Title Fulfillment:
[Generated Title]: Microsoft: AI Hype vs. Cold, Hard Numbers
The Hype Train vs. Reality
Microsoft (MSFT) is riding high on the AI wave, no question. The stock's been buoyed by Fed Governor Waller hinting at rate cuts and Jefferies maintaining a $675 price target, citing accelerating Copilot adoption. Optimism is palpable. But let's take a cold, hard look at what's actually happening under the hood, beyond the press releases and analyst upgrades.
Jefferies points to Microsoft's impressive 15.59% revenue growth over the last twelve months, attributing it to Copilot. Okay, fair enough. But is that growth solely attributable to AI? Or is it a rising tide lifting all boats, with AI just being the shiniest new toy? We need to isolate the specific contribution of AI to that revenue figure, and that’s where the data gets murky. Microsoft isn’t exactly breaking down revenue streams with that level of granularity.
Then there's the capex elephant in the room. Rothschild & Co Redburn downgraded both Amazon and Microsoft, arguing that the market is overestimating the returns related to AI investment. They estimate that each GPU generates about $10 billion of revenue per gigawatt of power, but costs firms around $40 billion in capex per gigawatt. That's a 4x difference (or a 75% loss if you look at it that way).
And here’s the kicker: these AI chips have a short lifespan. Replace them every three years, and Rothschild’s estimates suggest projects become "value destructive." So, it's not just about if the AI bet pays off, but how quickly it pays off. Time is quite literally money here.
The Gates Factor & Market Sentiment
Adding another layer of complexity, Bill Gates' Foundation Trust has been steadily selling off its Microsoft holdings since the end of 2023. Now, the article is quick to dismiss this as just routine portfolio management, pointing out that "Gates doesn't control the day-to-day operation of his trust." Maybe. But it's hard to ignore the optics of a founder, even indirectly, reducing their stake when the stock is trading near all-time highs. (And yes, I know, correlation doesn't equal causation, but still...).
The article also mentions that Microsoft stock trades at nearly 13 times sales. And while it has achieved that valuation a few times in recent years thanks to the AI craze, the last time Microsoft stock reached this high of a price-to-sales ratio was more than two decades ago, during the dot-com bubble.
I've looked at hundreds of these filings, and this particular comparison to the dot-com bubble is, frankly, tiresome. It’s lazy analysis. Every tech boom gets compared to the dot-com bust. It’s almost a cliché at this point, and it rarely provides any real insight.
So, what does this all mean?
It means that while the AI narrative is compelling, the underlying economics are far from certain. It means that the market is pricing in a lot of future growth, and any stumble could lead to a significant correction. And it means that even the insiders, like the Gates Foundation, seem to be taking some chips off the table.
Is This Time Really Different?
The question isn't whether AI will be transformative (it almost certainly will be). The question is whether Microsoft's current valuation accurately reflects the risk-adjusted return on its AI investments. The market seems to be saying "yes," but the numbers, at least as presented by Rothschild, suggest otherwise. We're in the "1st half of the 1st inning," according to Jefferies. If that’s true, a lot can change. According to a recent article on Investing.com, Microsoft stock price target maintained at $675 by Jefferies on AI growth By Investing.com.
