The S&P 500 can sit near a record while most of the market is not having a record year. That is not a footnote for technicians. It is the story, and the people paid to soothe you would rather you treat the index level as the economy.
Thirty-four percent of large-cap stocks have beaten the S&P 500 over the past year. Goldman Sachs — not a gold newsletter — says breadth is the weakest since the dot-com bubble. Semiconductors are 17 percent below their peak. Large banks are down 13 percent. The approved sentence is still that stocks are near highs, so things must be fine.
A cap-weighted index is not a portrait of American business. It is a portrait of whatever has gotten biggest. For several years that has meant a small club of technology companies, carried by an artificial-intelligence story Wall Street decided was too important to price like a business. Technology and chip shares are now under pressure because OpenAI’s revenue growth is no longer an article of faith, and because Chinese competition in semiconductors has migrated from the panel discussion to the tape. Deutsche Bank has downgraded the sector it was happy to celebrate, citing crowded positioning and earnings trouble. Too many professionals were on the same side of the same trade, and the profits have to show up.
The serious objection deserves to be stated plainly. American technology firms have earned extraordinary profits. Narrow markets can keep rising. A wobble in chip stocks is not 2000, and a company selling the picks and shovels of a real computing shift is not Pets.com. Goldman’s own strategist, having made the dot-com comparison, still expects further S&P gains over the next one to three months. Breadth is not a crash date. Nobody honest should pretend otherwise.
But “it can go higher” is not “the foundation is broad.” In the late 1990s the index looked magnificent because a handful of names did all the lifting, and plenty of serious people explained why this time the concentration was the point. The wreckage arrived later for people who thought they owned “the market.” Using that comparison as a curiosity, then recommending another three months of gains, is how complacency dresses itself as sophistication.
There is a harder fact under the trading note about China. Semiconductor weakness is being filed as market color. It is industrial strategy. The chips that train and run artificial intelligence are not a consumer fad. They are a large part of the next generation of military power, intelligence, and economic leverage. Beijing has spent years trying to close the gap — by subsidy, by scale, by theft of know-how, and by the simple fact that we spent a generation treating a strategic rival as a vendor. When “Chinese competition” shows up as a reason U.S. chip shares are 17 percent off their highs, the market is admitting what the ribbon-cuttings still soften.
This is not a brief for a new religion of industrial policy. Washington already wrote large checks. Some of that was a legitimate national-security correction. Some of it was corporate welfare with a flag on it. Neither patriotism nor a subsidy guarantees a business can outrun a rival state. If Chinese firms are real competitors, the test is whether we can design, fabricate, and defend the advantage — not whether the Nasdaq-100 is still up about 5 percent since September.
The same test applies to artificial intelligence. Doubt about OpenAI’s revenue is not Luddism. It is the minimum requirement of capitalism. A technology can be real and the promotion around it can still be ahead of the cash. When a few firms and a single narrative carry the index, the country is not “invested in innovation.” It is leveraged to a story. Stories get revised on earnings, which is exactly when, as John Kolovos noted, we learn whether weakness stays in the leaders or spreads.
The financial press will keep leading with the level of the S&P. That number is easy, flattering, and useful to anyone who wants the public to feel richer than the median stock. Look at the breadth. Look at the chips. Look at whether the AI miracle is producing revenue or just market cap. A record built by a handful of companies, under pressure from a strategic rival and from the rude question of profits, is not proof the country is strong. It is proof we have gotten very good at mistaking a crowded trade for a commonwealth.
How it may affect me
A standard S&P 500 index fund is not an even bet on 500 companies. The largest technology names dominate both the returns and the risk, so the fund can sit near a high while most of the stocks inside it lag — and a chip or tech sleeve can already be well off its peak before the headline index notices. Check your 401(k) or IRA’s actual top holdings, and whether a target-date fund has stacked the same names twice. That is a concentration check, not a sell signal.
