CNX Resources (CNX) — Stock Analysis 2026 [3.7]

Analysis as of 6 September 2026. This is a point-in-time snapshot, not an evergreen guide. Fundamentals are from CNX Resources’ fiscal-2025 Annual Report (10-K, year ended 31 December 2025, filed 10 February 2026) and its Q2 2026 results, released 30 July 2026; market data is as of the 4 September 2026 close and will move. Rating: ★★★½, Solid — Overvalued (wide band) on a mid-cycle gas deck → the operating base is the best in the basin and the price is not. Price deck used in the valuation: base Henry Hub US$3.00/MMBtu — the representative trailing average (3-month US$2.94, 6-month US$2.93) snapped to the fixed US$2.00–4.00 natural-gas grid — with every grid price run as a scenario, from US$2.00 to US$4.00; the EIA’s US$3.49 (2027) forecast is a 0%-weight cross-check and no spot deck is carried. 10% discount rate, sensitised 8–12%. Refreshed on each annual report and on material events. For information only, prepared with AI assistance — see the disclaimer at the end.

CNX Resources is the quiet, unfashionable end of Appalachian gas: no dividend, 390 employees, almost no liquids, and a business built on owning four million acres of Pennsylvania, West Virginia, Ohio and Virginia outright and squeezing gas out of them for fifteen cents a thousand cubic feet. The thesis in one line: a low-cost, long-life, self-funded dry-gas producer that has bought back a tenth of its own shares and generated free cash flow for twenty-five consecutive quarters — but which is shrinking, carries the heaviest leverage of its peer set, and is already priced for the mid-cycle gas price it needs. Why look now: the 2026 hedge book is struck at US$2.74/mcf against a market near US$3.50, and that gap closes mechanically in 2027–2028. To screen CNX against every North American upstream name on production, cost, reserve life and free-cash-flow yield, go to Metal Pilot.

1. Snapshot & thesis

CNX Resources Corporation (NYSE: CNX) is an independent natural gas exploration, development and production company headquartered in Canonsburg, Pennsylvania, with an integrated midstream arm. It is a producer/operator by archetype and an energy producer by sector, operating in two reportable segments — Shale (Marcellus and Utica in Pennsylvania, West Virginia and Ohio) and Coalbed Methane (the Pocahontas #3 seam in Virginia) — plus a small Other segment. In FY2025 it produced 1.72 billion cubic feet equivalent per day (bcfe/d), 92% dry natural gas, from 4,488 net producing wells across 3.97 million net acres, of which 99% of the unproved position is held by production. (mcf = thousand cubic feet; mcfe = thousand cubic feet equivalent, with liquids converted at 6 mcf per barrel; bcfe = billion cubic feet equivalent; Tcfe = trillion cubic feet equivalent.)

Figure 1. CNX Resources in numbers

Figure data: CNX Resources FY2025 Form 10-K (production, reserves, costs, debt, buyback); market data and analyst consensus as of the 4 Sep 2026 close. Rating per Section 9; valuation read per Section 7.

Table 1. CNX Resources in numbers

Source: CNX Resources FY2025 Form 10-K for all operating and financial figures; market data, net debt, share count (147.94 m) and analyst consensus as of the 4 Sep 2026 close. EV = market cap + net debt. Free cash flow is operating cash flow (US$1,029 m) less capital expenditures (US$495 m). Leverage uses the company’s trailing-twelve-month adjusted EBITDAX of ~US$1.32 bn at 31 Mar 2026. Listed: Public (NYSE: CNX).

Thesis in brief. Bull: you are buying the lowest-cost operating base in Appalachia — US$0.15/mcfe to lift gas, a 15.4-year reserve life, 99.1% of reserves operated and 99% of the acreage held by production — at roughly 10× its FY2025 free cash flow — though at the base deck only 5.0% of today’s price comes back as free cash flow this year (Section 7.4) — run by a management team that has retired a tenth of the share count in a single year and has US$2.4 billion of buyback authorization left. Bear: it is a pure price-taker on one commodity with no liquids buffer, its volumes are guided down in 2026, it carries the heaviest leverage of the Appalachian peer set, and at US$37.49 the shares trade at 2.5× a net asset value built from the company’s own audited after-tax reserve disclosure, with US$2.37 billion of net debt standing in front of it. What tips it: whether Appalachian gas demand — data-centre load and LNG pull — lifts realized prices enough to justify paying above the reserve value for the inventory behind it, and whether the hedge book that protects the downside also caps the upside being paid for. The full rating and its rationale are in Section 9.

How to read this analysis: here for the verdict? The statcards above and the rating in Section 9 are the whole story. Want the evidence? Read Section 2 (the assets) through Section 7 (the valuation).

2. Assets & operations

Everything about CNX runs off one price. Henry Hub averaged US$2.94/MMBtu over the three months to August 2026, with the U.S. EIA forecasting US$3.49 for 2027 — a market held down by record dry-gas production (111 bcf/d in 2026) even as LNG exports climb to 17.4 bcf/d and power-sector burn reaches a record. For the full picture of how gas is priced, who produces it and how the Appalachian basis discount works, see the Natural Gas — A Complete Market Guide . This section spends its words on the company. For how CNX compares with the other four large US gas producers on one construction — scale, cost, reserve value, balance sheet and the nine-point scorecard — see US Upstream Natural Gas Producers Compared .

2.1 Portfolio overview & map

CNX’s portfolio is not a set of discrete projects but one contiguous Appalachian land position, developed from a shared gathering and water network. What makes it unusual is the tenure: of 3.52 million net unproved acres, 3.48 million — 99% — are held by production or fee, meaning CNX pays nothing to keep them and faces no drilling clock. Leases with expiry inside two years are about 1% of the net unproved position.

Table 2. Asset base by segment, 31 December 2025

Source: CNX Resources FY2025 Form 10-K , Summary of Properties and segment MD&A. WI = working interest. Reserves are net of royalty on an SEC (SPE-PRMS) basis, effective 31 Dec 2025, at an average first-day-of-month Henry Hub price of US$3.387/MMBtu. Margin is realized price less total production costs including depreciation, depletion and amortization. Listed: Public (NYSE: CNX).

The concentration is stark: 91.5% of proved reserves and 94% of production sit in the Shale segment, where CNX holds approximately 557,000 net Marcellus acres and 612,000 net Utica acres (about 341,000 of the Utica coinciding with the Marcellus), plus 52,000 Upper Devonian acres it does not currently drill. The Virginia CBM position — 283,000 net acres, 3,784 net wells — is 6% of volumes and loses money on a fully-loaded basis; it is in the portfolio as the platform for the remediated-mine-gas business (Section 5), not because the gas pays.

2.2 Revenue split by commodity & business line

Two cuts of the same revenue base show why CNX has no shelter from the gas price. By commodity, the company is almost purely dry gas: US$1,736 million of the US$1,914 million of hydrocarbon revenue — 90.7% — came from natural gas, with NGLs at US$168 million (8.8%) and oil/condensate at just US$8 million (0.4%). The liquids uplift is trivial and shrinking: excluding hedges, liquids added only US$0.05/mcfe to the average gas price in 2025, down from US$0.17 in 2024. What little there is prices off crude — CNX realized US$21.30/Bbl on NGLs and US$55.26/Bbl on condensate, so the Oil guide covers the market that sets it. Where Antero Resources books 32% liquids and can lean on propane and ethane when gas is soft, CNX cannot.

Figure 2. FY2025 hydrocarbon revenue by commodity

Figure data: CNX Resources FY2025 Form 10-K ; derived from FY2025 volumes (580.6 bcf gas, 7,907 mbbl NGL, 153 mbbl oil/condensate) and realized prices (US$2.99/mcf gas, US$21.30/Bbl NGL, US$55.26/Bbl oil), which reconcile to the reported US$1,914 m of natural gas, NGLs and oil revenue.

By business line, the picture is a little less monolithic. Beyond the two producing segments, CNX earns a genuine non-hydrocarbon strip: US$78 million from sales of environmental attributes, US$69 million of third-party gathering revenue, US$22 million of excess firm-transportation income and US$15 million of water services — US$184 million in total, about 9% of operating revenue and disproportionately high-margin.

Figure 3. FY2025 operating revenue by business line

Figure data: CNX Resources FY2025 Form 10-K , segment MD&A and Other Revenue and Operating Income. Excludes the US$97 m gain on commodity derivative instruments and US$45 m of purchased gas revenue, which are pass-through or mark-to-market items rather than operating revenue.

Read together: CNX is a dry-gas company with a small, valuable services and attributes tail. That tail is what separates it from a commodity clone — it is the part of the business that does not simply track Henry Hub — but at 9% of revenue it cannot carry a thesis on its own, and environmental-attribute revenue fell 18% in 2025.

2.3 The Shale segment

The Shale segment is CNX. It produced 590.8 bcfe in FY2025, up 15.5% year-on-year, and holds 8,844 bcfe of proved reserves, 73% developed. Volume growth came from the US$518 million acquisition of Apex Energy II, LLC, completed on 27 January 2025, which added a central-Pennsylvania upstream and midstream position, and from the timing of new wells turned in line — offset by normal declines. Realized gas prices rose 52.6% to US$2.93/mcf, but hedges gave back US$0.31/mcf, leaving the segment’s total realization at US$2.70/mcfe, up only 3.1%.

The cost structure is the argument for owning this business. Shale lease operating expense is US$0.12/mcfe, production and ad valorem fees US$0.04, transportation, gathering and compression US$0.54 (low because CNX owns the gathering), and DD&A US$0.83 — total production costs of US$1.53/mcfe against a US$2.70 realization, for a US$1.17/mcfe production margin, up 9.3% year-on-year even in a mediocre price year. Development activity, however, is falling: 18.9 net development wells drilled in 2025, down from 25.7 in 2024 and 30.8 in 2023, with 10.0 net drilled-but-uncompleted wells and 2.0 net completions waiting to be turned in line at year-end. No exploratory wells were drilled in any of the three years.

The key asset-level risk is the one the cost structure cannot fix: Appalachian gas realizes below Henry Hub, and CNX’s average gas price of US$2.99/mcf against a US$3.387/MMBtu SEC benchmark is the size of that discount in a single number.

2.4 The Coalbed Methane segment

The CBM segment produced 37.8 bcf in FY2025, down 3.3% on normal declines, and holds 813 bcfe of proved reserves, 61% developed, from approximately 283,000 net acres over the Pocahontas #3 seam in Virginia. It realized US$3.61/mcf — a premium to Shale, because CBM gas is high-Btu and sells into a tighter local market — but its costs are far higher: lease operating expense of US$0.64/mcf, transportation, gathering and compression of US$1.69/mcf and DD&A of US$1.56, for total production costs of US$4.06/mcf and a negative US$0.45/mcf margin. The segment lost US$17 million before tax in 2025, an improvement on the US$26 million loss in 2024.

CNX drilled no CBM wells in 2025, 2024 or 2023. The reason the segment survives is that it is the platform for remediated mine gas (RMG) capture from active and abandoned coal mines in the region — methane that would otherwise vent — which is the source of the environmental-attribute revenue in Section 5. Read honestly, CBM is a declining, cash-negative gas business attached to a small, higher-margin attributes business. The single asset-level risk is that the attributes market, not the gas, determines whether it is worth keeping.

2.5 Other assets & the development pipeline

Beyond the two producing segments, CNX holds three things worth naming. First, approximately 2,600 miles of owned and operated gathering pipelines plus processing facilities, part of which sit inside CNX Midstream Partners LP (detailed in Section 4.3). Second, an option secured with the Apex transaction granting CNX the right to acquire Utica Shale oil and gas rights beneath the legacy Apex Energy footprint, paid for in three annual instalments of US$16 million from 2026 — a second bench under an already-owned position. Third, a water sourcing, delivery and disposal business serving CNX’s own completions and third parties, which generated US$15 million in 2025.

The “pipeline” in the conventional sense is the undrilled land itself: 710,414 net unproved Shale acres, 1.86 million net CBM acres in other states with no drilling plan, and 946,000 net acres of shallow rights, almost all held by production — against only 26,092 net acres booked as proved undeveloped. That gap is where the long-run optionality lives, and where a valuation must decide what to pay for land that generates nothing today.

2.6 Production, reserves & costs (consolidated)

At the group level CNX produced 629 bcfe in FY2025 — 92% natural gas, 8% liquids; 94% Shale, 6% CBM. Proved reserves stood at 9,662 bcfe (9.66 Tcfe) at 31 December 2025 on an SEC (SPE-PRMS) basis, 89.5% natural gas, 72.2% proved developed and 99.1% operated, up 13.2% from 8,538 bcfe a year earlier — an increase driven substantially by the Apex acquisition rather than the drill bit, since only 18.9 net wells were drilled. The resulting reserve life is 15.4 years at 2025 production rates, against roughly 12.8 years for Antero Resources on its 19.1 Tcfe of proved reserves and 1,497 bcfe of 2026 guidance. Group lifting cost, excluding ad valorem and severance taxes, was US$0.15/mcfe, up from US$0.13 in 2024 and US$0.11 in 2023.

The uncomfortable half of the group profile is the direction of travel. 2026 guidance is 605–620 bcfe — a 1.4% to 3.8% decline on 2025 — on base capital expenditure of US$540–570 million, of which US$390–410 million is drilling and completions. This is a company converting a long-life asset into cash and shares, not one growing volumes.

Figure 4. Production by fiscal year, FY2023–FY2026E

Chart source: CNX Resources FY2025 Form 10-K . FY2025 volume (629 bcfe) is as reported; FY2023 and FY2024 volumes are derived by dividing reported natural gas, NGLs and oil revenue (US$1,302 m and US$1,186 m) by the reported average sales price excluding derivatives (US$2.32 and US$2.15/mcfe). 2026 guidance of 605–620 bcfe (midpoint plotted) per the Q1 2026 results . Total realized price including hedges moved US$2.61 → 2.66 → 2.75/mcfe (Table 4) — that second series is carried in the table rather than overlaid.

2.7 Peer positioning

CNX’s natural comparison set is the four large Appalachian-focused natural gas producers: EQT Corporation (NYSE: EQT), the basin’s largest; Expand Energy Corporation (Nasdaq: EXE), the largest US gas producer by volume; Antero Resources Corporation (NYSE: AR), the liquids-rich Appalachian name; and Range Resources Corporation (NYSE: RRC), the closest analogue in scale and southwest-Pennsylvania geography. Every “vs. peers” claim in this analysis — each scorecard star, the cost read, the valuation multiples in Section 7 — uses that set.

Table 3. Quality-metric peer positioning, mid-2026

Source: company guidance and results as reported — EQT Q2 2026 results (2026 guidance 2,375–2,450 bcfe), Expand Energy Q2 2026 results , Antero Resources FY2025 results and 2026 guidance , Range Resources Q2 2026; CNX per its FY2025 Form 10-K and Q1 2026 guidance. Leverage ratios are on each company’s latest reported basis and are not strictly comparable (GAAP debt/EBITDA vs. net debt/adjusted EBITDAX); “n/d” = not disclosed on a comparable basis in the sources used. Figures are approximate and should be refreshed at publish.

Where CNX sits: the smallest and most levered of the group, with the longest reserve life and the lowest lifting cost, and the only one guiding volumes down. It is roughly a quarter of EQT’s scale and a fifth of Expand’s, with none of Antero’s or Range’s liquids optionality. That is the crux the scorecard has to quantify — genuinely excellent unit economics and asset tenure, attached to a sub-scale, shrinking, single-commodity, levered balance sheet.

3. Financials & balance sheet

FY2025 was CNX’s best year since the 2022 gas spike, and the headline flattered it. Net income was US$633 million (US$3.98 diluted), reversing a US$90 million loss in 2024 — but US$278 million of that swing was an unrealized mark-to-market gain on commodity derivatives, and a further US$97 million came from gains on asset sales. The cash story is cleaner and less dramatic: operating cash flow of US$1,029 million less US$495 million of capital expenditure gave US$534 million of free cash flow, and the company has now generated positive free cash flow for 25 consecutive quarters through Q1 2026.

Table 4. Five-year financial summary

Source: CNX Resources FY2025 Form 10-K for FY2023–FY2025 income statement, cash flow and per-unit figures; stockanalysis.com (S&P Global Market Intelligence, drawn from prior CNX filings) for FY2021–FY2022 revenue, net income, EPS, cash flow and net debt, and for the five-year net-debt series on one consistent basis. “—” = not disclosed on a consistent basis within the FY2025 filing window. Total revenue includes gains and losses on commodity derivative instruments, which is why it swings by more than 170% in a year while realized prices move by cents — the natural gas, NGLs and oil revenue line is the better read of the underlying business. Free cash flow is operating cash flow less capital expenditures. CNX pays no dividend.

The balance sheet is the weak link. CNX ended 2025 with US$2,421 million of debt — US$600 million of 7.25% notes due March 2032, US$500 million of 6.00% notes due January 2029, US$500 million of 7.375% notes due January 2031, US$400 million of CNX Midstream 4.75% notes due April 2030, and US$209 million of 2.25% convertible notes due May 2026 — against just US$13 million of cash. Interest expense was US$171 million, up from US$151 million, and net debt rose because CNX funded the US$518 million Apex acquisition and US$524 million of buybacks in the same year. Liquidity is nevertheless comfortable: a US$1.4 billion senior secured revolver with a US$2.4 billion borrowing base, redetermined semi-annually. The convertible resolved cleanly — on 28 January 2026 CNX elected to settle it in shares rather than cash, removing the 2026 maturity but lifting the share count (up 9.05% year-on-year by mid-2026). Net debt at 31 March 2026 was approximately US$2.53 billion, roughly 1.9× the trailing adjusted EBITDAX of US$1.32 billion — heavier than EQT (0.84× debt/EBITDA), Range (0.73×) or Expand (~0.5×).

Hedge & treasury posture. CNX runs one of the most aggressive multi-year hedge programmes in the basin, using physical supply contracts and financial over-the-counter swaps to manage both outright price and in-basin basis. As of 8 January 2026 the book covered 448.8 bcf of estimated 2026 production at US$2.74/mcf, 379.3 bcf of 2027 at US$3.28/mcf, 186.5 bcf of 2028 at US$3.25/mcf and a nominal 2029 volume. Against the US$3.00 base deck this analysis runs — the representative Henry Hub trailing average — and the EIA’s US$3.67 forecast for 2026, the 2026 book is struck materially below both — the programme cost CNX US$0.31/mcf in 2025 and the mark-to-market on open commodity derivatives at 31 December 2025 was a net liability of US$296 million. This is the honest trade: CNX has bought cash-flow certainty for its capital programme and its buyback, and paid for it with upside. The strikes step up in 2027 and 2028, which is the single most mechanical source of margin improvement in the story.

Capital returns are all buyback, no dividend. CNX repurchased 16,869,709 shares in 2025 at an average price of US$31.00, for US$528 million including excise tax — about 11% of the shares outstanding at the start of the year, and 99% of the year’s free cash flow. As of 31 December 2025, US$428 million of the US$2.9 billion cumulative authorization remained; on 29 January 2026 the board added US$2.0 billion, taking available capacity to approximately US$2.4 billion — roughly half the current market capitalisation.

Figure 5. Free cash flow by fiscal year, FY2021–FY2025

Chart source: Table 4, this analysis; CNX Resources FY2025 Form 10-K and stockanalysis.com for FY2021–FY2022. Total revenue (which swings on derivative marks) and net debt (rising from US$2.27 bn to US$2.60 bn) are read from Table 4 rather than overlaid as additional series.

4. Management, strategy & corporate structure

4.1 Management & governance

CNX is led by President and Chief Executive Officer Alan K. Shepard, promoted from the Chief Financial Officer’s chair with more than twenty years in the energy sector — a finance-trained operator at a moment when the strategy is explicitly about per-share value rather than volume. The board is chaired by Ian McGuire and has eight members — compact for a company of this size — overseeing strategy through standing committees including a dedicated Environmental, Safety and Corporate Responsibility (ESCR) Committee, which governs the environmental-attribute strategy and operating standards. The CEO also serves as the chief operating decision maker for segment reporting, reflecting an integrated view of the upstream and midstream businesses. Governance practice is anchored in a Quality Management System (QMS) standardising health, safety and environmental controls across a workforce of just 390 employees — 44 in midstream, 54 in the Virginia CBM operations — with no collective bargaining agreements.

The governance question a reader should weigh is concentration of judgement: a first-time chief executive, an eight-person board, and 2.78% insider ownership against 98.44% institutional holding. There is no disclosed related-party conflict, but there is also little insider capital standing behind the buyback-led strategy.

4.2 Strategy & capital allocation

The stated strategy is to blend cash flow from current operations with responsible development of the Appalachian resource to create long-term per-share value — and the capital-allocation stack makes that concrete: fund a disciplined US$540–570 million 2026 programme (US$390–410 million of drilling and completions, US$150–160 million of non-D&C), then split the residual between share repurchases and debt reduction. Three named sub-strategies sit on top. First, commercialisation of internally developed proprietary technologies intended to cut both cost and emissions during development — a genuine differentiator in intent, though the 10-K states plainly that there has been no material impact to the financial statements from these activities to date. Second, growing the volume and value of environmental attributes, monetising remediated mine gas through the Pennsylvania Alternative Energy Portfolio Standard and voluntary carbon markets. Third, leveraging owned midstream and water infrastructure to serve both its own development and third parties.

The forward targets are unusually modest and unusually specific: 605–620 bcfe of 2026 production, US$1,265–1,315 million of adjusted EBITDAX (revised down at Q1 2026 from US$1,310–1,360 million), and approximately US$525–550 million of free cash flow, including about US$45 million from expected asset sales and roughly US$20 million from 45Z clean-fuel tax credit sales. This is not a growth plan; it is a cash-conversion plan.

4.3 Ownership & corporate structure

Three structural facts define the corporate entity. The most material is the acquisition of the natural gas upstream and associated midstream business of Apex Energy II, LLC on 27 January 2025 for approximately US$518 million in cash, which expanded the central Pennsylvania footprint and drove the 15.5% rise in Shale volumes, and which carried the option over Utica Shale rights beneath the legacy Apex acreage (three annual payments of US$16 million from 2026). Second, CNX Midstream Partners LP operates part of the gathering network as a separately-financed subsidiary with its own US$600 million revolver and US$400 million of 4.75% senior notes due April 2030, neither guaranteed by CNX — a ring-fence that flatters recourse leverage. Third, CNX separated its coal business in 2017 into what is now Core Natural Resources, Inc. (formerly CONSOL Energy Inc.); the two carry mutual indemnification obligations for certain historical liabilities, including the UMWA 1974 Pension Plan settlement — a residual claim from a business CNX no longer operates.

On the equity side, the structure is straightforward: 142,590,509 shares issued and outstanding at 31 December 2025 (down from 148,879,640 a year earlier on buybacks), no preferred stock outstanding, and the 2.25% convertible senior notes due May 2026 — conversion price US$12.84 — settled in shares in 2026, which is why the count rose 9.05% year-on-year despite the repurchase programme. Short interest stands at 15.83 million shares, 11.19% of shares outstanding — an unusually crowded short for a free-cash-flow-positive producer, and a structural fact worth knowing before reading the price action.

5. ESG & sustainability

CNX’s environmental profile has one genuinely distinctive feature and a set of otherwise conventional programmes. The distinctive feature is remediated mine gas (RMG) capture: CNX holds the right to capture methane from active and abandoned coal mines across its Virginia CBM footprint and, on a more limited basis, in West Virginia, Pennsylvania, Ohio, Illinois, Indiana and New Mexico. That methane would otherwise vent to atmosphere as third-party mining progresses, so capturing it is a direct abatement of a high-global-warming-potential gas — and CNX monetises it through the Pennsylvania Alternative Energy Portfolio Standard (AEPS) programme, other compliance schemes and voluntary offset buyers, generating US$78 million of environmental-attribute revenue in 2025, down from US$95 million in 2024 on lower volumes sold and lower prices.

The governance and social layers are conventional but real. Sustainability is overseen by the board’s Environmental, Safety and Corporate Responsibility Committee, and operating standards run through the Quality Management System, which unifies health, safety, environmental and quality controls with internal and external audits and company-wide stop-work authority for every employee and contractor. Named social programmes include continuing-education assistance, a diversity commitment focused on the Appalachian home region, and emergency-preparedness drills run with local municipalities and responders, reviewed biannually. In April 2026 CNX marked its first full year of “dynamic” ESG reporting, having replaced the conventional annual sustainability report with a continuously-updated format.

The balanced read: the RMG business is a rare case of an oil and gas producer earning money from abatement rather than merely disclosing it, and it is governed at board level — but the attribute revenue fell 18% in a year, the filing itself warns that these markets are “volatile” and carry “significant risk associated with eligibility, qualification and compliance,” and the shift to dynamic reporting makes year-on-year comparison against peers harder, not easier. The 10-K discloses no quantified emissions-intensity or safety-frequency targets, so this analysis does not attribute any to the company.

6. Risks

Table 5. Risk register

Source: CNX Resources FY2025 Form 10-K risk factors, MD&A, segment disclosures and Supplemental Gas Data; net debt per stockanalysis.com at 31 Mar 2026. Likelihood and impact are the author’s assessment.

The through-line: CNX has engineered away almost every operating risk a gas producer can control. It owns its land outright, operates 99.1% of its reserves, gathers its own gas, and lifts it for fifteen cents. What it has not engineered away — and structurally cannot — is one commodity price and one balance sheet. Its biggest single vulnerability is a sustained Henry Hub price below roughly US$3.00, at which point the hedges that currently look like a drag become the only thing protecting the buyback. Its most idiosyncratic exposure is the hedge book itself: having sold 449 bcf of 2026 production at US$2.74 against a US$3.00 base deck, CNX has already given away most of this year’s price recovery. And its most stubborn structural issue is scale — at 621 bcfe against Expand’s 7.5, it has the highest leverage and the least room to absorb a bad year.

Two of those risks compound rather than offset: leverage and price are the same risk seen twice. At the US$3.00 base deck, the forward build carries about US$1.1 billion of EBITDA against US$2.37 billion of net debt — roughly 2.1× — and leaves US$275 million of free cash flow after all capital for the buyback (Section 7.4). One grid step down, at US$2.50, the same build leaves about US$70 million, US$1.39 a share less (Section 7.6), and the discretionary cash that funds the repurchase — the main reason to own the equity — is squeezed first. Nothing in the capital structure breaks: the US$2.4 billion borrowing base and the 2029 maturity wall mean there is no forced event. The thesis simply stops working. Which price regime arrives is a macro question rather than a company one — Commodities Across the Cycle sets out the regimes in which gas leads and lags. And declining production is what makes price risk harder to grow out of: a company adding volumes can offset a weaker price, and CNX has chosen not to.

Figure 6. Risk heat-map

Figure data: this analysis; risks per the register above.

7. Valuation

Valuation as of 6 September 2026, in US dollars. Horizon: spot fair value. Price deck: base Henry Hub US$3.00/MMBtu — the representative trailing average snapped to the fixed US$2.00–4.00 grid (3-month US$2.94, 6-month US$2.93, 12-month US$3.59, all to end-August 2026 on the U.S. EIA monthly spot series, the twelve-month window carrying a January 2026 spike, so the shorter windows are the representative ones) — with every grid price run as a scenario (deep bear US$2.00 / bear US$2.50 / base US$3.00 / bull US$3.50 / deep bull US$4.00); the EIA’s US$3.49 (2027) forecast carried as a 0%-weight cross-check; no spot deck is carried, so the section does not age with the daily quote. Realized gas = Henry Hub less the US$0.397/Mcf Appalachian differential FY2025 booked against the reserve report’s own US$3.387 benchmark. Discount rate 10%, the E&P convention for a single-basin producer and the rate the standardized measure is struck at, sensitised 8–12%. Share price US$37.49 (4 September 2026 close), 147.94 m shares, balance sheet as of 30 June 2026.

CNX is valued on the E&P producer archetype, run as a sum-of-the-parts over three claims — the proved developed reserve base, the proved undeveloped tranche the reserve report funds, and the non-hydrocarbon business the reserve report does not cover at all. The method behind the numbers is set out in the How to Value Commodity Stocks guide; this section applies it to CNX. The headline is a deck-to-value map, not a single number: the blended fair value is US$17.72/share at the US$3.00 base price, US$8.83 at US$2.50 and US$27.03 at US$3.50, and each US$0.50/MMBtu of Henry Hub is worth about US$6.8 of NAV/share (US$13.6 per US$1.00). That figure is already net of the hedge book, which is the most consequential single line in this valuation and the reason the section can be built at all: CNX publishes volumes, strikes and tenor, so the book is re-marked in every column, and it swings from +US$694 m at US$2.00 gas to −US$492 m at US$4.00 — US$8.02 a share across the grid, and roughly a quarter of the company’s raw deck leverage handed back. The tiers frame the structure: the producing base, the non-hydrocarbon business and the whole bridge are worth US$10.61/share, the proved undeveloped tranche US$4.31 more, and the resource tier is empty. The section sets the current US$37.49 price against that map only in §7.6, where the rating and the flip prices are published.

7.1 Method selection

CNX is a producer, so the blend starts from the E&P-producer default set out in the valuation guide linked above (NAV/DCF 45% / EV/EBITDA 30% / a reserve- or flowing-unit read 25%). The third default method is not carried: the guide’s EV/2P anchor is struck on oil-equivalent barrels, and at a 6:1 conversion it implies US$1.67 per mcfe against the US$0.71 per mcfe CNX’s own audited PV-10 carries — 2.4× the company’s own reserve value, which would price the reserve base alone at US$16.1 billion against a US$7.9 billion enterprise value. It is set aside, reported unweighted in §7.5, and its absence logged as a data gap; a gas-basis dollar-per-mcfe anchor would let the method return. The substitute, in the guide’s order, is a cash-flow read built from disclosed lines — P/CF at the archetype’s own 4.5× anchor, rather than a free-cash-flow yield, because the guide publishes a P/CF anchor for this archetype and no yield anchor. That puts two methods in the cash-flow family, so they are held together at the 50% collinear ceiling and the NAV takes the balance: NAV/DCF 50% / EV/EBITDA 30% / P/CF 20%, a stated deviation driven by the substitution and the cap.

Table 6. Valuation method selection

Source: method-to-archetype mapping, anchors and the default weight set per The Commodity Investor, Part 11: How to Value Commodity Stocks , “The archetype is the unit of analysis” and “The valuation toolkit”; the E&P-producer default deviated to 50/30/20 for the reserve-multiple substitution and the collinearity cap, per “Putting it together”. Archetype per Section 1; target multiples derived in §7.3 from the archetype anchors, not from the Section 2.7 peer set, which stays a quality comparator. Input families: intrinsic 50% (single method), cash-flow 50% (two methods, at the collinear ceiling), asset & capacity 0% (anchor not sourceable for gas — see §7.6), transaction 0%.

7.2 Net asset value

Vehicle map. CNX holds its producing assets directly and its midstream inside a wholly-owned partnership, so nothing inside one line can reappear as another. What the map has to separate is not ownership but coverage: the reserve report values the gas and nothing else, and about 9% of operating revenue sits outside it.

Table 7. Vehicle map

Source: this analysis; segments, reserves, the non-hydrocarbon revenue lines, the abandonment and midstream capital inside the reserve report and the convertible settlement per the CNX Resources FY2025 Form 10-K , Item 2 (properties and reserves), the consolidated statements of income, Note 12 (long-term debt) and Note 22 (supplemental gas data).

Tax basis, overhead and abandonment. The reserve base is carried at the SEC after-tax standardized measure, so tax is inside the figure by construction and the balance sheet’s US$857.4 m deferred tax liability sits behind it — neither charged nor credited again. That schedule runs at 25.4% of pre-tax future net cash flows (US$4,191.6 m against US$16,488 m); an incremental dollar of gas price is taxed at the 25.7% combined statutory rate (21.0% federal plus 4.7% state net of federal, per the FY2025 rate reconciliation), which is the rate the deck adjustment uses. Corporate overhead is treated as in rows, and the evidence is arithmetic rather than assertion: the reserve report’s future production cost of US$1.078/mcfe sits above CNX’s FY2025 field cash cost of US$0.814/mcfe and just above the US$1.037/mcfe that field cost carries once selling, general and administrative expense is added — so the report appears to charge overhead already. If that reading is wrong, the NAV is overstated by the capitalised value of that line, US$801 m or US$5.42/share, and the read below is the more generous of the two treatments. Abandonment is inside the rows on the filing’s own words: US$1,017 m undiscounted, US$70 m on a PV-10 basis, against the US$163.1 m carried on the balance sheet; the relative legs in §7.3 still deduct it, because EBITDA does not.

Stage risk (n/a). No asset in the model is pre-production. The proved undeveloped tranche is a booked SEC reserve category and the US$2,221 m of future development costs that convert it are already charged inside the standardized measure, so it takes a 1.00 risk weight and neither the target P/NAV nor the discount rate carries a second charge. What is worth naming instead is the direction of travel: net development wells fell from 30.8 in 2023 to 18.9 in 2025, and only 26,092 net acres of the 3.52 m net unproved position are booked as proved undeveloped.

The per-asset NPV build. Every NPV the model carries is built here first, one block per claim, so the arithmetic arrives before the answer. The two reserve blocks are CNX’s own disclosed figures apportioned on the filing’s own lines and moved to the deck; the non-hydrocarbon block is built from two income-statement lines and one stated life.

Table 8. Per-asset NPV build — base case (US$3.00/MMBtu Henry Hub ≈ US$2.60/Mcf realized, 10%)

Notes to Table 8

1. The filing publishes the standardized measure in total only, and reserve volumes by category, so the after-tax figure is apportioned on the volume split with all of the development capital charged to the undeveloped tranche. The apportionment is exact by construction: 4,149 + 918 = US$5,066 m, the disclosed total. It assumes one discount profile across both tranches; because the undeveloped volumes produce later, the split overstates the undeveloped tier and understates the developed tier, while leaving the total unaffected.
2. Cross-check on the pre-tax step: 13,501 + 2,987 = US$16,488 m, the filing’s own pre-tax future net cash flows.
3. The same ratio implies a flat-equivalent life of 21.0 years for re-discounting — the annuity that reproduces 0.4120 at 10% — and that is the profile the rate rows of Figure 8 move both blocks on. It runs longer than the 15.4-year reserve life because the undeveloped volumes sit behind the developed ones.
4. The deck term: 8,647.6 Bcf of gas reserves (89.5% of 9,662.1 Bcfe) × (1 − 1.63% production, ad valorem and other fees, FY2025’s US$31.2 m against US$1,913.7 m of hydrocarbon revenue) × (1 − 25.7% combined statutory tax) × 0.4120 gives US$2,604 m per US$1.00/MMBtu, or US$0.270 per mcfe of reserves. Liquids are 8% of volume and are held at their FY2025 realisations. This is the one figure the filing lets a reader check directly: CNX publishes the standardized measure and the Henry Hub benchmark for three years, and the 2025-against-2023 pair (US$5,066 m at US$3.387 against US$3,110 m at US$2.637) implies US$2,608 m per US$1.00 — within 0.2% of the model.
5. The gas share of each tranche is not published separately, so the volume split is used; because the developed tranche is marginally gassier than the undeveloped one, this understates the developed sensitivity slightly.
6. Including US$1,017 m of undiscounted plugging and abandonment (US$70 m on a PV-10 basis) and US$157 m of midstream and water capital (US$133 m PV-10), both stated in the filing’s own footnote.
7. Ten years, not the 21-year reserve profile: environmental-attribute revenue depends on the Pennsylvania alternative-energy standard and voluntary markets, and it fell 18% in 2025. At a fifteen-year life the row is worth US$647 m and NAV/share US$15.76 instead of US$14.92; the ten-year figure is the conservative one. As a sanity check, the US$114.5 m pre-tax margin at the archetype’s own 5.0× EV/EBITDA anchor is US$573 m.
8. Unlike its peers CNX faces no drilling clock — 99% of the unproved position is held by production or fee — so the omitted inventory carries no holding cost and no expiry. But with no location count, no type curve and only 26,092 net acres booked as proved undeveloped, the row cannot be built from disclosure and is printed n/d rather than proxied; the company’s investor presentation would close it.

Source: the CNX Resources FY2025 Form 10-K , cited by the filing’s own Item and Note numbers with its printed labels quoted, so every figure can be found by searching the document. The filing reports in thousands; this table prints US$m. Filed marks a figure printed in the filing, Derived the arithmetic of rows above it, Input a parameter of this section. The value column is headed Value rather than US$m because a build that multiplies heterogeneous terms cannot hold one unit; the unit sits in the line item, and only the = rows are US$m. The one relationship the table cannot express is the parametric form the sensitivity grid runs across the deck and the rate: reserves(HH, r) = [5,066 + 2,604 × (HH − 3.387)] × AF(r, 21.0) ÷ AF(10%, 21.0).

Table 9. Per-asset model — base case (US$3.00/MMBtu Henry Hub ≈ US$2.60/Mcf realized, 10%)

Source: this analysis, from the CNX Resources FY2025 Form 10-K (Item 2 reserves and PV-10 reconciliation, Note 22 supplemental gas data, the consolidated statements of income) and the FY2026 guidance reaffirmed with the Q2 2026 results on 30 July 2026. Every NPV in the last column reproduces from its block in Table 8; this table adds the reserve, cost, capital, tax and discounting inputs behind those figures. One discount-rate treatment: both reserve blocks re-discount on the Note 22 timing and the non-hydrocarbon row on its own ten-year annuity, and the grid’s notes say so.

Table 10. NAV build-up and equity bridge (base case — US$3.00/MMBtu Henry Hub, 10%)

Source: this analysis; the derivative, lease, working-capital, goodwill and asset-retirement lines per the CNX Resources FY2025 Form 10-K consolidated balance sheets (pp.70–71), Note 7, Note 9, Note 12 and Note 19; net debt, share count and market data per stockanalysis.com , 4 Sep 2026 close, on the 30 Jun 2026 balance sheet. Bridge lines in the standard order, each printed even where empty on one of the five value-column states. The tiers sum to the published NAV/share: producing US$10.61 + development US$4.31 + resource US$0.00 = US$14.92 on unrounded inputs — and the producing tier alone sits 72% below the US$37.49 price, before the undeveloped tranche is counted. Values computed on unrounded inputs, printed to whole US$m and two decimals per share.

Figure 7. Sum-of-the-parts NAV build-up

Figure data: Table 10. Equity net asset value of US$2,207 m equates to US$14.92 per share; the producing tier alone is US$10.61. The hedge bar is the re-marked book at the base deck and is the one bridge line that moves with the price — see Table 11.

Figure 8. NAV/share sensitivity — Henry Hub price × discount rate

Notes to Figure 8

1. Checksum — the bear column (US$2.50) at the 10% base rate: proved developed = 4,149 + 1,879 × (2.50 − 3.387) = US$2,482 m; proved undeveloped = 918 + 725 × (2.50 − 3.387) = US$275 m; + the non-hydrocarbon business US$523 m = US$3,279 m enterprise; − 2,374 + 397 (the hedge book re-marked at US$2.50) − 101 = US$1,202 m ÷ 147.94 m = US$8.12.
2. Rate rows — they move both reserve blocks on the 21.0-year flat-equivalent profile the standardized measure’s own discount ratio implies (×1.186 at 8%, ×0.858 at 12%) and the non-hydrocarbon row on its ten-year annuity; net debt, working capital and the hedge book are held down each column. One cell prints < 0: at US$2.00 gas and a 12% rate the bridge returns negative equity of −US$0.20/share. It is the only such cell, so the equity is not treated as an option, but it marks how thin the cushion is.
3. Cost — a +10% shock to the US$0.764/mcfe of lease operating and gathering cost (US$47 m/yr) takes NAV/share to US$12.89 (−13.6%); a +10% Henry Hub move to US$3.30 lifts it to US$19.00 (+27.3%), and with a 5% cost lag applied to that price move, US$17.98 (+20.5%).
4. FX — n/a, CNX reports and trades in US dollars.
5. Stage risk — n/a, no asset in the model is pre-production; the proved undeveloped tranche is a booked SEC category at a 1.00 weight (§7.2).
6. Schedule slip — n/a, no development asset stands outside the reserve report’s own five-year schedule, and that schedule’s capital is already inside the standardized measure.

Figure data: this analysis’ model (Tables 8–10), every cell recomputed at that column’s Henry Hub price and that row’s rate, never scaled from the base cell. Price columns are the fixed natural-gas grid, grid version 2026-09 (US$2.00–4.00); base case US$3.00 at 10%. A one-step (US$0.50) Henry Hub move shifts NAV/share by US$6.80, or ~46%; the deck sensitivity is tabulated in Table 11.

Deck sensitivity — what one US$0.50 step of Henry Hub is worth. The grid above holds the recomputed values at each grid price; this table names the slope between them. The line to read first is the hedge book: it moves against the deck, by US$296 m a step, and that is what turns a raw reserve sensitivity of US$8.80 a share into the US$6.80 the NAV actually shows.

Table 11. Deck sensitivity — value per US$0.50/MMBtu step of Henry Hub (US$/share unless stated; base rate, target multiples held)

Source: this analysis, Tables 8–10 and 18. % of base is each line’s per-step move divided by its own base-price value — a leverage read; the hedge row carries no percentage because its base value crosses zero inside the grid. Linear over is the Henry Hub range on which the slope holds: ¹ the hedge book is linear because the disclosed instruments are swaps, and it crosses zero at about US$3.17; ² below ~US$2.34 the cash-tax line reaches zero and both cash-flow slopes flatten, and FCF/share crosses zero at ~US$2.33; ³ the ladder blend steps 6.25 → 8.89 → 9.31 → 10.70 because the multiples move ×0.10 per step with the deck. Every step is the difference between two recomputed grid prices of Figure 8, never a scaled figure. Forward EBITDA moves US$277 m per step (563.5 Bcf of gas volume, net of production taxes) and cash flow per share US$1.39. How to use it: start from the base-price values (NAV/share US$14.92, blended fair value US$17.72) and add or subtract the per-step figure for every US$0.50 of Henry Hub away from US$3.00 — a US$3.25 flat deck gives a NAV/share of ~US$18.3 and a held-multiple blend of ~US$20.9; for a reading that also lets the multiples move with the cycle, use the ladder columns of Table 18.

P/NAV ladder (unweighted). The NAV restated as a price map, straight off Figure 8’s base-rate row: for each of the five fixed P/NAV levels this series uses for producers and E&Ps, the share price it implies at every grid price — NAV/share at the deck × level, the base rate held. Read down a column for the deck you hold, across a row for the multiple you would pay; where today’s quote sits on the map is said once, by the market-implied deck in §7.5.

Table 12. P/NAV ladder — share price implied by each P/NAV level at each grid price (US$/share)

Source: this analysis, solved on Tables 8–10: each cell is the Figure 8 base-rate NAV/share at that column’s Henry Hub price (1.32 / 8.12 / 14.92 / 21.72 / 28.52) × the row’s P/NAV level. The levels are fixed for the series (0.50× to 1.50× in quarter steps), so two producers can be read column-for-column; CNX’s 0.82× target, derived in §7.3, reads US$12.23 at the base price, US$6.66 at US$2.50 and US$17.81 at US$3.50, between the 0.75× and 1.00× levels. Unweighted: it translates a multiple and a deck into a share price without reference to today’s quote — parity at the base price is US$14.92, and the whole map tops out at US$42.77, which is the sharpest statement in this section of what the reserve base alone can support.

7.3 Relative valuation

At US$37.49 and 147.94 million shares, CNX’s market capitalisation is ~US$5.55 billion and enterprise value ~US$7.92 billion. This section values CNX standalone: each target multiple is the fixed anchor for the E&P-producer archetype moved by the signed drivers this post’s own scorecard has already scored. No peer multiples are tabulated here — reading CNX against EQT, Expand, Antero and Range on observed reserve and cash-flow multiples is the job of the sector comparison , on one shared basis. Forward metrics are struck on the FY2026 guidance year at the base deck. The US$3.00 base sits 20% below Henry Hub’s five-year average of US$3.74 (September 2021–August 2026) — inside the ±25% band that marks a cycle extreme — so the subject is treated as near mid-cycle, and the scenarios in §7.6 flex both the deck and the multiples along the ladder.

Table 13. Target-multiple driver line (one line, applied to every multiple)

Source: this analysis; each term is tied to one scored dimension, capped at ±10%, and no fact is charged under two labels. The driver set is the standard one — asset quality, cost position, reserves and replacement, balance sheet and capital allocation; management, jurisdiction and ESG (Dims 7–9) sit outside it and are scored in Section 9. The line very nearly cancels, which is the correct read of a company that owns the best cost structure and the longest life in its peer group and carries the worst scale and the worst balance sheet: the target lands within two points of the archetype anchor. The one line is applied, unchanged, to every anchor:

Target P/NAV = 0.80× anchor × 1.02 = 0.816 → 0.82× · Target EV/EBITDA = 5.0× anchor × 1.02 = 5.100 → 5.1× · Target P/CF = 4.5× anchor × 1.02 = 4.590 → 4.6×. Rounded figures are the ones used in every table below.

Table 14. Forward EBITDA build — FY2026 guidance year at the base deck

Source: this analysis; volumes, unit costs and the capital outlook per the CNX Resources FY2025 Form 10-K and the FY2026 guidance reaffirmed with the Q2 2026 results on 30 July 2026. “Forward” is the next twelve months = the FY2026 guidance year; the volume ties to guidance with nothing added above it. Reconciliation: run at the reserve report’s own US$3.387 benchmark this build gives US$1,325 m, against the company’s own FY2026 adjusted EBITDAX guidance of US$1,270–1,320 m — 0.4% above the top of the range, on a measure that is unhedged where the guidance is hedged, so the build reproduces the company’s own economics. Cost-basis note: the NAV rows use the reserve report’s own US$1.078/mcfe of future production cost; this build states the cash lines separately — a definition, not a gap.

Table 15. Relative valuation — implied value per share (base case)

Source: this analysis; archetype anchors (E&P: P/NAV 0.80×, EV/EBITDA 5.0×) per the valuation guide linked in §7, “The valuation toolkit”. The implied EV crosses the same claims as the NAV bridge — plus the asset-retirement obligation, which the reserve report carries inside its development costs but EBITDA does not — and omits only the corporate overhead, which is deducted inside the EBITDA build. Values computed on unrounded inputs. To run the same reserve and cash-flow multiples across every North American upstream name, screen the sector on Metal Pilot.

The two reads part company by US$8.91, and the multiple is the higher one — the opposite of a long-life producer and the same shape Expand shows. The reason is arithmetic rather than opinion: a 5.1× multiple on one guidance year prices about six years of cash flow, while the reserve report runs 9,662 Bcfe through a 21-year discount profile and then charges US$2.2 billion of development capital to get the undeveloped quarter out of the ground. The reserve base itself is the thinnest per unit in this series — US$0.71 of PV-10 per mcfe — and behind it stands US$2.37 billion of net debt, which is why the NAV leg falls away so much faster than the multiple does.

7.4 Further weighted methods — P/CF support

The third weighted read is a cash-flow multiple on CNX’s forward cash flow, built from disclosed lines and taken at the archetype’s own P/CF anchor.

Table 16. P/CF support build — FY2026 guidance year at the base deck

Source: this analysis; interest, income taxes paid and depreciation per the CNX Resources FY2025 Form 10-K consolidated statements of income and Note 6; capital guidance per the Q2 2026 results. Every trailing line sits on FY2025, the latest reported fiscal year. The company’s own FY2026 free-cash-flow guidance of ~US$525 m is struck on its hedged book and its own cash-tax position; this build is unhedged at a US$3.00 deck with a statutory tax charge, and the gap between the two is almost exactly the hedge gain plus the tax the loss position defers.

The method lands at US$26.31, the highest of the three — and that ordering is the finding. A cash-flow multiple sees a company with the lowest lifting cost in Appalachia throwing off US$846 million a year; the NAV sees the same company with US$2.37 billion of debt in front of a reserve base worth US$0.71 per mcfe and shrinking. Both cash-flow reads are two views of one signal rather than two confirmations, which is what the collinear cap exists to contain — the family carries 50% between them, not 50% each.

7.5 Cross-checks (unweighted)

Nine diagnostics locate the blend; none carries weight.

Table 17. Cross-checks — reported, reconciled, never weighted

Source: this analysis; the market-implied and flip prices solved on the Tables 8–16 model; reserve, PV-10, hedge, acquisition and drilling figures per the CNX Resources FY2025 Form 10-K (Item 2, Notes 12, 19 and 22); Henry Hub history per the U.S. EIA monthly series, five-year window September 2021–August 2026; the trailing multiple and book-value series, consensus and analyst count per stockanalysis.com , read 6 Sep 2026 on the 4 Sep close. CNX pays no dividend, so no yield-support price is computable.

7.6 Scenarios & fair value

Every weighted method is re-run in every column of the Henry Hub grid — the same five grid prices as Figure 8 and Table 11, so the whole section reads on one price axis. Each column is its own world, and Table 18 lists what changes in it above the values it produces: the deck moves one step at a time; because the base sits inside 25% of gas’s five-year average (§7.3) the subject is near mid-cycle, so the three target multiples step ×0.80 / ×0.90 / — / ×1.10 / ×1.20 with the deck; the discount rate steps out to 12% and 14% on the downside and holds at the 10% convention on the upside. The hedge book is re-marked in every column from its disclosed strikes, which is why the deep-bear column is less severe than the reserve arithmetic alone would make it. The last memo row is the same blend with the multiples held at their targets, which is the linear version a reader can reproduce from Table 11.

Table 18. Scenarios & fair value — inputs, value per method and the blend by grid price (US$/share)

Source: this analysis; weights per §7.1, scenario names by offset from the base price. Base blend on a calculator: 0.50 × 12.23 + 0.30 × 21.14 + 0.20 × 26.31 = 6.12 + 6.34 + 5.26 = US$17.72 (on unrounded values, 17.719). Inputs behind the rows, by column: the flexed targets 0.656× · 4.08× · 3.68× / 0.738× · 4.59× · 4.14× / 0.82× · 5.1× · 4.6× / 0.902× · 5.61× · 5.06× / 0.984× · 6.12× · 5.52×; forward EBITDA US$556 m / 834 m / 1,111 m / 1,388 m / 1,665 m; cash flow per share US$2.61 / 4.33 / 5.72 / 7.11 / 8.50; cash tax US$0 m / 23 m / 94 m / 165 m / 237 m — it reaches zero below US$2.34; the hedge book re-marked at +US$694 m / +US$397 m / +US$101 m / −US$195 m / −US$492 m from the disclosed strikes. Risk weights are 1.00 in every column — no tranche in the model is pre-production. The deep-bear NAV leg is floored at 0.00: at US$2.00 gas and a 14% rate the bridge returns negative equity (−US$1.44/share before the P/NAV), and a negative value is never blended. Illustrative scenarios, not forecasts.

Figure 9. Value per share by method and scenario

Source: Table 18; each cell recomputed at its column’s deck, rate, hedge mark and multiple flex, never scaled; data-level ranked 0–9 across the whole grid. The methods keep their order in every column — P/CF highest, then the EBITDA multiple, then the NAV — and the fan widens on the downside rather than the upside, because the NAV is the only method carrying US$2.37 bn of net debt and the reserve report’s development capital. Current share price US$37.49 (4 Sep 2026); market-implied deck ~US$4.56/MMBtu. The bracketed figure under each value is its change against the same row’s base-case value.

Conclusion. The blended base-case fair value is US$17.72, inside a US$2.58 (Deep Bear, US$2.00) – US$37.73 (Deep Bull, US$4.00) range, against a US$37.49 price — an implied −52.7%, Overvalued, published as Overvalued “(wide band)” because the deep-bear blend sits 93% below the price. At the base price the guidance-year free cash flow of US$275 m is a 5.0% yield on the US$5.55 bn market capitalisation — against the company’s own ~US$525 m guidance, which is struck on the hedged book and the cash-tax position the loss balances defer. Rating-flip prices: with the multiples held at their targets, the base blend crosses up into Modestly overvalued above ~US$3.67/MMBtu Henry Hub (+22% from the base price) and into Fairly valued above ~US$4.26 (+42%); Overvalued is the bottom band, so no downward flip exists. The one assumption that drives the downside is Henry Hub at the US$2.00–2.50 grid prices with US$2.37 billion of net debt in front of a production base guided to shrink — though the hedge book, worth +US$694 m at US$2.00, is real protection in exactly that world.

The three methods spread by US$14.08 and the NAV is the lowest of them, which is the reverse of Range and the same shape as Expand. It is not a disagreement to average away: a 4.6× multiple on one year of cash flow values the lowest lifting cost in Appalachia and stops there, while the NAV carries the audited reserve value — US$0.71 of PV-10 per mcfe, the thinnest in this series — through a 21-year discount profile, charges US$2.2 billion of development capital, and then subtracts US$2.37 billion of net debt. One judgement is worth naming because it runs the other way: corporate overhead is treated as already inside the reserve report’s costs on the evidence of that report’s own US$1.078/mcfe, and if that is wrong the NAV is US$5.42 a share lower still — so the read below is the more generous of the two available treatments and would only harden. The framing that matters is the balance sheet: the producing base, the attributes business and the whole bridge are worth US$10.61/share and the booked undeveloped tranche US$4.31 more, and a buyer at US$37.49 is paying a US$22 premium to the sum of the two for Henry Hub at about US$4.56 held flat. What the company does well is not in doubt — twenty-five straight quarters of free cash flow, an eleventh of the share count retired in a year, and gas lifted for fifteen cents. What the price asks is that the gas those assets produce be worth roughly half as much again as the audited report says it is. The Street, at a Hold and a target equal to the share price, sees no gap in either direction; this section, on the reserve value, sees the largest one in the series. This is an analytical read of price against value, not a recommendation.

Assumptions box: valuation date 6 September 2026; balance sheet as of 30 June 2026 for net debt and shares, 31 December 2025 for the derivative, working-capital, lease, goodwill and asset-retirement lines (the latest in the source set); horizon spot fair value. USD throughout — trading currency = model currency, no FX. Price decks: base Henry Hub US$3.00/MMBtu (the 3- and 6-month trailing averages, US$2.94 and US$2.93, snapped to the fixed grid, version 2026-09), run across the US$2.00–4.00 grid; the EIA’s US$3.49 (2027) forecast as a 0% cross-check — no spot deck is carried; constant-price, unescalated deck and costs, matching the SEC reserve convention, so the 10% rate is real. Discount rate 10% — the E&P convention for a single-basin producer, and the rate the standardized measure is struck at — sensitised 8–12% on both reserve blocks; no jurisdiction premium (Dim 8 ★★★★★, 100% United States, so the band is +0%). Share basis 147.94 m (basic ≈ diluted, after the January 2026 convertible settlement in shares); values per share to two decimals, multiples to two significant figures, on unrounded inputs. Cycle: base 20% below the five-year average of US$3.74 (Sep 2021–Aug 2026), inside ±25%, so deck and multiples flex together on the ladder; anchors E&P P/NAV 0.80×, EV/EBITDA 5.0×, P/CF 4.5× per the valuation guide linked in §7, one driver line (×1.02); metric basis forward FY2026 (guidance year); EBITDA before all capital and after SG&A, unhedged; net debt includes finance leases and excludes operating leases; P/NAV form equity (market cap ÷ equity NAV); no peer multiples enter this section. Method weights NAV 50% / EV/EBITDA 30% / P/CF 20% — the E&P default deviated as §7.1 states. NAV provenance: CNX’s own disclosed after-tax standardized measure, apportioned to the proved developed and undeveloped tranches on the filing’s own inflow, production-cost and development-cost lines and moved to the deck on a term built from the filing’s own gas volumes, production-tax ratio, tax reconciliation and discount ratio — calibrated to within 0.2% of the company’s own 2023-to-2025 standardized-measure history; the non-hydrocarbon business at its disclosed margin over a ten-year life; tax basis the SEC schedule (basis 3); asset-retirement obligation and, on the evidence of the report’s own cost level, corporate overhead inside the reserve report’s own costs. Primary value yardstick P/NAV (equity form). Stage-risk placement n/a — no pre-production asset; every row at a 1.00 weight. Known data gaps: (1) whether the reserve report’s future production costs include corporate overhead is not stated — the US$1.078/mcfe cost level is the evidence for treating it as in rows; if it does not, NAV is overstated by US$801 m (US$5.42/share), and a segment-level cost breakdown in the reserve disclosure would close it; (2) drilling inventory beyond proved reserves n/d — 3.52 m net unproved acres, 99% held by production, but no location count or type curve; NAV understated, closed by the company’s investor presentation; (3) the gas share of each reserve tranche n/d — the volume split is used, which slightly understates the developed tranche’s price sensitivity; (4) the maintenance/growth capital split n/d — the whole US$571 m programme is treated as maintenance, which understates free cash flow before growth capital; (5) a gas-basis EV/reserve anchor n/d — the asset-family method is not carried (§7.1). Only the first would change the rating, and it would harden it. To run the same net asset value and multiples across every North American upstream name, screen the sector on Metal Pilot.

8. Near-term catalysts (1–3 years)

CNX’s forward upside over the next two to three years is unusually mechanical — it comes from contracts repricing and shares being retired, not from new production. That is both the strength and the limitation of the story.

Table 19. Near-term catalysts (1–3 years)

Source: CNX Resources FY2025 Form 10-K (hedge book, buyback authorization, Apex Utica option, debt maturities, the non-hydrocarbon revenue lines, New Technologies), the Q2 2026 results of 30 July 2026 (2026 guidance and 45Z credits) and the U.S. EIA Short-Term Energy Outlook . Timing reflects company guidance and is not guaranteed.

The common thread is that none of these require CNX to spend more money. The hedges reprice on their own; the buyback is funded from free cash flow; the demand growth is someone else’s capital expenditure. The swing factor is not execution — it is whether Henry Hub cooperates while the hedges roll.

9. Rating & verdict

CNX is scored on the Metal Pilot Company Scorecard — the same nine dimensions, on the same ★1–5 scale, that every North American upstream name in the series is scored on, so the peers are directly comparable. Each star is relative to the Appalachian gas peer set declared in Section 2.7 and substantiated below.

Table 20. The CNX Resources scorecard

Source: the Metal Pilot Company Scorecard; evidence in Sections 2–7. Peer basis: the four large Appalachian-focused natural gas producers named in Section 2.7 (EQT, Expand Energy, Antero Resources, Range Resources).

Weighted average = (0.45 + 0.60 + 0.75 + 0.125 + 0.45 + 0.60 + 0.1875 + 0.3125 + 0.25) = 3.73/5 → rounds to the published ★★★½, Solid. Weights follow the archetype-weighted scheme for the producer/operator archetype (dimensions 1/2/3/5/6 at 15% each, dimensions 4/7/8/9 at 6.25% each) per Table 2 of the Metal Pilot Company Scorecard playbook.

The two-axis verdict. Quality Solid (★★★½) × Value Overvalued (wide band) → the assets are better than the equity, and the price now assumes the gas is worth half as much again as the reserve report says. The quality axis is durable and genuinely bifurcated: three dimensions score at or near the top of the basin — reserve life, jurisdiction and cost position — while growth scores ★★☆☆☆ and scale, leverage and governance sit at the median. That is an unusual shape. It says CNX owns something very good and is not, at present, doing very much with it beyond converting it into cash and shares.

The value axis is the dated layer, and it is why the verdict is no longer neutral. At US$37.49 the shares sit at 2.5× a US$14.92 net asset value built from the company’s own after-tax standardized measure at a US$3.00 deck, and the blended fair value of US$17.72 implies −52.7% (Section 7.6). Two things drive that. The reserve base is the thinnest per unit in the peer set — US$0.71 of PV-10 per mcfe — and US$2.37 billion of net debt stands in front of it, which is why a P/B of 1.15× flatters: book value is 2.2× what the audited reserves are worth. What tips the verdict from bear to bull is the hedge roll: 449 bcf sold at US$2.74 for 2026 becomes 379 bcf at US$3.28 for 2027, and if Henry Hub holds near the EIA’s US$3.49 forecast, that repricing lifts realized margin without CNX drilling a single additional well. What tips it the other way is gas below US$3.00, where the same net debt in front of a shrinking production base stops being a detail — though the hedge book is worth +US$694 million in that world, and is the reason the downside is survivable. This is an analytical read of quality and price, not a recommendation.

To go from this single-name view to the whole peer group — screening every North American upstream producer on production, cost, reserve life, net debt and free-cash-flow yield — explore Metal Pilot.

10. Sources, methodology & disclaimer

10.1 Sources, methodology & data vintage

Company fundamentals, structure, management, hedge posture, reserves, acreage, segment economics and risk factors are from CNX Resources Corporation — 10-K Filing / Annual Report — 2025 (fiscal year ended 31 December 2025, filed 10 February 2026, audited by Ernst & Young LLP), including its Summary of Properties, Detail of Operations, segment MD&A, Note 5 (Stock Repurchase), Note 12 (Long-Term Debt), Note 19 (Derivative Instruments) and Note 22 (Supplemental Gas Data, unaudited). Forward guidance is from the company’s Q1 2026 results (30 April 2026). Reserve figures are SEC (SPE-PRMS) proved reserves net of royalty, effective 31 December 2025, at an average first-day-of-the-month Henry Hub price of US$3.387/MMBtu; PV-10 is the company’s disclosed non-GAAP pre-tax measure and the standardized measure is the GAAP after-tax equivalent.

Market data (share price US$37.49 at the 4 September 2026 close, 147.94 million shares outstanding after the January 2026 convertible settlement in shares, market cap ~US$5.55 billion, net debt ~US$2.37 billion at 30 June 2026, short interest 10.02% of shares outstanding, and the five-year P/OCF and P/B series used as cross-checks in Section 7.5) and the 12-analyst Hold consensus with its US$37.73 target are from stockanalysis.com , sourced from S&P Global Market Intelligence, read on 6 September 2026. FY2021–FY2022 financial history falls outside the FY2025 filing window and is drawn from the same provider on prior CNX filings, marked as such in Table 4. The Henry Hub trailing averages behind the price deck — 3-month US$2.94, 6-month US$2.93, 12-month US$3.59, and the five-year average of US$3.74 for September 2021–August 2026 — are from the U.S. EIA monthly spot series; the 2027 forecast used as a 0%-weight cross-check is from the U.S. EIA Short-Term Energy Outlook . Peer operating figures are from each company’s own results releases as cited in Table 3.

Methodology and its limits. The net asset value starts from the company’s disclosed after-tax standardized measure of US$5,066 million — not the pre-tax PV-10 — apportions it to the proved developed and proved undeveloped tranches on the filing’s own inflow, production-cost and development-cost lines, moves it to the base deck on a term built from the filing’s own gas volumes, production-tax ratio, tax reconciliation and discount ratio, and then bridges to equity. That deck term reconciles to within 0.2% of CNX’s own 2023-to-2025 standardized-measure history, and the forward EBITDA build run at the reserve report’s own benchmark lands within 0.4% of the top of the company’s FY2026 adjusted-EBITDAX guidance. Two lines are author constructions and both are printed as terms: the non-hydrocarbon business (environmental attributes, third-party gathering, firm transport and water) at its disclosed margin over a stated ten-year life, and the hedge book, re-marked in every price column from the filing’s own disclosed volumes, strikes and tenor. Abandonment and, on the evidence of the reserve report’s own US$1.078/mcfe cost level, corporate overhead are treated as already inside the report’s costs; if overhead is not, the net asset value is US$5.42 a share lower, and that judgement is logged in Section 7.6. The drilling inventory beyond proved reserves is carried at n/d rather than proxied — 3.52 million net unproved acres with no location count or type curve — and the company’s investor presentation is the document that would close it. The asset-map figure is omitted — a proportional-symbol map of net acreage by state would not render legibly at this scale, so Table 2 and the §2.1 concentration prose carry the by-state footprint instead; every published figure is an inline HTML/CSS component. 9 September 2026 — vintage alignment of Sections 1, 3 and 6 to the September market data and the US$3.00 base deck; no valuation figure, star or rating moved. The trailing free-cash-flow multiple in the thesis is restruck at the 4 September price (about 10× FY2025 free cash flow, against 5.0% of the price returned as free cash flow at the base deck this year); the hedge paragraph reads against the base deck rather than a spot quote; and the risk section’s price-stress read is rebuilt on the base deck, the forward EBITDA build and the 30 June net debt. Data as of 6 September 2026; refreshed on each annual report and on material events. Provenance: CNX Resources Corporation — 10-K Filing — 2025.

10.2 Disclaimer & disclosure

This analysis is for informational purposes only and is not investment advice; do your own research or consult a licensed advisor. It is a point-in-time snapshot as of 6 September 2026 — share prices, multiples, analyst targets and the valuation read move, and reserve, production and net-asset-value figures are estimates as of the stated dates. The two-axis verdict is an analytical read of quality and price, not a personal buy or sell instruction. This report was prepared with AI assistance; figures were sourced from CNX Resources’ filings, the U.S. EIA and market data and reviewed, but readers should verify before acting. The author holds no position in CNX Resources as of the date of writing.