The $678 Billion Ledger Microsoft's About to Open

Microsoft handed Wall Street the strongest opening hand in the AI era, and traders still can't stop squinting at it skeptically. Here's the setup two days out from fiscal Q1 FY2027 earnings on October 28, and why I think the skepticism is running on fumes.

Let's set the board. MSFT closed at $532.60, up a hair on the day, sitting at a $3.96 trillion market cap. It's up 35.40% on the quarter, up 25.37% half-year, and it clawed its way to a record $543.45 back on October 13 before trimming. The FactSet consensus average price target is roughly $600, with BNP Paribas at $604, Cantor at $608, Oppenheimer hiked to $625. The stock isn't chasing the thesis. The thesis is chasing the stock.

Now the part nobody's going to let you forget: this thing already beat into print. On July 29, Microsoft reported fiscal Q4 with $90 billion in revenue, up 18%, GAAP EPS of $4.81 against roughly $4.24 consensus. Azure accelerated to 43%, pushed past $100 billion in annualized revenue for the first time. Microsoft 365 Copilot blew past 30 million paid seats, GitHub Copilot hit 50 million users. And then it dropped the number that should end the debate: commercial remaining performance obligations surged 84% to $678 billion.

Six hundred and seventy-eight billion dollars in contracted cloud demand, largely from customers outside the frontier AI labs. That's not a roadmap. That's a purchase order.

Here's where I actually have a problem with the prevailing mood. The bear case, repeated with relish, is capex anxiety. Q4 capital spending was $41 billion, and management guided Q1 FY2027 to clear $50 billion a quarter. Free cash flow dipped 23% year-over-year. Legacy lines — Xbox down 10%, personal computing down 4% — aren't helping the vibe. And yes, the stock got hit hard earlier in 2026 on exactly this fear.

But look at the valuation math, because it's the crux. MSFT trades at a 29.7x trailing P/E and 22.6x forward. The trailing multiple is below its five-year average. Someone could buy this business at a discount to its own history while Azure grows at 40%-plus and carries a backlog that makes most companies' lifetime revenue look like a rounding error. That is not a bubble pricing in perfection. That is a market still bracing for a blow-up that hasn't happened.

And the blow-up risk on the downside is more contained than the narrative suggests. Microsoft restructured its OpenAI deal in late April, killing the revenue-share and locking licensing through 2032. Sure, OpenAI committed to spend over $300 billion on Azure and "doesn't have that money today." Fine — but Microsoft isn't begging them to spend it. That's demand Microsoft gets first dibs on as capacity comes online. The monetization question that spooked everyone in the spring is now answered by a $678 billion contract book and Copilot seats jumping 50% in a single quarter.

So what do I actually think about Thursday?

The tape has already done the heavy lifting. The stock is up 35.40% off the quarter and sitting just under its record. The analysts who matter have all reiterated buy ratings with targets in the $600s. The real event isn't whether the story holds — it's whether management, facing a market that's still watching their capex line like a hawk, finally gives the street something to stop worrying about.

My read: this is a "show me, but I'm tired of being scared" moment. The backlog and the Azure re-acceleration mean the fundamental engine is hotter than the multiple implies. If they guide Azure in line with that ~45% constant-currency number and stop making capex the star of the show, the ceiling toward $600 isn't a fantasy — it's the consensus for a reason.

The one thing that could humiliate this thesis isn't weak demand. It's management telling you they're going to spend even more than expected on hardware while revenue guidance disappoints. That's the exact fear that already knocked this down 30% earlier this year.

Bottom line: Microsoft handed the market a $678 billion ledger and a 22x forward multiple. The bears are still picking through the capex. I think they're reading the footnotes and missing the deal.