Case Study 04 — Reliance · Dhirubhai Ambani · US→India Scrap Trade Playbook

§1

The short version

Reliance begins as an import–export desk in a 350 sq ft Masjid Bunder office trading spices (export) and polyester yarn (import). It takes eight years to build the first textile mill, then uses the export business to fund backward integration — from yarn to fabric to petrochemicals to a refinery that became the world's largest grass-roots refining complex. In the twenty-first century the same company — from the same trading discipline — built an integrated Indian consumer economy: telecom (Jio), retail, media. The headline: FY25 revenue of ₹10,71,174 crore, net profit ₹81,309 crore, the first Indian company past a ₹10 lakh crore net worth.

The transferable insight is not “be twice the GDP of a small country.” It is the earliest sentence: in 1958 this was precisely the business our §9 plan describes — thin capital, thick information, arranging flows between markets where the seller and buyer do not know each other.

§2

Where it started — the trading desk again

1958
Reliance Commercial Corporation: commodity trader in a 350 sq ft office — exports spices, imports polyester yarn; the leanest possible founding model.
1966
Reliance Textiles Industries — first textile mill at Naroda, Gujarat; the Vimal polyester brand follows. Trade profits now fund a factory, not the reverse.
1977
The equity cult: Reliance's IPO is massively oversubscribed (reported ~7×), turning ordinary Indians into shareholders — the first mass ‘reverse diaspora’ capital raise.
1980s–90s
Backward integration, twice: Patalganga (petrochemicals), Hazira (cracker), then Jamnagar — the world's largest grass-roots oil refinery (single train, built 1996–99; 25th anniversary celebrated in 2025).

The logic was always the same and always the trader's logic: own the input that your customer's customer depends on. First polyester yarn, then the base chemical, then the crude barrel — the same “control your input cost” reading that Agarwal reached with copper, reached earlier by Ambani with thread.

§3

The consumer pivot — what a trading empire looks like at 60

2016
Jio launches — free-data pricing rewrites Indian telecom; the network went from zero to options for most of the country's phone users in a decade.
2010s
Reliance Retail scales: ~19,340 stores, 77.4M sq ft, 349M registered customers, 1.39B transactions in FY25 (gross revenue ₹3,30,870 cr).
2020s
Green/tech third act: 1,000+ in-house scientists; new energy capex; consumer businesses now contribute >50% of group EBITDA.

Read the fiscal relics of the trading desk in every segment name: the O2C business (oil to chemicals) still produced ~51% of FY25 revenue. The empire never stopped being a downstream of the original import–export trade; it just integrated every layer of it.

§4

The numbers — FY25 (Reliance Industries, official)

Group

Revenue ₹10,71,174 cr (~$125B), +7.1% YoY
EBITDA ₹1,83,422 cr, +2.9%
Net profit ₹81,309 cr — India's most profitable company
Net worth >₹10 lakh cr — first Indian company to cross it

Consumer engines

Jio: 488M subscribers, 191M True5G, ARPU ₹206.2; FY25 revenue ₹1,50,270 cr, PAT ₹26,109 cr
Retail: gross revenue ₹3,30,870 cr; 19,340 stores; 349M customers
O2C still ~51% of revenue — the trading core, integrated

Source: Reliance's FY25 results (Apr 2025) and the 48th Annual Report (Aug 2025), as reported by Groww, Upstox and company disclosures.

§5

The honest ledger

Raw-material volatility, again: fabric margins thrashed with yarn prices; the refinery margins with crack spreads. The response was integration — the same chapter as Vedanta's. Volatility is the recurring reason traders integrate.
Political-capital dependence: Reliance's growth phases track Indian policy regimes (licence raj exit, diesel price decontrol, telecom licence awards). The maharaja privilege cuts both ways — regulatory risk is business risk.
Debt-funded entrances: Jio's entry burned tens of billions before profit; the consumer pivot is a rolled bet, not a gradual climb.
Family succession: the 2004–05 brothers' split is the standard cautionary scroll on family-governed empires that outgrow their governance.

For us: the wrong lesson is “swing for the fences with leverage.” The right lesson is the one at the start — a paper-thin trading desk with obsessive price information and a plan to integrate what it learns.

§6

What this means for our playbook

Stage 11 customers are Reliance-scale: Indian mills we would sell scrap to include steel groups that are Reliance's peers in procurement habit — price, spec compliance, reliability. The case explains why they pay a premium to reliable traders: their own volatility history.
The catalog moat has a Reliance version: Reliance's moat was vertical (own the chain); Chem-Impex's is horizontal (own the catalog). Both began as desks. Our §6 models give us the choice — the cases say either can work.
Equity as fuel: Reliance's 1977 IPO turned customers into owners. For us: not now; the 104-company database and §9 debt-light model are our equity phase.
Reinvention cadence: every ~15 years Reliance changed its core (yarn → petchem → energy → consumer). A trading company that refuses reinvention stays a trading company; one that integrates when the trade pays for it does not have to.

Verdict

The desk was never small — it was phase one. Reliance is the full 67-year arc of the trade-first thesis: a ₹15,000 commodity desk became India's most profitable company by integrating, in turn, every input and output of its own trade. For our §9 plan, the case sets the ceiling and the floor: the floor is the 1958 desk (which we are building now); the ceiling is integration funded by trade, nothing more exotic, and nothing less honest.

§7

Reliance — Dhirubhai Ambani