Join institutional investors, analysts and strategists from the world's largest banks: Subscribe today
Already have an account? log in
‘Inline Is Not Good Enough’: Microsoft’s ‘Solid’ Quarter Scrutinized
Microsoft reported overall sales and cloud revenue that basically matched estimates on Tuesday afternoon in the US, following a rocky session during which the company grappled with connection issues and outages.
The top-line grew 15% (16% ex-FX impact) to $64.7 billion in fiscal Q4, slightly ahead of the $64.5 billion consensus expected.
For the full fiscal year, sales were $245.1 billion. So, Microsoft raked in a quarter trillion in revenue over the 12 months to June 30. (I didn't quite get t
Revenue +16.6% but no operating leverage. Where is the AI-driven acceleration in Azure? I see positive inflection in Gaming and nice growth in Office and Windows, but no-one cares. The LinkedIn acceleration must be the stealth white-collar recession. Waiting for guide on call, but I think the “beat and blah” I thought now looks more like “inline and blah”.
Looks like Wall St group think has caught up to the idea that there are no immediate material P&L impacts from AI. I’d expect this trend to continue for the rest of earnings season.
From my experience working in/with and consulting for software companies, it takes 18 months to go from concept to implementation/in market. But these are AI applications that are at the margins of any given offering and not core offerings. If you’re talking about truly transformational change, something that will change the business model of a company you’re taking about 18 months to built/get to market and 3-5 years to prove it in market.
That said, eventually, Wall Street will wake up to the reality that demand for the chip makers will not continue in a linear fashion upward as the largest buyers of chips slow new orders as they struggle to realize P&L impacts from the AI CapEx spend and their stocks get punished for it. As of now, the chip maker story is in tact (great beat and guide from AMD with Nvidia following higher after hours), but eventually demand will plateau, or worse decrease. Valuations of chipmakers is based on unrealistic/unsustainable upwardly linear growth. Once this narrative unravels indexes will struggle as likely will the entire economy.
I should clarify by “ Wall St group think” I mean equity prices
Thougt I saw something on an earnings scroll on M-Watch that reported that a company exec said that AI buisiness was 8% of some total at Azure. (The quote did not make it clear, but I suspect ot was revenue.)
If so, that’s something to ponder, aint it?
“In the latest quarter, artificial intelligence accounted for 8 percentage points of the company’s cloud computing growth, more than last quarter.” NYT
Assuming it was rather small in Jun 2023 quarter, you can back into the dollar amount of AI cloud revenue in Jun 2024 quarter.
@derek, I looked into that some more.
The actual quote from the earning call is: “Overall, server products and cloud services revenue grew 21% and 22% in constant currency. Azure and other cloud services revenue grew 29% and 30% in constant currency, in line with expectations and consistent with Q3 when adjusting for leap year. Azure growth included 8 points from AI services”.
So my guess is:
– I estimate “Azure & other cloud services” revenue in Jun 2024 qtr grew about +$4.6BN growth, to about $20BN (+29% from about $15.4BN in Jun 2023 qtr). Note when MSFT’s script says only “Azure” that means they are excluding “other cloud services”, but I don’t know what the breakout between “Azure” and “other cloud services” is, so to be generous I’ll ignore the latter.
– The most generous interpretation of “8 points of growth” is 8% of Jun 2023, or $15.4BN x 8% = $1.2BN. The less generous interpretation is 8% of the total growth, or $4.6BN x 8% = $368MM.
– Presuming MSFT’s AI services revenue was very minor in Jun 2023, that suggests total AI services revenue within Intelligent Cloud in Jun 2024 qtr was somewhere between half a billion and modestly over one billion.
Another interesting note from the call is “even in the capital spend, there is land and there is data center build, but 60-plus percent is the kit. That only will be bought for inferencing and everything else if there is demand signal . . . There is definitely spend for training. Even there, of course, we will only be scaling training as we see the demand accrue . . . roughly half of FY 2024’s total capital expense as well as half of Q4’s expense, it’s really on land and builds and finance leases.”
So half the AI capex is land and buildings, most demand to date is training, and MSFT is waiting for “demand signal” for inference before kitting out those DCs with hardware. I guess MSFT investors like that – AMD investors should too.