To The Moon

One of the best ways to find great businesses isn’t to ask:

“Who has the biggest market share?”

Ask this instead:

“What happens if customers can’t use them?” 👀

That’s where the real monopolies and oligopolies show up.

🏰 MONOPOLIES / NEAR-MONOPOLIES

$ASML Holding NV(ASML)$ — EUV lithography
$Taiwan Semiconductor Manufacturing(TSM)$ — advanced semiconductor manufacturing
$VeriSign(VRSN)$ — .com domain registry
$CoStar(CSGP)$ — U.S. commercial real estate data
$Fair Isaac(FICO)$ — credit scoring
$CME Group Inc(CME)$ — futures & derivatives infrastructure
$MSCI Inc(MSCI)$ — global indexes & benchmarks

ASML is probably the cleanest example.

It is currently the only manufacturer of EUV lithography systems, the machines required to manufacture the most advanced chips.

That’s not just market share.

It’s a technological bottleneck.

⚔️ DUOPOLIES

$Visa(V)$ $MasterCard(MA)$ → global card network
$Boeing(BA)$ $AIR → commercial aircraft
$Coca-Cola(KO)$ $Pepsi(PEP)$ → beverages
$Advanced Micro Devices(AMD)$ $NVIDIA(NVDA)$ → GPUs
$Apple(AAPL)$ $Samsung Electronics Co., Ltd.(SSNLF)$ → smartphones
$Cadence Design(CDNS)$ $Synopsys(SNPS)$ → chip design software
$Zoetis(ZTS)$ $IDEXX Laboratories(IDXX)$→ veterinary diagnostics

The interesting part about a duopoly is that you don't necessarily need monopoly pricing.

You just need two companies that are extremely difficult to displace.

🧱 OLIGOPOLIES

$Amazon.com(AMZN)$ $Microsoft(MSFT)$ $Alphabet(GOOGL)$ → cloud
$Broadcom(AVGO)$ $CrowdStrike Holdings, Inc.(CRWD)$ $Palo Alto Networks(PANW)$ $Fortinet(FTNT)$ → cybersecurity
$Linde PLC(LIN)$ $Air Products & Chemicals(APD)$ $L'Air Liquide SA(AIQUY)$ → industrial gases
$Micron Technology(MU)$ $NVIDIA(NVDA)$ $SK Telecom(SKM)$ → memory/HBM ecosystem
$JPMorgan Chase(JPM)$ $Bank of America(BAC)$ $Citigroup(C)$ $Wells Fargo(WFC)$ → U.S. banking
$Meta Platforms, Inc.(META)$ $Alphabet(GOOGL)$ $Trade Desk Inc.(TTD)$ $Snap Inc(SNAP)$ → digital advertising/social
$United Parcel Service Inc(UPS)$ $FedEx(FDX)$ → parcel delivery
$GE Aerospace(GE)$ $RTX Corp(RTX)$ → aircraft engines
$Moody's(MCO)$ $S&P Global(SPGI)$ $Fitch → credit ratings
$Equifax(EFX)$ $Experian plc(EXPGY)$ $TransUnion(TRU)$ → credit data

And I’d add one category that deserves much more attention:

🔥 THE AI SUPPLY-CHAIN OLIGOPOLY

$ASML Holding NV(ASML)$ $Taiwan Semiconductor Manufacturing(TSM)$ $NVIDIA(NVDA)$ $Micron Technology(MU)$ $SK Telecom(SKM)$ $Lam Research(LRCX)$ $Applied Materials(AMAT)$ $KLA Corporation(KLAC)$ $Linde PLC(LIN)$ $Air Products & Chemicals(APD)$

AI gets most of the attention at the application layer.

But underneath the models are a handful of companies controlling the machines, fabs, memory, equipment, gases and infrastructure required to build the compute.

Take HBM.

It’s not really a duopoly. Samsung, SK hynix and Micron dominate the market, making it one of the clearest semiconductor oligopolies.

And then there’s industrial gas.

Linde, Air Liquide and Air Products sit in another critical bottleneck. These businesses don't necessarily look exciting on the surface, but semiconductor fabs, steel plants, chemical facilities and other industrial customers need extremely reliable supplies of specialty and industrial gases.

That creates a very different kind of moat.

🔒 The common denominator

The strongest monopolies usually have at least one of these:

• Extremely high switching costs
• Regulatory barriers
• Proprietary technology
• Network effects
• Decades of accumulated data
• Customer certification requirements
• Massive capital requirements
• Mission-critical infrastructure
• Long-term contracts
• Economies of scale

And that’s why I wouldn't simply rank these companies by market share.

A company with 30% share and an unbeatable moat can be more powerful than a company with 70% share in a commodity market.

The real question is:

Can someone realistically replace you?

If the answer is no, you may have found the real moat. 🏰

Which monopoly, duopoly or oligopoly would you add? 👇