Math Models

Both models were built from each company's actual revenue mechanics and checked against reported numbers before being handed to you: the HCL model reproduces FY26 revenue within 0.2% and EBITDA within 0.7%; the Sunlite identity reproduces the Q1 FY27 business update within ~1%. Drag the sliders — everything recomputes live. Defaults are the base case.

Sunlite — spread machine

Revenue = volume × ₹-copper (pass-through) · Profit = volume × EBITDA/t

The copper price sets revenue optics and inventory gains; the spread (EBITDA/t) sets profit. EBITDA/t anchors: Sunlite H1 FY26 ~₹17.5k → H2 FY26 ₹28.6k → your FY27 band ₹29–34k; peers Q1 FY27: Jain ₹31k, Pondy ₹48.5k, Gravita/RMIL ₹55k. PAT bridge uses the delivered ~67–68% EBITDA→PAT conversion (D&A+interest ₹8 cr, 25% tax), plus aluminium PAT.

Copper volume (MT/yr) — Q1 ran 5,74325,000 MT

Blended EBITDA/t — H2FY26 ₹28.6k, gates ₹25k/₹30k29,000 ₹/t

LME avg ($/t) — revenue optics only13,400 $/t

USD/INR — FY26 avg 87.9, spot 95.2595.50

Aluminium (SAPL) PAT (₹ cr)5 cr

Revenue (₹ cr)

3,253

metal pass-through — optics

Copper EBITDA (₹ cr)

72.5

vol × spread

Consol. PAT (₹ cr)

53.4

67–75% conv. + SAPL

EPS ₹ · P/E @ ₹477

38.4 · 12.4x

FY26 actual: ₹29.1 · 16.4x

Reference points auto-shown: FY26 actual (25.7k MT, ₹23.2k/t blended, PAT ₹40.1 cr) and the corrected version of your note's base case. Revenue is informational — judge the P/E row.

Hindustan Copper — price machine

EBITDA = MIC × (LME × f × (1 − royalty) − cash cost) × INR

f is the realization factor — reported revenue per tonne ÷ (avg LME × INR). FY26 implied f ≈ 1.10–1.13 (Q4-weighted dispatches into rising prices + by-product credits + e-tender premiums); 1.05 is the conservative default. Royalty is ad valorem (~8.4%, slider up to 12.6% for the 1.5x-royalty risk Anand Rathi flags). Cash cost $5.7k/t calibrates to FY26's all-in ex-royalty cost. PAT = (EBITDA − D&A ₹230 cr + other income ₹60 cr) × 75%. Shares 96.65 cr.

MIC sales (t) — FY26: 27,367; plan implies ~30–33k30,000 t

LME avg ($/t) — deck: base ’26 $13.6k → ’27 $13.3k13,400 $/t

USD/INR95.50

Realization factor f — FY26 implied ~1.131.05×

Royalty (% of realization) — 1.5x risk = 12.6%8.4%

Cash cost ($/t MIC) — guided $5–6k5,700 $/t

Revenue (₹ cr)

4,031

FY26 actual: 3,078

EBITDA (₹ cr)

2,059

FY26 actual: 1,462

PAT (₹ cr)

1,417

FY26 actual: 921

EPS ₹ · P/E @ ₹540

14.7 · 36.8x

TTM: ₹9.5 · 56.9x

Calibration: at FY26 actual inputs (27,367t, ~$11,350 avg LME, INR 87.9, f 1.13) the model returns revenue ₹3,085 cr vs actual ₹3,078 cr and EBITDA ₹1,472 cr vs ₹1,462 cr. Sensitivities at base: +$1,000 LME ≈ +₹265 cr EBITDA; +₹1/USD ≈ +₹22 cr EBITDA; +1,000t MIC ≈ +₹69 cr EBITDA.

The structural contrast these two models encode: Sunlite's profit is volume × spread — copper price only leaks in through inventory and payability, so a crash dents it but doesn't destroy it (and cheap scrap can even help). HCL's profit is volume × (price − cost) — price is ~everything, in both directions: the same $13.4k→$10.8k move that trims Sunlite's EPS ~25% cuts HCL's roughly in half. That asymmetry is why the pair actually diversify each other despite both being "copper stocks".