By Mari Hansen 
Between the Strait of Hormuz headlines and a trader’s phone alert about another record close, there’s a moment when you have to pause and consider whether something has truly gone wrong with the way markets are meant to function. The price of oil is more than $95 per barrel. Earlier this year, Brent surpassed $100. At dinner, no one wants to name the locations where American carriers are stationed. Nevertheless, the S&P 500 continues to close higher, occasionally shrugging and other times practically smiling. It’s difficult to ignore it.
According to the old rulebook, geopolitical shocks result in discounts. A tense Strait, a missile exchange, and a flare-up in the Middle East were all expected to push money into gold and Treasuries, with stocks taking a few weeks to stabilize. That still occurs, but only momentarily. When news of direct U.S.-Iranian engagement surfaced in early March, the CBOE Volatility Index did rise above 25. Gold did increase in value. Defense names did appear. However, the rotation began to reverse within hours, and by Friday, the index was reaching new highs once more. Nearly no one is paying attention to Goldman’s strategists’ repeated warnings that a protracted oil disruption could significantly reduce 2026 earnings growth by percentage points.
| Topic Overview | The S&P 500’s Defiance of Geopolitical Risk in 2026 |
|---|---|
| Index | S&P 500 (^GSPC) |
| Index Operator | S&P Dow Jones Indices LLC |
| Founded | March 4, 1957 (expanded to 500 companies) |
| Components | 500 leading US-listed large-cap companies |
| Approx. Market Cap Covered | ~80% of total US public equity |
| Recent High | Above 7,000 (April 2026) |
| Top Weighted Constituents | Nvidia, Alphabet, Apple, Microsoft, Amazon |
| 2025 Annual Return | ~18% |
| Year-to-Date 2026 | Roughly +7% |
| Primary Driver | AI capital expenditure cycle |
| Key Risk Cited | Oil-driven GDP drag, Middle East tensions |
| Live Quote Source | Yahoo Finance ^GSPC |
| Analyst Models Referenced | Goldman Sachs top-down EPS model |
There is a structural component to the explanation. In the most recent reporting season, about 85% of S&P 500 companies exceeded earnings expectations. At a scale that, to be honest, shouldn’t be mathematically feasible for a company that size, Meta reported a 33% increase in revenue. Instead of plateauing, cloud spending is increasing at Alphabet and Amazon. Not too long ago, AMD increased by nearly 19% in a single session. These are cash flows, not games of valuation. It’s the kind of stuff that, when piled high enough, turns a war headline into background noise.
However, there is also a stranger component, which is the aspect that no one is quite sure how to value. Artificial intelligence infrastructure is becoming less of a discretionary tech wager and more of a strategic national asset in the eyes of investors. The message was clear when Nvidia increased 3.7% in response to news of a domestic data center contract during the same session that the Dow fell 400 points. Capital is divided into two categories: those that are negatively impacted by geopolitical conflict and those that may profit from it. AI is firmly in the second box, which is a re-rating with implications that no one has yet fully comprehended.
However, those who notice it are bothered by the quieter story that lies beneath. Only 52% of the S&P’s components were above their own 50-day moving averages, despite the S&P closing 7.7% above its 50-day moving average on a recent Friday. According to Jonathan Krinsky of BTIG, that combination has virtually never occurred in thirty years of data. On a day when the index reached a new high, there were more new 52-week lows than new highs. Since 1990, it had happened three times. There is a real narrowness. While much of the market silently drifts sideways or worse, a small group of names doing exceptional work are carrying the headline number.
One might be tempted to label this as complacency as it develops. Perhaps it is. The S&P is trading at about 21 times forward earnings, which is significantly higher than its long-term average of about 16. The margin for error is very small. However, when you consider real revenue growth, real capital expenditure, and a global rewiring of compute that may be more significant than any one crisis, complacency seems like the wrong word. Investors appear to think that the AI cycle is accomplishing something that the previous models are unable to fully explain. This is belief, not certainty.
Whether that conviction will hold the next time something truly breaks is still up in the air. The market seems to be deciding which risks to take seriously and which to put off. Gravity is on hold for the time being. Only the next shock will reveal whether that’s wisdom or a story we’ll tell awkwardly in retrospect.