VS. NOTHINGBY HERITAGE HOLDINGS

HOUSE UNDERWRITING · WORKING PAPER · AUGUST 14, 2026

SHADOW MODE · NOT A LIVE CALL

RRC clears the first screen.
Now attack the assumptions.

Chronometer Partners argues that committed LNG exports and gas-fired power demand will outrun deliverable U.S. supply after 2028. HOUSE does not need certainty to identify opportunity. At $39.91, RRC provisionally clears the proposed return and downside rails; EXE needs a lower entry.

THESIS ORIGIN CHRONOMETER PARTNERS · “I GOT GAS” · JUNE 22, 2026 + JULY UPDATE TEST THE CLAIMS
PHYSICAL THESIS · CREDIBLE, UNPROVEN HOUSE POSITION · NONE YET BENCHMARK · XOP UNCHANGED CLOCK · NOT STARTED
CURRENT HOUSE ACTIONADVANCE RRC

Build the full Call packet for RRC. The rough four-year cases are $31 bear, $68 base, and $112 bull against a $39.91 close. EXE remains a watch near $64. CRK stays outside the core book until it funds drilling internally and reduces financing risk.

WHY SHADOW MODE?

The source's decisive asset-by-asset inventory model is proprietary. No HOUSE basis, weights, or result are sealed. This page records the strongest case, the public evidence, the disagreement, and the work required before a Call can begin.

01

INVESTMENT CONCLUSION FIRST

Plausible shortage. Unpriced edge.

OBSERVATIONCONDITIONAL PASS

LNG exports and power burn are rising into a system with slow infrastructure. Exhausted storage by 2030 is not independently verified.

SECURITY SELECTIONNARROW

RRC and EXE deserve full work. CRK is a higher-leverage satellite. Midstream, nuclear, solar, and shorts belong in separate Calls.

VALUATIONRRC · PROVISIONAL PASS

The first explicit cases clear the proposed 11% base return and 25% downside rails. The delivered-price and inventory model still needs full replication.

HOUSE POSITIONNONE YET

Advance RRC to Call review. Do not convert a provisional screen into a position or force EXE and CRK into a basket.

THE CLAIM WE MAY EVENTUALLY LOCK

From a future seal date, an equal-weight basket of eligible low-cost gas producers with durable inventory and egress should outperform XOP held unchanged over 48 months because delivered gas tightens from 2028 while those companies retain open price exposure and capital discipline.

THE THESIS IS NOTAmerica is running out of gas.
THE THESIS ISEconomic inventory plus transport may not flex quickly enough to satisfy committed demand.

ATTRIBUTION RAIL. Chronometer supplied the physical-system model, shortage forecast, and candidate map. HOUSE sets the public-evidence standard, valuation work, eligibility rule, benchmark, horizon, and challenges. The source letter is paraphrased here and is not republished.

02

HOW THE THESIS MAKES MONEY

Follow delivered gas into free cash flow.

01 · LOADDemand starts

Financed LNG trains and gas-fired power plants pull more molecules from the same physical network.

02 · FLEXSupply responds slowly

Declines, acreage quality, capital, gathering, processing, and pipeline capacity limit deliverable growth.

03 · BUFFERStorage absorbs the gap

Inventories fall and weather produces larger price moves because the system carries fewer days of cover.

04 · CASHOnly some producers capture

Open volumes with economic wells and egress realize higher prices without spending the windfall away.

ALL FOUR LINKS MUST HOLD

The macro thesis can be directionally right while the shortage, timing, or equity return is wrong.

  1. Approved is not consumed. Nameplate capacity must start on time, secure feedgas, and run at a high enough utilization rate.
  2. The supply ceiling must be real. The proprietary well-inventory model must beat public models that allow more production growth.
  3. Price cannot solve the gap too early. Higher prices can add drilling, destroy demand, alter dispatch, delay LNG, or provoke export policy.
  4. The windfall must reach owners. Hedges, basis, service inflation, debt, acquisitions, and reinvestment can consume commodity upside.
03

MAP THE SECURITIES

Two core candidates. Two different risk books.

These names share gas-price exposure; they are not diversification. The selection edge must come from inventory quality, realized basis, egress, balance-sheet resilience, and the discipline to convert price into per-share cash.

RRC · CORE CANDIDATEINVENTORY

Appalachian gas and liquids with low net debt, marketing access, and a source claim of unusually durable Marcellus drilling depth.

WHAT MUST PROVE · ENGINEERED WELL INVENTORY AND EGRESS SUPPORT HIGH-RETURN VOLUMES AFTER 2028
EXE · CORE CANDIDATESCALE

North America's largest gas producer across multiple basins, with low leverage and substantial later-curve exposure after near-term hedges roll.

WHAT MUST PROVE · TWIN EAGLE INTEGRATION, INVENTORY QUALITY, AND BUYBACKS CREATE PER-SHARE VALUE
CRK · CONVEX SATELLITETORQUE

Haynesville and Western Haynesville exposure offers direct Gulf Coast sensitivity, but current drilling exceeds internally generated cash.

WHAT MUST PROVE · DEEP-WELL RETURNS COVER CAPITAL AND $3.1B OF DEBT BEFORE THE CYCLE TURNS
TOU CN · CANADIAN BOOKESCAPE VALVE

Low-cost Montney and Deep Basin inventory may help the continent, with different currency, basis, tax, transport, and LNG exposure.

WHAT MUST PROVE · CAD UNDERWRITING AND CANADIAN EGRESS BEAT THE U.S. PRODUCER SET
04

PRICE IS PART OF THE THESIS

The curve looks calm. The equities are not free.

This is a first cash screen, not a valuation verdict. Quarterly cash flow is highly sensitive to commodity prices, hedges, working capital, and drilling cadence. Annualizing one quarter exposes the starting yield; it does not forecast 2029.

NAMEAUG 13 CLOSECURRENT COMPANY BASEROUGH SCREENWHAT THE PRICE DEMANDS
RRC$39.91Q2 2.30 Bcfe/d; $333M CF before working capital; 2026 capex $650-$700M~7.0% annualized Q2 post-capital yieldLong-duration inventory and egress must offset Appalachian basis. Q2 pre-NYMEX-hedge gas realization was $2.42/Mcf.
EXE$94.68Q2 7.48 Bcfe/d; $343M adjusted FCF; $3.08B net debt~6.2% annualized Q2 adjusted FCF yieldScale, integration, later-curve exposure, and repurchases must overcome volatile cash: first-half adjusted FCF was $2.05B.
CRK$13.58Q2 1.24 Bcfe/d; $189M CF before working capital; $390M E&D capex~($202M) Q2 pre-acquisition cash deficitGas must rise enough to fund deep Western Haynesville development and service $3.05B net debt without diluting the equity.
TOU CNNOT SCREENEDCAD model, hedge book, basis, and filings pendingNOT ELIGIBLEA source endorsement is not a valuation. Canada enters only after a complete local-currency model.
XOP$179.17Doing NothingBenchmarkThe selected gas producers must beat broad upstream exposure after taking concentration, balance-sheet, and basis risk.

PRICE BASIS. Settled closes for August 13, 2026, retrieved August 14 from Nasdaq historical data. These are research observations, not sealed Call bases.

COMPANY SOURCES. RRC Q2 · EXE Q2 · CRK Q2.

ROUGH-SCREEN MATH. RRC uses four times Q2 cash flow before working capital less the midpoint of annual capex, divided by 236.2M shares at the observed price. EXE uses four times Q2 company-defined adjusted FCF divided by 234.35M issued shares at the observed price. CRK compares Q2 cash flow before working capital with Q2 exploration and development capex. These definitions are not comparable GAAP valuation measures.

PHYSICAL BEAR · $3.50-$4 HH

Approved capacity starts late or runs below nameplate, supply flexes, and storage stays normal. The shortage does not arrive; gas equities lose their scarcity premium.

PHYSICAL BASE · $5-$6 HH

The public EIA path: demand grows, production responds, and prices rise into the early 2030s without system failure. Stock selection and capital discipline matter more than beta.

PHYSICAL BULL · $8+ HH

Storage falls, weather exposes fragility, and the price response becomes nonlinear. Producers win first; demand destruction, export intervention, and new supply cap the duration.

These are physical-market frames, not equity return cases. Every candidate still needs production, basis, hedge, cost, capital, balance-sheet, tax, share-count, and exit-value work in each scenario.

05

THE CAPITAL RULE

Own molecules that can reach a buyer.

STEP 01Model delivered price

Start with production by basin, benchmark and basis, transport commitments, hedge book, royalty burden, and realistic volume growth.

STEP 02Model replacement capital

Separate maintenance from growth. A commodity windfall is not free cash if keeping volumes flat consumes it.

STEP 03Reverse today's equity

Solve for the Henry Hub path, production, capital, and terminal value required by the current enterprise value.

STEP 04Apply one hurdle

Base-case annualized return must exceed XOP by at least 3 points, with bear-case total downside no worse than 25%.

FIRST TO UNDERWRITERRC · EXE

Cleaner balance sheets and broader evidence. Neither is eligible until the 2029 per-share cash cases clear.

CONVEX SATELLITECRK

More direct Haynesville torque with substantially more financing and execution risk. Never equal-risk by assumption.

SEPARATE BOOKSTOU · MIDSTREAM

Canada and toll-road infrastructure require different bases, currencies, mechanisms, and benchmarks.

COMPARISON BOOKXOP · 100%

No HOUSE Gas allocation. This is the research benchmark, not a statement of actual holdings.

The 3-point / 25% rails are proposed underwriting rules, not sealed methodology. They make the decision auditable. Howard's methodology must approve or replace them before any Call locks.

06

IS THE CALL REALLY BETTER THAN DOING NOTHING?

Test selection, not just gas beta.

HOUSE GAS · IF ELIGIBLE 1-3 PRODUCERS EQUAL STARTING WEIGHTS · NO REBALANCE · 48 MONTHS
VS.FROM THE SAME FUTURE CLOSE
DOING NOTHING XOP BUY ONCE · HOLD UNCHANGED · SAME 48 MONTHS
WHY 48 MONTHS?

The source's mechanism begins in 2028 and culminates around 2030. Quarterly checks observe evidence; they do not resolve the Call early.

WHY XOP?

It is the liquid alternative to selecting specific upstream winners. SPY and Henry Hub can add context; neither replaces the sealed selection grade.

MONEY GRADE

Return of one starting dollar in the fixed eligible basket versus one starting dollar in XOP, using the same close and corporate-action treatment.

BREADTH GRADE

Show every producer's contribution, median excess return, and how many beat XOP. One levered winner cannot hide poor selection.

NO REBALANCE

Hold the starting lots. Rebalancing would add an unregistered commodity-timing strategy and conceal drift.

EVIDENCE GRADE

Demand, deliverable supply, storage, price, and company conversion are graded separately. A correct shortage never rescues a losing money verdict.

07

CHALLENGE THE THESIS

The hardest question is the missing model.

FRONTIER AI PANELWAITING FOR THE SEALED PACKET

The models cannot vote on a proprietary conclusion or moving valuation. Once the public demand bridge, company cases, benchmark, and kill conditions freeze, every model receives the same packet blind and before the outcome.

ATTACK 01 · REPRODUCIBILITY

What if the ceiling is wrong?

The decisive 132 Bcf/d maximum depends on private well-level acreage, decline, spacing, economics, and infrastructure assumptions HOUSE cannot inspect.

ATTACK 02 · UTILIZATION

What if approved is not consumed?

Chronometer uses roughly 35 Bcf/d of approved 2030 nameplate. EIA expects 27.7 Bcf/d of capacity by 2030. Construction, contracts, utilization, and feedgas are different facts.

ATTACK 03 · REFLEXIVITY

What if price prevents crisis?

Higher gas prices stimulate drilling and imports, reduce dispatch and industrial demand, improve competing economics, delay LNG, and invite export intervention.

ATTACK 04 · CAPTURE

What if producers spend the upside?

Basis, hedges, service inflation, replacement capital, acquisitions, debt, and multiple compression can leave shareholders with little of the commodity move.

08

REALITY GRADES

Demand is public. The shortage is not.

The public evidence supports rising LNG exports, rising power demand, infrastructure friction, and higher long-run gas prices. It does not independently reproduce Chronometer's well inventory, prove a 5+ Bcf/d 2030 deficit, or establish exhausted working storage.

PROPOSED KILL CONDITIONS · NOT YET SEALED

The hard questions stay visible before the basis exists. Final series, seasonal treatment, dates, and thresholds must freeze with the methodology packet.

  1. SUPPLY CEILING BREAKS. U.S. dry gas production reaches 132 Bcf/d before 2029 without Henry Hub sustaining above $6 and without abnormal storage draws.
  2. LNG DEMAND BREAKS. Actual U.S. LNG exports remain below 24 Bcf/d through Q4 2029 because projects are delayed, cancelled, curtailed, or underutilized.
  3. STORAGE BREAKS. Working gas finishes two consecutive withdrawal seasons after 2028 at or above its contemporaneous five-year average.
  4. CAPTURE BREAKS. At least two eligible producers show two consecutive quarters of falling per-share free cash flow despite a higher realized delivered-gas price.
  5. POLICY BREAKS. Binding export limits, emergency allocation, or durable demand destruction remove the demand required by the sealed balance.
  6. THE TRADE LOSES. At the sealed 48-month horizon, the fixed eligible basket fails to outperform XOP. The money verdict stands even if gas tightens.
09

WHAT REMAINS BEFORE CAPITAL

Rebuild the system. Then price the stocks.

01 · PHYSICAL MODELReproduce the 2026-2030 balance from public series.

Project-level LNG startup and utilization, power burn, industrial demand, pipeline trade, production by basin, and seasonal storage.

02 · COMPANY MODELSBuild three per-share cases for RRC, EXE, and CRK.

Volumes, basis, hedges, costs, maintenance and growth capital, debt, tax, buybacks, dilution, and exit value.

03 · CANADAUnderwrite TOU in local currency.

AECO and Dawn basis, egress, LNG Canada exposure, royalties, taxes, currency, return policy, and the correct benchmark.

04 · ELIGIBILITYName only producers that clear.

No source-prestige allocation. No minimum basket size. No equal weights across unequal balance-sheet risk.

05 · METHODOLOGYSeal XOP, 48 months, and the evidence series.

Same close, no rebalance, money plus breadth, corporate actions, seasonality, and kill-condition treatment.

06 · LOCKUse the next eligible close—never this page's date.

Freeze one packet for the AI panel. If price or facts move first, rerun the underwriting instead of backfilling a basis.