No crash occurred. There was never a day when the anchors began talking over one another and the screens went red. Rather, something more subdued occurred, which is more intriguing in certain respects. Wall Street has been methodically reducing its most exuberant bets on artificial intelligence over the past year—not in a panic, but with the kind of slow, deliberate patience that only institutional money can afford. If the AI bubble ever existed, it didn’t burst. It simply let out a breath.
Thrivent’s chief investment officer, David Royal, summed it up nicely when he said that everything came down in a fairly orderly manner. He was specifically referring to Nvidia, which is still the most iconic brand in the AI industry. Despite earnings continuing to rise, the company’s stock has hardly moved in the last three quarters. As a result, the forward price-to-earnings multiple has shrunk from the low 30s to about 20. It’s not a collapse. The repricing is under control. Additionally, it seems that the majority of regular investors hardly noticed it.
| Key Information: Goldman Sachs & The AI Market Outlook | Details |
|---|---|
| Organization | Goldman Sachs Group, Inc. |
| Founded | 1869 |
| Headquarters | 200 West Street, New York, NY 10282 |
| Key Analyst | Peter Oppenheimer — Chief Global Equity Strategist |
| 2026 S&P 500 Target | 7,600 |
| AI Infrastructure Share of S&P 500 EPS Growth (2026) | ~40% |
| Current IT Sector Forward P/E | Below Consumer Discretionary & Industrials |
| Mag Seven Correlation Status | Broken down; rising dispersion among hyperscalers |
| Key Risk Identified | AI disruption fears suppressing long-term growth estimates |
| Goldman’s Characterization | “Technology Value Opportunity” — a once-in-a-generation entry point |
| Reference | Goldman Sachs Insights |
Peter Oppenheimer of Goldman Sachs, who has spent decades observing tech cycles, recently described the current situation in stark terms: the technology sector has recently experienced one of its worst periods of relative underperformance in comparison to the larger global market since the early 1970s. Today, the IT industry is trading at a forward multiple below consumer discretionary, industrials, and even staples. It’s difficult not to pause and consider that for a moment. That statement would have sounded like science fiction eighteen months ago.
The selloff was motivated by a question rather than panic. What precisely are the major cloud providers receiving in exchange for all that expenditure? This is a costly, persistent, and increasingly difficult question to ignore. As a percentage of operating cash flow, the hyperscalers have been investing in AI infrastructure at historically high levels. Unfortunately, infrastructure builders do not fare well in history. The people who rode on top of railroads, telegraph networks, and early internet backbones made enormous wealth; the people who laid the tracks seldom benefited. It appears that investors have finally begun to read those chapters.
After being a monolith of correlated bets for a long time, the Magnificent Seven have now broken up. Amazon, Google, Meta, Microsoft, and Oracle are no longer collaborating. According to Goldman’s data, there has been a significant divergence between these names and a sharp decline in the three-month realized correlation. Fear of disruption from within the home was part of what broke the unity. Investors were forced to reevaluate how long-lasting these companies’ competitive advantages actually are as newer, less expensive, and more effective large language models—DeepSeek being the most striking example—were released one after the other. The word “Kodak” began to appear in serious analyst notes for the first time in years.
Ben Snider, a Goldman strategist, was noticeably pessimistic in his writing just last week. He cautioned that uncertainty regarding AI disruption and long-term growth projections might last for months or even years. Growth stocks that are at risk of this kind of disruption must provide concrete proof that their current business models aren’t being subtly undermined, not forecasts or roadmaps. Whether that evidence shows up this year or much later is still up in the air.
Observing all of this, there’s a sense that the stock market bottom for AI-linked stocks may be closer than the headlines indicate. According to Oppenheimer, this is a once-in-a-generation chance to purchase technology stocks that have been pricey for decades. Goldman’s goal of 7,600 for the S&P 500 is still in place, but getting there now seems more like a cautious ascent through uncertainty than a straight line. It’s really difficult to say at this point whether that’s cause for optimism or simply well-dressed caution.
