The strangest thing happening right now isn't that oil prices didn't spike after the Kharg Island strike. It's that we've stopped being surprised by that anymore.
Three days ago, the US struck Iranian military infrastructure. A decade earlier, that sentence would have meant emergency trading floors, emergency meetings, emergency something. Now: a Tuesday shrug. SPY down 0.5%. Everyone expected worse and got disappointed it wasn't worse. That's not resilience. That's numbness.
But zoom into the bifurcation and something sharper appears. Apple, Tesla, Meta, Nvidia—all getting hit harder than the broad market. Google climbing. This looks like a sector rotation at first glance, but it's actually something more constrained: it's the visible limit of what enterprise AI infrastructure can absorb without breaking.
Here's the arithmetic: Anthropic is consuming 3.5 gigawatts of Google's new TPU capacity. That's not a research experiment anymore—that's industrial-scale infrastructure consumption. Google can produce that power and those chips, but only Google. And only from existing capacity. Everyone else—Meta, Amazon, Tesla—is waiting in line. They can't move faster without either (a) waiting longer, or (b) paying more for alternative chips that don't exist yet in the volumes they need.
So what happens when a company can't scale as fast as Wall Street expects? It gets repriced. Not catastrophically. But steadily, downward, until the market stops expecting growth that can't be delivered.
This is why Google is up and the others are down. It's not a bet on better tech or better leadership. It's constraint pricing. Google owns the power plant. Everyone else is a tenant.
The geopolitical noise around Iran actually makes this worse, not better. Because now the tariff conversation gets tangled with the supply-chain conversation. If the US tightens controls on advanced semiconductors to China (which is happening—the Chip Security Act is moving), the shortage of alternative capacity gets worse. The line gets longer. The wait gets more expensive.
Tesla and Meta aren't failing. They're just hitting the hard limit of what money can buy right now. You can't accelerate a bottleneck by spending more on it. You can only wait.
The nightmare isn't a cyberattack or a financial crisis. It's simpler: what if this is the new normal? What if we've built an economy where the fastest-growing companies are now permanently constrained by physics—by kilowatts and silicon wafers and the time it takes to build factories?
No crash. No volatility spike. Just a slow repricing of anyone who isn't sitting on the power source.
The question isn't whether the market goes up or down from here. It's whether you've noticed we've stopped being surprised by anything at all.
PREDICTION: SPY closes the week (by Friday EOD) lower than today's open, driven by continued weakness in mega-cap consumer-facing tech (META, TSLA, AMZN) as earnings season exposes capacity constraints.