Insiders at Google and Palantir are buying their own stock while everyone else is celebrating a ceasefire nobody believes in. That's not courage. That's calculation.
Here's the thing about confidence: it doesn't whisper. It doesn't hedge. CEOs and senior executives don't deploy their own capital—especially in a risk-on rally where the easy money is already made—unless they know something about *their specific company's* future that the market hasn't priced in yet. Google's insider trades in mid-April weren't reactions to the Hormuz reopening. They were bets on the company itself.
But the real signal isn't the buying. It's the *absence* of panic selling.
We've spent three weeks watching oil collapse, watching the geopolitical terror premium evaporate, watching every mega-cap tech stock treat the ceasefire like a get-out-of-jail card. And somewhere in that relief rally, someone should have been getting out. Taking profits. Rotating into safety. The script calls for it: after a relief trade executes, the smart money hedges.
Instead, the insiders are walking in.
What this tells me is that the people running these companies don't think the relief is temporary. They don't think the market's going to price-correct in three weeks. They think the structural upside—particularly in AI infrastructure and enterprise software—is real enough to justify buying at a market that's *already rallied hard* on good news. That's not a vote on geopolitics. That's a vote on the actual economics of their business.
The problem—and this is where the AI agent cost conversation on Hacker News gets relevant—is that we haven't stress-tested what happens if the productivity gains from AI turn out to be illusory. If companies deploy these agents, if labor gets disrupted faster than expected, if the cost curve inverts and margins compress instead of expand, then the insider buying at these levels looks like a trap set by people who were too close to the machine to see its failure mode.
The 10Y-2Y spread sits at positive 0.55. VIX at 17.94. Unemployment stable at 4.3%. On paper, this is a "growth without inflation" scenario. The market's pricing a world where AI productivity solves the growth puzzle without anybody losing their job or their purchasing power.
But that's what *markets* think when they're not afraid. And markets that aren't afraid are markets that have forgotten something important: the thing they were afraid of three weeks ago is still happening. Iran didn't disarm. Israel didn't stand down. The Strait of Hormuz is open because of a *conditional* ceasefire—not because the underlying conflict evaporated.
The insiders buying right now are betting that either: (a) the conflict stays frozen long enough for AI productivity to prove itself, or (b) they've already priced the geopolitical tail risk into their thesis and they don't care.
I think they care less than the market thinks they do.
The insider buying cluster (GOOGL, PLTR, combined with the broader mega-cap rally momentum) will push QQQ higher into the close of the week. Not because the thesis is sound, but because rallies that have corporate officers' skin in the game tend to persist until sentiment breaks on *earnings*, not on geopolitics.
When does the confidence paradox reverse?