Summary

Ormat Technologies is the world's largest geothermal power operator (1,340MW, ≈8% global share) and largest binary ORC equipment manufacturer. The company owns 35 power plant complexes across the US, Kenya, Indonesia, Guatemala, Honduras, and Guadeloupe, sells electricity under long-term PPAs averaging 14 years remaining, and is expanding into grid-scale battery storage.

The investment narrative centers on PPA recontracting: ≈340MW of legacy contracts priced at $60-80/MWh expire between 2026-2033, with replacement PPAs now pricing above $100/MWh. New hyperscaler demand (Google 150MW, Switch 13MW) validates geothermal as the only firm, 24/7, zero-carbon baseload technology — a unique characteristic in a grid increasingly dominated by intermittent renewables.

The financial reality is more complex. Three years of flat GAAP EPS ($2.02-2.08) despite 11% annual revenue growth, negative cumulative free cash flow of $953M, interest expense consuming 84% of operating income, and a negative effective tax rate (-19.2%) that accounts for approximately 70% of reported net income. The company survives its growth phase on tax credit monetization and project finance — both of which function as designed, but create fragility if either disrupts. The stock has beaten consensus EPS in four consecutive quarters, but this reflects low-bar management of estimates, not earnings acceleration.

This memo examines each dimension of the business — operations, financials, competitive dynamics, management, factor exposures, and implied expectations — sourced from the 10-K filed February 26, 2026, Q4 2025 earnings transcript, counterparty filings, and market data. What follows is factual assessment, not recommendation.

1. Business Overview

Three Segments, One Core

Electricity (70% of revenue, $694M): The engine. Geothermal plants sell power under fixed-price PPAs to utilities and, increasingly, to data center operators. Capacity factor of 84% — geothermal runs 24/7, unlike wind (25-35%) or solar (20-30%). Revenue has two components: energy payments (per MWh delivered) and capacity payments (fixed for availability). The fleet is predominantly US-based (72% of segment revenue), with Kenya's Olkaria III complex (150MW) as the largest international asset.

International operations punch above their weight: 28% of Electricity revenue but 39% of gross profit and an estimated 49% of segment net income. Lower-cost operations abroad make Kenya disproportionately important to profitability.

Top three customers — SCPPA (17.8% of total revenue), NV Energy (13.8%), and KPLC-Kenya (11.9%) — account for 43.5% of revenue. SCPPA is a municipal authority with long-dated contracts through 2043-2051. NV Energy is a Berkshire Hathaway subsidiary. KPLC is a Kenyan government utility with deteriorating payment patterns (discussed below).

Product (22%, $217M): Designs, manufactures, and sells binary ORC power plant equipment. Revenue recognized on percentage-of-completion, making it lumpy and estimate-dependent. Backlog of $352M includes the ≈$100M Topp2 New Zealand project recognizing in Q1 2026. Manufacturing is in Yavne, Israel. Revenue is 95% international. Gross margins improved to 21.2% from 18.4% but competition is explicitly intensifying (per 10-K: "increase in competition... has started to affect our ability to secure new purchase orders").

Energy Storage (8%, $79M): Grid-scale BESS (415MW/1,038MWh) across California, Texas, PJM, and Hawaii. Fastest-growing segment (+109% YoY) with margins inflecting from 6.4% (2023) to 36.4% (2025), driven by PJM pricing and scale. Storage margin inflection from single-digit to 36% is underweighted by most models — if sustainable, this becomes material to consolidated margins as the segment scales. Pipeline of 410MW/1,540MWh under development. Operates primarily in merchant markets — revenues fluctuate with energy pricing.

Unit Economics

Geothermal development is a 4-7 year cycle: explore ($10-50M), develop and construct ($150-400M), then operate for 30+ years on contracted revenue. Most resource rights are secured through BLM leases on federal land (78% of domestic acreage). Geothermal resources decline 1-5 degrees F per year, requiring periodic make-up well drilling. The 10-K discloses specific decline rates for each plant — a useful indicator of ongoing capex needs.

New PPAs are signing at $100+/MWh versus legacy fleet pricing of $60-80/MWh — a 25-67% step-up. This is the central repricing thesis. The Google 150MW PPA (15-year term, deliveries 2028-2030) is the marquee hyperscaler contract — at an assumed $100+/MWh and 84% capacity factor, full delivery implies ≈$110M+ in annual revenue, or roughly 11% of current consolidated revenue. The Switch 13MW PPA (20-year term) is smaller but confirms the hyperscaler demand channel. Blend-and-extend discussions are underway on ≈40MW of existing capacity.

2. Financial Profile

The Earnings Stagnation Problem

Revenue has grown 11% annually over three years ($720M to $990M). Net income has not moved: $124M in 2023, $124M in 2024, $124M in 2025. EPS has actually declined slightly: $2.08, $2.04, $2.02. The revenue-to-earnings disconnect is explained by three forces:

Interest expense growth. Net interest grew from $99M (2022) to $142M (2025) — a 43% increase in three years — on $2.86B of total debt at 4.8% average cost. Interest consumes 84% of operating income ($142M / $169M). This ratio overstates cash-flow stress because $66.7M in tax benefit transaction income sits below the operating income line — effectively, the company earns $236M including tax monetization, against which $142M interest is 60%. Still elevated, but the tax benefit income is the load-bearing offset. At 6% cost of debt, interest would consume 100% of operating income or 73% of operating income + tax benefit income.

Electricity margin compression. The core segment's gross margin fell from 36.6% (2023) to 28.5% (2025), a 810bps decline. Drivers: curtailments costing $18.6M in 2025 (grid maintenance at McGinness Hills and other Nevada plants), Puna wellfield issues (-$13.9M), rising D&A from new plant investments (+$20M), and Nevada property tax increases (+$8.3M). D&A and property taxes are permanent. Curtailments are guided to moderate ($4-6M in 2026).

Revenue mix shift. Electricity's share fell from 80% to 70% as Product and Storage grew. Since Electricity carries the highest margins historically, mix shift alone pressures consolidated profitability.

Tax Credits: The Load-Bearing Wall

Pre-tax income in FY2025 was $105.7M (10-K, line item "Income from operations before income tax and equity in earnings of investees"). Reported net income was $123.9M. The difference — and then some — comes from a -19.2% effective tax rate, driven by ITC/PTC monetization. Tax benefit transaction income was $66.7M in FY2025. Combined with direct tax provision benefits ($20.3M), tax-related items contributed roughly $87M — approximately 70% of net income.

The OBBBA (enacted July 4, 2025, replacing the IRA) extends 100% PTC/ITC eligibility for geothermal and storage projects starting construction by December 2033, with phase-down through 2035. Ormat has 7 active tax monetization partnerships (including two hybrid tax equity structures with Morgan Stanley totaling $62M). Tax credit transferability allows direct third-party sales without requiring tax equity.

The tax runway is secure through 2033. But the structure creates a fragile dependency: at a 0% effective tax rate (still favorable), GAAP EPS drops to ≈$1.72 and the P/E expands to 60x. At +15%, EPS falls to ≈$1.46 and P/E reaches 71x.

Free Cash Flow: Persistently Negative

Cumulative free cash flow from 2022-2025: negative $953M. Operating cash flow of $335M in 2025 was consumed by $620M in capex. The gap is funded by project finance, tax equity, and green bonds — financing cash flow was +$466M in 2025. Management guides $675M capex for 2026, with the Topp2 asset sale (≈$100M) as partial offset.

Maintenance capex is approximately $55M annually. Growth capex runs $520-620M. This is a capital-intensive development business that cannot self-fund its growth program from operations.

Balance Sheet

Total debt of $2.86B against $582M EBITDA yields 4.4x net debt/EBITDA (ex restricted cash). $304M in principal repayments are due in 2026. Available liquidity of $680M provides buffer, but the debt stack requires ongoing refinancing access. Construction-in-process for exploration and development projects ballooned to $286.9M from $193.7M (+48% YoY) — representing capitalized costs that convert to impairment charges if projects fail commercial viability.

3. Competitive Position

The Moat Is Real — In Conventional Geothermal

Ormat's competitive advantages in conventional geothermal are genuine and difficult to replicate:

Resource rights. 434,430 acres of domestic geothermal leases, 78% on federal BLM land. These are exclusive, long-term resource rights on proven geothermal areas. No competitor can access the same subsurface resources.

Vertical integration. Ormat is the only company that explores, develops, manufactures equipment, constructs, and operates. Competitors either develop projects (no equipment capability) or sell equipment (no operating track record). This integration reduces execution risk and provides a cost advantage.

Subsurface knowledge. Decades of proprietary drilling data, temperature decline curves, and reservoir management expertise across 35 operating complexes. Each geothermal reservoir is unique geology. This institutional knowledge is irreplaceable.

Baseload positioning. Geothermal is the only firm, 24/7, zero-carbon generation technology. Hyperscalers seeking round-the-clock clean energy have limited alternatives — nuclear (long lead times, regulatory complexity), gas + CCS (not zero-carbon), or overbuilt solar + storage (expensive for true baseload). The Google PPA validates this unique positioning.

In the US, Ormat has no public competitor of comparable scale in conventional geothermal. The closest was Cyrq Energy, which Ormat acquired in 2025.

The Threat at the Margins

EGS competition. Enhanced Geothermal Systems represent a potential 100x expansion of the addressable market — geothermal deployable anywhere, not just near tectonic boundaries. The emerging competitive landscape:

• Fervo Energy (private): Devon Energy owns ≈15%, Baker Hughes supplying 5 ORC plants for 300MW Cape Station. Uses horizontal drilling adapted from oil & gas. The most advanced EGS developer.
• Baker Hughes is explicitly positioning as a surface power generation competitor to Ormat, providing "comprehensive subsurface-to-surface solutions." Five ORC plants for Fervo represent a direct assault on Ormat's equipment market.
• SLB (Schlumberger) has a JV with Ormat for EGS development. CEO Le Peuch (Q4 2025 transcript): "Put together two leading companies... Ormat leaders into geothermal power plants understanding full life cycle."
• Oil & gas supply chain broadly entering: Helmerich & Payne (geothermal rigs), Vallourec (insulated tubing for closed-loop), DynaEnergetics (perforating charges for EGS wells), Devon (investor in Fervo).

Ormat's hedge: $25M investment in Sage Geosystems (Series B), SLB JV, and a key executive hire — Daniel Moelk, formerly EVP European Operations at Eavor Technologies (an EGS competitor), now running Ormat's drilling and EGS operations. Smart strategic response, but EGS is a newly added risk factor in the 2025 10-K, with management explicitly acknowledging "substantial technical, operational, and geological uncertainties."

Product segment erosion. Competition is intensifying from Baker Hughes, Mitsubishi/Turboden, Chinese manufacturers (Kaishan, TICA/Exergy), and Turkish players (Egesim/Atlas Copco). The 10-K states this "has started to affect our ability to secure new purchase orders."

Energy Storage: no moat. Highly competitive market against NextEra, AES, Fluence, Vistra, and dozens of developers. Ormat has no differentiated advantage. This segment is a capital allocation decision, not a competitive franchise.

4. Management & Governance

Insider Activity

Zero open market purchases (Form 4 code P) in the last 12 months. The selling pattern:

• November 7, 2025: Coordinated same-day sales by CFO Ginzburg ($2.25M at $112.73), Director Angel ($2.03M), Director Barniv ($197K), Officer Benyosef ($58K). Stock has since fallen 8.0%.
• September 2025: Director sales (Stern $60K, Sharir $95K). CEO Blachar and General Counsel Woelfel exercised RSU conversions (not open market purchases).

Insider ownership is below 1%. Management's financial exposure comes primarily through equity grants, not personal capital at risk.

ORIX Exit

ORIX Corporation acquired 22.1% in 2017 for ≈$627M as a strategic partner. In December 2024, ORIX sold 3.7M shares at $76.20 ($282M). They retain 12.9% (≈7.24M shares, ≈$750M at current prices). ORIX transcripts (Q1-Q2 FY2026) book "gains on sale of Ormat shares" as profit contributions and are simultaneously exiting other renewable positions (Greenko Energy). This is programmatic portfolio rotation, not Ormat-specific distress — but the remaining 12.9% creates supply overhang.

Leadership Transition

Shimon Hatzir, a 36-year Ormat veteran who led the Electricity segment, Resource/Drilling, R&D, and Energy Storage, retired in June 2025. His role was split between two external hires: Aron Willis (EVP Electricity, from TransAlta) and Daniel Moelk (SVP Resources/Drilling/EGS, from Eavor Technologies). The Moelk hire from a direct EGS competitor signals strategic commitment to next-generation geothermal. The loss of Hatzir's institutional knowledge in a subsurface-expertise-driven business is a legitimate risk.

Compensation and Capital Allocation

CEO total comp is ≈$3.8M (60% PSUs, 40% RSUs). PSU targets include MW capacity additions — incentivizing growth over profitability. The board awarded 93% bonus payout in a year of flat EPS and margin compression.

Capital allocation priorities in practice: (1) growth capex $520-620M, (2) debt service $304M, (3) interest $142M, (4) maintenance $55M, (5) dividends $29M, (6) buybacks $0. No buybacks since a $18M program in FY2022, including during the $64 low. Management issues convertible debt and sells stock personally.

Stewardship grade: B-. Tax monetization execution is excellent ($180M collected vs $160M guided). EGS hedging strategy is smart. Growth program is thesis-aligned. But four years of negative FCF, 84% interest consumption, and aggressive insider selling undermine near-term confidence. They're building the right thing; the financial bridge is shaky.

5. Factor Profile

The Standard Model Lies

Using the standard factor model (SPY + XLU + MTUM), ORA appears to have 81% idiosyncratic variance — comfortably above the 75% target. Alpha looks like 27.9% annualized. This is misleading.

ORA does not trade like a utility. Correlation with XLU is only 0.37 — barely higher than with SPY (0.38). Correlation with ICLN (iShares Global Clean Energy ETF) is 0.58 — the highest of any factor tested.

The Correct Model

Adding ICLN to the regression changes the picture materially:

R-squared rises from 19% to 37%. Alpha drops from 27.9% to not statistically distinguishable from zero (point estimate 0.0% annualized, wide confidence interval — this means the model can't detect alpha, not that alpha doesn't exist). Idiosyncratic variance drops from 81% to 63% — below the 75% target.

ICLN is the dominant factor. 22% of ORA's daily return variance is explained by clean energy sentiment alone. This loading is persistent: rolling 60-day ICLN beta has never fallen below 0.14 over two years, averaging 0.52 and currently elevated at 0.69.

Meanwhile, XLU beta has collapsed from an average of 0.34 to a current 0.10. The market no longer treats ORA as a utility.

Rolling Analysis (2Y, 60-Day Windows)

True idiosyncratic variance averages 55.6% over two years (54.6% over the last six months), with a range of 38-80%. Rolling alpha averages +0.2% annualized. The 1Y return of +49.3% is predominantly explained by ICLN's +67% move (at ≈0.5-0.7 beta), not stock-specific alpha.

Implications

ORA is a borderline factor bet. Approximately 37% of return variance comes from systematic factors — primarily clean energy sentiment — in which a company-specific thesis provides limited edge. The PPA repricing story is genuine idiosyncratic alpha, but it lives within ≈63% of the variance, with the remainder driven by macro clean energy flows.

The recent 1M drawdown (-19%) confirms this classification: XLU was +10.2% over the same period (utilities rallied), while ICLN was -4.3% (clean energy sold off). ORA's excess decline versus ICLN reflects idiosyncratic selling pressure — insider sales, ORIX overhang, earnings disappointment — compounding the clean energy factor drawdown.

Technical note: RSI is 24.5 — deeply oversold. The last time ORA traded at this RSI level was at the $64 trough. This doesn't change the fundamental picture but contextualizes the timing: any catalyst (PPA announcement, curtailment normalization, storage margin beat) would be hitting a technically compressed base.

6. Forward Expectations Gap

What the Price Requires

At $103.70, ORA trades at 51.3x GAAP diluted EPS ($2.02) or 47.6x TTM adjusted EPS ($2.18, per yfinance). Either way, the stock prices one of two things: a sustained premium multiple (35-40x forward) with consensus earnings, or earnings acceleration well above consensus. Working backward through terminal multiple analysis (3Y horizon, 8% discount rate):

Current price works at 40x terminal P/E with consensus estimates. Any P/E contraction requires earnings growth that consensus does not project.

Consensus Estimates

Eight analysts cover ORA. FY2026E: $1,133M revenue (+14.5%), $2.39 EPS (+9.6%). FY2027E: $1,165M revenue (+2.9%), $2.69 EPS (+12.5%). The FY2027 EPS range is wide ($2.38-$3.09), reflecting low conviction on outyear trajectory.

Q1 2026 is anomalous: $345M revenue (+50% YoY) driven by Topp2 NZ product delivery, not recurring growth. Q2 normalizes to +7.3%.

Management guides FY2026 segment revenue of $715-730M Electricity, $300-320M Product, and $95-110M Storage — summing to $1,110-1,160M — with adjusted EBITDA of $615-645M (+8.2% at midpoint). No explicit consolidated revenue guidance was provided; the segment sum implies ≈$1,110-1,160M before intercompany eliminations. Consensus revenue of $1,133M sits at the upper end of consolidated guidance. No formal EPS guidance is provided. Ormat has beaten consensus EPS in each of the last four quarters.

Six Specific Gaps

Gap 1: Revenue cliff in FY2027. Growth collapses from +14.5% to +2.9% after Product backlog clears. Electricity grew only +3.3% organically in FY2025 excluding acquisitions. The recontracting wave adds revenue gradually over 7 years, not in a step function. Consensus doesn't address where FY2028+ growth comes from. Potential offsets: new greenfield commissioning (first in ≈7 years expected from the development pipeline), the Google 150MW PPA ramping 2028-2030, and Indonesia's Telaga Ranu (up to 182MW in pipeline). But all are 2028+ deliveries — the FY2027 gap is real.

Gap 2: Margin expansion required but headwinds mount. Consensus needs +70bps net margin in FY2026 and flat margins in FY2027 to deliver +12.5% EPS growth on only +2.9% revenue. Our research found active compression drivers: resource degradation at three complexes (Olkaria, Sarulla, Brawley), BLM solar permit freeze blocking auxiliary cost reduction, and interest expense still growing. PPA repricing provides a tailwind but materializes gradually.

Gap 3: Tax credit dependency. Tax-related items contribute ≈70% of net income. Pre-tax income is $105.7M. At a 0% effective tax rate (still favorable), GAAP EPS drops to ≈$1.72 and P/E expands to 60x. OBBBA provides runway through 2033, but the structure is fragile to legislative change, FEOC restrictions, or credit transfer market pricing. Additionally, Pillar 2 global minimum tax now applies ($1.9M impact in 2025, potentially growing).

Gap 4: PPA repricing magnitude. The full 340MW recontracting wave at +$30/MWh yields +$76M revenue and +$0.44 EPS over 7 years. That is approximately $0.06 per year in EPS accretion. Meaningful in aggregate but incremental annually. Analyst models from Baird and Oppenheimer incorporate some of this; it is partially priced in the $126.80 mean price target.

Gap 5: KPLC risk is unmodeled downside. Kenya Power (11.9% of revenue, disproportionately higher share of net income from lowest-cost international operations) had $29.5M overdue as of December 31, 2025. The 10-K discloses KPLC "recently requested more favorable rates on its existing PPAs." If Kenya PPAs reprice downward, it directly contradicts the US PPA repricing bull case. No sell-side model incorporates Kenya rate renegotiation risk. Partial mitigant: Olkaria PPAs include capacity payments (fixed for availability regardless of dispatch), which provides a floor — but energy payments, the larger component, remain at risk in renegotiation.

Gap 6: Interest expense as constraint. At current 4.8% cost of debt, interest consumes 81% of operating income. At 5.5%, consumption rises to 93%. At 6.0%, interest exceeds operating income. Debt maturities in 2027-2028 introduce refinancing risk. Every dollar of debt-funded growth capex widens this ratio.

Peer Context

ORA trades at 47.6x trailing P/E and 18.9x EV/EBITDA. NextEra Energy (NEE), the premier US clean energy utility, trades at 28.4x P/E with 11.2% ROE and 2.66% dividend yield. AES trades at 11.4x. The ICLN clean energy ETF trades at 21.6x. ORA commands a 70% premium to NEE on less than half the ROE.

7. Key Risks

Geothermal resource degradation. Three complexes disclosed wellfield problems in FY2025: Olkaria (Kenya, reduced generation from well field performance decline), Sarulla (Indonesia, wellfield issues and equipment failures reducing profitability), and Brawley (California, non-cash impairment from continuous wellfield issues). Simultaneous problems at three sites may indicate growth-over-maintenance tension in capital allocation. Resource decline is existential for geothermal operators — make-up well drilling costs are real and ongoing.

Balance sheet fragility. $2.86B debt, 4.4x net debt/EBITDA, $953M cumulative negative FCF over four years, 84% interest consumption of operating income. The growth program is entirely debt-funded. $304M principal due in 2026, with additional maturities in 2027-2028. Capital access is not optional — it is operational.

Tax credit concentration. ≈70% of net income derives from tax-related items. OBBBA provides structural runway through 2033, but political modification, FEOC supply chain restrictions, or credit transfer market disruption would immediately impair earnings.

Kenya sovereign exposure. KPLC represents 11.9% of revenue with $29.5M overdue receivables and active rate renegotiation requests. Honduras (ENEE) has $20.3M overdue with only $1.0M collected in January-February 2026. Combined international receivables overdue: $49.8M, or 5% of annual revenue.

BLM solar permit freeze. BLM has not issued new or renewed permits for solar projects on federal lands. This blocks Ormat's hybrid solar strategy to reduce auxiliary power costs at existing geothermal plants. Most of Ormat's Nevada fleet sits on BLM land. The 10-K warns of "material delays in development timelines, increased project costs, loss of site control."

Capitalized exploration costs. CIP for exploration and development reached $286.9M (+48% YoY from $193.7M). These are capitalized costs that become impairment charges if projects fail commercial viability assessment. A 20% failure rate implies ≈$57M in potential write-downs, equivalent to 26% of annual net income.

ORIX overhang. 12.9% ownership (≈$750M at current price) with demonstrated willingness to sell. ORIX is programmatically exiting renewable positions.

Clean energy factor exposure. 22% of return variance is explained by ICLN (clean energy ETF), not company-specific factors. Any reversal in clean energy sentiment — policy change, rate shock, tariff escalation — flows through to ORA regardless of fundamentals.

8. What to Watch

Galena 1 PPA recontracting (2026). The first live test of the repricing thesis. This Nevada Power Company contract expires this year. Terms of replacement PPA will validate or weaken the $100+/MWh recontracting narrative. Monitor for 8-K filing.

Q1 2026 earnings (May 6). $345M consensus revenue includes ≈$100M Topp2 product delivery. The quality question: is Electricity segment organic growth accelerating or still stalling? Strip out Product lumpiness and Storage merchant volatility to see the core business.

KPLC rate negotiation outcome. Any 10-Q or 8-K disclosing Kenya PPA amendment terms. Downward repricing in Kenya would be an anti-thesis data point — the highest-margin customer moving pricing in the opposite direction of the US.

Curtailment trajectory. Management guided $4-6M for 2026 versus $18.6M in 2025. If curtailments normalize, Electricity margins should recover 150-200bps. If grid maintenance issues persist, the margin compression thesis strengthens.

ORIX secondary activity. Form 4 filings or S-3 shelf registrations indicating further liquidation of the 12.9% stake.

September 2026 Analyst Day. Management has flagged longer-term targets for this event. The 2028 capacity target of 2.6-2.8 GW and any FY2027-2028 EBITDA framework will set the reference for multiple justification.

EGS milestones. Sage Geosystems commercial progress, SLB JV pilot results, and Fervo Energy Cape Station (300MW) operational data. If EGS demonstrates commercial viability, Ormat's positioning as equipment/EPC supplier and JV partner becomes significantly more valuable. If EGS stalls, the $25M Sage investment and associated R&D spend become impairment candidates.

Debt refinancing (2027-2028). Terms on upcoming maturities will determine whether the interest consumption ratio improves or deteriorates. The difference between 4.5% and 5.5% cost of debt is approximately $29M annually in interest — equivalent to the entire dividend payment.

Sources

• ORA 10-K filed 2026-02-26 (FY2025), SEC/EDGAR
• ORA Q4 2025 Earnings Transcript (2026-02-26)
• SLB Q4 2025 Earnings Transcript (2026-01-23)
• ORIX Corporation (IX) Q1-Q2 FY2026 Earnings Transcripts
• yfinance market data, factor analysis, analyst estimates (2026-02-28)
• Cross-corpus transcript search: Fervo/EGS mentions in DVN, BKR, BOOM, VLOWY, HP transcripts
• ORA 10-K competitive landscape disclosures (Item 1, pp. 35-37)
• ORA 10-K risk factors (Item 1A, pp. 46-67)
• ORA 10-K critical accounting estimates (MD&A, pp. 78-79)