Samsung Sells You a Phone, Insures Your Life, and Built Your Apartment. One Family Holds It All by a Thread.

It started with dried fish. In 1938, Lee Byung-chul opened a trading company in Korea that dealt in dried fish, locally-grown groceries, and noodles.⁹ Black-and-white televisions came thirty-one years later, when Samsung Electronics was incorporated as a separate entity in 1969 and built its first sets in partnership with Sanyo.⁵ Today the group that grew from that grocery trade sells you a smartphone, fabricates the memory chip inside it, insures your life, and may well have poured the concrete of your apartment block. In 2024 its affiliates booked 331 trillion won — about 13% of all the economic activity in South Korea.³ That is the kind of breadth other companies dream about. It is also a trap with one door.

The official story is that Samsung is a sprawling, diversified empire — too big to fail, hedged against any single shock by the sheer range of things it does. The real story is the opposite. The very structure that ties chips to insurance to construction also funnels control of all of it through a single family line, which means the group's strategic coherence has at times depended on whether one man stayed out of prison.

The breadth is real, and so is the cash that powers it

Diversification at Samsung is not a portfolio of unrelated bets thrown together to look big. It is a machine where the profit centre subsidises the strategic ones. Samsung Electronics is the engine: in fiscal 2023 it posted KRW 258.94 trillion in consolidated revenue, generating KRW 6.57 trillion in operating profit even in a brutal year for memory chips.¹² That electronics cash flow is what makes the breadth viable. The semiconductor business — the world's biggest in DRAM and NAND¹⁰ — throws off enough capital to keep capital-hungry, slow-cycle businesses like shipbuilding and heavy construction patient through their down years. No single one of those businesses would survive the cycle alone. Inside a group, the cycles cancel out. That is the genuine logic of chaebol breadth: it is not diworsification, it is internal capital allocation at national scale.

The scale distorts the country it sits in. Samsung accounts for about 8% of South Korea's manufacturing output and roughly 17% of the main stock index's total market capitalization.⁴ When the group's largest division has a good quarter, the national figures move with it. This is what 'too big to fail' actually looks like: not a bank propped up in a crisis, but an economy whose growth rate is partly an arithmetic function of one company's chip prices.

Where the breadth quietly becomes a single point of failure

Here is the part almost everyone gets wrong. People picture Samsung as one giant corporation with a chairman at the top issuing orders down a clean org chart. It isn't. The group's many subsidiaries are legally distinct companies, each with its own board and its own shareholders. There is no formal holding company sitting above them. Instead, control is exercised indirectly, through a circular cross-shareholding web: Samsung C&T, Samsung Life Insurance, and Samsung Electronics hold interlocking stakes in one another, looping back around again.⁸ Credit Suisse analysts, trying to chart it, gave up on tidiness and called it 'extremely complicated with some circulars within the affiliates.'⁸ The point of the loop is leverage: the founding family directly owns only a sliver of the empire, but because each affiliate owns pieces of the others, a small stake in the right node controls the whole chain. Samsung Everland — now folded into Samsung C&T — historically served as the de facto apex of that web.⁸

A circular shareholding web is an elegant way to control an empire on a thin slice of equity. It is also a transmission line for risk. Because the chain ultimately resolves to the family's grip on a few critical nodes, anything that threatens that grip — a court case, a succession, a regulatory ruling on cross-holdings — threatens the coherence of the entire group at once. The breadth that was supposed to diversify away risk has quietly re-concentrated it in the one place no balance sheet lists: the personal legal standing of the heir.

When succession ran through a courtroom

The 2015 merger of Samsung C&T and Cheil Industries was not really an industrial deal. It was a succession mechanism — the move that consolidated the Lee family's control over the group's apex node and smoothed the handover to heir Lee Jae-yong. And it is precisely there that the structural fragility stopped being theoretical.

Read that sequence as a corporate event, not a personal one, and the fragility is impossible to miss. Lee was convicted in 2017, served eighteen months, was paroled in 2021, and formally pardoned in August 2022.⁶ Then in February 2024 a Seoul court acquitted him of a separate set of charges — stock price manipulation and accounting fraud — directly tied to the 2015 merger, ruling that prosecutors had failed to prove the deal was conducted unlawfully.⁷ Three distinct legal phases, collapsed in most retellings into a single felon-runs-Samsung caricature. The honest framing is stranger and more revealing: for years, the question of who could legally direct the strategy of 13% of an economy was being decided in a criminal court. No diversified conglomerate is supposed to have a load-bearing dependency on a single individual's parole hearing. Samsung's does.

Isn't the breadth still the thing that saves them?

The fair objection is that the breadth obviously works — Samsung is enormous, durable, and has survived every crisis thrown at it, including its own heir's imprisonment. If the structure were truly fragile, wouldn't it have broken? It's a real point, and it concedes more than it means to. The group did keep running while Lee was in prison; the affiliates have their own boards precisely because they can operate independently. But notice what the survival actually proves. Diversification protects against operational and market shocks beautifully — a bad memory cycle, a flopped phone, a shipbuilding glut. What it cannot hedge is a shock to the control structure itself, because every affiliate hangs off the same family thread. The group survived 2017 because the courts ultimately let Lee back, not because the structure was indifferent to whether they did. And there's a precedent the group rarely advertises: after the founder died in 1987, the original empire fractured into five separate groups — Samsung, Shinsegae, CJ, Hansol, and JoongAng.⁹ Breadth does not prevent splintering when control is contested. It just sets the size of the thing that splinters.

Samsung is not the world's largest technology company — it ranks well below the top by overall revenue, first only in narrower lanes like memory chips and smartphone units. The empire's true distinction was never being biggest. It was reaching across an entire economy from a single root, the way a banyan tree drops branches that look like trunks until you realise they all feed one stem. That breadth is a genuine moat: chips funding insurance funding steel, cycles cancelling cycles, capital allocated at national scale. But cut the stem and the whole canopy is in question. Samsung spent eight decades growing the most diversified industrial canopy on earth — and a single courtroom proved how little that breadth could do to protect the one root it all still grows from.