'Stay the Course' Was Advice for a Different Market
If you have money in a retirement account, you have been told to stay the course. Buy the broad index, hold it, ignore the noise. You have heard it so often it stopped sounding like advice, and for most of the last forty years it was good advice.
The advice is not wrong. The market it was written for is gone.
When the advice was written, the index was a sliver
When “just buy the index” hardened into conventional wisdom, index funds held a small fraction of all American shares. Rob Arnott and Que Wu, independent index researchers at Research Affiliates, laid out the arc in a July 2026 study: index investing went from roughly 3 percent of US market value in 1990, to about 10 percent in 2000, to 25 percent in 2010, to more than half by 2025.
That is not more people doing the same thing. It is a different thing, and it changed what buying the index actually does.
An index used to measure the market. Now it moves it.
An index started life as an instrument. It read the market the way a thermometer reads a room. It reported the temperature. It did not set it. When almost every buyer and seller was weighing what a company was worth, the index just totaled up their verdict.
Michael Green, a market-structure strategist, has made this argument in public for years. Money going into a cap-weighted index fund is price-insensitive: the dollar does not ask what anything is worth. It buys the largest companies in proportion to how large they already are. Rising prices make larger weights. Larger weights pull in more of the next dollar. More dollars hold the price up. The one signal that is supposed to discipline that loop, the question of what a business is actually worth, matters less every year that the share of money which never asks it grows.
You can see the fingerprint in the data. Over the 35 years from 1989 to 2025, Arnott and Wu found that the 500 companies ranked just below the very largest grew their operating cash flow at about 9 percent a year, against about 6.5 percent for the top 500. Yet those smaller, faster-growing companies delivered only about 0.6 percent a year more in return. The giants kept pace not by out-growing the field but because investors paid steadily higher multiples for them. By 2025 the largest 500 stocks traded at roughly an 80 percent premium on price-to-cash-flow over the next 500, a gap that did not exist in 1989, when the giants actually traded at a discount.
Arnott is blunt about why. He can “see no other driver for such odd and profoundly inefficient market behavior than the immense flow of capital into index funds”. He calls that an inference rather than a proven fact, and I am not going to upgrade it for him. The premium is a fact. Its cause is an argument.
Then it reached the people paid to judge
Once the index became the yardstick everybody is measured against, looking different from it became the career risk. A manager who deviates and is wrong for a year or two loses clients. A manager who stays close to the index and is wrong loses nobody, because everyone else was wrong in exactly the same way. So the safe move is to look like the index.
The people whose job was to judge what a company is worth are now graded on how closely they match the thing they were supposed to be judging. They got to match the market.
What the machine produces
I pulled the fund’s own published holdings rather than take somebody’s estimate for it. As of August 21, 2026, the six largest companies in the S&P 500 were about 30 percent of the entire index, and the ten largest were more than a third of it. Ten holdings out of 504. That is our own count, and it agrees independently with UBS’s estimate for mid-2026. The energy sector, the companies that physically fuel the country, was about 3.4 percent of the index on August 3, 2026.
On August 21, NVIDIA was the largest holding in the index, ahead of Apple, and three of the ten largest were semiconductor companies. The S&P 500 has become an AI trade.
No committee decided the American economy is thirty percent artificial intelligence and three percent energy. Nobody voted on it. That is the shape you get when a large share of the money in the market has stopped asking what anything is worth and buys the biggest names in proportion to their bigness.
An index fund still holds all 504 companies. What it no longer holds is the economy’s proportions. It holds market capitalization’s proportions. Those two tracked each other closely enough for decades that nobody had to tell them apart, and they have come apart. Energy’s 3.4 percent of the index is not 3.4 percent of what the country runs on.
The index did not shrink the physical economy. It shrank its picture of it.
This is your index fund
This is not somebody else’s portfolio. If you own a fund that tracks the S&P 500 in your 401(k) or your IRA, this is the shape you own. About thirty cents of every new dollar you put in goes to those six companies, and more than a third of it goes to ten. You did not choose that. The index chose it for you, and it tightens every time the largest names rise, because rising is exactly what makes the index buy more of them.
“Stay the course” never anticipated that. At 3 percent of the market, the index really was a broad cross-section of the economy, and holding it really was prudent diversification. Past half, the same phrase increasingly describes concentration in whatever is already largest.
The strongest case against this
The best argument against everything above belongs to Jim Bianco.
The macro researcher reads the same concentration and draws the opposite conclusion. On August 3, 2026, when the headline index sold off on the day’s news, Bianco pointed out that “the 459 non-AI stocks were up almost 1% on Monday”. Two markets inside a single trading day. But where I see a machine manufacturing concentration out of flows, he sees a genuine technological revolution being priced in real time, and priced roughly right. In his view AI is “the biggest technology innovation that we have ever seen,” on the scale of the railroads. The market is not in a bubble but perhaps “a couple of years off,” because it is held back right now by a shortage of computing power rather than drowning in excess. On that reading the giants are large because they are winning, the premium is a forecast that may prove right, and the flow mechanism is a sideshow.
I will not tell you I know which of us is correct. Both forces are almost certainly at work, and their proportions are genuinely unknown. If the revenue the optimists expect arrives, much of the premium will have been earned after all. And the plainest evidence on Bianco’s side is that standing apart from the giants has cost money. The same study found that being underweight them, even while being right that their premium was flow-driven, paid off by only fractions of a percent a year, with a bumpier ride, and the most recent decade favored the giants outright.
The opposite trade is not the answer either. Respected contrarians now argue those positions have become crowded themselves. In one analyst’s phrase, the energy and hard-asset positions that were lonely a few years ago have become a consensus “hubris trade.” Being early and being consensus are not the same thing, and by the time everyone can recite the argument, the early part is usually gone.
The course moved
None of this is a prediction. I am not telling you the index falls, on any date, by any amount. I am not telling you to sell your index fund, or to buy anything else. Reasonable investors can weigh this same evidence and reach different conclusions, and what any of it means for you depends entirely on your own circumstances.
“Stay the course” is sound advice that quietly assumes the course has not moved. It moved. When the phrase became gospel, buying the index meant buying a broad slice of the American economy. Today it increasingly means holding a concentrated position in whatever is already the largest, assembled by a machine that never asks what anything is worth. That may work out perfectly well. But it is a decision you are making, and the phrase gets repeated as though the market underneath it never changed.
If you would like to see what your own index funds actually hold today, bring a statement. Whether to change anything is a separate question, and the honest answer often is nothing.
References
- Arnott, Robert, and Que Wu. “Membership Has Its Privileges: Who Pays the Premium?” Research Affiliates, July 2026. Figures cited (the 1990-2025 index-ownership arc; the Next-500 vs top-500 cash-flow growth, return, and price-to-cash-flow premium; the “no other driver” quotation) verified against the retrieved text of the published paper.
- Draho, Jason, et al. “The AI Economy: A Roadmap.” UBS Chief Investment Office, Global Wealth Management, June 30, 2026. The share of S&P 500 market capitalization held by the six largest AI-linked mega-caps is UBS’s own estimate as of that report.
- SPDR S&P 500 ETF Trust (SPY) sector weightings, retrieved August 3, 2026 (SPY used as the S&P 500 proxy). Energy weighting 3.38 percent as of that date.
- Bianco, Jim (President, Bianco Research). Remarks on a public interview, August 3, 2026; transcribed and archived in the firm’s research corpus. Quotations verbatim from the archived transcription.
- SPDR S&P 500 ETF Trust (SPY) published fund holdings, State Street Global Advisors, as of August
21, 2026 (SPY used as the S&P 500 proxy; 504 total holdings). The top-six and top-ten weights are
Long Point Wealth Management’s own additions of the issuer’s published holding weights: 29.70
percent and 37.16 percent respectively, stated in the text as “about 30 percent” and “more than a
third.” NVIDIA at 7.87 percent was the largest single holding; NVIDIA, Broadcom and Micron were
the three semiconductor companies among the ten largest. Cross-checked against two independent
public holdings aggregators, StockAnalysis.com and Yahoo Finance, which give top-six weights of
29.90 percent and 30.19 percent as of their own retrieval dates; all three rank NVIDIA first. The
0.49 percentage-point spread is consistent with differing timestamps rather than disagreement.
Substantiation:
compliance-vault/Registers & Logs/Claims Substantiation/2026-08-25 S&P 500 Concentration/.
This article reflects the views of Long Point Wealth Management as of the date written and those views are subject to change. It is intended for educational purposes only and should not be construed as personalized investment, tax, or legal advice, or as a recommendation to buy or sell any security or to adopt any investment strategy. References to specific companies, sectors, indices, and third-party research are for illustration and attribution only and are not recommendations. Index data referenced is unmanaged and cannot be invested in directly. Figures attributed to third-party research are those sources’ own and are stated as of the dates given; they are not the performance of any Long Point account and no client performance is shown. All investing involves risk, including the possible loss of principal. No strategy assures a profit or protects against loss, and diversification does not guarantee a profit or protect against loss. Past performance does not guarantee future results. For guidance tailored to your situation, consult a qualified professional. Long Point Wealth Management, LLC is an investment adviser registered in Massachusetts and New Jersey, serving clients nationwide consistent with applicable state requirements and exemptions; registration does not imply a certain level of skill or training. Additional information is available in the firm’s Form ADV Part 2A.
Fees, minimums, and who we work with →