For & Against

1. Extreme valuation gap against cash-generation reality. 8.94x P/E and 6.09% dividend yield on a business still doing 48% RoCE, ₹29,200 Cr operating cash flow, and ₹99,105 Cr of equity with ₹9,146 Cr of debt (9% D/E). Quant's peer scatter makes the point: HINDZINC earns 61% RoCE and trades at 21x; CIL earns 48% and trades at 9x. The market is charging a ~55% transition discount on a business where coal is still 75% of Indian electricity and thermal capacity is growing.

2. The dividend floor is structural, not discretionary. The Government of India needs the ₹10,300 Cr CIL pays it annually at 63.13% ownership — Sherlock flagged dividends as effectively ring-fenced by fiscal math. Even through FY25's earnings decline, DPS held at ₹26.5. At ₹435, a continuation of the ₹26.5 DPS is a 6.09% yield with AAA credit behind it, and CIL has never cut the dividend. That is a real floor, not a narrative floor.

3. The wage risk is finally quantifiable and partly pre-funded. Sherlock confirmed management has already taken a ₹2,201 Cr executive pay provision in Q3 FY26 — a preview, not a one-time. Historian's guidance-miss table shows management tells the truth about costs even when they miss on volumes. The June 2026 NCWA-XI settlement will be a shock, but it will also clear a five-year overhang — the last wage-reset cycle (FY17-18) preceded the FY22-24 earnings tripling.

4. A real near-term catalyst: e-auction normalising off a crushed base. Warren's read is that 48% premium in FY25 was near the floor of the normal 40–70% band; 9M FY26 has already ticked back up to ~57-60%. If Q4 FY26 prints e-auction realisation above ₹2,600/t (still well below the ₹3,321 Q3 FY25 peak), operating margin should rebuild 300-500 bps and the 22% 9M PAT decline stops being a trend.

1. Volume miss is now a pattern, not an excuse. Historian's guidance table is damning: FY22 missed by 7%, FY25 missed by 7%, and 9M FY26 production is down 3% YoY against 875 MT guidance implying a ~12% miss before Q4 monsoon risk. 1 BT has been pushed FY24 → FY25 → FY26 → FY27 → FY29. Meanwhile captive/commercial miners are on track for 320 MT by FY30 — directly eroding CIL's share from 83% to 75% and falling. Volume was supposed to be the one variable CIL fully controls; it isn't.

2. RoCE has halved in two years and that math propagates. Quant's RoCE series: 78% (FY23) → 64% (FY24) → 48% (FY25). Each 10-point RoCE drop on ₹99,105 Cr equity is ~₹9,900 Cr of earnings. If FY26 prints in the mid-30s (plausible given Q3 margin of 27% plus the June wage reset), the 8.9x P/E isn't cheap — it's fair for a ~30% RoCE business in secular decline. Warren's "cash cow runway" question becomes a number, not a narrative.

3. Unprovisioned ₹35,000 Cr MMDR royalty contingency is ~13% of market cap. Historian flagged this as the quiet risk that moved from intensity 2 (FY21) to 5 (FY25). Management refuses to provision — states have begun filing demands. If even a third of this crystallises over FY27-28, it wipes out a full year of dividend capacity. This is the kind of overhang that doesn't move the stock on the news — it suppresses the multiple permanently until it's resolved.

4. Diversification capex is dilutive by design. ₹16,000 Cr/year into 9.5 GW renewables, coal gasification, critical minerals, Hindustan Copper MoU, DVC thermal JV — Warren priced these at sub-10% returns versus the 48% RoCE core. Sherlock noted zero management equity upside if it works. This is a government-directed pivot away from the 48% RoCE business into 10% RoCE businesses, funded by the cash that would otherwise be dividend. It is the exact mechanism by which "fortress balance sheet" becomes "mediocre returns" over a decade.

5. Governance ceiling is permanent, not fixable. Sherlock's C+ grade rests on a structural fact: CIL itself told SEBI it "cannot comply" with LODR independent-director requirements because appointments sit with the President of India. The board has one independent director with real finance experience, none with mining/energy/capital-markets depth, and a CMD whose personal holding is zero. The multiple discount versus HINDZINC is not irrational — it's the market pricing that shareholders ride along on dividends, not decisions.