A cautious investor distrusts concentration: when a handful of US giants, the Magnificent Seven, account for a third of the New York market, they look for the way out. The obvious emergency exit: emerging markets, reputedly distant, uncorrelated, "something else." They open the door, step through… and find themselves in the same room. Because the emerging index has itself become a bet on the very same semiconductors. This piece explains why diversifying by country no longer protects against what really matters: the common risk factor.
1 The move
Fleeing US concentration.
The intention
Getting out of the Magnificent Seven
The reasoning looks sound. The US market has become dependent on a handful of technology mega-caps, the Magnificent Seven, which concentrate an unprecedented share of it. To avoid putting every egg in one basket, the investor seeks exposure elsewhere: to different economies, other currencies, other cycles. Emerging markets tick every box: China, India, Brazil, Taiwan, Korea, South Africa. Buying the emerging index is supposed to be the diversification move par excellence, the exact opposite of a concentrated bet on US tech.
2 The surprise
Nearly 30% of the index in three foundries.
The finding
Three companies, almost a third of the index
Lift the hood of the emerging index and the surprise is considerable. Three chipmakers dominate it: Taiwan's TSMC (about 14.5%), South Korea's Samsung (about 7.8%) and its compatriot SK Hynix (about 6.6%). Together they approach 30% of the whole index. Taiwan and Korea between them account for nearly half of it. The investor thought they were buying the emerging world in all its variety; mostly they bought three semiconductor foundries. The emergency exit opened onto a very narrow corridor.
TSMC alone in the index
≈ 14.5%
more than half of Taiwan's entire weight.
The three foundries combined
≈ 30%
TSMC + Samsung + SK Hynix.
3 The common factor
The real risk is not the country, it is the factor.
The key idea
What makes stocks move together
Naive diversification reasons by labels: a US stock and a Taiwanese stock look different because they fly different flags. But what makes a stock rise or fall is not its nationality: it is the risk factor it is exposed to. And US mega-caps and Asian foundries share the same one: artificial intelligence demand and the semiconductor cycle. Nvidia designs the chips, TSMC manufactures them, Samsung and SK Hynix supply the memory: they are links in a single chain. When the AI bet wobbles, they wobble together, whatever their passport. The common factor is the true denominator of risk.
4 The same room
Out through one door, back through another.
The metaphor
Two doors, one room
Here is the paradox. The investor leaves the US market to escape its dependence on AI, and buys the emerging index; but that index has in turn become a bet on AI, through the chipmakers. They thought they had changed rooms; they merely changed doors. The emergency exit and the main entrance open onto the same space. A shock hitting US AI stocks would also hit, by ricochet, "their" emerging diversification. The protection sought is an illusion: both pockets rise and fall to the beat of the same engine.
The key idea
Diversifying by country offers no protection if every basket depends on the same factor. The real question is not "where do my stocks come from?" but "what do they all depend on?"
6 The illusion
Apparent diversification against real diversification.
The distinction
Many lines, one risk
An emerging index holds hundreds of companies across dozens of countries. On paper, that is diversification itself. In reality, much of its performance depends on a small number of stocks exposed to the same factor. This is the difference between apparent diversification, counted in number of holdings, and real diversification, measured by the independence of the risks. Holding a thousand stocks that rise and fall together is not being diversified: it is holding the same bet a thousand times. The count reassures; the common factor decides.
7 Managers trim
Large investors are cutting exposure.
The market's reaction
When the weight passes the reasonable threshold
The phenomenon has not escaped professionals. As the combined weight of the three foundries approached, then passed, 30% of the index, major managers began deliberately cutting their exposure to these stocks, judging the concentration excessive. The move is telling: to stay genuinely diversified, one must now actively deviate from the index, that is, do the opposite of what passive management does. The "safety" of the benchmark has become, on this precise point, a risk to be corrected by hand.
8 The counterpoint
What this reading should not overstate.
The analysis's honesty
Concentration is not a crash
Alarmism should be avoided. TSMC, Samsung and SK Hynix are not empty bubbles: they are highly profitable companies at the heart of a real, growing global industry. Heavy concentration does not announce a collapse; it signals heightened sensitivity to a single factor, which is different. Moreover, diversifying by country is not useless: currencies, rates and policies still vary from one region to another. Finally, solutions exist: emerging indices excluding technology, caps on the weight of the largest holdings, active management. The message is not "flee emerging markets," but "know which bet you actually hold."
9 What it changes
For how one diversifies, and reads an index.
The effects
Look at factors, not just labels
The episode invites a change of lens. Diversifying is not about multiplying geographies, but about spreading risk factors: a portfolio can be "global" and yet rest on a single theme. This revives interest in equal-weighted or capped indices, which limit the weight of giants, and in active management able to deviate from the benchmark. It also recalls a limit of passive management: by faithfully tracking market value, it embraces the very concentrations it is supposed to dilute. Reading an index is not counting its countries; it is identifying what it truly depends on.
10 A factor, not an address
The conclusion: risk lodges in the factor, not the flag.
The meaning of the episode
Diversification is judged by risks, not borders
The lesson goes beyond emerging markets. In a world where a few themes, AI above all, irrigate the entire planet, geography is no longer a rampart: the same factor can hide under different flags. To believe one is diversifying because one changes continent is to mistake the address for the risk. True diversification means holding bets that do not depend on the same causes; it is work on factors, not on borders. The emergency exit leads elsewhere only if it does not reopen onto the same room. Which leaves, for anyone reading what comes next, a simple question: what, fundamentally, does my portfolio depend on?
The compass
①
The exit leads to the same room. Bought to flee US concentration, the emerging index has become a bet on TSMC, Samsung and SK Hynix, nearly 30% of it (Taiwan + Korea ≈ half).
②
The risk is the common factor. US mega-caps and Asian foundries share the same engine, AI and chips: they rise and fall together, whatever their flag.
③
Diversifying spreads factors, not countries. A thousand stocks depending on the same cause is not diversification. This piece frames a debate; it is not advice.
Read alongside: The index is no longer a cushion (concentration inside an index) and The silicon shield (TSMC, leverage and dependence). See also The alarm signal we fund anyway (the AI bet). Neighboring notion: index concentration. Reference: abbreviations & acronyms (MSCI, ETF, AI).