alirezarezvani/claude-skills / finance/skills/stock-analysis
stock-analysis
UnverifiedAnalyze, research or value a listed stock or company (US, global, NSE or BSE) with verified data, ratios, scores and valuation. Use whenever asked if a stock is a buy, what it is worth, or how a company looks.
references/sectors/oil-gas.md
View on GitHub# Oil and gas — sector playbook
Use this when: the company's earnings are driven by hydrocarbon prices, refining or marketing spreads, or regulated gas throughput — Indian upstream (ONGC, Oil India, Vedanta's Cairn segment), OMCs and refiners (IOC, BPCL, HPCL, MRPL, CPCL, RIL's O2C), gas transmission and city gas distribution (GAIL, GSPL, IGL, MGL, Gujarat Gas, Adani Total Gas, Petronet LNG), and global integrateds, independent E&Ps, pure refiners, LNG players and oilfield service names.
Three facts govern everything below. The company is a **price taker** — revenue is a market price it does not set multiplied by a volume it can only partly control. Upstream assets are **depleting** — production without reserve replacement is liquidation reported as profit. And the accounting is **non-comparable** — successful-efforts vs full-cost, LIFO vs weighted-average inventory, and unit-of-production DD&A mean two companies with identical physical operations report materially different earnings. Every generic ratio breaks on at least one of these. Express margin in physical units ($/bbl, ₹/litre, $/boe, ₹/scm), value on cash flow and asset value rather than earnings, and normalise to mid-cycle before applying any multiple.
If the name is a **regulated gas transmission or city gas distribution** business, treat it as infrastructure with a volume ramp, not as a commodity producer — the relevant logic is regulated-asset-base returns and EBITDA per scm, and it is a category error to screen it alongside an E&P. Sections below flag where CGD/transmission diverges.
## Contents
- [Why the generic ratio set fails here](#why-the-generic-ratio-set-fails-here)
- [The metrics that actually matter](#the-metrics-that-actually-matter)
- [How to value companies in this sector](#how-to-value-companies-in-this-sector)
- [Peer set construction](#peer-set-construction)
- [Sector-specific red flags](#sector-specific-red-flags)
- [Cycle and structural context](#cycle-and-structural-context)
- [India vs global notes](#india-vs-global-notes)
- [Checklist](#checklist)
## Why the generic ratio set fails here
**OPM, net margin and revenue growth — near-meaningless.** Revenue is price × volume, not a business decision. A 40% revenue jump can coincide with falling production. For refiners and marketers, revenue is a pass-through of feedstock plus excise and duties: an Indian OMC's topline includes excise duty and road cess, so a 3% "OPM" and a 9% "OPM" can represent the identical physical margin at two different crude prices. Worse, upstream margins and refining margins are **inversely correlated** — a blended "energy OPM" benchmark averages two opposing signals into noise. Always restate margin per physical unit: $/bbl GRM, ₹/litre marketing margin, $/boe netback, ₹/scm for CGD.
**P/E — inverted and actively dangerous.** Cyclicals print their *lowest* P/E at the peak (peak earnings, imminent mean reversion) and their highest or a negative P/E at the trough. Screening for "low P/E" in this sector systematically buys the top of the cycle. Trailing EPS carries essentially no information about mid-cycle earning power.
**EPS — not comparable across companies, by construction.** Successful-efforts vs full-cost accounting for E&P produces materially different DD&A, capitalised exploration and net income for identical physical operations. Full-cost companies take ceiling-test write-downs when prices fall; successful-efforts companies expense dry holes as incurred. DD&A is unit-of-production based, so it moves with reserve revisions rather than with economics — a positive reserve revision lowers DD&A per barrel and raises EPS without anything happening in the field.
**Reported refining margin — polluted by inventory accounting.** US refiners commonly use LIFO, which insulates reported margins. Ind AS and IFRS prohibit LIFO, so Indian and European refiners book large inventory *gains* when crude rises and losses when it falls. Reported GRM must be split into core (current-cost) GRM and inventory gain/loss before any cross-border or year-on-year comparison. Comparing an Indian refiner's reported GRM against a US LIFO refiner's unadjusted is a standing error.
**D/E and interest coverage — understate true leverage.** Asset retirement / decommissioning obligations, long-term take-or-pay LNG and pipeline capacity commitments, rig and FPSO charters, prepaid offtake and volumetric production payments, and reserve-based lending covenants are all economically debt-like and sit outside a simple D/E. Interest is capitalised during multi-year refinery and petrochemical construction, flattering coverage. Indian OMCs fund seasonal crude swings with buyer's credit and letters of credit, so quarter-end short-term debt can be a fraction of average debt through the quarter. Use net debt / mid-cycle EBITDA, gearing, debt per boe of 1P, and net debt *including* ARO instead.
**ROCE — distorted in both directions, and perversely by write-downs.** Historic-cost legacy fields (a field discovered fifty years ago and long since depreciated) show inflated ROCE; recently acquired or newly commissioned assets carried at fair value show depressed ROCE for reasons unrelated to quality. A full-cost ceiling-test write-down shrinks the capital base and mechanically raises next year's ROCE — a punishment that reads as an improvement. And a single-year ROCE is uninformative without knowing where in the cycle it sits.
**FCF — can be manufactured by liquidating the asset.** Reserves deplete, so upstream cash flow is partly return *of* capital. A producer that cuts capex below maintenance level prints spectacular FCF for two or three years while its reserve life shrinks. FCF is only interpretable alongside reserve replacement ratio and reserve life. The inverse also holds: deeply negative FCF at a trough from counter-cyclical capex can be the most value-creating thing management does.
**P/B and book value — close to irrelevant upstream.** Reserves are carried at historical finding cost, not value; the gap between book equity and NAV runs to multiples in either direction. Book value has residual use only as a cyclical band indicator (own 10–15 year P/B range), and mainly for refiners where book at least reflects real physical plant.
**Current ratio, inventory days, cash conversion cycle — polluted.** Compulsory strategic and pipeline stock, 30–60 day crude voyages, and the price effect on inventory value all distort them. Inventory days can fall simply because crude got dearer.
**India-specific breakages.** (a) Retail petrol and diesel pricing is *de jure* deregulated but *de facto* administered — in FY23 OMCs absorbed large marketing losses through the crude spike, so earnings reflected government policy, not management. (b) LPG under-recoveries and subsidy receivables shift between years depending on budget compensation. (c) The SAED windfall levy (July 2022 to December 2024) capped domestic crude realisations and taxed diesel/ATF exports, breaking the historic Brent-to-domestic-EPS link. (d) APM/administered gas pricing under the Kirit Parikh formula (a percentage of the Indian crude basket, with a floor and ceiling) caps upstream gas realisation regardless of Henry Hub or JKM. (e) Cross-holdings — the OMCs hold listed stakes in each other and in ONGC, GAIL and Petronet LNG — make consolidated P/E meaningless without SOTP and a holdco discount. (f) PSU dividend policy is a function of the government's fiscal position, so dividend yield is not evidence of capital discipline.
## The metrics that actually matter
Ranges are **indicative only**. They vary by market, sub-sector, cycle position and period, and any of them can be wrong for a specific company in a specific year. Comparison against a tight peer set at the *same price deck*, and against the company's own 5–10 year history, overrides every absolute band below. Metrics marked (U) are upstream, (D) downstream/refining-marketing, (G) gas/CGD.
| Metric | Definition / how to compute | Indicative healthy range | Why it matters |
|---|---|---|---|
| **Reserve Replacement Ratio (RRR), organic, 3-yr average** (U) | Proved (1P) reserves added via extensions, discoveries and revisions ÷ production for the year. Strip acquisitions for organic RRR; use a 3-year rolling average because single years are lumpy. Exclude price-driven revisions — those are not barrels found. | >100% sustained; >120% for a genuine growth story; <80% for three consecutive years signals liquidation. | The upstream equivalent of revenue sustainability. A producer that does not replace what it produces is converting a reserve base into cash and calling it profit. This is the single check that tells you whether reported FCF is earnings or self-liquidation. |
| **Reserve Life Index (R/P) and reserve quality mix** (U) | 1P (and separately 2P) reserves ÷ current annual production = years of runway. Read alongside proved developed vs proved undeveloped (PUD) split, and oil vs gas mix. | Integrated majors 8–13 yrs; conventional international E&P 10–15; US shale 6–9; Indian PSU upstream historically ~8–12 on 1P. PUD above ~40% of 1P is aggressive; SEC rules require PUDs to be drilled within five years. | Sets the runway and, with the PUD share, tells you how much of the reserve base still needs capital before it produces a barrel. A long R/P built of undeveloped gas in a stranded basin is worth far less per boe than a short R/P of developed, pipeline-connected oil. |
| **F&D cost per boe and recycle ratio** (U) | (Exploration + development capex) ÷ reserves added, 3-year rolling. All-in adds acquisitions and unproved property spend. Recycle ratio = netback per boe ÷ F&D per boe. | F&D $8–15/boe for competitive conventional and good-rock shale. Recycle ratio >1.5x, ideally >2x. | F&D is the true cost of goods sold for a producer, and the recycle ratio is the cleanest measure of whether the company creates or destroys value per barrel. A company can look profitable on the income statement while replacing barrels at a cost above what it sells them for. |
| **Cash lifting cost and all-in cash cost per boe** (U) | Lifting cost per boe, then the fuller stack: lifting + royalty + cess/production taxes + transport + cash G&A. India: royalty, OID cess (₹/tonne) and NCCD materially raise the all-in figure. | Best-in-class Middle East and legacy onshore <$10/boe; majors $6–12/boe lifting and $20–30/boe all-in cash cost; deepwater and mature offshore higher. Should not drift up faster than inflation, ex price effects. | Cost position determines who survives the trough and who issues equity at the bottom. It is the only durable competitive advantage available in a commodity business — and it is completely invisible in OPM. |
| **Netback per boe and FCF breakeven oil price** (U) | Netback = realised price − royalty − production taxes − opex − transport, per boe. FCF breakeven = the Brent price at which operating cash flow covers maintenance capex, and separately at which it also covers the dividend. | Organic FCF breakeven including dividend below $40–50/bbl Brent for a resilient major; below $35–40 is strong. Post-dividend breakeven above ~$70 is a fragile balance sheet in disguise. | Converts the entire cost and capital structure into one number denominated in the only variable that matters. It answers directly: at what price does this company start borrowing to pay you? |
| **Realisation vs benchmark (differential capture)** (U) | Realised $/bbl and $/mmbtu against Brent/WTI/Dubai and against any administered price, as a discount or premium. Include crude quality, location basis, marketing arrangements, and any windfall levy or price cap. | Discount should be stable and explicable by quality and logistics. India: check gas realisation against the APM ceiling, and whether new/deepwater/HPHT fields qualify for the higher free-market price. | Two producers with identical costs can have very different economics purely through differentials and fiscal capture. In India this is decisive — the APM ceiling and the 2022–24 windfall levy severed the Brent-to-realisation link, so modelling Indian upstream off Brent alone gives the wrong answer. |
| **Production growth and base decline rate** (U) | boe/d output growth split organic vs acquired, plus the underlying annual decline of the existing base before new wells — the maintenance treadmill. | Conventional base decline 4–8%/yr; US shale base decline 25–40%, with first-year well declines of 60–70%. Maintenance capex should be disclosed and reconcilable to flat production. | Decline rate determines how much of capex is running to stand still. A shale producer with 35% base decline and a "growth" budget may be spending 80% of it on maintenance. For PSU upstream that has struggled to hold volumes flat despite rising capex, that *is* the story — not the reported profit. |
| **GRM: reported, core (ex-inventory), and premium over benchmark** (D) | Product revenue − crude and feedstock cost, per barrel of throughput. Decompose into current-cost/core GRM and inventory gain or loss. Benchmark: Singapore complex for Indian and Asian refiners, NWE for Europe, USGC 3-2-1 crack for the US. | Mid-cycle Singapore complex GRM roughly $4–7/bbl. A complex refiner should sustain a $2–5/bbl premium over benchmark; the most complex Indian private refining has historically run a mid-to-high single-digit premium. | GRM is the entire refining P&L in one number, and the premium over benchmark is the only part attributable to management — configuration, crude sourcing, product slate, yield optimisation. The benchmark is the weather; the premium is the skill. A GRM that beats only because crude rose during the quarter is an inventory gain, not a business. |
| **Nelson Complexity Index and crude slate flexibility** (D) | Secondary conversion capability (cokers, hydrocrackers, FCC, hydrotreaters) relative to simple distillation, plus achieved share of heavy/sour and discounted crude, and distillate yield. | Simple hydroskimming 3–6; typical Indian PSU refineries 8–12; the most complex refining complexes globally run in the mid-to-high teens and above. Heavy/sour share above 60–70% with strong distillate yield indicates real conversion capability. | Complexity is the structural moat in refining: it lets a refiner buy the cheapest, dirtiest barrel and still sell high-value distillates, earning a premium GRM in every environment and staying cash-positive when simple refiners go negative. Since 2022 it also determined who could process discounted Russian Urals. |
| **Utilisation / capacity factor and fuel-and-loss** (D) | Throughput as % of nameplate capacity; energy consumed plus processing losses as % of throughput; turnaround schedule and unplanned outage days. | >90% healthy; Indian refiners frequently run 100–110% of nameplate via debottlenecking. Fuel & loss 6–9% depending on complexity — a rising trend signals ageing units. Below 80% in a non-turnaround quarter needs an explanation. | Refining is a high fixed-cost business, so at the margin utilisation drives unit economics more than headline crack. Utilisation also flags reliability problems and deferred maintenance, which surface as an earnings surprise one or two quarters later. |
| **Marketing margin per litre and throughput per outlet** (D) | Gross marketing margin in ₹/litre on petrol and diesel (net realisation − refinery gate / import parity price − transport − dealer commission), plus market share, volume growth vs industry, throughput per retail outlet (KL/month), and network count. | Normalised gross marketing margin roughly ₹3.0–4.0/litre through the cycle; below ~₹1.5 implies the marketer is absorbing losses. Throughput per outlet materially above industry average signals network quality rather than network size. | For the Indian OMCs the marketing segment often contributes more mid-cycle EBITDA than refining and is far less volatile — except when policy intervenes. Tracking ₹/litre against crude tells you immediately whether the company is being used as a price-stabilisation instrument, which is the dominant risk in Indian downstream. |
| **Unit EBITDA by segment and segment split** | Normalised EBITDA per boe (U), per tonne of throughput (D), per scm (G), with consolidated EBITDA explicitly split into upstream, refining, marketing, petrochemicals, gas transmission and trading. | No universal level. The test is stability of the unit figure versus peers at the same price deck, and whether disclosure is granular enough to model at all. India CGD: EBITDA of roughly ₹5–8/scm has been the working band. | Integrated companies are natural hedges — upstream and refining margins move inversely — so consolidated numbers hide everything. Collapsing refining and petrochemicals into one reported segment is a live example: less granularity means you can no longer see which leg is carrying the result. |
| **Net debt / mid-cycle EBITDA, gearing, debt per boe of 1P** | Leverage against *normalised* EBITDA at a mid-cycle deck, not trailing. Gearing = net debt ÷ (net debt + equity). Debt per boe = (net debt + ARO + leases) ÷ 1P reserves. | Net debt / mid-cycle EBITDA below 1.5–2.0x; gearing 20–25% is the majors' stated target band; debt including ARO below roughly $4–6 per boe of 1P for a conventional producer. | Trailing EBITDA at a peak makes every energy balance sheet look pristine and at a trough makes every one look distressed. Mid-cycle normalisation and debt-per-barrel are the only leverage measures that survive the cycle — and debt-per-boe is what reserve-based lenders actually underwrite against. |
| **Reinvestment rate and maintenance vs growth capex** | Capex ÷ cash flow from operations, with maintenance capex (the spend needed to hold production and refinery reliability flat) explicitly separated from growth and energy-transition capex. | Below 60–70% of CFO through the cycle for a shareholder-return model; sustained above 100% means external funding. Capex persistently below DD&A for three or more years is under-investment. | This is where energy companies destroy value most reliably: pro-cyclical capex at the peak, starvation at the trough. It is also the test of the FCF story — high FCF with capex below maintenance level is a temporary accounting artefact, not a durable cash yield. |
| **Carbon intensity, methane intensity and ARO** | Scope 1+2 emissions per boe produced or per tonne refined; methane intensity of gas production; flaring rate; and the undiscounted *and* discounted asset retirement obligation with its assumed discount rate and retirement timing. | Upstream Scope 1+2 of roughly 15–20 kgCO2e/boe is upper quartile; methane intensity below 0.20%; routine flaring approaching zero. Compare ARO against market decommissioning cost estimates for the same water depth and asset vintage. | In developed markets this is a direct cost-of-capital and cost-of-operation input (EU ETS, CBAM, methane rules, lender and insurer exclusions), not a reporting nicety. ARO is real, senior, unavoidable cash that most screens ignore entirely — for a mature offshore portfolio it can rival net debt in size, and a small change in the discount rate moves it materially. |
| **Regulated tariff, RAB and volume ramp** (G) | For transmission: approved capital base, PNGRB tariff order, achieved vs approved utilisation, and tariff review cycle. For CGD: geographical areas won, minimum work programme commitments, CNG stations and domestic PNG connections added, volume/day by segment, and exclusivity expiry dates. | Utilisation of transmission capacity trending toward the level assumed in the tariff order; CGD volume CAGR in the low-to-mid teens during the ramp; capex funded within operating cash flow after the initial build. | These businesses earn a regulated or quasi-regulated return on an approved asset base, so value accrues through volume ramp and capex execution, not through the oil cycle. Marketing-exclusivity expiry and network-exclusivity expiry are the two dates that change the competitive structure of a CGD; both are knowable years in advance and are routinely under-modelled. |
| **Hedge book and forward sales position** | Percentage of next 12–24 months' production hedged, instrument type (swaps, collars, three-ways), average floor and ceiling, and whether realised hedge gains/losses sit inside or outside headline margin. | Levered producers typically hedge 40–70% of near-term production; an unhedged, highly levered producer is a directional bet, not a business. | Hedging determines whether the balance sheet survives a 40% price drop, and it is a common site of margin flattery — hedge gains booked inside EBITDA make a high-cost producer look low-cost for exactly as long as the hedge lasts. |
## How to value companies in this sector
Because earnings are cyclical and accounting is non-comparable, this sector is valued on **cash flow, asset value and normalised mid-cycle earnings** — essentially never on trailing P/E.
**Normalise the denominator first.** Before any multiple, build EBITDA / DACF at a mid-cycle deck — for example $65–75 Brent, $5–6/bbl Singapore complex GRM, ~₹3.5/litre normalised marketing margin — and apply the multiple to *that*. Applying a "cheap" multiple to peak earnings is the classic energy value trap and the most common single error in the sector.
**Cash-flow multiples (the working tools).**
- **EV/DACF** is the standard upstream multiple. DACF = cash flow from operations before working capital + after-tax interest expense. It neutralises differences in leverage and in the tax shield, making a levered and an unlevered producer comparable. Typical 3–6x; below ~3.5x on mid-cycle assumptions is historically cheap, above 7x demands growth or premium rock.
- **EV/EBITDAX** (EBITDA before exploration expense) exists specifically to neutralise successful-efforts vs full-cost, since exploration cost is expensed under one and capitalised under the other. Typical 3–6x for E&P, 4–7x for refiners, 7–10x for regulated gas transmission and CGD, which are annuity-like.
- **P/CF** is preferred to P/E throughout the sector because it steps around DD&A, which is unit-of-production based and moves with reserve revisions rather than economics.
**Asset-based valuation (the anchor).**
- **NAV / DCF of the production profile** is the primary upstream method: discount after-tax free cash flow field by field over the producing life at 10% nominal (the PV-10 convention embedded in SEC disclosure) or 8–12% by risk; add risked or unrisked value for probable/possible reserves and exploration acreage; subtract net debt, ARO and corporate G&A.
- Watch the convention: the **SEC standardised measure is PV-10 on 1P at trailing 12-month average prices** — a disclosure convention, not a valuation. Canadian and Australian practice quotes NPV10 pre-tax on 2P. Always rerun at strip or at your own deck before using either.
- **P/NAV** is the resulting yardstick: integrated majors and large caps typically 0.9–1.2x NAV; single-asset and junior E&Ps 0.4–0.7x, reflecting funding and execution risk. A persistent sub-0.5x for a producing, funded company usually signals a governance, fiscal-regime or decommissioning problem rather than an opportunity.
- **Transaction/screening comparables:** EV per boe of 2P reserves, and EV per flowing barrel ($ per boe/d of production). Both need adjusting for reserve life, oil/gas mix and fiscal regime, but they frame transaction value fast.
- **Replacement-cost benchmarks downstream:** EV per barrel/day of refining capacity against greenfield build cost (far higher in developed markets than in India or the Gulf), EV per *complexity-barrel* (capacity × Nelson index) to give credit for conversion, and EV per retail outlet or per tonne of marketing throughput.
**Normalised and cycle-aware earnings.**
- **Mid-cycle P/E** on normalised margins is legitimate; trailing P/E is not. The operating rule is inverted — a high or negative P/E at trough conditions is often the entry point, and a low P/E at peak margins is often the exit.
- **P/B against the company's own 10–15 year band** works as a crude cycle-position indicator, most usefully for refiners where book reflects real physical plant.
- **FCF yield stated at a specified crude price** ("X% FCF yield at $70 Brent") is the standard framing for the majors and large US independents, because it makes the price assumption explicit rather than burying it inside a trailing multiple. Quote at two or three prices, not one.
**Indian conventions that differ.**
- **SOTP is effectively mandatory for the Indian integrateds.** The OMCs each hold large listed stakes (upstream, gas transmission, LNG regas), conventionally valued at market with a 20–40% holdco discount, plus pipelines valued separately on EV/EBITDA, plus refining and marketing on segment EV/EBITDA. Consolidated P/E on these names is close to meaningless.
- **Conglomerate energy names are pure SOTP:** O2C on EV/EBITDA, telecom on EV/EBITDA or per subscriber, retail on EV/EBITDA or EV/sales, upstream on DCF, and new-energy build-out largely at invested capital until it generates cash.
- **Regulated gas is valued on regulated-return logic, not the oil cycle.** PNGRB sets transmission tariffs to deliver a post-tax return on capital employed (12% for natural gas pipelines, 14% for petroleum product pipelines) on an approved asset base, so RAB-style DCF plus steady EV/EBITDA and P/E apply. CGD is valued on EBITDA per scm, volume growth and P/E — infrastructure with a ramp.
- **Indian upstream must explicitly model the APM gas price formula ceiling and any windfall levy regime,** and should carry a governance/policy discount for PSU capital allocation. Standard practice: value the standalone E&P on DCF or EV/EBITDA, then add listed investments separately at a discount.
- **Disinvestment/privatisation optionality and government shareholding** create event-driven gaps no multiple captures. Treat as a separate scenario, not as a haircut baked into the base multiple.
**Do not use:** trailing P/E on its own, PEG, EV/Sales (revenue is a pass-through of feedstock and duties), P/B as a level judgement upstream, or any cross-company EPS comparison that ignores successful-efforts vs full-cost and LIFO vs weighted-average.
## Peer set construction
A valid comparable shares **sub-sector, fiscal regime, asset quality and price exposure** — not merely the label "energy". Mixing across these lines produces confidently wrong conclusions, and the correlation structure makes it worse: upstream and refining margins move in *opposite* directions, so a mixed peer set averages a signal against its own inverse.
**Splits that must not be mixed:**
- **Upstream vs refining vs marketing vs gas transmission/CGD vs oilfield services.** Four different business models, four different valuation frameworks. A CGD utility and an E&P do not belong in the same table under any circumstance.
- **Integrated vs pure-play.** Integrateds are internally hedged, so their consolidated volatility is structurally lower and their multiple structurally different. Compare integrated to integrated, and compare *segments* to pure-plays.
- **Conventional vs shale vs deepwater vs oil sands.** Base decline of 4–8% vs 25–40% changes the meaning of every capex, FCF and growth number. Capital intensity, cycle time and breakeven differ by more than the differences between many separate sectors.
- **Oil-weighted vs gas-weighted producers.** Different price benchmarks (Brent/WTI vs Henry Hub/JKM/TTF/APM), different transport economics, different demand cycles. Convert to boe only for scale, never for valuation.
- **Simple vs complex refiners.** A hydroskimmer at Nelson 4 and a deep-conversion refinery at Nelson 12+ have structurally different margins in the same crack environment. Benchmark complexity-adjusted.
- **PSU vs private (India) and NOC vs IOC (global).** State-controlled companies carry policy-driven pricing, forced acquisitions, government dividend demands and appointment-driven governance. A PSU at 5x EV/EBITDA is not "cheap versus" a private peer at 8x — the gap is largely a governance and policy discount, and it is usually justified.
- **Fiscal regime.** Royalty/tax regimes, production sharing contracts with cost recovery, and service contracts distribute upside completely differently at high prices. Two identical fields under different PSCs have different equity value.
- **Regulated vs merchant gas.** PNGRB-tariffed transmission and a merchant LNG trading book have almost nothing in common.
**Also align:** the **price deck** (comparing a company's FY at $95 Brent against a peer's at $70 is a comparison of years, not companies — restate to a common deck or a common period); accounting regime (successful-efforts vs full-cost, LIFO vs weighted-average, US GAAP vs IFRS vs Ind-AS); reserve reporting standard (SEC 1P at trailing-12-month prices vs SPE-PRMS 2P at forecast prices — these are not the same barrels); fiscal year end (Indian April–March vs calendar); and hedging posture.
Aim for 5–8 peers. State the basis explicitly, and benchmark every metric twice — against peers at a common deck, and against the company's own full-cycle history including its last trough.
## Sector-specific red flags
- **Reserve replacement below 100% for three consecutive years**, or a reserve base held up mainly by positive *price* revisions and PUD bookings rather than by the drill bit. Check whether additions came from extensions and discoveries or from revisions, and whether PUDs are rebooked year after year without being drilled — the SEC five-year rule exists precisely because of this abuse.
- **Reported GRM with no core / ex-inventory breakdown**, or a quarter where the entire beat is an inventory gain from rising crude. The mirror image is equally telling: a refiner that never discloses inventory *losses* in a falling-crude quarter is disclosing selectively.
- **Capex persistently below DD&A while the company markets a high FCF yield.** This is harvesting the asset; the reserve life index will confirm it. Related: a dividend or buyback funded by asset sales, hybrid issuance or incremental debt rather than by operating cash flow, especially near a price peak.
- **Aggressive capitalisation.** Capitalising dry holes or unsuccessful exploration under a loose "successful efforts" reading; capitalising workover and maintenance spend; or switching accounting policy between successful-efforts and full-cost. A large full-cost ceiling-test write-down is itself a flag, and the post-write-down jump in ROCE and asset turnover is an artefact, not an improvement.
- **Under-provisioned or opaque ARO.** A discount rate quietly raised to shrink the liability, retirement dates pushed further out, or an ARO small relative to peers with comparable water depth and vintage. For mature offshore portfolios this is the largest hidden claim on equity.
- **Off-balance-sheet and quasi-debt commitments outside D/E:** long-term take-or-pay LNG and pipeline capacity commitments, FPSO and drilling-rig charters, prepaid offtake and volumetric production payments, and — India — heavy reliance on buyer's credit and LCs that make quarter-end short-term debt look far lower than average debt through the quarter.
- **A trading arm that never loses money**, or opaque related-party crude sourcing and product placement. Trading profits with no disclosed VaR, no volume-versus-margin decomposition and no losing quarters are a classic site of smoothing and, occasionally, of fraud.
- **Hedging problems in either direction:** an unhedged, highly levered producer at a price peak; or a producer that locked in large volumes at trough prices and books the resulting derivative gains *inside* EBITDA. Check where realised hedge gains and losses sit, and how much of next year's production is already sold forward.
- **Project cost overruns and stagnant CWIP.** Refinery expansions, cokers and crackers routinely run over budget and late. A capital-work-in-progress balance that grows for years without commissioning is capital earning nothing — and it flatters current ROCE by sitting outside the denominator in some presentations while depressing future returns.
- **India — government policy masquerading as an earnings trend.** Marketing margins suppressed ahead of elections or during crude spikes; price hikes deferred without compensation; LPG under-recoveries not booked while awaiting a budget subsidy; ballooning subsidy receivables from the government. The inverse error is equally common: extrapolating a windfall-margin period as a structural re-rating.
- **India — fiscal-regime shocks and levies.** SAED windfall tax on domestic crude and on diesel/ATF exports (July 2022 to December 2024), OID cess, NCCD, and the APM gas ceiling. Any model that runs Indian upstream off Brent alone, without ceiling and levy mechanics, will be systematically wrong in a predictable direction.
- **PSU capital-allocation flags:** forced acquisitions of other state entities funded with debt, large dividends and buybacks paid to the government while the company simultaneously borrows for capex, and cross-holdings used as a fiscal instrument. These are governance costs justifying a discount, not a bargain narrative.
- **Reduced segment granularity or definitional change.** Collapsing refining and petrochemicals into a single segment, redefining GRM, changing the comparison benchmark, or restating a "normalised" margin without reconciliation. Loss of disclosure almost always precedes deterioration in the leg that got hidden.
- **Concentration and contract risk:** single field, single refinery or single-basin exposure; a PSC nearing expiry with uncertain extension terms; cost-recovery disputes with the regulator (Indian DGH disputes have run for years); non-operated interests where the company cannot control capex timing; and resource-nationalism or fiscal-terms risk in the host country.
- **Working capital and cash-conversion breakdown:** receivable days rising against state distributors or government entities, an inventory build not explained by price, and operating cash flow persistently lagging EBITDA. In gas distribution and marketing, receivables from state-owned counterparties are a recurring and under-priced credit exposure.
## Cycle and structural context
**Locate the cycle before writing a single conclusion.** State explicitly where the current price deck sits relative to the last 10 years, and whether reported margins are above, at, or below mid-cycle. Every metric above changes meaning with that answer. The classic trap sequence is: peak prices → peak EPS → low trailing P/E → screen flags "cheap" → mean reversion. The inverse — negative EPS and no P/E at the trough — is frequently the entry point. Cross-check with own-history P/B band, inventory levels, and rig-count or refinery-margin trends.
**Supply-side drivers to check each time.** OPEC+ spare capacity and quota discipline (spare capacity is the shock absorber; when it is thin, prices are convex to any disruption); US shale response elasticity, which has fallen since operators shifted to capital discipline and shareholder returns; and multi-year underinvestment in long-cycle conventional supply, which tightens supply with a lag of five to ten years.
**Refining is a capacity cycle, not a price cycle.** Cracks are set by the balance between global capacity and demand. Closures in developed markets, new mega-capacity in Asia and the Middle East, and the timing of turnarounds drive margins far more than crude direction. Complexity and location determine which refiners survive the down leg — simple refiners in high-cost jurisdictions go cash-negative first and close permanently, which itself sets up the next up leg.
**Sanctions, war and trade reroutes create differentials, not just prices.** Since 2022, the discounted-crude opportunity rewarded refiners with the complexity and the freedom to process heavy sour barrels. Differential capture became a bigger earnings driver than the flat price for several Asian refiners. Treat differentials as a distinct, separately-modelled line, not as noise around a benchmark.
**Secular demand threats, sequenced by timing.** Gasoline demand faces EV penetration first (developed markets already past peak in several); diesel and jet are more durable because heavy transport and aviation electrify slowly; petrochemical feedstock demand is the longest-lived leg, which is why refiners are pushing crude-to-chemicals. Gas is the transition fuel with a long runway in Asia — Indian gas demand is policy-supported toward a higher share of the primary energy mix — but LNG price spikes destroy demand in price-sensitive markets, which is why volume ramps in CGD depend on gas being *cheap enough against liquid fuel alternatives*, not just available.
**Terminal value is the hard part.** For upstream, the depleting asset base means a conventional perpetuity growth assumption is wrong; run the DCF to reserve exhaustion plus a risked exploration value instead. For refining, ask whether the asset is a survivor or a closure candidate in 15 years, and let that drive the terminal assumption. For CGD and transmission, terminal value depends on regulatory continuity and exclusivity renewal, which is a policy question, not a financial one.
**Regulation as a cost of capital input.** EU ETS and CBAM, methane regulation, lender and insurer exclusion policies, and disclosure regimes (TCFD/ISSB) all feed into discount rates and access to capital in developed markets — a European or US major faces a genuinely higher cost of capital for long-cycle oil projects than it did a decade ago. In India the binding constraints are different: PNGRB tariff orders, DGH cost-recovery, the APM formula, and administered retail pricing. Model the regime you are actually in.
## India vs global notes
- **Disclosure.** India: Ind-AS financials, Schedule III format, figures in crore/lakh, quarterly results plus a concall and investor presentation where GRM and marketing margin are usually disclosed. Annual report includes CARO (fixed-asset verification, loans to related parties, statutory dues, CWIP), related-party disclosures, and promoter/government shareholding. US/global: 10-K on EDGAR, with the **SEC Subpart 1200 reserve disclosures** — proved reserves, standardised measure (PV-10), reserve roll-forward by category and a third-party reserve auditor's report. There is no Indian equivalent of that reserve schedule at the same granularity, so Indian upstream reserve analysis leans on DGH data, annual report disclosures and management guidance.
- **Reserve standards.** SEC = 1P (proved), priced at trailing 12-month average, PUDs must be drilled within five years. Canada (NI 51-101) and Australia/UK practice commonly report 2P under SPE-PRMS at forecast prices. Never compare an SEC 1P number against a PRMS 2P number without restating.
- **Inventory accounting.** US GAAP permits LIFO and most large US refiners use it; Ind-AS and IFRS prohibit it, so Indian and European refiners report weighted-average-cost inventory with large gains and losses flowing through GRM. This single difference makes headline GRMs non-comparable across the two blocks.
- **Pricing regimes.** India: retail petrol/diesel nominally deregulated but administered in practice; LPG subsidised with under-recovery and budget-compensation timing risk; domestic gas priced under the APM/Kirit Parikh formula with floor and ceiling, and a separate higher price band for HPHT/deepwater/new fields. Global: merchant pricing off Brent/WTI/Dubai and Henry Hub/TTF/JKM, with long-term LNG contracts often oil-indexed at a slope to Brent.
- **Regulators.** India: PNGRB (pipeline and CGD tariffs, exclusivity, authorisation), DGH (upstream, PSC administration and cost recovery), MoPNG (policy and PSU control), SEBI/exchanges (disclosure), plus state VAT on fuels alongside central excise — note that petroleum products remain outside GST, which is why the tax stack is visible in OMC revenue. Global: SEC (reserve and disclosure rules), EPA/BOEM and equivalents, EU ETS and CBAM, host-country petroleum ministries and NOC partners.
- **Ownership and governance.** India: government promoter holding in the PSUs, with disinvestment optionality and periodic government-driven dividend, buyback and acquisition decisions; cross-holdings among the energy PSUs. Monitor pledge, promoter changes and related-party transactions as standard. Global: dispersed institutional ownership with activist pressure typically pushing toward shareholder returns and away from growth capex — the opposite direction from the Indian PSU pressure toward strategic capex.
- **Units and conventions.** India: crore/lakh, ₹/litre for marketing, ₹/scm for CGD, MMSCMD for gas volumes, MMTPA for refinery capacity, ₹/tonne for OID cess. Global: $/bbl, $/mmbtu, boe/d, bcf, MMTPA for LNG, barrels/day for refining. Standardise units *before* computing anything, and state the FX and conversion assumptions (a boe conversion of 6 mcf per barrel is an energy convention, not a value convention — gas-weighted producers look artificially large on a boe basis).
## Checklist
- [ ] State where the current price deck sits versus the 10-year range, and whether reported margins are above, at, or below mid-cycle — before any other conclusion.
- [ ] Refuse to report OPM, trailing P/E or EV/Sales as signals; restate margin per physical unit ($/bbl, ₹/litre, $/boe, ₹/scm).
- [ ] Identify the sub-sector (upstream / refining / marketing / gas transmission / CGD / integrated / services) and use only that sub-sector's framework.
- [ ] Upstream: check 3-year organic reserve replacement, reserve life, PUD share, and whether additions came from the drill bit or from price revisions.
- [ ] Upstream: compute F&D per boe and the recycle ratio (netback ÷ F&D); below ~1.5x the company is not creating value per barrel.
- [ ] Upstream: compute all-in cash cost per boe and the FCF breakeven Brent price including the dividend.
- [ ] Upstream: check realisation vs benchmark, and in India apply the APM ceiling and any windfall levy explicitly — never model off Brent alone.
- [ ] Upstream: separate base decline and maintenance capex from growth capex before believing any growth narrative.
- [ ] Refining: split reported GRM into core and inventory gain/loss; benchmark the *premium* over Singapore/NWE/USGC, not the absolute.
- [ ] Refining: check Nelson complexity, heavy/sour slate share, utilisation and fuel-and-loss trend.
- [ ] Marketing: track ₹/litre gross margin against crude to detect policy absorption; check throughput per outlet, not just outlet count.
- [ ] Gas/CGD: use RAB and tariff-order logic, EBITDA per scm and volume ramp — never oil-cycle logic; note exclusivity expiry dates.
- [ ] Identify the accounting regime: successful-efforts vs full-cost, LIFO vs weighted-average, Ind-AS vs IFRS vs US GAAP — and say so before comparing EPS or GRM.
- [ ] Compute net debt / *mid-cycle* EBITDA, gearing and debt per boe of 1P; add ARO, leases, charters and take-or-pay commitments to debt.
- [ ] Check reinvestment rate and whether capex is below DD&A or below maintenance level while an FCF yield is being marketed.
- [ ] Value on EV/DACF, EV/EBITDAX, P/CF and NAV/PV-10 at your own deck; state FCF yield at two or three named crude prices.
- [ ] Use SOTP for any Indian integrated or conglomerate energy name, with a stated holdco discount on listed stakes.
- [ ] Build the peer set within sub-sector, fiscal regime and price exposure, restated to a common price deck; never blend upstream and refining.
- [ ] Scan for the flatter list: inventory-gain-driven beats, PUD rebooking, ARO discount-rate changes, trading profits without losing quarters, hedge gains inside EBITDA, stagnant CWIP, reduced segment granularity.
- [ ] For PSUs, quantify the policy and governance discount explicitly rather than treating the low multiple as an opportunity.
- [ ] Test the terminal assumption: for upstream run to reserve exhaustion plus risked exploration; for refining ask survivor-or-closure at 15 years; for CGD ask regulatory continuity.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| add-product-e2e-evalpaperclipai/paperclipAdd or extend a Paperclip full-stack runner E2E workflow, fixture, matcher, or report evidence path for local or Daytona execution. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| add-runner-evalpaperclipai/paperclipAdd or extend a Paperclip Runner protocol evaluation definition, roster, assertion, or report fixture with provenance and narrow validation. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| diagnose-why-work-stoppedpaperclipai/paperclipCompatibility entry point for failed, stalled, looping, or over-recovered Paperclip work. Use fix-failed-paperclip-run for the full diagnosis, regression, repair, real verification, and green-PR workflow; honor an explicitly requested diagnosis-only scope. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |
| fix-failed-paperclip-runpaperclipai/paperclipDiagnose and fix failed, stuck, slow, or incorrectly blocked Paperclip tasks and runs. Use with a task/run URL or failure report to trace the root cause, prove a regression red then green, check second-order effects, verify the real workflow, and deliver a PR with passing checks. | Stars: 115k | Updated | Claude CodeCodex | Verified | Install |