Broadcom: Hock Tan's M&A Compounding Machine — Disciplined 'Combining' at Its Peak

Movement II: Not Every "Combining" Destroys Value

Throughout Movement I, we kept looking at "splitting" — GE, Abbott, Ferrari, one empire after another being broken apart to release locked-up value. By this point, you might have come away with an impression: M&A (combining) is bad, and spin-offs (splitting) are good.

But that would be a colossal misreading. From the very first piece, this series has stressed: neither combining nor splitting is the answer — timing and discipline are. What Movement I criticized was never the act of "combining," but "combining without discipline" — buying to make the empire bigger, buying a pile of things with no real synergy, and ending up as the conglomerate-discounted GE.

So in Movement II, we turn the coin over to look at the companies that truly know how to combine: they prove that when you are the right owner and you have iron financial discipline, serial M&A need not be a value-destroying imperial dream — it can be a compounding machine. And no one has played this out more extremely, or more coldly, than Broadcom's Hock Tan.

Who Is Hock Tan? An M&A Machine That "Grew Up Through Reverse Mergers"

First, a piece of trivia you may not have noticed: today's trillion-dollar Broadcom, the hottest name in the AI wave, is not actually "the original Broadcom."

The protagonist of the story is Hock Tan, born in Malaysia and educated at MIT. The company he ran was originally called Avago — the former semiconductor division of Hewlett-Packard (HP), which after various twists was bought out by private-equity firms KKR and Silver Lake and re-listed in 2009 under the ticker AVGO. Avago was not big to start with, but Hock Tan had a knack no one else did: using M&A to swallow rivals bigger than himself, one bite at a time.

2016 was the decisive battle — Avago bought the then-larger and better-known "original Broadcom Corporation" for about US$37bn, and then did a very Hock Tan thing: it kept the other company's name (Broadcom) and brand for itself, while keeping its own Avago ticker (AVGO). A snake swallowing an elephant, then putting on the elephant's skin. From then on, the world remembered the name "Broadcom," but the true soul and playbook were those of Hock Tan's Avago.

He's Not a Tech Dreamer — He's a Cold-Blooded "Businessman"

To understand Broadcom, you first have to understand the temperament of Hock Tan the man. He came up as an engineer (MIT mechanical engineering, later a Harvard MBA), and spent early years in industry and private equity, but at his core he is not the Silicon Valley "change the world" tech dreamer — he is more like a businessman and capital allocator so calm he is almost cold-blooded. In an industry where everyone chases growth stories and burns cash to grab share, he deliberately runs the other way, asking one question over and over: how much cash can this business reliably spit out?

This is his "turnaround philosophy": after buying a company, he carves it up like a surgeon into "core" and "everything else" — the handful of product lines that serve big customers, have pricing power and fat cash flows are kept and fed with resources; the peripheral projects that sound sexy but burn cash with no visible return are ruthlessly cut. He doesn't care about being called "short-sighted," doesn't care about sacrificing reported revenue growth — he cares about one thing: is every dollar of capital placed where it earns the highest return? To sum up his belief in one line — "I'm not trying to build the biggest company; I'm trying to build the best-returning one." That line is the very spirit of Movement II.

A Staggering Acquisition List

Lay out Hock Tan's M&A over the past decade and you'll understand what a "serial-acquisition machine" really means:

From semiconductors (LSI, Broadcom, Brocade) all the way into enterprise software (CA, Symantec, VMware) — Hock Tan's empire gradually expanded from a "chip company" into two pillars of "semiconductors + infrastructure software." Each deal chased not a sexy startup dream but mature, sticky, cash-generating "rent-collecting" assets that customers can't leave.

Breaking Down the Model: Buy, Cut, Focus, Harvest

Why does Broadcom's M&A create value, instead of destroying it the way most M&A does? Because it has a repeatedly proven, almost industrialized "Broadcom model." Broadly, it's four moves:

(1) Only buy things customers "can't switch away from." Hock Tan doesn't buy growth stocks with hype but no moat; he targets the mature franchises with deep customer dependence and extremely high switching costs — once your company's IT systems run on VMware or CA software, replacing them is bone-breaking engineering. That "stickiness" is the source of pricing power.

(2) Cut costs hard. After buying, Broadcom slashes overlapping headcount, administrative and marketing expense, pushing operating margins up. It buys not the "story" but the "cash flow" — so its first priority is to maximize the profitability of that business.

(3) Concentrate R&D on the core. It doesn't spread resources evenly across a pile of research projects; it concentrates them hard on the few most profitable, most strategic product lines, cutting the peripheral experiments. This maximizes the return on every dollar of R&D — but it is also its most controversial trait (see below).

(4) Raise prices, harvest cash. Leaning on the stickiness customers can't escape, Broadcom often adjusts pricing and licensing after an acquisition to push the asset's earning power to the limit, then uses the relentless cash flow to repay the acquisition debt, pay dividends, buy back stock, and build a war chest for the next, bigger deal.

The Essence of This Case: The Cash-Flow Flywheel

If you can only remember one thing about Broadcom, remember this — it is not a series of isolated acquisitions, but a self-accelerating "cash-flow flywheel." The four moves above are not independent; they chain into a loop that spins faster and faster:

The essence of the flywheel is "compounding": the cash squeezed from each acquisition becomes the ammunition for the next, bigger one; so the targets get larger and the cash flow rolls up thicker, until it has the firepower to swallow a giant on the order of VMware. This is why a semiconductor company could roll its market cap up to a trillion dollars within a decade — not thanks to some blockbuster product, but thanks to this discipline-driven, ever-self-reinforcing cash machine. Understand this flywheel and you understand all of Broadcom.

The Best Owner Test: It Buys You In, and Makes You Earn More

A System, or a Single Person?

At this point we have to ask the most lethal question of any "serial acquirer" — the very one this series has pursued since the "neither-fish-nor-fowl framework": is this magic a transferable "system," or a "personal talent" tied to one genius brain? After the founder is gone, is it still there?

For Broadcom, the honest answer is: half and half.

The half already systematized — "the turnaround technique." "How to cut after buying, how to focus, how to maximize the cash flow from core customers" — this "Broadcom model" has been executed repeatedly and standardized into a replicable playbook. Each acquisition runs the same process — this part really has been internalized into the organization, like a well-drilled production line that runs no matter who operates it.

The half not yet proven transferable — "the capital-allocation judgment." But the truly hardest and most valuable piece — which company to buy, what price to pay, when to chase, when to cut and run in time (like the Qualcomm stop-loss) — is highly concentrated in Hock Tan alone. This is a near-artistic judgment, and there is as yet no evidence that it has been dissected, turned into teaching material, and internalized by a whole bench of successors the way Danaher's DBS has.

Let the Numbers Speak: A Compounding Machine's Report Card

Note: figures are approximate, in US dollars, as of 2026; semiconductor and AI stocks are highly volatile, and past returns do not indicate future performance.

The most elegant part is that Broadcom doesn't only "harvest" old assets through M&A — its core semiconductor business happens to sit right in the sweet spot of the AI wave: it designs custom AI chips (ASICs) for cloud giants like Google and supplies the most critical networking chips in the data center, with demand exploding alongside the AI arms race. On one side, a "cash cow harvesting mature assets"; on the other, a "growth engine riding the AI tailwind" — together they have pushed it into the trillion-dollar club.

The Honest Side: This Model Has Its Shadow Too

Painting Broadcom as flawless would be dishonest. Its "harvesting" M&A model has always come with enormous controversy and risk.

In other words, Broadcom is an extraordinarily sharp double-edged sword: it has created stunning value for shareholders, but its methods — harvesting, raising prices, focusing — also make it a perennially controversial presence in the eyes of customers and regulators. Acknowledging this is the only way to understand the true cost and limits of "disciplined combining."

Disciplined Combining vs. Empire-Building Combining: Broadcom Against GE

Put Broadcom side by side with Movement I's GE, and you'll fully understand why this series says "combining isn't the problem — combining without discipline is."

Both pursued serial M&A and both grew ever bigger, yet the endings differ by a world. The difference isn't the act of "combining" but the discipline of combining: GE bought "scale" and "prestige," Broadcom bought "cash flow" and "returns." This is the core Movement II sets out to prove — "combining" in itself is neither right nor wrong; whether you can combine, and whether you combine with discipline, is the watershed.

M&A Discipline Scorecard: Broadcom

Movement I used a "post-spin-off scorecard" to assess "splitting"; in Movement II, we switch to a corresponding "M&A Discipline Scorecard" to assess "combining":

Overall verdict: the most extreme exemplar of "disciplined combining" — five of six checkpoints fully pass, the only "△" being the high leverage that needs continuous watching. It proves serial M&A can be a compounding machine, but don't forget its sharp other edge (raising prices, harvesting, controversy). This is a machine of awesome power, but one whose costs must be understood.

Lessons for Taiwan: M&A Creates Value Through Discipline, Not Scale

Up Next: From Chips to a Luxury Empire

Broadcom demonstrated the "harvesting" style of disciplined M&A. In the next piece, we fly to Paris to see a completely different yet equally peerless kind of "combining" — how LVMH's Bernard Arnault used M&A to combine one independent luxury brand after another into an unshakeable empire of luxury.

Both are "combining," but where chips are about cash and efficiency, luxury is about brand and desire. See you in the next piece, at LVMH.