Business

Claude View

Know the Business

Coal India is a regulated-price quasi-utility, not a commodity cyclical. Roughly 88% of volumes go to state-owned power utilities at long-term Fuel Supply Agreement (FSA) prices set administratively, with only ~10% sold at spot e-auctions where global coal prices bleed through. The market prices it at 8.9x earnings and 6.1% yield because the "moat" — 74% of India's coal output, FSA contracts covering ~629 MT/year — is inseparable from two ceilings: a government that uses CIL as a social-policy instrument (FSA prices, wage revisions, dividends) and a grid that is supposed to wean off coal. The honest question is not "is it cheap" but "how long does the cash cow last, and who gets the cash."

**Bottom line.** Volume growth, not price, is the story. FSA realisation rises ~2% a year by design; e-auction premium fell from 72% in FY24 to 48% in FY25 to ~60% today — so the incremental tonne now earns less per rupee. The stock discount reflects energy-transition risk + government capital allocation, not weak economics. ROCE of 48% and a fortress balance sheet say the core engine is fine; the question is the reinvestment runway.

1. How This Business Actually Works

CIL digs non-coking thermal coal out of 310 mostly open-cast mines in eastern India and sells it to power plants under decade-long contracts. The economics are unusual for a miner: the input is cheap coal in the ground (India has 212 billion tonnes of proven reserves, CIL sits on most of it), the output price is administered (FSA tariffs are revised infrequently, linked to notified prices set by CIL itself subject to government oversight), and the cost structure is dominated by people and contractors — wages are ~43% of cash opex, contractual mining is another ~28%. This is a cost-plus utility wearing a miner's overalls.

The three revenue streams — and which one moves the P&L

No Results

9M FY26 (Apr–Dec 2025). FSA realisation grinds up because of pass-throughs and periodic notified-price revisions; e-auction realisation tracks seaborne benchmarks and domestic availability.

The margin lever is e-auction, not FSA. E-auction premium collapsed from 72% over notified price in FY24 to 48% in FY25 as domestic supply normalised and Newcastle prices fell. In 9M FY26 it was still only ~60%. Every ₹100/tonne lost on e-auction is worth ~₹550 Cr of profit — which is exactly why FY25 PAT slipped and 9M FY26 is down 22%.

Where profit comes from: the subsidiary truth

CIL is not one company. It is seven coal subsidiaries + a design arm, and three of them earn almost everything. MCL (Mahanadi, Odisha), NCL (Northern, MP/UP) and SECL (South Eastern, Chhattisgarh) sit on thick, low-stripping-ratio seams near power demand centres. ECL, BCCL and WCL are high-cost legacy operations in old mining belts — BCCL actually flipped to a loss in 9M FY26.

Read the chart this way: MCL + NCL = 77% of consolidated PBT from 35% of production. Those are the reserves worth owning. The rest is labor-and-pension drag with optionality on coal gasification / land / FMC infrastructure.

The cost side — what makes ROCE 48%

Employee costs (~₹47,000 Cr annualised), contractual mining (~₹30,000 Cr), and stripping activity adjustment (a non-cash accounting credit that flatters operating profit by ~₹5,000 Cr) dominate. Depreciation is tiny because the ore body is owned outright and most capex is incremental. Other income is material — ₹9,930 Cr in FY25 — because CIL carries ₹99,000+ Cr of cash at any given time. Strip the cash and the "real" operating ROCE is still 35-40%, an exceptional number driven by almost zero working capital (WC days = –14) and near-zero debt.

2. The Playing Field

CIL has no domestic peer at scale — it is 10x NMDC's coal-equivalent output and ~20x NLC India's. The useful comparison is to other Indian PSU natural-resource businesses where the sovereign owns the resource, sets the price, and takes most of the cash out as dividends.

No Results

What the peer set actually tells you

Two honest observations. First, CIL and Hindustan Zinc are the only two businesses in this set earning ROCE north of 40% — both are captive-resource monopolies. Hindustan Zinc gets 21x earnings because zinc is a transition-metal tailwind with export pricing; CIL gets 9x because thermal coal is a transition-metal headwind with administered pricing. Second, every stock in this peer set trades below a "normal" materials multiple — the market discounts Indian PSU mining for capital-allocation risk (the sovereign uses them as piggy banks). CIL gets the steepest discount because the runway question is sharpest.

3. Is This Business Cyclical?

Yes, but not the way a global miner is. CIL's cycle is domestic power-demand × FSA price revisions × e-auction premium, not Newcastle coal prices. Volumes barely move (they only go up, by policy); earnings cycle through auction spreads and employee-cost resets.

The history tells three stories:

FY17–FY18 collapse (margin 11%): wage revision hit, e-auction premium dried up, imports flooded in. This is the template downside — a wage reset combined with weak e-auction is a 50%+ earnings hit.
FY22–FY24 boom (margin 33%): Russia/Ukraine pushed global coal to $400/t; e-auction premium spiked to 72%; CIL's spot revenue per tonne doubled even though FSA barely moved. This was once-in-a-decade and is now unwinding.
FY25 onward: volumes hit records (781 MT), but realisation is softening. Revenue flat YoY, PAT down 6%. 9M FY26 PAT down 22% because of a ₹2,201 Cr one-time executive pay upgradation + auction pricing normalising.

**The real cycle is labor, not coal.** Pay revisions arrive every ~5 years (the last NCWA wage agreement was 2017; a new one is overdue). When they land, employee cost steps up 15–25% at once. This is the single biggest recurring P&L shock — bigger than any likely coal-price move.

4. The Metrics That Actually Matter

Forget P/E and EBITDA — for CIL the five numbers below tell you everything about whether a given year is good or bad.

No Results

Two of these deserve a visual

Left chart: auction premium normalising is the reason FY24 was peak earnings. Right chart: CIL's share of India's coal has slipped from 83% to 75% as commercial mining auctions bring private capacity online — this is the quiet structural erosion the market is pricing in.

5. What I'd Tell a Young Analyst

This is not a "cheap coal miner" trade. It is a policy trade with an energy-transition clock on it. Three rules:

Stop modeling coal price; model the FSA-mix shift. Every year, more contracted volume gets signed at higher notified prices and less at spot, because that is what the government wants. The incremental tonne is worth ~₹1,500/te, not the blended ₹1,645. Extrapolating current EBITDA is wrong — the mix is deteriorating.

Watch the wage agreement, not the coal market. NCWA-XI expired in 2021; negotiations are underway. A 20% employee-cost step-up is a 25% earnings hit in the landing year and is not in consensus. The executive-cadre ₹2,201 Cr provision in Q3 FY26 was a preview — the workmen revision is 10x bigger.

The real thesis either goes "volumes to 1 BT" or "diversification actually works." Neither is free. Volumes need land, rail, and faster clearances — visible progress (FMC projects, mechanised loading up to 32%), but 1 BT by FY29 requires 5-6% CAGR from here and OBR has to rise faster still. Diversification (9.5 GW renewables, coal gasification, critical minerals REE block, Hindustan Copper MoU, thermal JV with DVC) is ₹16,000 Cr capex/year chasing sub-10% returns — directionally right, but earnings-dilutive for at least 5 years.

What would change the thesis, bullishly: a move to mandatory partial deregulation of FSA pricing (the "market-linked pricing" proposal that surfaces periodically); e-auction premium sustaining above 60% into a wage cycle; visible critical-minerals revenue by FY28.

What would break it: a wage agreement above 25%; loss of FSA volume share to private miners below 70% of India output; a dividend cut to fund diversification (this would be the tell that the government sees coal-demand deterioration faster than disclosed).

**The one-line test.** If you can't tell me next year's FSA volume, e-auction premium, and wage-agreement status, you don't have a view on COALINDIA — you have a view on P/E 8.9 vs 15, which isn't a view.