Ormat Technologies, Inc. (NYSE: ORA) — Scarce Heat, Expensive Growth

Public Research Memo. Report date: 2026-08-30. Fresh coverage. Sections 1–15 carry no recommendation or price target.

⚡ Claude’s Take

The following is a subjective opinion, separated from the evidence-based body; not investment advice. Sections 1–15 below carry no position.

Verdict: HOLD / do not initiate here; accumulate only on material weakness. Medium conviction. My reasonable-value zone is approximately $75–95, with better risk/reward below approximately $80. At $103.04, Ormat has already fallen 28.9% from its June peak, but the valuation still asks investors to pay approximately 18.5× TTM EBITDA, 14.7× management’s 2026 adjusted-EBITDA midpoint, 16.8× a more conservative operating denominator, and 1.66× book enterprise capital for a business whose filing-derived normalized ROIC has been around 3%. The upper end of my zone gives credit for contracted geothermal scarcity and management’s 2.6–2.8 GW 2028 portfolio ambition; the lower end recognizes that five years of capex have materially exceeded operating cash flow, storage margins are temporarily flattered by PJM merchant prices, and EGS has not yet produced a commercial pilot.

The framing is real scarcity, weak consolidated returns, and a large option already partly capitalized. Ormat’s proven geothermal resources, permits, interconnections, plant know-how, and long PPAs are difficult to recreate locally. That is a narrow advantage. It is not a wide corporate moat: Electricity operating income was only 2.1% of segment assets in 2025, Product’s Q2 2026 gross margin fell to 9.7%, and Storage has less than 1% U.S. share in a market with 890 GW queued. The market appears to be valuing Ormat less like a conventional IPP and more like a scarcity platform for 24/7 clean power, data centers, and next-generation geothermal.

The tape does not yet provide a reason to rush. ORA is below its 21-, 50-, and 200-day averages after a roughly 25% three-month decline; the Solar and Clean Energy factors are in hostile 63-day regimes. The stock is not fully broken—its one-year return remains positive—but positive momentum is absent. I would become more constructive after project-level evidence that incremental capital earns above the cost of capital, or at a price that assigns much less value to EGS and merchant storage. I would become more negative if 2027 conventional projects slip, tax-credit cash fails to reconcile, PJM storage margins normalize below management’s implied level, or EGS pilots require repeated capital without stable flow and cost data. Tag: the geothermal scarcity premium is deserved; the current growth premium still needs proof.

📈 Stock Price Action — Five-Year Event Map

Arc. ORA’s adjusted close rose from roughly $65 in August 2021 to above $100 in late 2022, fell to $58.73 in November 2023, recovered through 2024–2025, reached a five-year high of $146.39 on 2026-06-03, and closed at $103.04 on 2026-08-28. The latest price sits below the 21-day EMA of $107.41, the 50-day EMA of $110.29, and the 200-day EMA of $111.15. Price moves are facts from the AZI series; event attributions are interpretations unless directly tied to a filing or company announcement.

The February 2026 break and April–June surge matter more than the exact daily attribution. On February 26, the stock-specific return was approximately -7.8% around the FY2025 release. By June, investors had absorbed a 150 MW geothermal portfolio PPA supporting Google load, a Switch PPA, a $1 billion low-coupon convertible issue, an EGS partnership narrative, and a strong Q1. The subsequent reversal occurred without an equivalent collapse in reported guidance: Q2 raised the 2026 EBITDA range. That pattern is consistent with a narrative/multiple correction rather than a fundamental liquidity event.

Factor data support that reading. ORA’s model loadings are largest to Utilities, the broad market, Solar Energy, Clean Energy, and Dividend Yield; Momentum is zeroed as immaterial and Quality is negative. Over 63 trading days, Solar Energy returned -12.0% with a -2.38 z-score and Clean Energy returned -9.0% with a -1.54 z-score. ORA’s five-year annualized volatility was 31.9%, its five-year maximum drawdown was 41.2%, and annualized idiosyncratic volatility was 27.5%. The label “utility” should not be confused with regulated-utility risk.

1. Executive Summary

Ormat is a hybrid of three businesses. Electricity owns and operates geothermal, recovered-energy, solar, and related plants, usually under long-term PPAs. Product designs and manufactures binary-cycle equipment and delivers geothermal plants to third parties. Energy Storage develops, owns, and dispatches batteries in U.S. markets under a mix of tolling, resource-adequacy, and merchant arrangements. In 2025 those segments generated $693.9 million, $216.7 million, and $79.0 million of revenue, respectively. The mix matters because a contracted geothermal plant, a competitive equipment contract, and a merchant PJM battery do not deserve the same multiple or risk budget. Ormat 2025 10-K

The current operating news is strong on growth and weaker on cash conversion. Q2 2026 revenue increased 10.6% to $258.8 million and adjusted EBITDA increased 6.9% to $143.9 million. Management raised full-year revenue guidance to $1.15–$1.20 billion and adjusted EBITDA to $630–$650 million. Storage revenue nearly tripled, Product had a large first-half delivery/sale contribution, and new PPAs improved the Electricity outlook. Yet H1 operating cash flow fell to $129.2 million from $184.9 million while capex was $251.8 million; another $449 million was budgeted for the balance of 2026. The company funded the program partly with a $1 billion convertible issue and expects approximately $90 million of 2026 tax-credit monetization proceeds. Q2 2026 results, Q2 2026 10-Q

The moat is real but narrower than the corporate story. At a de-risked geothermal site, resource rights, successful wells, gathering infrastructure, permits, interconnection, operating data, and a long PPA form a locally scarce bundle. Ormat reported an 84% geothermal capacity factor, and 93.8% of 2025 Electricity revenue had a fixed-price component with roughly 14 years of weighted-average PPA life. But contracts do not remove reservoir decline, curtailment, availability penalties, or recontracting risk. Electricity segment operating income was only 2.1% of year-end segment assets in 2025; Brawley was impaired after persistent wellfield problems, Puna had reduced generation, and U.S. curtailment reached 277,923 MWh. These facts reject a wide-moat conclusion.

The industry offers both scarcity and a capital-cycle warning. U.S. geothermal nameplate capacity was only 3,969 MW in 2024, and DOE counted 26 PPAs representing more than 1,000 MW signed since 2021. That supports prices for firm renewable power and makes Ormat’s installed base meaningful. The same DOE work presents 90 GW by 2050 only under aggressive EGS cost reductions; it is a scenario, not an addressable-revenue forecast. Meanwhile U.S. batteries reached roughly 52 GW by mid-2026, developers expected another 54 GW through 2028, and 890 GW sat in interconnection queues. Ormat’s 495 MW operating storage portfolio is small in that supply wave. DOE geothermal market report, EIA battery storage, Berkeley Lab queue study

At $103.04, equity value is approximately $6.34 billion. Adding $3.41 billion of debt and $146 million of NCI and subtracting $514 million of cash produces about $9.38 billion of enterprise value. That is 18.5× TTM EBITDA and 14.4–14.9× 2026 adjusted-EBITDA guidance. Reported 2025 operating cash flow was $335 million against $620 million of capex; 2024 was $411 million against $488 million; 2023 was $309 million against $618 million. Growth can justify negative free cash flow, but only if incremental returns exceed the cost of capital. Public filings do not disclose the project-level unlevered return needed to prove that proposition.

The central variant is therefore not whether geothermal is valuable. It is whether Ormat can convert a scarce conventional fleet, PPA repricing, Storage build-out, and EGS partnerships into common-owner cash returns, rather than EBITDA growth financed by debt, tax-benefit sales, and outside capital. Evidence supports the former at individual de-risked plants; consolidated returns and cash flow have not yet proved it at the portfolio level.

2. Business Overview

2.1 Corporate map

Ormat describes itself as the only vertically integrated company that designs, develops, builds, owns, manufactures, and operates geothermal and recovered-energy plants. The economic virtue of integration is reduced interface risk: resource engineers, equipment designers, EPC teams, and operators can share field data and optimize the plant around a reservoir. The economic cost is that Ormat retains capital and execution risk at more steps of the chain. Integration is an advantage only when avoided failures and better plant output exceed the capital committed to exploration, manufacturing, and ownership.

At year-end 2025, Ormat owned or operated 35 plants and complexes totaling 1,340 MW; geothermal was approximately 1,089 MW. By Q2 2026, management described an approximately 1.85 GW total portfolio, including 1,355 MW of Electricity and 495 MW/1,358 MWh of Storage. The company targets a 2.6–2.8 GW total portfolio by year-end 2028, a 15–18% capacity CAGR from 2025. The disclosed construction path includes 202 MW of Electricity projects through 2028 and 497 MW/1,888 MWh of Storage under construction or development. Those are management plans with project and timing risk, not installed capacity. Ormat corporate profile, Q2 2026 results

2.2 Electricity: contracted price, retained operating risk

Electricity accounted for 70.1% of 2025 revenue and most of the asset base. The segment owns the resource and generation asset and sells output, normally to utilities or public-power buyers. The model is capital-intensive upfront and relatively predictable after a reservoir, plant, and interconnection are proven. PPAs can contain fixed energy pricing, availability payments, escalators, curtailment rules, and performance penalties. In 2025, substantially all geothermal output was contracted; 93.8% of Electricity revenue had a fixed-price component and weighted-average remaining PPA life was about 14 years.

That contract visibility is valuable but incomplete. A geothermal PPA does not make subsurface temperature, pressure, permeability, fluid chemistry, scaling, or well productivity contractual. Minimum-performance failures may reduce capacity payments, trigger liquidated damages, or permanently reset contracted capacity. Some newer PPAs require compensation for replacement-power cost. In 2025, curtailment reduced U.S. output by 277,923 MWh, Puna suffered wellfield-related generation weakness, and Brawley was impaired for chronic wellfield underperformance. Price is hedged better than physical volume.

Concentration also matters. SCPPA, NV Energy, and Kenya Power represented 17.8%, 13.8%, and 11.9% of 2025 consolidated revenue, or 43.5% combined. International Electricity represented only about one-fifth of consolidated revenue but contributed a disproportionate share of gross profit, net income, and adjusted EBITDA. Kenya plant operations, receivables, sovereign/regulatory conditions, and currency therefore matter more than a revenue-only map suggests.

2.3 Product: bankability without captive demand

Product designs organic Rankine cycle/binary equipment, turbines, pumps, and complete plants and provides EPC services. Its customer proposition is lower execution risk: Ormat’s installed references, operating feedback loop, integration, and balance sheet make it a credible counterparty for a technically complex plant. Existing customers may also prefer compatible parts and service. Those are search-cost and modest switching-cost advantages.

They do not create captivity on the next tender. Named competitors include MHI/Turboden, TICA/Exergy, Baker Hughes, Kaishan, Fuji, Toshiba, and others. Customers are sophisticated and can select a steam or binary design and competing OEM/EPC team before construction. Product backlog was $351.7 million in February 2026, but roughly $100 million related to the one-time TOPP2 asset sale; underlying backlog was closer to $252 million. At Q2, Product revenue fell 21.6% and gross margin fell to 9.7% from 27.7% because of project cost and foreign-exchange effects. The margin volatility is inconsistent with strong pricing power.

2.4 Storage: a different business under the same ticker

Storage buys standardized battery systems, develops sites, secures interconnections and permits, finances projects, and earns from capacity, resource adequacy, tolling, ancillary services, and merchant energy spreads. The potentially scarce elements are land/interconnection positions and contracts; cells, inverters, and basic EPC are purchasable and competitive. Ormat grew from 83 MW/176 MWh in 2021 to 415 MW/1,038 MWh at year-end 2025 and 495 MW/1,358 MWh by Q2 2026.

Current profitability is not a normalized run rate. First-half 2026 Storage revenue increased to $87.7 million from $32.2 million and gross margin rose to 57.7% from 22.2%, principally due to higher PJM energy rates plus new facilities. New projects accounted for only $11.9 million of the revenue increase. Management expects second-half gross margin to normalize toward 30–40% and full-year margin toward 40–50%. The segment also recorded a Pomona impairment and abandoned-project write-offs. Storage is simultaneously a growth engine, a merchant-price exposure, and a capital-cycle risk.

2.5 The accounting map

Consolidated EBITDA overstates neither value nor cash by definition, but it can obscure ownership and reinvestment claims. Project debt is senior to common equity. Tax-equity investors and noncontrolling interests receive allocations of tax attributes and sometimes cash. Tax-credit transfer income and negative effective tax rates can lift reported earnings. Depreciation is large because the assets are large; the difference between economic maintenance capital and accounting depreciation is a decisive valuation input.

The June 2026 balance sheet reported approximately $3.4 billion of debt and $146 million of NCI. Management expects about $17 million of 2026 adjusted EBITDA attributable to minority interests. A consolidated EV/EBITDA multiple is consistent only if both NCI and consolidated EBITDA are included. A common-owner multiple should subtract both the NCI claim and its EBITDA. This report uses the consolidated convention for peer and history comparisons and separately flags the common-owner bridge.

3. Industry Dynamics

3.1 A tiny firm-power market with renewed demand

Geothermal is valuable because it combines renewable attributes with high capacity factor and firm or flexible output. EIA reports that U.S. geothermal generated about 16 billion kWh in 2025, only 0.4% of utility-scale generation, concentrated in seven states. DOE reports U.S. installed nameplate capacity of 3,969 MWe in 2024, up 8% from 3,673 MWe in 2020; the number of projects under development remained 54, but 26 PPAs representing more than 1,000 MWe had been signed since 2021. EIA geothermal use, DOE geothermal market report

Global conventional deployment is likewise small and slow. IRENA reports 15,674 MW of geothermal capacity in 2025 versus 14,451 MW in 2021, approximately 2.1% annualized growth. Ormat’s comparable operating geothermal fleet represented roughly 6.6% of global installed capacity in 2021 and 6.9% in 2025. The definitions are not exact, but the stable-share direction supports relevance. Stable share with low excess returns, however, is not a moat test victory. IRENA Renewable Capacity Statistics 2026

Data-center and industrial load have made firm clean power more valuable. Ormat signed a 15-year portfolio PPA of up to 150 MW through NV Energy to serve Google load and a 20-year approximately 13 MW direct PPA with Switch. The CD4 blend-and-extend amendment increased pricing about 27% on 15 MW and extended the contract through 2037. These transactions are direct evidence of improved pricing at particular assets. They do not prove that every expiring PPA receives the same uplift. Switch PPA, CD4 amendment

3.2 Capital goes first into holes in the ground

Conventional geothermal’s critical bottleneck is resource confirmation. The developer must secure land and resource rights, characterize heat and fluid, drill exploration and production wells, prove sustainable flow, engineer the gathering system and plant, obtain permits and interconnection, and contract the output. The largest destruction risk occurs before or during drilling: an unsuccessful well can consume capital without creating a cash-producing asset.

Once de-risked, a resource basin can exhibit efficient scale. A rival cannot cheaply place an identical reservoir and interconnection next door; the incumbent has operating and subsurface data and shared infrastructure. But geology also limits growth replication. The incumbent must move to another basin and repeat the risky development process. This combination—high local barrier, low replication rate—explains why geothermal assets can be scarce while corporate ROIC remains modest.

The Product value chain avoids direct resource ownership but competes for equipment and EPC awards. Equipment bankability and references matter because plant failure is expensive. Yet manufacturers and EPC providers can enter tenders, and oilfield-service firms bring subsurface capabilities. Vertical integration reduces Ormat’s interface cost, but it does not eliminate external competition.

3.3 EGS: much larger technical potential, much less proof

Enhanced geothermal systems engineer permeability by stimulating hot rock where natural fluid pathways are inadequate; closed-loop approaches circulate fluid through sealed wells. DOE scenarios reach at least 90 GW of U.S. geothermal by 2050 under aggressive EGS cost reductions and as much as 300 GW under broader next-generation assumptions. Those are technology/deployment scenarios, not current TAM or backlog. DOE geothermal basics, DOE EGS

Ormat’s surface plant, binary-cycle, reservoir-operation, and utility-contracting skills are relevant to EGS. They do not solve drilling cost, fracture connectivity, induced seismicity, water management, thermal decline, or repeatability. The SLB collaboration sensibly assigns subsurface/well work to SLB and plant/full-life-cycle work to Ormat. SLB’s January 2026 call described one or two assets as learning projects before safe and economic scaling. Ormat’s 2026 program includes two 2–4 MW pilots, with drilling/evaluation extending into 2027–2028, and only about $20 million of remaining-2026 EGS spend. The correct classification is a call option with explicit technical gates.

3.4 Storage supply is responding aggressively

U.S. utility-scale battery capacity was 43.6 GW at year-end 2025 and approached 52 GW after another 8.3 GW in the first half of 2026. EIA reports capacity had grown roughly 70% annually over the prior three years and developers expected another 54 GW through 2028. Berkeley Lab counted 890 GW of active storage in interconnection queues at year-end 2024. Queue projects have a low completion rate, but the direction is unambiguous: high returns invite capital. EIA battery storage, Berkeley Lab queue study

Ormat’s year-end 2025 fleet was approximately 1.0% of the installed U.S. base; the Q2 2026 fleet was still below 1%. NextEra and AES deploy several times Ormat’s capacity, while specialist developers compete for the same interconnections, tolls, and merchant spreads. FERC Order 841 reduced barriers to storage participation in organized markets. This is not a network business. Individual sites can be valuable, but segment-wide excess returns are vulnerable to entry. FERC Order 841

3.5 Policy changes the relative economics, not the moat

Section 48E provides a technology-neutral clean-electricity investment credit with a 6% base rate and up to 30% when labor requirements are met, plus potential bonuses; standalone storage is eligible and credits may be transferable. The 2025 law accelerated wind/solar termination while leaving qualifying geothermal and storage comparatively advantaged. It also imposed prohibited-foreign-entity and material-assistance restrictions that can raise battery sourcing cost. IRS Section 48E, Public Law 119-21

The policy is economically material. First-half 2026 included $33.2 million of income attributable to tax-benefit sales and a $25.1 million income-tax benefit, producing a negative effective tax rate. Yet a tax credit available to every compliant developer is a shared subsidy, not an Ormat-specific barrier. It improves project economics and financing capacity while potentially encouraging more supply. FEOC restrictions may also offset part of the benefit for Storage.

Industry verdict. Conventional geothermal offers local scarcity, contracted firm-power demand, and slow supply; EGS offers large technical optionality with unproven economics; Storage offers fast demand growth with even faster prospective capital entry. The favorable demand regime is real. The investment question is whether Ormat’s next dollar is deployed into the scarce part of the chain or the crowded part.

4. Competitive Position

4.1 Electricity under Greenwald

Greenwald’s durable sources are customer captivity, production advantages, and economies of scale within a protected market. Ormat’s Electricity segment has limited customer captivity: a utility is locked into an asset-specific PPA during its term, but can run a new procurement at expiry or for the next project. There is no network effect. The strongest advantage is instead a localized production/efficient-scale bundle—resource rights, proven wells, field data, gathering infrastructure, permits, interconnection, operations, and a customer-approved asset.

The evidence supports a narrow, project-local moat, not a company-wide wide moat. The fleet’s 84% geothermal capacity factor, long contracts, repeat utility relationships, and stable global capacity share validate capability. Returns provide the counterevidence. 2025 Electricity operating income was $114.6 million on $5.34 billion of segment assets, just 2.1%; even adding back D&A gives a 6.8% EBITDA/asset ratio before sustaining wells, interest, and tax. Segment capital additions were $446.8 million versus $250.8 million of D&A. Reservoir issues, impairments, and curtailment remain recurrent.

The historical Morningstar Narrow-moat label is therefore directionally plausible only at the asset/local-platform level. It should not be interpreted as evidence of 10 years of corporate excess returns. The current filing-derived consolidated ROIC check is roughly 2.6% using normalized tax and year-end invested capital, below a reasonable cost of capital.

4.2 Product under Greenwald

Product has search-cost advantages. A customer selecting a geothermal OEM/EPC team cannot cheaply learn whether an unproven vendor will deliver the specified output for decades. Ormat’s references, installed base, operating feedback, and integrated offering reduce that uncertainty. Service and compatible replacement parts can add modest switching costs after installation.

But new demand is contestable, and the tender resets economics. Product operating income improved from $3.5 million in 2023 to $23.4 million in 2025, yet Q2 2026 gross margin fell to 9.7%. The 2021 10-K cited a 78% share of installed binary generation, but the 2025 filing no longer quantifies it. Without a current denominator, stable high margins, or evidence that customers cannot switch, the historical share cannot carry the moat conclusion. The segment is weakly advantaged on reputation and bankability, with current evidence of margin erosion.

4.3 Storage under Greenwald and the capital cycle

Storage has no demonstrated segment moat. Hardware is purchasable; market rules allow rival participation; off-takers are sophisticated; Ormat’s share is below 1%; and much larger developers possess procurement and financing scale. Scarce interconnections, permitted sites, dispatch knowledge, and tolls can create project-level rents, but they do not protect the next project from entry.

Marathon’s capital-cycle lens is the more useful framework. Storage capital additions were 5.4× segment D&A in 2025, after 5.1× in 2024 and 8.5× in 2023. Industry queues and developer plans imply enormous capacity growth. Q2’s 56.2% gross margin benefited from very strong PJM merchant prices, which management already says are normalizing. High current returns plus long investment lags attract supply; when projects arrive together, merchant spreads and capacity economics can compress. Storage deserves a cycle-adjusted return, not an extrapolated peak margin.

4.4 Competitive map

Conventional geothermal competitors include Calpine’s Geysers assets, Pertamina Geothermal, Star Energy, Kaishan, and country-specific developers. EGS and next-generation entrants include Fervo, Sage, Eavor, and Quaise, often partnered with oilfield-service companies. Product competitors include the OEMs named above. Storage competitors include NextEra, AES, Plus Power, Jupiter, Broad Reach, Key Capture, Eolian, and technology/integration vendors such as Fluence.

No single public peer spans Ormat’s mix. NextEra and AES are capital-cycle comparisons; Brookfield Renewable and Algonquin are contracted-infrastructure comparisons complicated by large NCI and different geographies; Fluence is an equipment/integrator comparator with negative TTM EBITDA. Peer multiples should therefore be used as a boundary check rather than a precise price-setting device.

Competitive verdict. Ormat has scarce, locally defensible geothermal assets and a credible integrated development platform. Product has bankability, not captivity. Storage has sites and execution, not structural protection. The corporation has a narrow collection of asset advantages rather than one scalable moat.

5. Growth History and Forward Opportunities

5.1 The five-year record

Revenue increased from $663.1 million in 2021 to $989.5 million in 2025, an approximately 10.5% CAGR. EBITDA increased from $352.3 million to $461.3 million, approximately 7.0% annually. Yet gross margin declined from about 39.9% to 27.6%, and operating margin declined from about 25.5% to 17.1% on a segment-operating-income definition. Net income attributable to Ormat increased from $62.1 million to $123.9 million, partly reflecting tax-credit income and a lower effective tax burden.

The segment history is uneven. Electricity revenue was $666.8 million in 2023, $702.3 million in 2024, and $693.9 million in 2025; operating income declined from $168.8 million to $114.6 million over that period. Product revenue increased from $133.8 million to $216.7 million and operating income from $3.5 million to $23.4 million, but deliveries are lumpy. Storage revenue increased from $28.9 million to $79.0 million and operating income from a $5.8 million loss to $31.2 million, with merchant prices driving part of the latest result.

Growth has required external financing. Capex was $618.4 million in 2023, $487.7 million in 2024, and $619.8 million in 2025, versus operating cash flow of $309.4 million, $410.9 million, and $335.1 million. Debt rose as assets grew. The record demonstrates access to capital and execution at scale; it does not yet demonstrate self-funded compounding.

5.2 Conventional Electricity pipeline and recontracting

Management expects 2026 Electricity revenue of $710–$725 million and lists 202 MW of Electricity projects through 2028, all with PPAs: 87 MW geothermal and 115 MW solar. Its broader year-end 2028 portfolio target implies additional Storage and project execution. At Q2, two Caribbean projects were delayed one to two months, reducing the Electricity outlook by about $5 million. This is a reminder that “contracted” does not mean “on time.”

Recontracting may be the highest-return opportunity. Management says recent blend-and-extend agreements should add roughly $14 million of annual revenue during 2026–2030 with minimal incremental capital. Approximately 190 MW of PPAs expire in 2031–2034 at a weighted-average price near $86/MWh, below management’s indicated current market above $100/MWh. The CD4 amendment provides specific evidence of higher pricing. The opportunity is attractive because it monetizes scarce existing assets without repeating full exploration and construction risk. The open question is how much output carries similar counterparties, contract terms, and replacement-power value.

5.3 Product backlog and execution

Product’s August 2026 backlog was approximately $203 million, down from the February headline because TOPP2 revenue was recognized in Q1. Management expects 2026 Product revenue of $300–$320 million, but Q2 gross margin fell to 9.7% and management expects about 15% in the second half and 18% for the year. The growth issue is therefore not backlog alone; it is whether bids include enough contingency for equipment, labor, foreign exchange, and project complexity.

Product could benefit from wider geothermal adoption and the new Ormega100 binary unit designed for larger conventional and EGS applications. But the unit is not a moat or forecast until it wins third-party orders at attractive margin. A high-capacity surface unit can expand addressable projects while also inviting competing OEM responses.

5.4 Storage build-out

Ormat targets 950–1,050 MW and 2.5–2.9 GWh of Storage by year-end 2028. It reports 497 MW/1,888 MWh under construction or development and a 2.5 GW/10 GWh pipeline across roughly 25 prospects. The installed/under-construction figures are more decision-useful than the pipeline. The pipeline carries attrition, interconnection, permitting, battery sourcing, contract, and economics risk.

Management’s strategic target is about half contracted and half merchant: California and Texas projects seek tolling or resource-adequacy arrangements, while PJM carries greater merchant exposure. The mix can create upside in scarcity periods but weakens cash-flow duration. A normalized valuation requires revenue and EBITDA by tolling, RA, ancillary, and merchant category and the tenor of each contract; public disclosure is not yet sufficient.

5.5 EGS and data-center optionality

Ormat’s two pilots with SLB and Sage are expected to be only 2–4 MW each. SLB/Desert Peak drilling is planned for late 2026, Sage permitting and procurement continue, and management expects pilot work/evaluation through 2027 with possible connection during 2027–2028. The explicit technical questions are stable water circulation, fracture connectivity, thermal decline, well cost, induced seismicity, and repeatability.

The Google and Switch agreements validate customer willingness to contract firm geothermal. They do not validate EGS. The 150 MW Google-related portfolio PPA uses Ormat’s existing/conventional portfolio; the Switch contract starts in 2030 after a Salt Wells upgrade. The market may connect these contracts to an enormous EGS opportunity, but the commercial evidence belongs to conventional assets. EGS should enter the base case only after pilot flow, cost, and decline data meet predetermined thresholds.

Growth verdict. Recontracting existing geothermal is the highest-quality opportunity; contracted conventional additions are next; Product depends on execution and mix; Storage offers high growth with capital-cycle risk; EGS is an option. The 2.6–2.8 GW target is achievable only with sustained external capital and should be evaluated on per-share cash return, not installed MW.

6. Financial Quality

6.1 Five-year operating and cash record

Revenue compounded 10.5% over four years, but gross profit was essentially flat and operating income ended 2025 almost exactly where it began. Gross margin declined 1,230 basis points and operating margin about 840 basis points. Cumulative operating cash flow of $1.595 billion funded only 59% of $2.709 billion of capex, leaving negative $1.113 billion of conventional free cash flow. These figures come from the five annual SEC statements, not a vendor’s FCF field. 2021 10-K, 2022 10-K, 2023 10-K, 2024 10-K, 2025 10-K

Negative conventional FCF does not automatically mean destruction. Growth projects are long-lived and can rationally be funded with long-duration debt and tax equity. The problem is evidentiary: capex/D&A ran from 1.86× to 2.83× each year, while consolidated operating income did not grow. The company does not disclose mature-cohort post-tax unlevered returns. Investors can see the spending and MW, but not the return bridge that distinguishes investment lag from value destruction.

6.2 Segment quality and mix

The core Electricity gross-profit pool declined from $248.8 million in 2021 to $197.9 million in 2025 despite operating capacity increasing from 1,012 MW to 1,310 MW. Product and Storage supplied the incremental consolidated gross profit. That shift lowers average revenue duration and introduces more tender and merchant exposure. It also means an investor using the historical Electricity-heavy multiple without adjusting for mix is paying the wrong price for the current company.

Storage’s 2025 operating margin of 39.5% is particularly misleading as a steady-state indicator because segment operating expense was negative $2.4 million. Q2 2026 gross margin of 56.2% was driven by favorable PJM prices and should be normalized toward management’s own 30–40% second-half expectation. Product’s large first-half growth includes the TOPP2 transfer; its Q2 margin shows that equipment execution remains volatile.

6.3 Quality of earnings

Adjusted EBITDA increased from $401.4 million in 2021 to $582.0 million in 2025. In 2025, the reconciliation excluded $19.4 million of stock compensation, $12.1 million of impairment, $1.4 million of unsuccessful exploration/storage write-offs, $2.3 million of M&A expense, and other items. This is a useful operating and covenant metric. It is not owner earnings: stock compensation is recurring; unsuccessful projects are inherent to exploration/development; and D&A is added back when annual capex has been roughly twice D&A.

GAAP net income has a different distortion. Income attributable to tax-benefit sales was $61.2 million in 2023, $73.1 million in 2024, and $66.7 million in 2025. The income-tax line was a benefit in 2024 and 2025. Credits are legally valid project economics and should not be deleted wholesale, but their timing, eligibility, transfer discount, partner allocation, and expiration mean that headline P/E does not cleanly describe operating power.

H1 2026 illustrates the normalization problem. Revenue rose 42.9% to $662.7 million and adjusted EBITDA increased 18.9% to $338.8 million. Yet operating cash flow fell 30.1% to $129.2 million and capex was $251.8 million. The period included $105.1 million of TOPP2 Product revenue and a $24.8 million gain, a $9.6 million Hoku bargain-purchase gain, a $34.4 million induced-conversion expense, $8.4 million of asset write-off, $8.7 million of abandoned exploration/storage write-offs, a $4.3 million investment fair-value gain, supplier recovery, and tax-credit benefits. A clean run rate must separate each category rather than labeling the entire reconciliation “one-time.”

6.4 Maintenance capital is unknown

The 2025 10-K budgeted approximately $675 million of 2026 capex and explicitly labeled only $55 million as maintenance. It also included $240 million of construction/enhancements, $170 million of exploration/drilling/development/enhancements not fully released, $10 million for EGS, $180 million for Storage, and $20 million for land, business development, and manufacturing. Historical filings do not report actual maintenance versus growth capex. Major maintenance is also partly expensed.

It would therefore be incorrect to calculate “maintenance FCF” as operating cash flow less $55 million. Geothermal resources cool and pressure declines; wells are added; pumps, gathering systems, and surface equipment require work. The 2025 plant table discloses temperature declines at several fields, and Brawley has suffered repeated wellfield impairments. Economic sustaining capital includes some drilling and enhancements management may classify inside broader development buckets. The unknown is large enough to drive valuation.

6.5 Returns on capital

Using operating income taxed at 21% divided by average parent equity plus NCI plus debt less cash produces approximate normalized ROIC of 4.0%, 3.2%, 3.2%, 3.0%, and 2.7% from 2021 through 2025. Average-equity ROE was approximately 3.4%, 3.5%, 5.9%, 5.2%, and 5.0%. Including tax-benefit financing liabilities and proportionate unconsolidated debt would reduce an all-capital return.

This is the decisive Greenwald check. Scarce assets can exist without a corporate moat if their scarcity rents are consumed by drilling failures, construction, curtailment, overhead, financing, or reinvestment. Returns below a reasonable 7.5–9% cost of capital indicate that the current asset base has not produced consolidated economic profit. The optimistic interpretation is investment lag; the pessimistic one is size creation. Project-cohort disclosure would resolve it.

6.6 Balance sheet and structural claims

At June 30, 2026, Ormat reported approximately $513.7 million of cash, $144.4 million of restricted cash, and $3.412 billion of debt. Net debt was about $2.9 billion before restricted cash and about $2.75 billion including it; management reported roughly 4.3× net debt/adjusted EBITDA using its convention. About 88% of consolidated long-term debt was fixed-rate. Liquidity was approximately $1.1 billion.

Project-finance labels need care. FY2025 consolidated non- or limited-recourse project debt was approximately $739 million. Debt-service tests and reserve accounts can trap subsidiary cash, and completion, capital-infusion, indemnity, or deficiency commitments can create parent exposure. Unconsolidated Sarulla and Ijen debt adds approximately $148 million of proportionate economic exposure not fully visible in consolidated debt. NCI, a tax-benefit-sale liability, and other financing obligations also compete with common equity.

The March 2026 financing issued $825 million of 1.5% Series A notes and $175 million of 0% Series B notes. The initial conversion price is about $140.40; maximum initial potential issuance is roughly 9.26 million shares. Series B can be put back in March 2027. Low coupons and a 2031 stated maturity are attractive, but the financing funded growth and refinanced only part of the old notes; it did not delever the economic enterprise. Convertible financing release

Financial-quality verdict. Contracted assets and credit monetization support cash flow above net income, but core Electricity margins deteriorated, reported operating income was flat over four years, and externally funded capex has not yet demonstrated excess returns. Adjusted EBITDA and GAAP net income each answer only part of the question; the missing statement is project-level cash ROIC.

7. Capital Allocation

7.1 Growth over self-funding

Management has chosen to maximize capacity growth while financing the gap with project debt, tax equity, common equity, convertibles, and asset recycling. This is coherent. Long-lived contracted assets can support leverage, and transferable credits monetize tax attributes Ormat cannot efficiently use immediately. The shareholder test is whether the mature asset return exceeds the all-in capital cost after partner claims.

The present evidence is inconclusive-to-unfavorable. Five-year conventional FCF was negative $1.11 billion, diluted shares increased 8.8% from 2021 to 2025, and debt expanded. A 2023 common issuance of 4.14 million shares at $82.60 raised $341.7 million. The 2026 converts can create substantial dilution if the stock appreciates, while the Series B put creates a nearer liquidity claim. The flat $0.48 annual dividend costs roughly $29 million and is not covered by conventional FCF.

7.2 Acquisition record

The acquisitions share a rational pattern: operating contracted assets plus interconnection and expansion optionality. Yet purchase-accounting goodwill, assumed obligations, and low initial yields mean the return depends on upgrades, recontracting, or development. Blue Mountain is a specific diligence stop: the 2025 acquisition note says its PPA ends in 2029 while the plant table says 2047. No terminal value should use the later date until the conflict is reconciled.

TOPP2 illustrates asset recycling. Ormat constructed and sold the 50 MW New Zealand plant, recognizing $105.1 million of first-half Product revenue and a $24.8 million gain. That is evidence the integrated platform can monetize development. It is not recurring Product demand and should not be capitalized as such.

7.3 Tax equity and credit monetization

Tax-equity investors generally receive most tax attributes; Ormat typically retains most pre-flip cash and 97.5% of post-flip cash under disclosed structures. 2025 tax-monetization cash proceeds were $152 million. The Shirk project used a Morgan Stanley hybrid structure to monetize a 40% ITC. These structures can lower the effective funding cost and are normal in renewable development.

They also complicate “owner” economics. Income from selling tax benefits, transfer proceeds, NCI losses, deferred liabilities, and cash allocations appear in different statement lines. The correct capital-allocation scorecard follows total project construction cash, partner contributions, all cash distributions, debt service, taxes/credits, and residual value. Consolidated adjusted EBITDA alone cannot measure it.

7.4 Incentives and governance

CEO Doron Blachar received $3.62 million of 2025 compensation and CFO Assaf Ginzburg $1.98 million. Annual incentives emphasize revenue and adjusted EBITDA. Long-term equity is 60% performance stock units and 40% RSUs, with PSUs split between relative TSR and MW-capacity growth. The 2023 capacity tranche paid at the 200% maximum after 740 MW of three-year growth. 2026 proxy

The design has safeguards—absolute negative TSR caps, a partly organic-growth requirement, clawbacks, and anti-hedging/pledging rules—but no explicit ROIC, FCF-per-share, or capital-charge metric. It can reward buying/building MW even when per-share economic returns deteriorate. That is the central governance weakness.

Formal governance is otherwise strong: eight independent directors, an independent chair, no CEO board seat, annual majority voting, independent committees, and no reportable related-person transaction since January 2025. Executive/director ownership is below 1%. Say-on-pay support was about 86%.

7.5 Insiders and shareholder returns

The five-year mirror contains 210 original Forms 4 and 20 amendments. Across 615 original transaction rows, there were no P-coded open-market purchases. CEO Blachar sold 14,994 shares under a 10b5-1 plan in September 2025; CFO Ginzburg made exercise-related and other sales, including 17,776 shares near $133–135 in May 2026; independent chair Isaac Angel sold 18,018 shares in November 2025; other executive sales included planned trades. ORIX’s 3.7 million-share secondary was a strategic-holder exit and produced no cash for Ormat.

This is a weak-to-negative alignment signal, not a standalone short thesis. Planned, tax, award, and strategic transactions should not be conflated with discretionary sales. The more important signal is what did not occur: no insider used open-market cash to buy shares during the five-year review.

Capital-allocation verdict. The strategy is coherent and the financing toolkit is sophisticated. The economic outcome remains unproven because consolidated returns have fallen while capacity, debt, and potential dilution grew. Compensation compounds the concern by rewarding scale without an explicit capital-return hurdle.

8. Changes and Headwinds — Last Two Years

8.1 Business mix changed faster than the headline

The company entering H2 2026 is not the 2024 company. Storage scaled rapidly and merchant PJM exposure became a material earnings driver. Product reported a large TOPP2 transfer. Ormat acquired Blue Mountain and Hoku, expanded the development portfolio, refinanced with $1 billion of converts, and advanced EGS partnerships with SLB and Sage. Electricity remained the largest segment but did not produce the growth investors might infer from total MW.

The quality of the 2026 step-up is mixed. Contracted additions and PPA repricing are durable positives. TOPP2’s gain is transactional, Hoku’s bargain gain is accounting, merchant storage prices are cyclical, and the negative tax rate depends on project-credit timing. Management raised guidance, but a dollar of adjusted EBITDA from a long PPA is economically different from a dollar of merchant arbitrage.

8.2 Serial project slippage

Quarterly filings show a repeated, not isolated, delay pattern. Zunil’s 5 MW expansion moved from Q4 2024 to 2025 and then drilling shifted into 2026; construction is complete but the latest filing gives no COD. Bouillante moved from Q4 2025 to Q2 2026 and then Q4 2026. Stillwater’s remaining upgrade shifted to Q1 2027. Salt Wells moved from end-2025 through several 2026 dates. Dominica reached COD in July 2026 only after weather and utility-transmission delays. Bird Dog, Shirk, Arrowleaf, Lower Rio, and Bottleneck also moved.

The interpretation is not that every project fails. It is that drilling, interconnection, permitting, weather, battery supply, and customer transmission create recurrent schedule risk. A valuation that assumes every announced COD and immediate full run rate is structurally optimistic. Pipeline figures require probability and delay haircuts.

8.3 Scarcity improved recontracting

The countervailing change is power scarcity. The Google-related portfolio PPA, Switch agreement, and CD4 amendment show that firm geothermal is receiving better customer attention and, in at least one disclosed case, 27% better pricing. Management expects blend-and-extend agreements to add around $14 million of annual revenue through 2030 with little incremental capital. This is the cleanest positive change because it can raise returns on sunk assets.

Several conditions remain. Company disclosures said the Google/NV Energy portfolio and Jersey Valley agreements required Nevada approval; approval status at the report date remains an open question because no primary docket determination was obtained. The Puna legacy contract remains partly oil-linked until a new regulator-approved structure becomes effective. Blue Mountain’s PPA term is inconsistent within the filing. “Signed” must be separated from approved, in service, and cash generating.

8.4 Funding became longer but more complex

The March converts lowered near-term cash coupon and extended much of the maturity profile. At the same time, potential dilution increased, $175 million can be put in March 2027, and roughly $191 million of old 2027 notes remained after the induced conversion. H1 cash looks strong partly because proceeds arrived before construction spending. Liquidity is adequate; the capital structure is not simpler.

8.5 Regulation created relative advantage and supply risk

The 2025 tax law preserved longer-duration credit eligibility for geothermal and storage while constraining wind/solar. That improves their relative position. Prohibited-foreign-entity rules and tariffs can raise battery cost and reduce suppliers, especially where China dominates the chain. Ormat safe-harbored some equipment, but public disclosure does not quantify how much of the 2028 Storage pipeline is protected.

8.6 One-time-item ledger

Recent statements contain recurring categories of “nonrecurring” items: insurance/supplier recoveries, unsuccessful exploration and storage write-offs, impairments, M&A expense, tax-credit transfer gains, fair-value marks, and financing charges. Brawley was impaired in both 2022 and 2025; abandoned-project write-offs appear repeatedly. These are intermittent rather than exceptional to a development company and should receive a through-cycle allowance.

Change verdict. ORA gained a stronger firm-power narrative, better recontracting evidence, more Storage earnings, and credible EGS partners. It also entered a larger and more complex funding cycle while execution slippage, resource decline, and accounting noise persisted. The last two years improved growth visibility more than they improved proof of economic return.

9. Risk Analysis

KPLC receivables overdue more than 60 days were $37.4 million at June 30; $11.7 million was paid in July. Honduras’s ENEE owed $13.1 million overdue and paid $3.7 million in July. The three largest Q2 customers represented 41.9% of revenue, and primary customers represented 50% of trade receivables. These are current credit exposures, not generic boilerplate.

Resource risk is equally concrete. Brawley recorded a $30.5 million impairment in 2022 and another $7.2 million in 2025. Several plant fields report annual temperature declines; Olkaria required additional wells; Sarulla has experienced resource and equipment failures. Economic maintenance must absorb this reality.

Project debt can protect the parent from some losses but can also block cash upstream and permit foreclosure on pledged assets. Unconsolidated JV leverage adds exposure beyond consolidated debt. Insurance carries deductibles, exclusions, and sublimits, while Ormat does not control all transmission infrastructure.

Risk verdict. The highest-probability risks are ordinary features of the model—geology, delays, concentration, merchant Storage, and funding—not tail events. That makes normalized returns and margin of safety more important than a long list of low-frequency catastrophes.

10. Valuation Discussion — Embedded Expectations

10.1 Live enterprise-value bridge

TTM revenue and EBITDA through June 2026 were approximately $1.188 billion and $505.7 million, producing 7.9× EV/revenue and 18.5× EV/EBITDA. Management’s 2026 adjusted-EBITDA guidance is $630–650 million, including approximately $17 million attributable to NCI. On the consolidated convention, EV/2026E adjusted EBITDA is 14.4–14.9×. If the NCI claim is removed from EV, its EBITDA must also be removed from the denominator; mixing conventions would overstate common-owner value.

TTM P/E is around 50× at the current price, but it is not the primary diagnostic. First-half earnings include a negative effective tax rate, TOPP2, tax-credit income, financing charges, a bargain-purchase gain, write-offs, and fair-value items. EBITDA is cleaner for comparing enterprise claims but ignores the unresolved economic-maintenance requirement. Valuation must triangulate rather than crown one multiple.

10.2 Own history

ROIC.ai’s reconciled annual series shows average EV/TTM EBITDA of approximately 17.6× in 2021, 19.2× in 2022, 17.5× in 2023, 15.7× in 2024, and 17.5× in 2025. The current 18.5× TTM multiple is above the five-year average despite the 29% stock-price correction. The apparent contradiction is explained by higher debt and a TTM EBITDA base that has not yet captured management’s full-year guide.

Quarter-end 2025–2026 multiples show the re-rating more directly: approximately 17.5× in Q2 2025, 18.8× in Q3, 20.8× at year-end, 19.7× in Q1 2026, and 19.3× at June 2026 prices. The current price unwound part of the enthusiasm but did not create an own-history bargain. The AZI valuation_index percentile is unavailable because the authenticated endpoint could not run in this session; no percentile is inferred.

10.3 Imperfect peer boundary

The table does not justify averaging the numbers. ORA deserves a premium to weak-return generic IPPs if its geothermal scarcity and recontracting are durable. It does not deserve NEE’s multiple merely because both are “renewable utilities”; ORA lacks a regulated monopoly. Brookfield’s reported enterprise structure is too NCI-heavy for a clean read. The cross-section says only that ORA currently trades near the upper end of conventional IPP/utility boundaries and below scarcity/platform comparators whose structures are different.

10.4 Greenwald asset value and EPV

Year-end book invested capital—debt plus parent equity plus NCI less unrestricted cash—was approximately $5.64 billion; subtracting $428 million of goodwill/intangibles produces a tangible proxy near $5.21 billion. Current enterprise value is 1.66× book and 1.80× tangible invested capital. Book understates failed exploration, inflation, permits, interconnections, and accumulated know-how, so it is not a liquidation ceiling. It still establishes that the market capitalizes about $3.74 billion of franchise and growth value beyond recorded invested capital.

EPV is unusually sensitive to economic sustaining capex. The sensitivity starts with $550 million of normalized operating EBITDA: the $640 million guide midpoint less approximately $20 million of recurring stock compensation, $10 million of normalized exploration/project failures, and $60 million of tax-benefit-sale income that is real but finite/project-linked rather than perpetual operating earnings. After $70 million of normalized cash tax, three sustaining-capex interpretations produce very different results at an 8.5% discount rate:

These are diagnostics, not point estimates. D&A may overstate economic wear because young assets drive the charge; the company-labeled case likely understates wellfield drilling and enhancements. Even the most generous case is about 47% below the $9.38 billion enterprise value and below book invested capital. Under Greenwald, EPV below asset value means current assets have not demonstrated a company-wide franchise return. The market price requires a future return inflection from growth, recontracting, Storage, and EGS.

10.5 Reverse DCF

The central reverse DCF starts with the $640 million 2026 adjusted-EBITDA midpoint, an 8.5% discount rate, 2.5% terminal growth, and FCFF/EBITDA conversion improving from 20% in 2026 to 55% in 2031. The conversion ramp is already favorable relative to five years of negative conventional FCF; it assumes today’s construction cycle matures.

Solving to the $9.38 billion current EV requires approximately 14.3% annual EBITDA growth through 2031, reaching about $1.25 billion, while cash conversion reaches 55%. At a 7.5% discount rate and 65% terminal conversion, required EBITDA growth falls toward 6%; at an 8.5% rate and only 45% conversion, it rises to about 19%. The current price therefore capitalizes not merely the 2028 capacity target but a successful transition to harvest cash after it.

10.6 Scenario analysis — enterprise value, not a target

The downside assumes project slippage, sub-30% Electricity margin, 25–30% normalized Storage margin, and no commercial EGS. The base assumes released PPAs are largely delivered with ordinary delays, Product stabilizes in the high teens, Storage normalizes to 30–40%, and EGS contributes little before 2030. The upside requires de-risked data-center/EGS projects, better contracted Storage economics, and tax/FEOC compliance without offsetting equipment-cost inflation.

The scenarios are intentionally broad. They are not management forecasts and do not produce a price target in the institutional body. Current EV is above the base result and requires some combination of faster growth, higher mature cash conversion, a lower capital cost, or more EGS/Storage option value. Terminal value supplies most of every scenario, reinforcing the need for humility.

10.7 Directional conclusion without a target

ORA is cheaper than its June narrative peak but not cheap relative to no-growth earnings power, book enterprise capital, or conventional IPPs. The valuation recognizes geothermal scarcity and requires substantial execution of the 2028 plan plus better cash conversion after it. It does not require full DOE 90–300 GW EGS scenarios, but it assigns more than zero value to next-generation positioning and data-center demand.

The best valuation support comes from low-capital PPA repricing and contracted conventional additions. The weakest support comes from capitalizing peak Storage margins or technical EGS potential. A defensible underwriting case should be able to work without commercial EGS and should treat merchant Storage at normalized, not first-half 2026, economics.

11. Variant Perception

Consensus narrative

Ormat is the only vertically integrated pure-play geothermal platform at a moment when utilities and data centers need clean, firm 24/7 power. Conventional resource scarcity, a growing contracted fleet, strong Storage growth, tax-credit durability, and partnerships with SLB/Sage create a long runway. The 2.6–2.8 GW 2028 target and EGS could turn a niche company into a scalable power platform.

Variant

The scarcity is real, but the public financial record has not shown scarcity rents at the consolidated level. From 2021 to 2025, revenue rose 49% while operating income was flat; Electricity gross profit declined; normalized ROIC fell toward 3%; and conventional FCF was negative every year. Storage’s surge is partly merchant and sits inside an extreme supply response. EGS expands technical potential but shifts the bottleneck to subsurface drilling and stimulation, where Ormat relies on partners.

The market may be confusing three different forms of evidence:

1. Conventional PPA evidence — Google, Switch, and CD4 validate demand and pricing for proven geothermal output.
2. Storage cycle evidence — high PJM revenue validates a favorable current market, not a moat.
3. EGS option evidence — partnerships and pilots validate interest, not commercial flow or returns.

Separating them produces a less promotional and more useful thesis. The existing geothermal fleet can be valuable even if EGS fails; EGS can work technically without Ormat capturing outsized returns; Storage can grow MW while reducing portfolio quality. The stock should be underwritten on common-owner cash conversion first and optionality second.

Why the variant could be wrong

Reported ROIC may be depressed by a large immature construction base and development spend ahead of COD. Geothermal asset lives can far exceed accounting schedules, making D&A conservative relative to sustaining capex. PPA repricing could lift returns on old assets with almost no new capital. Ormat’s vertical integration, references, and resource data may create more pricing/cost advantage than segment statements reveal. If EGS pilots succeed, the company may control a scarce surface/operations layer in a much larger market.

Variant scorecard

12. Fact vs. Interpretation

13. Open Questions

1. What are post-tax unlevered IRR and mature-cohort ROIC by conventional Electricity, Storage tolling, merchant Storage, and Product project?
2. How much historical capex was sustaining wellfield drilling, reservoir mitigation, plant overhaul, battery replacement, and true growth?
3. What percentage of 2026–2028 Storage revenue/EBITDA is tolling, resource adequacy, ancillary, and merchant, and what is weighted contract life?
4. How much of first-half PJM Storage revenue reflected unusual scarcity prices versus sustainable capacity/optimization value?
5. What are the precise economics, contract approval, COD, and asset sourcing behind the Google/NV Energy 150 MW portfolio?
6. Which 190 MW of 2031–2034 expiries can plausibly reprice near current indications, and what capital is required?
7. Is Blue Mountain’s PPA expiry 2029 or 2047, and what terminal/recontracting assumption underpinned purchase accounting?
8. What is the current COD, cost to complete, well result, and PPA economics for Zunil after repeated delays?
9. What are project-level covenant headroom and distribution restrictions under non-/limited-recourse debt?
10. How much of Storage equipment was safe-harbored and FEOC compliant, and what cost premium applies to remaining procurement?
11. What EGS pilot milestones define technical success: drilling time/cost, flow, pressure, water loss, temperature decline, seismicity, and months of stable operation?
12. How are EGS economics and IP divided among Ormat, SLB, Sage, land/resource owners, and off-takers?
13. What collections are expected from KPLC and ENEE, and how much support is unconditional?
14. How will Ormat fund the March 2027 Series B put and remaining 2027 notes under a low-stock-price scenario?
15. Will compensation add ROIC, FCF-per-share, or a capital-charge metric rather than relying on EBITDA, revenue, MW, and TSR?

14. What Must Be True

The most important falsification date is not a single quarter. By the end of 2028, much of the currently funded program should be operating and the first EGS pilots should have measurable data. If consolidated and project-cohort returns still do not improve, “investment lag” becomes a weak explanation.

15. Source Appendix (Summary)

The research used a 60-month SEC census and a mirrored primary corpus of 314 documents, reconciled one-for-one to the intended download manifest. It includes five Forms 10-K, fifteen Forms 10-Q, thirty-six Forms 8-K, five definitive proxies, five Forms 3, 210 original Forms 4, and twenty Forms 4/A. All annuals, quarterlies, proxies, material current reports, and the complete mirrored Section 16 corpus were reviewed; 54 indexed Forms 144 were outside the mirror and were not reviewed.

Primary operating sources are Ormat’s 2025 Form 10-K, Q2 2026 Form 10-Q, 2026 proxy, 2021–2024 annual filings, recent 8-Ks, company releases, and Q1/Q2 2026 calls. Industry evidence prioritizes DOE, EIA, USGS, IRENA, Berkeley Lab, FERC, IRS/Treasury, and enacted statutory text.

AZI supplied price history; FactorsToday supplied factor and volatility diagnostics; ROIC.ai supplied transcripts and a valuation/fundamental cross-check. SEC statements supersede vendor fields where they differ. ROIC.ai’s capex field was null and its FCF equaled OCF, so it was rejected for cash-flow analysis. The AZI valuation percentile was unavailable and was not inferred.

Public evidence is limited to the SEC, company-filed exhibits, government and industry primary sources, and disclosed quantitative datasets.

Detailed Public Source Index

Primary — SEC filings

Annual and quarterly anchors

• FY2025 Form 10-K, filed 2026-02-26 — audited financials, segments, assets, PPAs, resources, acquisitions, tax equity, project debt, capex, risks, and 2026 budget.
• FY2024 Form 10-K, filed 2025-02-27 — Enel acquisition, supplier settlement, tax-credit accounting, project status, and 2022–2024 statements.
• FY2023 Form 10-K, filed 2024-02-23 — 2021–2023 financial and segment history.
• FY2022 Form 10-K, filed 2023-02-24 — Brawley impairment, prior operating history, and risks.
• FY2021 Form 10-K, filed 2022-02-25 — Terra-Gen acquisition, historic capacity/product claims, tax equity, and financial base year.
• Q2 2026 Form 10-Q, filed 2026-08-06 — current statements, segments, Hoku/TOPP2, converts, tax benefits, capex, debt, receivables, project status, and risks.
• Q1 2026 Form 10-Q, filed 2026-05-07 — financing, project and run-rate comparison.

The quarterly sequence also includes every Form 10-Q from Q3 2021 through Q3 2025. These were used for project-timeline changes rather than relying on management’s latest schedule alone.

Proxy, governance, and ownership

• 2026 definitive proxy, filed 2026-04-21 — board independence, compensation metrics/outcomes, ownership, say-on-pay, related-party review, and governance protections.
• SEC company filing page — filing index and Section 16 entry point.
• CEO Doron Blachar Form 4, filed 2025-09-30 — plan-based sale example.
• CFO Assaf Ginzburg Form 4, filed 2026-05-15 — recent sale example.
• Chair Isaac Angel Form 4, filed 2025-11-10 — director sale example.
• ORIX Form 4, filed 2024-12-13 — strategic-holder secondary, not management selling.

The complete mirrored Section 16 corpus comprises five Forms 3, 210 original Forms 4, and twenty Forms 4/A. Amendments were inspected but excluded from transaction-row totals to avoid double-counting. The broader SEC index includes 54 Forms 144 that were not mirrored or reviewed; insider conclusions are therefore explicitly limited to Section 16 reports, not the entire proposed-sale tape.

Material event filings and exhibits

• Q2 2026 earnings Form 8-K and SEC-filed Exhibit 99.1, filed 2026-08-05 — reported results, guidance, portfolio, and project highlights.
• FY2025 earnings Form 8-K and SEC-filed Exhibit 99.1, filed 2026-02-25 — FY2025 and initial 2026 guidance.
• Google/NV Energy Form 8-K and SEC-filed Exhibit 99.1, filed 2026-02-17 — up-to-150 MW portfolio PPA announcement.
• Convertible-notes Form 8-K, filed 2026-03-23 — proceeds, rates, maturities, repurchase, conversion, and dilution terms.
• ORIX secondary Form 8-K, filed 2024-12-13 — 3.7 million-share secondary and governance-right termination.
• 2025 board-change Form 8-K, filed 2025-03-13 — board change and size.

SEC-hosted Exhibit 99.1 filings control press-release facts where available.

Primary — government, statutory, and industry

• DOE 2025 U.S. Geothermal Market Report — U.S. installed nameplate capacity, projects, PPAs, and private investment.
• DOE Enhanced Geothermal Systems — hydrothermal/EGS/closed-loop technical distinctions and bottlenecks.
• DOE Geothermal Basics — 90–300 GW technical/deployment scenarios. These are scenarios, not revenue forecasts.
• IRENA Renewable Capacity Statistics 2026, March 2026 — global geothermal capacity series.
• USGS Electric-Grade Geothermal Energy Resources, June 2025 — identified, undiscovered hydrothermal, and EGS technical-resource estimates.
• EIA geothermal use — 2025 U.S. geothermal generation and geographic concentration.
• EIA U.S. battery storage capacity, 2026-08-07 — installed and developer-planned battery capacity. Planned additions are not a forecast.
• Berkeley Lab, Queued Up: 2025 Edition, April 2025 — interconnection-queue capacity with low-completion caveat.
• FERC Order No. 841 — organized-market storage participation rules.
• IRS Section 48E — credit base/bonus rates, eligible storage, and transferability.
• Public Law 119-21, enacted 2025-07-04 — clean-energy credit amendments and phase-down framework.
• Treasury/IRS FEOC guidance, 2026-02-12 — material-assistance guidance.

Primary — peer filings

• NextEra Energy 2025 Form 10-K — battery capacity/additions and large-developer capital-cycle comparison.
• AES 2025 Form 10-K — Clean Energy portfolio, storage construction, and sourcing.
• Fluence Energy 2025 Form 10-K — integrator competition, entrants, and price/total-cost pressure.

Peer facts are segment-specific; regulated-utility economics are not imposed on ORA.

Transcripts and company communications

• Q2 2026 results release, released 2026-08-05 — guidance and reported highlights; SEC Exhibit 99.1 controls reported facts.
• Q1 2026 results release, released 2026-05-06 — initial project/EGS/financing update.
• Switch PPA release, released 2026-01-12.
• CD4 blend-and-extend release, released 2026-03-10.
• SLB–Ormat partnership release, released 2025-10-27.

Q1 and Q2 2026 transcripts were obtained through ROIC.ai. Speaker-attributed guidance, Storage normalization, EGS timing, and target commentary are management hypotheses rather than independent corroboration.

Quantitative third-party sources

• AZI public price CSV, downloaded 2026-08-30 — split/dividend-adjusted price history, OHLC, EMAs, beta, and alpha. Returns, extrema, and drawdowns were recomputed from the local CSV. Verified column order: date/open/close/high/low.
• ROIC.ai, NYSE:ORA, accessed 2026-08-30 — transcript, income statement, balance sheet, ratios, enterprise value, and own-history valuation cross-check. SEC statements control. ROIC.ai’s capex field was null and its FCF therefore equaled OCF; that FCF was rejected.
• FactorsToday ORA loadings, stock information, leaderboard, specific volatility, and related stocks, accessed 2026-08-30 — model date 2026-07-31. Statistical context only.
• FactorsToday methodology — L1 sparsity, factor construction, orthogonalization, and volatility scaling.
• Yahoo Finance ORA statistics — unofficial 2026-08-14 short-interest snapshot used only as positioning color; no exchange-primary file was retrieved.

The AZI authenticated fundamentals endpoint was unavailable; the valuation_index percentile was not obtained and was not inferred. FactorsToday’s six-month annualized return does not reconcile to the raw AZI 126-day return; the memo uses the recomputed AZI value.