← all companies · 2026-07-07 · moat: narrow · raw .md

Reliance Industries: The Company That Learned to Break Molecules

Begin with a drop of crude

Hold an imaginary drop of crude oil in your palm. It is black, sticky, and — this is the surprising part — almost useless. You cannot run a scooter on it. You cannot make a shirt from it. You cannot even burn it cleanly in a lamp.

Yet that drop contains extraordinary treasure, and the treasure is chemical. Crude oil is not one substance but a crowd of thousands of different molecules jumbled together, almost all of them built from just two kinds of atoms: carbon and hydrogen. Chemists call them hydrocarbons. Some are short and light — a handful of carbon atoms holding hands in a chain. Those make cooking gas and petrol. Some are long and heavy — chains of twenty, thirty, forty carbons. Those make diesel, wax, tar.

Nature mixed them; industry must sort them. The first sorting trick is beautifully simple: boil the crowd. Short molecules are light sleepers — they evaporate at low temperatures. Long molecules need serious heat before they let go of each other. So a refinery's first machine is essentially a giant kettle called a distillation column: crude goes in at the bottom, heat rises, and different molecules step off at different floors, like passengers leaving a lift — gases at the top, petrol a few floors down, diesel below that, tar at the ground floor.

But here is the problem that built Reliance. The world wants far more of the short, valuable molecules than nature put into the barrel. What do you do with the long, cheap ones?

You break them.

A carbon–carbon bond is a strong piece of engineering; a molecule will not snap just because you ask politely. You must either hit it with ferocious heat, or — the cleverer route — introduce a catalyst: a specially designed material, often a mineral riddled with molecule-sized pores, on whose surface the big molecule is held, stressed, and cracked into smaller pieces, at hundreds of degrees, again and again, tonne after tonne, day and night. This is called cracking, and it is the heart of every modern refinery. A refinery is best understood not as a factory but as a chemical city: kilometres of steel piping, reactors taller than buildings, its own power plants and port, all arranged so that a molecule entering as sludge leaves as petrol, or as the raw material for polyester, plastic, and paint.

Once you see this, you understand why the business exists. Reliance's founding machine — its Jamnagar complex on the Gujarat coast, the largest refining site on Earth — is a molecule-breaking city. And you can already guess the economics: cities are staggeringly expensive to build. Physics and chemistry set the entry price, and the entry price is measured in lakhs of crores.

Why breaking molecules pays

Society pays Reliance for a simple reason: civilisation runs on the short molecules and dresses in the medium ones. Fuel for transport. Gas for kitchens. And — less obvious but just as important — petrochemicals: the plastics in your phone, the polyester in most Indian clothing, the packaging that gets milk to your door. Roughly everything in a modern room that is not metal, wood, or cotton began life in a cracker.

But the value is not spread evenly along the chain, and the unevenness is chemical. Pumping crude out of the ground is geology's business, and the profit there swings wildly with world prices. Selling petrol at a pump is a retail business with regulated, wafer-thin margins. The interesting middle — where Reliance chose to live — is complexity. A simple refinery can only digest expensive, high-quality crude. A highly complex one, with more cracking and treating units, can buy the world's cheapest, dirtiest barrels and still squeeze out the full menu of valuable products. The difference between what cheap crude costs and what its cracked products sell for is the refiner's margin, and the more complex your chemical city, the wider that gap. Complexity is bought with capital and engineering skill — once bought, it is a cost advantage a simpler rival cannot copy without spending a decade and a fortune.

Notice, though, what this margin is not. It is not pricing power. Petrol, diesel, and polyester are commodities: no customer pays extra because a molecule came from Jamnagar. Reliance's original business wins by being a lower-cost producer of things everyone sells, never by charging more. Keep that thought; it explains the company's whole second act.

From a yarn trader to a chemical city

The oil industry is barely 160 years old. It began with lamps, not cars — early refiners wanted kerosene for light and threw the petrol away. The twentieth century reversed that: the motor car, the aeroplane, and then the Second World War's demand for fuel drove furious invention, including catalytic cracking itself. By mid-century, oil was the bloodstream of the world economy, and countries without it — India prominently — learned the pain of importing their energy.

Reliance's own story enters from an unexpected door: cloth. Dhirubhai Ambani, the founder, began as a trader and then a maker of polyester yarn. Polyester is a hydrocarbon product; every step of its making depends on chemicals refined from crude. Rather than remain a customer of those chemicals, the company marched backwards down the chain, decade by decade: from fabric, to fibre, to the chemicals that make fibre, and finally, in the year 2000, to the refinery itself at Jamnagar — later doubled, making it the world's largest. Few companies anywhere have integrated backwards so far, so deliberately. The family of the founder still owns about half the company — 50.00% at last count — which tells you the architects still live in the building.

Then came the strangest chapter. In the 2010s, Reliance took the enormous cash its molecule-city generated and spent it on something with no molecules at all: a telecom network. Jio launched in 2016 with a nationwide 4G network and prices so low they rearranged the entire industry. A dozen operators became roughly three. Alongside it grew Reliance Retail, now among the country's largest store networks. The refinery had financed two new empires.

The price of admission

Suppose a well-funded rival decided today to become Reliance. What is the bill?

To replicate Jamnagar alone: land on a coast, a private port, years of government clearances, and a construction project that would consume lakhs of crores and most of a decade before the first rupee of profit — all to enter a business with commodity margins. To replicate Jio: nationwide spectrum bought at auction, hundreds of thousands of towers and kilometres of fibre, and the stomach to lose money for years against an incumbent famous for price war. To replicate the retail arm: thousands of stores and warehouses assembled over decades.

For scale: the whole company's shares can be bought today, in principle, for about ₹17.9 lakh crore, and even its existing borrowings stand at ₹4.03 lakh crore against reserves of ₹8.9 lakh crore. The admission price to this industry is not a barrier that clever engineering can tunnel under; it is the physics. Heavy molecules need heavy steel; radio waves need towers every few hundred metres. In both worlds, scale is the advantage, and scale is bought with sums that make new entrants extremely rare.

Charlie's ledger: three businesses, one balance sheet

Now let us reason the way Charlie Munger taught — several disciplines at once, starting with incentives.

Half the company belongs to one family. That cuts both ways. The good side: decisions are made for decades, not quarters — nobody builds Jamnagar or endures Jio's early losses to please next year's analyst. The other side: a conglomerate run from one chair depends utterly on the judgement in that chair, and succession — the handing of three empires to the next generation — is a risk no spreadsheet can price.

Next, opportunity cost — the discipline of asking what else the money could have done. Here the record is honest but mixed. Over ten years, sales compounded at 15% a year: from ₹3.74 lakh crore to ₹10.56 lakh crore, a near-tripling. Profit compounded at 10%. Yet return on equity — the profit earned on each rupee owners left inside — has sat at about 9% in every window you check: last year 9%, three years 9%, five years 9%, ten years 9%. Nine percent is what the arithmetic of giant, capital-hungry projects does: each new empire swallows the profits of the last. An owner must ask Munger's question: is management's next lakh crore likely to earn more than the owner could earn elsewhere? For thirty years the family's answer has been "trust us, the next machine will pay," and — to be fair — Jio and retail were exactly such machines.

Where are the true moats? Not in refining: cost advantage there is real but rests on commodity spreads the company cannot set. The durable structures are the new ones. Jio sits inside a three-player industry where the physics of spectrum and towers forbids easy entry — an oligopoly, which is to say a market where a few sellers face each other like cautious chess players rather than price-cutting brawlers. Retail enjoys scale in buying and logistics that a regional chain cannot match. And between the businesses run feedback loops: the network carries the commerce, the stores recharge the phones, one identity spans them all. There is optionality too — the declared push into new energy — though an honest analyst records that today it is an option, not an earnings stream.

Invert, always invert. What would kill Reliance? A world that rapidly stops burning transport fuel would hollow out the founding business. A second telecom price war — or a regulator deciding three players is one too few — would squeeze Jio. And the oldest conglomerate disease, capital spread too thin across too many ambitions, would show up exactly where it always does: in that stubborn 9% return on equity, refusing to rise.

What could management realistically do this decade to widen the moat? Three things. First, let the consumer businesses — where pricing power actually lives — grow to dominate the profit mix. Second, prove capital discipline by letting return on equity climb toward the mid-teens instead of launching a fourth empire. Third, convert the green-energy option into a real business before the old molecule business ages, not after.

Warren's turn: the owner's arithmetic

Imagine, as Warren Buffett likes to, that you bought the entire company — every share, about ₹17.9 lakh crore. What do you own?

You own last year's profit of ₹95,754 crore. That is roughly a 5.4% starting yield on your purchase — the market charges about 23 times earnings (a price- to-earnings ratio of 23 means you pay ₹23 today for each ₹1 of current annual profit). You own ferocious cash generation: ₹1.92 lakh crore came in from operations last year, twice reported profit — though an owner must immediately note that in this company cash has never rested; it is forever being poured into the next machine, so the "owner earnings" you could actually take home have historically been slim. The dividend payout confirms it: about 10% of profit in almost every one of the last ten years.

Predictability is split down the middle. The consumer half — telecom subscriptions, groceries, clothing — is about as predictable as Indian business gets: hundreds of millions of small payments, repeated monthly. The molecule half is hostage to world commodity cycles the company cannot influence. Pricing power follows the same divide: some in Jio and retail, none at the refinery gate. Operating margins have nonetheless been steady — 17% last year, within a point or two of that for a decade — which for a business this size is quiet evidence of competence.

Shareholders have not been quietly diluted: earnings per share nearly tripled over the decade, from ₹17 to ₹60 (the share count grew mainly through bonus issues, which split the same pie into more slices for the same owners). And the register tells its own story: the number of shareholders grew from 35 lakh to 44 lakh in under three years, with domestic institutions steadily replacing foreign ones.

Is this a business that compounds value for decades? The honest answer: parts of it, yes. The verdict here is a narrow moat — wide around Jio's oligopoly and retail's scale, nonexistent around the founding molecule business, and the blended owner's return of 9% on equity is the arithmetic proof of that mixture. A wonderful collection of assets; not yet, taken whole, a wonderful compounding machine.

What could crack the cracker

The structural risks — the ones that matter over decades, not quarters:

The energy transition. Electric vehicles need no petrol. Fuel demand in India will grow for years yet, but the direction of the century is set. Petrochemicals — plastics, fibres — will outlive fuels, and Jamnagar can tilt further toward them, but the founding river narrows.
Regulation. Spectrum policy, telecom tariffs, retail rules for foreign competition, and energy pricing are all written in Delhi. A company this central to the economy lives permanently in the state's peripheral vision.
Commodity dependence. Refining margins are set by global spreads; crude supply is set by geopolitics. Half the company's engine room takes its weather from abroad.
Capital misallocation. The signature risk of every conglomerate with heroic ambitions and one decision-maker. The green-energy build-out will test discipline at lakhs-of-crores scale.
Succession. Three empires, one family, one transition to manage.

The half-century view

Look out twenty or thirty years and reason from structural forces. India will likely be the world's largest growth market for energy, data, and organised retail simultaneously — the only country where all three curves are still steep. The molecule business will slowly change shape: fuels fading, chemistry enduring, because a decarbonising world still wants polymers, and someone must make them efficiently at scale. The data business sits in front of the century's surest trend — every Indian service, from banking to education to the coming wave of artificial intelligence, travels over networks of which the country has, in effect, three. And retail rides the simplest force of all: several hundred million households getting richer.

Reliance's founding skill was never oil. It was the nerve to build the biggest machine of whatever era it was living in, and to pay for the next machine with the last one's cash flow. Whether that skill survives its transfer to a new generation — and whether the returns on the next machine finally reward the owners as richly as the ambition — is the question the next few decades will answer.

An Omaha Investments chapter. Educational material, not investment advice. Figures from Screener.in and NSE data via Angel One as of the date above.