Goldman Sachs (NYSE:GS) has highlighted an unusual split inside the U.S. stock market that, in the bank’s view, looks uncomfortably close to the late-1990s technology boom. A record share of S&P 500 companies now display negative beta, meaning they have recently tended to fall when the benchmark rises and rise when it falls.
Over a short recent window, that group has approached half the index.
Over a one-year horizon the share is smaller but still historically elevated.Beta is a simple statistical measure of how a stock moves relative to the market.
A reading above 1 implies amplified moves in the same direction.
A negative reading implies the opposite.
That relationship is usually rare among large-cap stocks.
Its sudden prevalence is less a sign that hundreds of companies have suddenly become “defensive hedges” than a reflection of how concentrated the index itself has become.
A small cluster of mega-cap technology and artificial intelligence names now accounts for a large portion of the S&P 500’s market value.
Those stocks have been the main engine of index gains.
When they advance, the headline benchmark can grind higher even if a wide swath of industrials, consumer names, energy producers, and other sectors lag or decline.
The result is a market that looks healthy at the index level while the typical constituent tells a different story.Goldman has compared the pattern with the technology bubble, when a handful of high-flying names likewise pulled the major averages away from the experience of the average stock.
Concentration today is visible in other statistics as well: the largest names carry higher betas than the rest of the index, and the gap between the S&P 500 and the median stock’s distance from its own 52-week high has been among the widest in decades.
The backdrop for the non-leaders is less friendly.
Higher bond yields, a stronger dollar, elevated energy prices, and tighter financing conditions weigh more heavily on companies whose fortunes are tied to the broader economy rather than to AI infrastructure spending.
Passive index funds automatically add more of the winners as their weights grow, reinforcing the same loop.
Market commentators have been careful not to treat the statistic as an automatic crash signal.
Extreme divergence has historically been worth watching because it can reverse suddenly if leadership stocks stumble.
It also complicates ordinary portfolio construction.
A manager who simply avoids the mega-caps can end up looking more defensive than intended, not because of a deliberate hedge but because the benchmark itself is increasingly defined by a few names.
The same concentration affects how investors interpret risk.
If “the market” is increasingly a handful of companies, a stock’s beta of 0.6 may say less about its inherent volatility and more about its lack of exposure to the themes currently driving the index.
Valuation work, capital-allocation decisions, and hedging strategies all rest on assumptions about what the market portfolio actually is.
None of this means the current cycle must end the way 2000 did.
Earnings for the leaders have been stronger than those of many late-1990s darlings, and the investment boom still has supporters.
But the internal market map has grown unusually lopsided.
The index can keep making new highs while large parts of corporate America trade to a different rhythm. That is the pattern Goldman flagged, and it is the reason the comparison to the last technology boom has returned to client notes.