Reliance Industries: The Energy Elephant That Danced | The Spotlight ft. Probal Sen

ELI5 / TLDR

Reliance is so large it pays roughly 3.5% of India’s corporate tax and supplies a chunk of the country’s gas, fuel, plastic, telecom, and groceries. The engine underneath all of it is a deeply unglamorous business: turning crude oil into chemicals at a single colossal site in Jamnagar. The trick was never one clever invention. It was owning every link in the chain — from the barrel of crude to the toothpaste tube — so the company captures a margin at each step and answers to no supplier. That cash machine quietly funded Jio and Retail.

The Full Story

Probal Sen covers energy at ICICI Securities, and Krish Kothari sits him down to explain the one part of Reliance nobody talks about at dinner parties: the oil-to-chemicals business. It is also the part that pays for everything else.

The scale, stated flatly

The numbers are almost rude. First Indian company past 10 lakh crore in revenue. Market cap around 16.4 trillion rupees. It alone pays more than 3.5% of India’s corporate tax — roughly 25,000–26,000 crore last year. Add duties and profit petroleum and the total handed to the exchequer is about 1.86 lakh crore, somewhere near 14% of India’s budget. It is 16% of merchandise exports, 28% of gas production, 27% of refining capacity, 40% of petrochemical capacity, half the telecom subscribers. Retail revenue is six times its nearest rival, DMart. Sen lists this without raising his voice, which is the right way to deliver it.

One idea, repeated for sixty years

The founding insight, traced back to Dhirubhai, was not a product. It was a structure: control the value chain. Capture the margin at every stage rather than buying intermediates from someone else and taking only the final markup.

“He immediately realized that you have to control the value chain if you have to basically capture margins at every stage of whatever industry you’re into.”

Petrochemicals make this unusually valuable because the sub-segments move out of sync. Polymer margins can be strong while raw-material prices are weak; polyester can be soft while something upstream is hot. A company present at every stage rides the whole curve instead of being squeezed at a single point. A player who only does final manufacturing has to buy his intermediate at whatever the market dictates — he controls nothing upstream.

Two more founding traits: an ambition for world scale that was genuinely strange for an Indian company in the 1970s, and a talent for raising money in a controlled economy. The famous AGMs in football stadiums were not theatre for its own sake — bank financing simply could not stretch to the scale Dhirubhai wanted, so he built the equity cult instead. Sen’s summary: he never let the constraints of the operating environment hold him back.

What “integrated margin” actually means

Here Sen earns his fee. Refining is, stripped down, distillation — heat a barrel of crude and different hydrocarbons boil off at different temperatures. The light, high-energy stuff (LPG, jet fuel, gasoline) at the top; diesel in the middle; the heavy residue (petcoke, bitumen) at the bottom. Cracking is the next step: breaking a simple molecule like ethane or propane — applying heat and a catalyst — into ethylene or propylene, which then become the polyethylene we call plastic.

Reliance feeds its petrochemical plants three ways at once: naphtha from its own refinery, ethane imported directly from the US, and refinery off-gas captured by a dedicated cracker. The integrated margin is the sum of the spreads at every handoff — raw material to intermediate to final product — all kept in-house instead of paid away.

The ethane move is a good example of opportunism with infrastructure. Four or five years ago, with US gas production surging and prices low, Reliance signed to buy ethane directly and built Very Large Ethane Carriers to ship it — ethane, like LNG, has to be liquefied to move economically. They saw a surplus coming and bought the ships to capture it.

The crude side is the same logic in reverse. Final product prices are pinned to a benchmark — Brent, WTI, Dubai — and nobody, not even Reliance, can change that. But if your crude costs a couple of dollars less than the benchmark, that gap is pure captured value. So Jamnagar is built to swallow inferior crude: heavier, higher in sulphur, harder to process, therefore cheaper — and still produce the same export-grade product.

“The refining complex at Jamnagar has till date processed some 216 different kinds of crude… nobody else has processed probably more than 50, 60 kinds. Nobody tries to do that.”

Why no one copied it

Not for lack of trying. ONGC’s OPaL petrochemical plant kicked off around 2004, finished five to seven years late, and the cost overruns broke the economics. Government oil marketing companies face structural handicaps Reliance never had: refineries placed in the Northeast or landlocked in Punjab for strategic reasons rather than commercial ones; a rigid crude-tendering process that cannot pounce on a stranded cargo the way Reliance can; and a product slate dictated by domestic demand rather than by spreads — if the country needs diesel, a state company makes diesel regardless of the margin.

Reliance also had the first-mover prizes: 4,000-plus acres at Jamnagar, the size of a city, beside a world-class port. It has taken roughly thirty years for anyone to catch up — HPCL is only now building a 9-million-tonne refinery that converts 30% straight to petrochemicals, the first of its kind among the state companies.

Speed as a moat

The recurring theme is execution velocity. The first Jamnagar refinery took three to four years; the second, around 29 million tonnes, was built in roughly 36 months — unheard of, then or now (no Indian greenfield refinery since has come up in under five years). A contractor, Bechtel, reportedly stood up 19 dedicated global offices to service that one project and said it had never deployed manpower on that scale. The same pattern recurs: Jio went from nothing in 2014 to 480 million subscribers in under a decade while making data among the cheapest on earth; Retail went from its first thousand stores by 2010 to roughly 19,000 stores and 79 million square feet today.

Sen is honest that he cannot fully explain how, and equally honest that it is not all founder magic — long-tenured operators like Manoj Modi and PMS Prasad are clearly instrumental. And it is not all success. The US shale acquisitions failed and were written off to the tune of $7 billion-plus; the first telecom foray was a famous misstep. The model produces more wins than losses because of the appetite for risk, not despite it.

The cash machine

This is the crux. Oil-to-chemicals is 40–45% of EBITDA and closer to 55–60% of operating cash flow. Margins are cyclical — three or four good years, three or four lean — and have swung from 9–10% to over 20%, sitting around 13–14% now. Petrochemicals are in their most prolonged weak patch in two decades, courtesy of enormous Chinese and Middle Eastern capacity additions colliding with a slumping Chinese property sector. Even so, because the bulk of the plant was built thirty or forty years ago at a fraction of today’s cost, O2C still throws off roughly 20%-plus returns while the consolidated number — dragged by all the new capex — sits nearer 7–8%.

The second engine is financial. Reliance’s credit rating has never wavered below the highest any Indian corporate gets, so it borrows huge sums cheaply — roughly 90,000–95,000 crore of capex a year over the last decade, nearer 1.4 lakh crore lately. The stated philosophy: if debt costs 300–400 basis points less than equity, raise the debt. And when investors ask for fatter dividends, the answer is that the company believes it can compound the money better than they can. It also runs a treasury sitting on 1.2–1.3 lakh crore of cash at any moment — Sen compares it, half-apologetically, to Apple’s finance arm — generating an other-income stream large enough to be a business on its own.

Where it goes next

The direction Mukesh Ambani has signalled is to push more of every barrel away from transport fuel — the bit threatened by EVs — and into petrochemicals and specialty chemicals, where demand is stickier (there is no green substitute at scale for the plastic in a toothpaste tube or a packet). The far edge of that is the Middle East model — SABIC-style plants that convert crude straight to chemicals with no refined fuel at all. Alongside sits the new-energy bet: solar, green hydrogen, fuel cells, with a 100-gigawatt aspiration that, if it lands, would dwarf the current O2C business.

On hydrogen Sen is sober. Green hydrogen costs around 5–6 (rupees/dollars) a kg today; to compete it needs to fall below 2, which requires cheaper renewable power and real storage. Globally the momentum has cooled. He thinks it still needs years and an inflection point like the one solar already had.

Asked the blunt question — will anyone ever catch Reliance — Sen splits it. In telecom, Bharti is holding its own. In energy, the state oil companies and especially Adani in renewables have serious ambition. But in retail, and as a total package, he cannot see anyone coming close in the next five to seven years. The structural advantage is the willingness to front-end enormous capex and then wait six years to monetize it — Retail went from 30,000 crore of quarterly revenue to 90,000 crore-plus on infrastructure laid in 2021–22. That requires a balance sheet, a risk appetite, and an execution muscle that are hard to assemble at once.

Key Takeaways

Reliance pays ~3.5% of India’s corporate tax; total payments to the exchequer (~1.86 lakh crore) are near 14% of India’s budget. It is ~16% of merchandise exports, 28% of gas, 27% of refining, 40% of petrochemical capacity.
The founding strategy is vertical integration to capture margin at every stage of the chain. It matters most in petrochemicals because sub-segment cycles move independently — an integrated player rides the whole curve.
“Integrated margin” = the summed spreads from raw material → intermediate → final product, kept in-house. A pure final-stage manufacturer can’t control his intermediate cost.
Cracking = breaking a simple molecule (ethane, propane) with heat and a catalyst into ethylene/propylene, which become polyethylene (plastic). Refining = distillation, separating crude into light/middle/heavy fractions by boiling point.
Reliance feeds petrochemicals three ways: own-refinery naphtha, directly imported US ethane (shipped in Very Large Ethane Carriers), and captured refinery off-gas (the ROGC cracker).
Crude-cost edge: final product prices are pinned to a benchmark (Brent/WTI/Dubai) nobody can move; buying cheaper inferior crude (heavy, high-sulphur) and still making export-grade product captures the gap. Jamnagar has processed 216 crude varieties vs ~50–60 for the largest global refiners.
State oil companies are handicapped by strategically-placed refineries, rigid crude tendering (can’t grab a stranded cargo), and a product slate dictated by domestic demand rather than spreads.
Jamnagar’s second refinery (~29 MT) was built in ~36 months; no Indian greenfield refinery since has been built in under 5 years.
O2C is 40–45% of EBITDA, ~55–60% of operating cash flow. EBITDA margins have swung 9–10% to 20%+, ~13–14% now; petrochemicals are in their worst cyclical patch in two decades (Chinese/ME oversupply + Chinese property slump).
Because the plant was built decades ago at low cost, O2C still earns ~20%+ returns even as consolidated ROE sits ~7–8% (dragged by Jio/Retail capex).
Financial engine: top-tier credit rating → cheap debt; ~90,000–95,000 crore/yr capex (recently ~1.4 lakh crore); a treasury holding 1.2–1.3 lakh crore cash. Capital-allocation philosophy: borrow when debt is 300–400 bps cheaper than equity; retain cash because management believes it compounds better than dividends would.
~$10–12 billion invested in Jio, ~$5 billion in Retail to date. US shale acquisitions were written off ~$7 billion+.
Forward strategy: convert more of every barrel from transport fuel into specialty chemicals (SABIC/Middle East model = crude straight to chemicals, no refined fuel), plus a 100 GW new-energy aspiration.
Green hydrogen needs to fall below ~2/kg from ~5–6 today to compete; global momentum has cooled and it likely needs years plus a solar-style inflection point.

Claude’s Take

This is a clean, high-signal analyst conversation with almost no fluff, and Sen is unusually willing to say “I don’t know how they do it” rather than manufacture a tidy thesis. The explanations of integrated margin, cracking, and the crude-discount play are the genuinely useful core — that is the mental model the episode delivers, and it delivers it well.

Two caveats keep it at a 7. First, it is admiring throughout; there is no real bear case beyond “petrochemicals are cyclically weak right now.” The governance, related-party, and conglomerate-discount questions that a sceptic would press on simply don’t come up — this is a how-the-machine-works episode, not an is-the-stock-cheap one, and that’s fine as long as you read it that way. Second, several numbers are delivered loosely from memory (“I think more than 1 lakh, 1 lakh 20,000, 1 lakh 30,000 crore”) — directionally right, not precise. As an explainer of why the O2C business is a structural cash machine, it’s excellent. As a complete view of Reliance, it’s one well-lit room in a large house.