Tech stocks don't price in war

In the last week of March 2026, the Nasdaq 100 sank into official correction territory, falling more than 11% from its October peak. Alphabet dropped 9%. Microsoft shed 7%. The Nasdaq composite lost 3.23% in a single week. The AI narrative, the one that was supposed to make Big Tech untouchable, didn't save anyone. This wasn't a rotation out of growth stocks. It was a reminder that tech companies, no matter how sophisticated their models, still exist in the physical world.

The numbers don't lie

The Nasdaq closed out the first quarter of 2026 down 7%, its worst quarterly performance since Q1 2022. The S&P 500 declined 5.1% in March alone, its worst month since March 2025. Bloomberg reported the Nasdaq 100 closed at 23,132.77 after falling 1.9% in a single session, the kind of move that would have been unthinkable six months ago when AI hype was at its peak. The catalysts were stacked. Iran tensions escalated into full-blown conflict. Meta faced a damaging legal verdict. And suddenly, the market remembered something it had been conveniently ignoring: geopolitical risk doesn't care about your earnings multiple.

Tech valuations assume a stable world

Here's the disconnect at the heart of this selloff. Tech stock valuations, especially among the Magnificent 7, are built on the assumption of a stable, growing, interconnected global economy. Steady supply chains. Predictable energy costs. Open markets. Wars, sanctions, and supply chain disruptions break that assumption overnight. As Angelo Kourkafas, senior global investment strategist at Edward Jones, put it: "Everything is getting hit in this environment, and tech is no exception." The market had been pricing in intelligence, compounding returns from AI, cloud expansion, and data-driven growth. What it hadn't priced in was chaos.

The Strait of Hormuz problem

On March 2, Iran's Islamic Revolutionary Guard Corps declared the Strait of Hormuz closed. Tanker traffic plunged by more than 90%. Gulf oil exports fell roughly 60%. The head of the International Energy Agency called it "the largest supply disruption in the history of the global oil market." About 20% of the world's oil and liquefied natural gas transits the strait daily, roughly 20 million barrels. When that flow stops, energy costs spike. And when energy costs spike, the infrastructure that powers AI gets hit first. This isn't abstract. Amazon, Microsoft, and Google have spent years building data centers across the Gulf, betting the region would become the next great hub for artificial intelligence. The undersea cables connecting those facilities to Africa, South Asia, and Southeast Asia pass through two narrow passages: the Red Sea and the Strait of Hormuz. Both are now effectively closed to commercial traffic. Iran has also directly targeted data center infrastructure in the Gulf, a strategic move that underscores just how vulnerable the physical backbone of AI really is. And the disruption extends beyond energy. Qatar, home to the world's largest proven helium reserves, halted production after strikes on its Ras Laffan facilities. Helium is critical for semiconductor manufacturing, and experts estimate repairs could take three to five years.

The Project Maven elephant

While tech stocks were falling on macro fears, a quieter story was unfolding that carries its own set of risks. On March 20, Reuters reported that the Pentagon would adopt Palantir's Maven Smart System as an official program of record, embedding AI-driven weapons targeting technology across the U.S. military. Project Maven started in 2017 as a pilot program using AI to sift through drone and satellite footage for intelligence purposes. It has since evolved into a comprehensive targeting and logistics system. Democracy Now reported on how the system is actively helping the Pentagon select bomb targets in the current Iran conflict. This creates a triple risk for tech companies involved in defense AI: legal exposure, ethical scrutiny, and reputational damage. Google famously pulled out of Project Maven in 2018 after employee protests. But the broader tech industry has moved closer to the Pentagon, not further away. That proximity is a risk that most valuations don't account for.

The pattern is clear

This isn't the first time the market has been caught off guard by the real world. In April 2025, Trump's tariffs triggered a similar shock. Tech stocks plunged as investors suddenly remembered that global supply chains are fragile and trade policy matters. The pattern repeats: during calm periods, tech valuations climb on narratives of exponential growth and AI-driven transformation. Then a real-world shock hits, tariffs, a pandemic, a war, and the market corrects because those narratives never accounted for geopolitical risk in the first place. Alphabet's stock surged 4.8% on March 31 when Trump signaled willingness to wind down hostilities. That single-day move tells you everything. The rally wasn't driven by a product launch or an earnings beat. It was driven by the mere possibility of less chaos. When a stock's best day in weeks comes from a geopolitical headline, that stock was never really priced on fundamentals alone.

What this means for builders

If you're building software, shipping products, or running infrastructure, this is a reality check worth internalizing. Don't build your entire stack on one cloud provider in one jurisdiction. The Gulf data center strategy that Big Tech pursued looked brilliant in peacetime. It looks fragile now. Geographic diversification isn't just a corporate finance concept, it's an engineering decision. The biggest risk to AI isn't alignment. It's geopolitics. The models will keep getting better. The chips will keep getting faster. But none of that matters if the data centers are offline, the cables are cut, or the energy supply is disrupted. And for investors, the lesson is simpler: every time there's a real-world shock, the market remembers that AI companies are still companies. They have supply chains, energy dependencies, regulatory exposure, and geopolitical risk. The premium you pay for a tech stock assumes those risks are manageable. Sometimes they aren't.

The market prices in intelligence. It doesn't price in chaos.

Tech stocks will recover. They always do. But the next time someone tells you that AI makes Big Tech recession-proof or geopolitics-proof, remember March 2026. Remember that the most sophisticated models in the world couldn't predict a strait closure, and the most valuable companies in the world couldn't avoid the fallout. The world is messier than any valuation model can capture. The sooner we price that in, the fewer surprises we'll face.