The S&P 500 Index plummeted as much as 2.3% on Monday over DeepSeek, a Chinese artificial intelligence startup that developed a model competitive with the US’s very best — and, supposedly, on the cheap. Venture capitalist Marc Andreessen called it a “Sputnik moment,” a reference to the Russian satellite that set off the 1957-1960s Space Race. Chip companies plummeted and so did many of the communications giants developing AI tools of their own. But the ostensible pandemonium in the world’s biggest stock market was not as widespread as you might imagine, and it seemed to abate as the trading day wore on. With DeepSeek hype still largely indistinguishable from reality, the main lasting lesson may be that diversification still matters.
Consider the following factoids about Monday, the worst intraday selloff of 2025:
At the time of writing, 328 stocks on the S&P 500 were up.
The median stock was up 0.7% and the average was down just 0.2%.
Among sectors, health care, consumer staples, real estate and financials were all positive on the day.
Information technology accounted for 95% of the index move.
Nvidia Corp., which is behind cutting edge AI chips that are also eye-poppingly expensive, accounted for about two thirds of the decline on its own.
I’ll admit it: betting against the cap-weighted index has been a losing proposition for the past decade and a half, but concentration risk has become a more acute problem for investors in the past two years. S&P 500 Index investors’ exposure to information technology and communication services companies is at its highest since the dot-com bubble.
And just seven companies account for about a third of the entire index by weighting. Nvidia alone had a greater weighting than five of the 11 sectors represented in the index.
That concentration is a big reason why a Goldman Sachs Group Inc. report in October suggested that the S&P 500 would deliver an annualized total return of just 3% over the next decade (or only about 1% per year if you adjust for inflation). “Our historical analyses show that it is extremely difficult for any firm to maintain high levels of sales growth and profit margins over sustained periods of time,” Goldman wrote at the time.
It’s always technically possible that today’s index giants continue to outperform, but history is working against us. Similarly, the research suggests that market concentration is associated with greater volatility going forward. If the market truly underwhelms over the next decade, it may well be in the form of a crash followed by a long, grueling recovery — rather than 10 years of nearly flat real returns.
Fortunately, elementary mitigation strategies are easy to implement, and you don’t even need options (in fact, tail hedges are very inefficient in slow-moving bear markets like the dot-com bust). The Goldman report suggested that the equal-weighted version of the S&P 500 could outperform the S&P 500 by 200-800 basis points in the decade. Additionally, the juicy income benefits of bond ownership may give new life to 60/40-type mixed asset class portfolios. And investors may finally take the opportunity to add some exposure to unloved international stocks, as well as small- and mid-capitalization US stocks that can still benefit from a strong macroeconomic backdrop.
Clearly, it’s hard to know where the DeepSeek panic will lead. Companies representing about 38% of the S&P 500 by weighting are expected to report earnings this week, and those announcements should provide some insight into how US executives are processing the developments and help us sort hype from reality. Even in a scenario in which the narrative proves well-founded, it’s entirely possible that a cheaper path ahead for AI turns into a net positive for many publicly traded US companies, including companies developing AI-related software, and end users.
But first, Monday’s market action may shake index tracking investors out of their complacency. For all the strengths of the US economy and stock market, the index’s composition is tilted strongly in favor of the spectacular AI story and the premise that we’ve correctly identified the market winners. Odds are that we have the narrative at least a little bit wrong, and investors should expect to pay for their lack of true diversification with ongoing volatility and perhaps even subpar total returns.
A message from Advisor Perspectives and VettaFi: To learn more about this and other topics, check out some of our webcasts.
Bloomberg News provided this article. For more articles like this please visit
bloomberg.com.
