Wall Street's biggest bank just bet $30 billion more on Google the same week its flagship AI model missed the launch window.
The Summary
- Bank of America doubled its Alphabet stake while setting a $430 price target — 16% above current levels — ahead of Q2 earnings on July 22
- Google's Gemini 3.5 Pro launch delay triggered a selloff from $370 to current levels, raising questions about AI execution just as Magnificent Seven earnings kick off with 28% expected EPS growth
- BofA is simultaneously expanding crypto infrastructure and recommending 1-4% digital asset allocation to clients — a signal that traditional finance sees AI and crypto infrastructure as complementary bets, not competing ones
The Signal
Bank of America's move is interesting not because a bank bought tech stock, but because of the timing and the context. They're doubling down the same week Alphabet's stock dropped from $370 on model delays. That's not panic buying. That's a thesis play on infrastructure, not product hype.
The Gemini 3.5 Pro delay matters because it breaks a 20-year pattern. Google has always been the company that ships. Now they're the company catching up. OpenAI ships. Anthropic ships. Google promises and then… recalibrates. For a stock trading on AI credibility, that's a margin compression event waiting to happen.
"Google just broke a 20-year funding habit, and the market is watching to see if capex discipline or AI ambition wins."
But here's where it gets interesting for Web4. BofA isn't just buying Google. They're building crypto rails at the same time. They're telling clients to hold 1-4% in digital assets while expanding their own blockchain infrastructure. That's not hedging. That's portfolio construction for a world where compute and ownership run on different stacks but need each other.
The Magnificent Seven earnings cycle starts with Google and Tesla on July 22. Analysts expect 28% EPS growth across the group. That number assumes capex keeps climbing and margins hold. But Google's spending pattern just changed. For two decades, they've poured money into infrastructure ahead of revenue. Now they're being careful. Either they see margin pressure coming, or they're reallocating capital to something they're not talking about yet.
Key data points:
- BofA's $430 price target implies 16% upside from current levels
- Alphabet dropped from $370 after the Gemini delay news
- 28% expected EPS growth across Magnificent Seven this quarter
The crypto angle isn't just about BofA's asset allocation advice. It's about what happens when the companies building AI infrastructure start thinking about tokenized compute markets. Google Cloud already sells compute. If they can't out-ship OpenAI on models, maybe they start thinking about decentralized inference markets where they're the liquidity provider, not the model maker. That's speculation, but it's also exactly the kind of pivot companies make when the obvious path gets crowded.
The Implication
Watch the July 22 earnings call for two things. First, how much capex guidance changes. If Google is pulling back on infrastructure spend while BofA is loading up on the stock, someone's wrong about the timeline. Second, listen for any language about "platform" or "marketplace" plays in AI. That's the tell that they're thinking beyond selling models to selling infrastructure.
For anyone building in Web4, this is a reminder that the AI stack and the ownership stack are converging faster than either side wants to admit. The same banks buying Google are building crypto desks. The same compute powering LLMs could power agent economies. The question isn't whether these worlds connect. It's who builds the bridges first and who gets stuck on the wrong side when they do.