO: sustained delivered output and cheaper capital could make current financing fears excessive. The variant must be demonstrated in net MW, realized cash receipts and ownership of those receipts—not inferred from Google's brand name.

ORA: investors could underestimate how much recurring cash can emerge from repriced contracts and completed projects. Conversely, a geothermal narrative could encourage investors to overpay for merchant-storage earnings and one-off plant-sale profits. September 8 can help distinguish those interpretations.

The moat is therefore a hypothesis about execution and capital efficiency. Land, permits, engineering, supplier relationships and operating experience can be valuable, but we have not established that either company has exclusive control of the technologies or all the economic surplus. Industrial learning can benefit customers through lower tariffs as well as producers through better margins.

Real Adoption, With Commercial Delivery Still Ahead

FRVO's September 1 agreement adds 396 MW of Google demand, with delivery expected in 2028. Company disclosures put executed commercial capacity at 1,054 MW afterward. The approximately 600 MW Google expansion option remains optional; the broader framework is not additional firm backlog. [5] [6]

Historical accounts: [7] IPO prospectus [3] Q2 filing. Historical revenue is ancillary revenue, not a mature power-sales run rate.

The Q2 presentation shows a 3 MW pilot and 500 MW under construction at Cape Station. The initial roughly 100 MW is split across three units; management targets first power from the first unit in Q4 2026. The 400 MW second phase is a later build. Do not equate first power with a fully ramped, reliably exporting 100 MW business. [8]

Two Different Transmission Constraints

Management's August call indicated $60–$80m of 2027 revenue, reflecting potential curtailment on a third-party transmission network. It believes that disruption is specific to 2027. Separately, filed disclosures identify 290 MW of Phase II transmission rights against 384 MW of expanded SCE/CPA obligations. These are different constraints and need separate resolution. [9] [7]

The investment implication is practical: a producing well can pass its engineering test while the equity still misses its revenue schedule. We need operator delivery dates, contractual remedies and net realized generation—not merely mechanical-completion announcements. Legal-firm allegations about disclosure timing are not findings of misconduct and do not establish investment value.

The Market Has Already Reacted

FRVO rose from $15.38 on August 31 to $19.75 on September 1, then closed September 4 at $18.16. The Google announcement was not ignored. The subsequent giveback also does not prove the selloff has finished. The August 12 call introduced a concrete deliverability debate; attributing the entire drawdown to one cause would require more evidence. [1]

The Cash Is Real. It Is Also Committed to a Buildout.

[3] Q2 filing [10] Option strike information. The dilution proxy applies the earlier $2.43 weighted option strike to June awards and August basic shares. Post-quarter exercises, vesting and ESPP terms can change it.

Management's H2 capex plan is $850–$900m. Deducting the $875m midpoint and an assumed repeat of H1 operating cash burn leaves approximately $1.19bn of year-end cash before financing and other movements. This is an illustrative bridge, not a survival runway forecast: 2027 spending and project-finance terms still matter. [11]

Phase I also has priority distributions and a subsequent royalty to financing partners. Its project debt has a 2031 maturity, not permanent financing. Those obligations belong in the cash waterfall; they cannot disappear because the business is described using consolidated EBITDA. A book-value claim adjustment is only a provisional substitute for valuing the actual payout schedule. [3]

Test the Cash Return Before Capitalizing the Megawatts

The prospectus presents general EGS economics using $115/MWh, an 83% capacity factor and $160,000 of annual operating cost per MW. Those are illustrative industry inputs, not disclosed tariffs and operating costs for the Google contract. The Q2 target of $5.5m/MW for the newer design is a management cost target, not an achieved fleet average. [7] [8]

To see what can remain after financing, the following illustrative first-year cohort adds 70% debt funding, 7.5% interest, 20-year scheduled amortization, 30-year straight-line depreciation, a 25% cash-tax assumption and a $55,000/MW sustaining reserve. There are no tax credits. Equity invested is $1.65m per MW.

Analyst sensitivity, not a project forecast or life-cycle IRR. Annual debt service is $377,655/MW. Cash to equity deducts debt service, assumed sustaining expenditure and simplified first-year taxes. It excludes corporate overhead and existing preferred/royalty waterfalls. The 20-year amortization is not FRVO's actual committed debt maturity. Fifteen-year contracts, refinancing, temperature decline and later cash taxes can materially change lifetime returns.

That spread is the point: the same technology and installed capacity can produce an unattractive or attractive return depending on tariff, financing and sustaining costs. Tax incentives could improve these cases, but eligibility, timing, transfer discounts and who retains the benefit must be modeled. We do not add both a full investment credit and a perpetual production credit to the same assumed project.

What $18.16 Requires in a Financed-Growth Model

At the dilution-proxy share count, today's equity value is $5.87bn. Earning 12% annually through December 2030 requires approximately $9.57bn of terminal common-equity value. We test that requirement using $676,142/MW of pre-credit project EBITDA, $100m annual corporate overhead and a 15x terminal EBITDA multiple.

Crucially, more capacity requires more capital. The funding proxy credits existing construction in progress against total installed costs, uses current cash, assumes $700m cumulative project cash at 1.1 GW (scaled with capacity), $500m cumulative corporate cash costs and a $250m ending cash reserve. That project cash is explicitly after interest, taxes and sustaining expenditure, but before debt-principal repayments and preferred distributions; existing debt and preferred claims remain in the terminal bridge. All additional funding is debt; no new tax-credit cash or common-equity issuance is assumed.

Cells show operating capacity required at year-end 2030 to meet the 12% equity-return hurdle from $18.16. These are algebraic requirements—not probabilities, capacity forecasts, lender commitments or a precise market-implied consensus. Terminal multiples are analyst assumptions; no perpetual growth rate is used.

The middle case requires about $5.39bn of new borrowing and leaves net financial/priority claims around 5.5x EBITDA. That is a warning about financing feasibility, not evidence that the borrowing will be available. Additional equity could reduce leverage but increases the share-count hurdle.

At management's pre-September 1 target of 1.1 GW by 2030, the same illustrative model yields $20.79 per share in December 2030, or $12.75 discounted at 12%. We do not label $12.75 fair value: contract tariffs, credits, project cash waterfalls, capacity timing, asset-allocation assumptions and value beyond 2030 remain insufficiently resolved. The new Google agreement may change the relevant construction plan. [9]

Normalize the Existing Business Before Paying for EGS

Historical financials: [12] Annual report [13] FY2025 release [4] Q2 filing. EBITDA is management-adjusted; EPS is GAAP. H1 is not a full year.

Revenue growth has not yet translated into sustained GAAP earnings-per-share growth. Electricity gross margin fell from 36.6% in 2023 to 28.5% in 2025. Part of the pressure was operational, including curtailments and wellfield issues, rather than a collapse in long-term demand. That distinction gives recovery a plausible basis but does not make the recovery automatic. [12] [13]

The H1 Headline Needs Two Adjustments

The TOPP2 plant sale contributed $105.1m of revenue and a $24.8m gain. Excluding it, H1 revenue growth was approximately 20.2%, rather than 42.9%, and adjusted EBITDA growth was approximately 10.3%, rather than 19.0%. The sale generated real cash, but that is not a recurring annual profit stream. [4] [14]

Storage contributed roughly 73% of the H1 increase in consolidated gross profit. Electricity gross profit declined. New storage assets helped, but favorable merchant pricing was also important. An underwriting that calls all of this “geothermal acceleration” would misidentify the earnings driver. [15]

Our normalization: keep the established electricity franchise and real storage expansion; remove the plant-sale uplift from a recurring run rate; avoid extending unusually favorable merchant margins indefinitely. Treat EGS as an execution-dependent future business, not an established earnings stream.

Answering the Free-Cash-Flow Question

Trailing operating cash flow is $279.4m; trailing cash capex is $544.2m. Their difference is negative $264.8m. Replacing all capex with management's $55m electricity-maintenance plan produces a 3.5% cash yield on current basic market capitalization—but that shortcut excludes other sustaining needs and owner-level adjustments. [12] [4] [16]

Here is a more explicit maintenance-only sensitivity. Normalized CFO of $300–$400m allows some working-capital recovery and is anchored between the recent trailing result and the stronger 2024 result. Sustaining expenditure of $75–$110m deliberately exceeds the narrow $55m electricity figure, allowing for uncertainty around wells, batteries and other assets.

All normalized rows are analyst assumptions, not company guidance. Cash interest and cash taxes are already included in CFO. These yields are before new growth investment and debt principal; they are not distributable-cash forecasts. A growth valuation must also fund growth capex. The forward scenario below models share dilution rather than subtracting an additional noncash SBC charge.

Two accounting cautions matter. TOPP2 sale proceeds sit in investing cash flow. Tax-equity financing is also not recurring operating cash generation. Separately, credit-transfer income and cash receipts need their own reconciliation; we do not subtract tax benefits again when CFO has already reversed the noncash item. This is why a full EBITDA-to-owner-cash bridge is more informative than an adjusted EPS headline.

The analysis does not prove that the business is structurally cash-poor. It establishes that today's price requires future cash improvement, and that the retained cash is competing with a large investment program. That is an answer investors can test.

Use All the Claims on the Earnings

[4] June 30 balance sheet and debt notes. Shares current as of August 1. Book minority interests are a proxy, not fair value. This table uses all consolidated adjusted EBITDA, so minority claims are included rather than also subtracting minority EBITDA.

The JV debt adjustment matters because reported EBITDA includes proportionate affiliate earnings while the corresponding debt is not consolidated. Operating leases are not added while the denominator remains after rent. The $141.1m asset-retirement obligation is excluded from the working valuation, pending explicit lifecycle cash modeling. The separate maintenance-only sensitivity does not establish that this liability is captured in the terminal model. Deducting its carrying value would lower illustrative equity value by roughly $2.24 per forward share, before discounting. [4]

At this working EV, ORA trades at approximately 15.4x the $640m midpoint of 2026 EBITDA guidance—or 16.0x after removing the TOPP2 gain. The conventional EV before selected tax/minority claims is about 14.9x headline guidance. Metric definitions matter more than a one-decimal comparison between websites. [15]

Near-term financing also deserves attention: $175m of the newer notes has a March 2027 cash put, and approximately $190.6m of older converts matures in July 2027. The new Series A conversion price is $140.40. Keeping principal in debt and adding the entire as-converted share count would double-count that claim; the model instead charges only conversion value above principal. [4]

A Cheaper Share Price Has Not Removed the Growth Hurdle

[17] CWEN close [18] CWEN results and dividend. ORA's indicated $0.48 annual dividend is about 0.46% of $105.45. Dividend yields are not total-return forecasts or directly comparable FCF yields.

CWEN lowered its 2026 cash-available-for-distribution guidance to $430–$470m, illustrating that contracted renewable portfolios still face operating and resource variability. We do not pool its CAFD, ORA's EBITDA and FRVO's potential project earnings into a peer median. A broad utility median would create more apparent precision than analytical value here. [18]

The public estimate snapshot for ORA shows approximately $1.18bn revenue and $2.63 adjusted EPS for 2026; at $105.45, that is about 40x adjusted earnings. Its reported average analyst target is $132.42. FRVO's snapshot shows roughly $72.9m of 2027 revenue and a $42 average target. Those are third-party aggregates, not our forecasts or independently reconstructed analyst models. [19] [20]

Our ORA base does not claim a hidden revenue surprise: its $800m 2028 EBITDA assumption is anchored to the midpoint of management's published $775–$825m year-end run-rate goal. An exit run rate is not the same as earned full-year EBITDA. We credit it at the end of 2028 only if the capacity is operating and the run rate is credible. [16]

The variant is the financing and cash-conversion judgment. A higher headline EBITDA target without the capex and funding bridge is insufficient. The 12x / 15x / 17x scenario multiples below are explicit analyst assumptions, not a sourced peer median. The 15x base broadly retains today's headline enterprise multiple rather than relying on a rerating.

Do Not Freeze Debt While Crediting Future Earnings

The base funds all modeled growth cash needs through additional debt or retained cash. It assumes no new tax-equity cash, asset-sale receipts or common-equity cash proceeds. This is deliberately transparent, not a prediction of the company's optimal funding mix. Refinancing maturities does not itself create new equity value.

H2 capex uses the August management budget; later spending and all cash forecasts are analyst assumptions. Cash interest and taxes are embedded in CFO. The $470m FY2028 CFO forecast is an independent analyst assumption; the model does not reconstruct full-year earned EBITDA or establish a comparable cash-conversion ratio. The $800m exit EBITDA is a year-end run rate, not full-year earnings. Source anchors: [4] [13].

The 63m pre-new-convert share assumption allows for award issuance and residual old-convert dilution. The $175m put and 2027 maturities are assumed refinanced, not forgiven. Only the $825m Series A remains convertible in the end-2028 model; when the share value exceeds $140.40, its excess conversion value is charged separately. No unverified capped-call benefit is credited. This is a pro forma sensitivity, not an exact future cap table.

These are conditional values, not formal price targets or calibrated probabilities. No probability-weighted “fair value” is presented. Forecast horizon is 2.319 years. Dividends are approximated at $0.48 per year and received at the horizon for discounting. No perpetual growth rate is used; terminal multiples implicitly require durable future earnings and continued reinvestment.

Bear mechanism: cash conversion disappoints, projects cost more and merchant-storage economics normalize; incremental borrowing rises while the multiple contracts. Base: the existing 2028 goal becomes an operating run rate, with cash conversion recovering. Bull: stronger recurring earnings and cash delivery support both a $900m run rate and a premium multiple. The upside therefore requires more than a new slide showing higher MW.

Hold Funding Fixed: What Changes the Share Value?

End-2028 values, using the base $4.157bn claims balance and the same conversion treatment. This is a one-variable sensitivity, unlike the coupled bear/base/bull cases above. Additional $100m of funding claims reduces value by roughly $1.59/share below the conversion strike.

What Would Make ORA Actionable?

The funded base scenario supports an entry ceiling of approximately $96.57 for a 12% annualized return hurdle. Requiring extra room for uncertainty brings the research-entry zone to roughly $85–$90. This zone is conditional on the earnings and funding thesis remaining intact; it is not a prediction that the stock will trade there.

Avoid reading the $849m threshold as a demand for management to issue that exact target. A lower funding requirement, better contracted cash economics or lower risk could change value too. Conversely, a higher target that requires more dilution may not improve per-share returns.

The opportunity cost is keeping capital available for setups that offer a wider verified gap between price and value. Neither ticker has a demonstrated squeeze edge in this work: current short interest, borrow availability, options positioning and effective tradable float were not verified. No options or event-day position is recommended.

September 8 Is an Evidence Event, Not a Guaranteed Rally

ORA's official calendar lists its Investor Day for Tuesday, September 8, 2026, at 8:00 a.m. ET / 5:00 a.m. Pacific. Confirm webcast timing on the event page before attending. [21]

Event and project timing: [21] [15] [8] [9]. Later regulatory approval was not verified in this review; H2 timing remains the company's expectation, not a confirmed decision date.

The Questions I Would Take to Investor Day

1. How much of the 2028 run rate comes from merchant storage, conventional geothermal, equipment and EGS? Show the bridge separately.
2. How much cumulative cash investment is required, and what remains after project debt, tax monetization, minority claims and corporate refinancing?
3. What is genuine economic maintenance spending, including battery replacement and makeup wells, rather than only the category called maintenance?
4. Which contracts reprice soon enough to affect owner cash before 2028, and what capital is needed to earn that repricing?
5. What net output and cost thresholds must the EGS pilots clear before the company commits to commercial-scale builds?

ORA's disclosed 40 MW blend-and-extend arrangements add $20–$30/MWh. At an assumed 90% capacity factor, that represents only about $6.3–$9.5m of incremental annual revenue. It is useful evidence of pricing power, but insufficient by itself to justify a large equity rerating. [16]

Observe the Failure Before It Becomes a Narrative

Management credibility should be scored against dated commitments: cost targets, commercial-operation dates, net production, reported cash receipts and financing. Our current review does not establish a forecasting hit rate. Current beneficial ownership, institutional positioning and insider-trading patterns also remain unverified; none is used to support the investment case.

What This Underwriting Establishes—and What It Does Not

It establishes a current capitalization framework, material accounting adjustments, modeled funding needs, explicit return hurdles and a dated decision event. It does not establish a private project-finance valuation, a precise terminal multiple, calibrated scenario probabilities or a guaranteed stock-price response.

• FRVO: individual tariff schedules, credit elections, timing of transfer proceeds, project allocation and preferred cash waterfalls still prevent a precise equity NAV. The September investor deck could not be accessed; newer backlog details from secondary summaries were not adopted.
• ORA: economic sustaining expenditure, future tax-equity cash, minority fair values and the post-2026 funding mix remain scenario assumptions. The $800m run rate comes from an existing management goal, not independently forecast consensus EBITDA.
• Valuation: no fake peer median and no fabricated probability-weighted target. The ORA base holds a roughly current enterprise multiple; the FRVO model explicitly funds capacity. Neither is a fully integrated three-statement forecast.
• Data: prices are the last completed trading-session closes, not live quotes. Balance sheets are June 30; later cash balances can differ materially. Public estimate snapshots are secondary sources, not direct terminal-provider access.

Next underwriting handoff: update ORA after Investor Day, keeping a before/after record of EBITDA, cumulative funding and implied per-share return. Advance FRVO when project-specific cash economics and transmission evidence become sufficient to replace the illustrative reverse model.

Source Register

Accessed September 6, 2026. Filed accounts control historical figures; company targets remain targets. Model assumptions and derived calculations are identified in the relevant sections. Dates below are document dates or market-data cutoffs.

1. FRVO historical pricesSeptember 4, 2026 close; public S&P-sourced price history. Market-data reference.
2. ORA issuer stock-price pageSeptember 4, 2026 close. Issuer-hosted market-data reference.
3. Fervo Q2 2026 Form 10-QAugust 13 filing; June 30 accounts; August 10 actual shares. Financial statements, liquidity, debt, equity awards and project claims. Primary.
4. Ormat Q2 2026 Form 10-QAugust 6 filing; June 30 accounts; August 1 shares. Statements, debt, cash flows, affiliate and tax-benefit notes. Primary.
5. Fervo–Google 396 MW agreementSeptember 1, 2026. Company announcement; delivery timing and options are forward-looking.
6. Fervo's official commercial-capacity updateSeptember 1, 2026. Company confirmation of 1,054 MW executed capacity; company-controlled social disclosure.
7. Fervo final IPO prospectusMay 14, 2026. Audited historicals and industry economics, including illustrative economics pp. 123–124; not actual project guidance. Primary; use current 10-Q for superseding balances.
8. Fervo Q2 investor presentationAugust 12, 2026. Construction and design roadmap, pp. 9–11 and 19–20. Management targets.
9. Fervo August 12 call transcriptPublished August 19, 2026. Third-party reproduction of management remarks; full Q&A used, not its automated summary. Not independently audio-verified.
10. Fervo S-8 fee exhibitMay 14, 2026. Historical weighted option-strike information; cross-date dilution proxy only. Primary.
11. Fervo Q2 resultsAugust 12, 2026. Quarter results and capital-spending guidance. Primary company release.
12. Ormat 2025 annual reportFY2025; audited 2023–2025 statements and notes. Primary; used for history, not current guidance.
13. Ormat FY2025 resultsFebruary 25, 2026. Historical margins, EPS reconciliation and operating explanations. 2026 guidance superseded by Q2.
14. Ormat Q1 2026 resultsMay 6, 2026. TOPP2, financing and adjusted earnings context. Primary company release.
15. Ormat Q2 2026 resultsAugust 5, 2026. Segment performance, 2026 guidance and EGS progress. Primary company release.
16. Ormat May 2026 corporate presentationPPA pricing, slides 9–10; 2028 run-rate goal; maintenance-capex plan, slide 42. Company claims; 2026 earnings guidance subsequently updated.
17. Clearway Energy historical pricesSeptember 4, 2026 close. Public market-data reference.
18. Clearway Energy Q2 resultsAugust 5, 2026. Dividend, CAFD guidance and definitions; secondary comparison, not a full CWEN underwriting.
19. ORA public analyst-estimate snapshotUpdated September 4, 2026; S&P/TipRanks attribution. Adjusted EPS basis; secondary aggregate.
20. FRVO public analyst-estimate snapshotRetrieved September 6, 2026; September price/rating context. Secondary aggregate; underlying forecast models unavailable.
21. Ormat official event calendarRetrieved September 6, 2026. September 8 Investor Day listed at 8:00 a.m. ET. Dynamic page; reconfirm before attending.

Research Levels, Not Personalized Instructions

Disclosure: This is independent, impersonal investment research for informational and educational purposes only—not individualized investment advice or a solicitation to buy or sell any security. Investments can lose substantial value or all capital. Development-stage energy companies and capital-intensive infrastructure businesses carry construction, technology, financing, dilution, regulatory, customer, liquidity and market risks. Prices and scenarios are research levels used to evaluate a thesis, not instructions tailored to any reader's financial situation. Readers should conduct their own due diligence and independently determine whether an investment is appropriate for them. Forward-looking statements and modeled scenarios are uncertain; actual results may differ materially.

Ownership Disclosure: The author does not own shares in Fervo Energy ($FRVO) or Ormat Technologies ($ORA).

Original report and latest update: September 6, 2026. Thesis status: ORA valuation-gated; FRVO project-economics and financing-gated. Next review trigger: September 8 Investor Day, or an earlier material filing or financing. This report does not create an automated monitor or execute a trade.