The 500 Doesn't Know What It Is Anymore

Sometime before Monday's open, a committee of investors who never touched a can of soup in their professional lives will finally finish evicting The Campbell's Company from the S&P 500. Pool Corp goes with it. In their place: Marvell Technology, up 287% over the past twelve months, and Flex, the contract electronics manufacturer that happens to be bolted to the same AI hardware supply chain. The rebalance is effective June 22. Every index fund in the country — every 401(k), every passive ETF, every institution running a closet-index equity sleeve — is now mechanically required to buy Marvell at whatever price it trades on Monday.

MRVL closed Thursday at $324.73. The CFO filed to sell approximately $65 million worth of stock at roughly $290 in the days prior.

Run that sequence again slowly. The chief financial officer of a company, with full knowledge that passive flows are about to flood into the name, and with the stock already up 71% in the last thirty days and 287% in the last twelve months — filed to dump nearly half his stake before the index bid arrives. That is not a man who thinks the inclusion pop is leaving money on the table. That is a man who thinks $290 is an acceptable exit.

And now every person holding VOO, SPY, or FXAIX wakes up Monday morning as a Marvell shareholder. No prospectus. No vote. No valuation committee. The S&P 500's methodology required it, so it happened.

This is the part where you are supposed to shrug and say indices rebalance, that's the point, the market is efficient, price discovery works. Fine. But step back from the individual trade and look at what the composition change is actually recording. Campbell's Soup — incorporated in 1869, a consumer staple that has survived the Great Depression, two world wars, the invention of the microwave, and the death of condensed soup as a cultural institution — is being removed from the benchmark representation of American capitalism because a semiconductor company that makes custom AI chips for hyperscalers finally crossed a cumulative GAAP profitability threshold. Pool Corp, which sells equipment to install swimming pools and has approximately nothing to do with artificial intelligence, is also gone.

The S&P 500 is not a snapshot of the American economy. It hasn't been for a while. It is a systematic momentum engine dressed in the language of broad market exposure.

Here is the machinery: companies are added to the index after their market caps have risen enough to meet inclusion criteria. Passive funds are required to buy them. The forced buying adds further price pressure. Higher prices feed back into the narrative that validates the original price appreciation. More passive AUM flows into the index. The constituents with the highest market caps capture the largest share of those flows. Three of eleven sectors beat the index this week. Tech is up 4.4%. The equal-weight version of the same index — the version that would tell you something about breadth, about the health of the underlying 500 businesses rather than the top ten — gained 0.1%. The cap-weighted version gained 1.4%.

RBC's wealth management team published data showing the current relative outperformance of cap-weighted versus equal-weight has now surpassed the extreme reached at the peak of the dot-com bubble in 1999 and early 2000. The only comparable period in the index's modern history. They were careful to note this is not a forecast. They are correct to be careful. Concentration can persist for longer than anyone expects, especially when the mechanics of passive flow reinforce the concentration with every fresh dollar of inflows.

But there's a specific fragility here worth naming. The profitability requirement that finally let Marvell through the door — the one that kept it out for years despite its market cap — exists because the S&P 500 is supposed to represent durable, profitable businesses. Marvell's revenue is roughly $8.19 billion on trailing figures. Gross margins near 51%. EBITDA margins above 50%. By those numbers, the business is genuinely exceptional. The question is not whether Marvell is a real company. It is. The question is whether paying 8x sales and a forward multiple consistent with flawless hyperscaler AI capex execution for the next several years, in the week that nine Fed officials penciled in a rate hike, is a prudent allocation for every passive investor regardless of their goals, time horizon, or risk tolerance.

The index doesn't ask that question. It can't. The index has no opinion. It has rules.

B. Riley raised its MRVL price target to $345 from $240 this week, citing the Nvidia partnership, the S&P inclusion tailwind, and reaffirmed Q2 guidance — while noting in the same breath that valuation is "now rich." That is analyst language for: we can't downgrade the thing because the momentum is running, but we are aware of what we are doing. The stock ripped 12.5% in a single session on the news. The inclusion-driven buying hasn't even happened yet.

What makes this worth sitting with rather than dismissing is the scale. Passive assets tracking the S&P 500 run somewhere north of $5 trillion. When those assets rebalance — whether through quarterly index reconstitution, or through JPMorgan's estimate of $165 billion in quarter-end equity selling from institutional rebalancers this week, or through the mechanical drip of fresh 401(k) contributions hitting the market every two weeks — the flow is not making a judgment. It is executing a rule. The rule says buy the largest companies in proportion to their size. The largest companies are large because people have been buying them. The buying makes them larger. The index adds more when they become large enough.

This is not inherently wrong. Markets have been doing this forever in various forms. But the scale of passive investment now means that the gap between price and value in the highest-concentration names can persist and widen for much longer than fundamental analysis would suggest, and then correct much more violently when the flows reverse. It's the kind of dynamic that feels stable right up until it isn't.

Campbell's Soup will survive the exclusion. The company still exists. It still makes soup. Nothing about Monday's rebalance changes the underlying business.

But the benchmark — the thing millions of people have been told represents a diversified slice of American corporate life — just told you something about itself. It replaced a 157-year-old consumer staples company with a chipmaker whose CFO sold $65 million of stock the week before passive funds were forced to buy it.

The 500 is a great index. It has made a lot of people a lot of money over a long time. It will probably continue to do so. Just be clear about what you're actually buying when you buy it. You are buying the largest companies, at the largest weightings, in the most concentrated benchmark since the turn of the millennium. And on Monday you are buying Marvell at a price the CFO apparently found satisfactory enough to exit at, before the bid arrived.