You Already Own It

So: what do you actually do with all this?

Here’s the joke of it. You spent four parts watching me catalogue trillion-dollar pay packages and term-limit votes and a man flipping a battlefield’s comms with a setting, probably feeling like a spectator to other people’s empires. You’re not. If you hold a plain S&P 500 index fund, a chunk of it is already the empire. The Magnificent Seven are somewhere around a third of the index now.1 You didn’t choose that concentration. The cap-weighting chose it for you, quietly, every time the winners won. The only real decision left to you is whether you hold it knowingly or by accident.

That’s the frame for this whole piece, and I want to be blunt up front: none of what follows is advice. I’m not going to name a single ticker, and where my own positioning would go I’ll leave an honest “here’s how I think about it” rather than pretend I’ve got a trade on. I work around markets, so I’ll write it like someone who does. But the useful thing I can hand you isn’t a buy list. It’s a lens for holding a danger you can’t actually escape.

You already own it

Start with the uncomfortable bit. The default “safe, diversified, set-and-forget” choice, the broad cap-weighted index, is a concentration bet wearing a diversification costume. When seven companies are a third of the index, your “500-stock fund” rises and falls disproportionately on a handful of AI-and-platform balance sheets.1 That isn’t a flaw in the product. Cap-weighting is supposed to do this: let winners compound and ride them. It just means the thing most people own as their cautious option is, right now, about as concentrated as the US market has been in decades.2

People who notice this usually reach for a hedge. And here’s the part the brochures skip: the hedges are trades, not free lunches. They swap one risk for a different one. They don’t make the risk smaller.

Take equal-weight. Instead of letting the biggest name count thirty times what the smallest counts, you weight every company the same. Feels prudent. But what you’ve actually done is buy a structural bet against mega-cap and toward the median sector.3 In years the giants lead, you lag, sometimes painfully. You haven’t reduced risk. You’ve rotated it from “concentration in seven names” to “concentration in the un-glamorous middle.” That’s a real view to hold. Just don’t tell yourself it’s neutral.

Same with going ex-US to dodge American mega-cap. Fine, except now you’ve taken on currency risk, different governance regimes, and (if you tilt to emerging markets) exactly the state-and-promoter risks Parts 3 and 4 were about. A different risk. Not a smaller one. Every hedge here is a position, and the honest investor says out loud what position they’ve taken instead of pretending they’ve stepped off the board.

Owning the rails (and the chokepoint behind each one)

If you can’t buy the federation or the sovereign fund directly, the next instinct is to own the rails: the sectors the concentrated capital flows through no matter who’s holding the levers this decade. This is where my markets brain actually engages, because the flows are legible even when the politics aren’t.

Follow the money from the earlier parts and it pools in a few places. Compute and AI, because every model and every ambition pays a toll to the same narrow stack of chip designers, fabs, and hyperscalers. Defence, where the same founder-king fusion shows up as the US Army rolling dozens of contracts into single multi-billion-dollar awards to a couple of vendors.4 Energy and the grid, because the AI build-out is, underneath the software, an electricity problem, and someone has to supply the power those data centres are projected to inhale.5 And luxury, the oldest one, because concentrated wealth has to land somewhere and it tends to land on a short list of brands that price like monopolies.

That’s the “own the rails” thesis in one breath. Now the guardrail, said in the same breath, because this is the part that makes it not-advice: every one of those rails is a chokepoint, and a chokepoint is a single point of failure. The thing that makes a sector a great rail (everyone has to pass through it) is the exact thing that makes it fragile (so does everyone else’s risk). The cleanest example sits behind every AI position you could possibly hold: the leading-edge chips the whole boom depends on are made, overwhelmingly, in one place, on one island, that one large neighbour claims.6 You cannot be long AI compute without being implicitly short the Taiwan Strait staying quiet. Most people holding an AI fund have never priced that tail. It’s there whether they price it or not.

So “own the rails” and “the rails are where the system breaks” are the same sentence. If that sounds like I’m talking myself out of the trade, good. That’s the lens working.

Following the smart money has hard limits

The other tempting move is to front-run the kings: watch where the sovereign and oligarchic money goes and ride its coattails. Tempting, and mostly a trap, for a reason that’s technical rather than moral.

In the US you can read a big fund’s holdings through its quarterly 13F filing. Sounds like a window into the throne room. It’s a keyhole. Saudi Arabia’s Public Investment Fund, the roughly 900-billion-dollar machine from Part 1,7 discloses a US equity book in its 13F that runs to a few names and a few billion dollars,8 a rounding error against its real, mostly private and global, exposure. And it arrives on a 45-day lag, so by the time you see the position, the people who put it on have had a quarter and a half to change their minds.9 You are not following the smart money. You’re following a heavily redacted, time-delayed photograph of a fraction of it. Coattail-riding works best when the coat is visible and the wearer hasn’t already left the room. Here, neither is true.

And then there’s the deeper limit, the one that actually matters: rent-seeking is not alpha. Suppose you correctly identify a company whose real moat is a regulatory capture, a political connection, a rule written in its favour. That moat is genuine economic value, no argument. But finance is brutally efficient about known edges. The classic research on politically connected firms finds the connection shows up as a one-time re-rating when it’s revealed, around events like a connected figure entering office.10 Once the market knows the moat is there and prices it in, the excess return from simply holding the rent-seeker drops toward zero. You don’t earn the rent. The person who bought before it was priced earned it.11 So the cold investor’s real question is never “who’s powerful?” Everyone can see that, and it’s in the price. It’s “what’s about to change in the structure of that power?”

Three flavours of risk, made practical

Part 4 split the world into three risk environments. Let me make them concrete, because this is the bit that should actually change how you’d size a position.

The US is slow risk. When concentrated power gets challenged in America, it grinds through courts and regulators and elections, and you can watch it develop over months and years. The Google antitrust case is the tell: the government won, proved the monopoly, and the remedy landed soft, behavioural rather than a breakup, and the stock went up on the ruling.12 As an investor that’s almost comforting. The risk is real but legible and slow. You get to read the docket. You get to reprice in daylight.

China is overnight risk. Here the state is the correction mechanism, and it doesn’t file a brief. The defining image is the Ant Group IPO in 2020: the largest share sale ever assembled at the time, around 34 billion dollars, suspended by the authorities barely 48 hours before it was due to price, after the founder gave one unwelcome speech.13 A national champion repriced by fiat, between dinner and breakfast, with no docket to read and no appeal that matters. Beijing frames this kind of intervention as legitimate anti-monopoly and data-sovereignty governance, reining in private power for the public good,14 and you should hold that rationale in view rather than treat it as pure caprice. But for someone holding the stock, the relevant fact is the shape of the risk: it’s a step-function, it arrives without warning, and the thing you can do least is see it coming.

India is the hybrid. It concentrates like an emerging market (promoter families, a strong central state) but it carries live brakes China doesn’t. Independent courts that actually bite: the same judge from Part 1 refusing to wave through a tycoon’s case on the government’s say-so, even as that same group’s own regulator at home, SEBI, cleared it of the original Hindenburg fraud allegations in 2025.15 Real elections that surprise the incumbent, like 2024, when the ruling party fell below a single-party majority.16 And a financial backstop that’s quietly become the most interesting structural fact in the market: the domestic SIP and DII flow. Indian retail investors now put money into equities through automatic monthly systematic plans at a scale that runs to roughly thirty thousand crore a month,17 and domestic institutions have grown into a large enough counterweight to absorb foreign investors heading for the exit.18 That’s a partial cushion, not a guarantee. But it means the Indian market has a homegrown floor under it that didn’t exist a decade ago. The risk here isn’t slow like America’s or instant like China’s. It’s both kinds stacked on top of each other, with a softer landing underneath.

The honest two-sidedness

I have to close the loop on the thing that’s been nagging this whole series, because if I don’t, this section reads as a stock tip and it isn’t one.

The entire argument for “own the concentration” rests on the concentration lasting. But Part 2 and Part 4 spent a lot of words on the opposite case: that this stuff is at least partly self-correcting. Promoter holdings in India’s broad index have actually been falling to record lows, around 49.5 percent of the Nifty-500 by early 2025, not tightening.19 American antitrust, however soft, is awake. Creative destruction keeps eating yesterday’s untouchable. And here’s the trap that follows directly from that: if concentration is even partly self-correcting, then buying the incumbents at peak-concentration multiples is, definitionally, buying near the top. You’d be paying the richest price for the names precisely when the mechanism that could erode them is most loaded.

I genuinely don’t know how to resolve that, and I don’t trust anyone who says they do. So the only honest posture is to hold both reads at once. The danger read: the kings are entrenched, the levers are real, own the rails. The mean-reversion read: that’s exactly the consensus, it’s in the price, and the rails crack right where everyone is crowded. A good investor isn’t the one who picks the right read. It’s the one who keeps both live and sizes the bet so that being wrong about which one wins doesn’t end them. That’s not a hedge in the product sense. It’s a hedge in the epistemic sense, and it’s the only one I actually believe in.

Closing the whole thing

I opened Part 1 with a question I said I couldn’t shake: in 2026, how much of the world can one person actually command? Five parts later I think the honest answer is “more than should be comfortable, and less than it looks from any single snapshot.” The man who flipped the Starlink setting is real. So is the federal judge who won’t let a tycoon walk. Both. At once.

What I didn’t expect, going in, was that the question would fold back on me, on us, the people standing underneath all this. It turns out you can’t opt out. The concentration is in your index fund whether you looked or not. The chokepoints are behind your “diversified” position whether you priced them or not. The kings’ risk environment, slow or overnight or stacked, is the weather your savings live in, and the only thing you get to choose is whether you understand the weather or get caught in it.

I came into this expecting to write something cleanly alarming, a tidy little doom essay about new emperors. I can’t, quite. The empires are real, the brakes are real, and they’re not taking turns. The most useful thing I learned is that holding both of those in your head, without collapsing into either cheerleading or panic, is the actual skill. In politics and in a portfolio alike. Live underneath the concentration knowingly. Invest in it knowingly. That’s not a triumphant ending and it isn’t meant to be. It’s just the difference between being a subject of all this and being a conscious one… and as endings go, I’ll take conscious over comfortable. We’ll see, in a few years, which of the kings are still standing. My honest hunch? Fewer than the index is currently betting on. But it’s only a hunch, and I’d want my own money to survive being wrong about it.

References

Footnotes

  1. Should Investors Be Worried That the “Magnificent Seven” Is Such a Big Part of the S&P 500? (The Motley Fool, Jan 2026); the Magnificent Seven sit at roughly 34–35% of the S&P 500. See also The Magnificent Seven’s Market Cap vs. the S&P 500 (The Motley Fool, 2026). ↩ ↩2

  2. The great narrowing: S&P 500 concentration (RBC Wealth Management, 2025/26); and S&P 500 concentration risk (Guinness Global Investors, 2025), which lays out both the record-concentration read and the earnings-backed counter. ↩

  3. RSP vs SPY: does equal-weight beat the cap-weighted S&P 500? (24/7 Wall St., June 2026); equal-weighting trades mega-cap exposure for a structural tilt toward mid-cap and the median sector. See also Why traders are managing concentration risk with equal-weight S&P 500 futures (CME Group, 2025). ↩

  4. US Army announces contract with Anduril worth up to $20bn (TechCrunch, Mar 2026); and the earlier Palantir lands $10 billion Army software and data contract (CNBC, Aug 2025), which rolled 75 separate contracts into one. ↩

  5. Top ETFs positioned to benefit from the AI power-demand boom (ETF.com, 2025), on the data-centre electricity build-out; see also The ETFs powering the AI supply chain (VanEck, 2025), a product piece, illustrative of the flow rather than an endorsement. ↩

  6. TSMC 2025 update: riding the AI wave amid global expansion (SemiWiki, 2025); TSMC makes upwards of 90% of the world’s leading-edge (3nm/2nm) chips, concentrating that supply in Taiwan. ↩

  7. Saudi wealth fund PIF’s assets under management up 19% in 2024 to $913bn (The National, Aug 2025). ↩

  8. PIF 13F filing index (13F.info, drawing on SEC data, 2026); the fund’s disclosed US equity book is a handful of names and a few billion dollars, a small fraction of its real exposure, reported on a 45-day lag. ↩

  9. Tracking Saudi PIF’s US portfolio rotation (Arab News, 2025), on the gap between PIF’s disclosed and actual exposure. ↩

  10. Politically Connected Firms (Mara Faccio, American Economic Review, 2006); political connections show up as a re-rating around the event that reveals them, not as a durable holding-period excess return. ↩

  11. Causes and consequences of rent-seeking (Econlib); the “once known and priced, the excess return is roughly zero” reading is analyst-level inference, not a direct claim from the source. ↩

  12. Department of Justice wins significant remedies against Google (US DOJ, Sept 2025); the remedies were behavioural rather than a breakup, and Google’s stock rose on the antitrust ruling (CNBC, Sept 2025). ↩

  13. The speech that scuppered the world’s largest IPO (Fortune, via Reuters, Nov 2020); Ant Group’s ~$34bn dual listing was suspended on 3 November 2020, roughly 48 hours before pricing, after founder Jack Ma’s 24 October Bund speech. ↩

  14. China in Xi’s “New Era”: The Return to Personalistic Rule (Susan L. Shirk, Journal of Democracy, 2018); Beijing frames such interventions as legitimate anti-monopoly and governance measures even as authority concentrates upward into the Party. ↩

  15. US federal judge questions DOJ decision to drop Adani charges (Al Jazeera, June 2026). The underlying charges were securities and wire-fraud conspiracy, not FCPA charges against the Adanis themselves. On the domestic side, SEBI cleared the Adani Group and dismissed the Hindenburg allegations (Business Standard, 18 Sept 2025), finding them “not established.” ↩

  16. 2024 Indian general election (Wikipedia); the BJP fell below a single-party majority, the clearest recent instance of an electorate repricing an entrenched incumbent. ↩

  17. India mutual fund AUM and SIP inflows, May 2026 (AMFI data) (StartupTalky, citing AMFI, May 2026); monthly systematic-investment-plan inflows run to roughly ₹30,000 crore. ↩

  18. DIIs now own more of the top-500 than FIIs (PublicMitra, citing NSDL data, 2026); domestic institutions have grown large enough to absorb sustained foreign outflows. The FII-vs-DII crossover stat should be re-verified against primary NSDL data before publishing. ↩

  19. Promoter holdings in Nifty-500 drop to record low of 49.5% in March 2025 (Tribune India, citing Motilal Oswal, Mar 2025); the direction (falling) is the load-bearing point. ↩