There’s one thing I most want to know this earnings season: after the giants have poured so much money into AI, can their balance sheets actually handle it?
For more than a year, cloud providers and AI labs have kept ramping up capital spending.
GPUs, HBM, advanced packaging, optical modules, data center power—the whole supply chain has been lifted by this spending.
It no longer matters how compelling the story sounds.
Now we need to see how much of that money comes back after it’s spent.
Earnings reports start coming thick and fast in the second half of October.
➢ Around the 21st and 22nd: IBM, Tesla, and Intel
➢ On the 28th: Microsoft, Google, and Meta all report on the same day
➢ On the 29th: Apple and Amazon
I’m most interested in the group reporting on the 28th.
If Microsoft, Google, and Meta keep raising their capex guidance without squeezing cloud margins, the case for the hardware and power supply chains will remain solid.
On the other hand, if spending growth slows, or free cash flow comes under significant pressure,
the first to be sold will be the high-valuation stocks propped up by a narrative.
//
As it happens, FOMC is also that same week, on the 27th and 28th.
Long-term yields are already at multi-year highs.
Oil prices have climbed back toward $100 on geopolitical concerns.
If the meeting minutes or dot plot remain hawkish, and oil prices stay elevated, real interest rates will be hard to bring down.
Tech stocks are already stretched on valuation; add this on top and volatility will be high.
━━━━━ • ━━━━━
The direction of money has actually become clear lately:
it’s moving toward areas with real supply-and-demand bottlenecks.
Memory, especially HBM and scarce DRAM; data center power; independent power producers and nuclear power; optical modules; advanced packaging. These areas have long order visibility and strong pricing power. They won’t be afraid of questions about where the money went this earnings season.
Companies relying purely on the AI story, but without capacity or locked-in customers, will be most vulnerable to repricing when earnings pressure and interest-rate pressure hit at the same time.
——
My view is this:
this earnings season will separate the winners from the rest.
Companies that can deliver orders and cash flow will keep attracting capital.
Those that can’t will give back some of the gains they made in recent months on the back of a narrative.
Tech stocks as a whole will be more volatile in the short term, especially with FOMC and earnings landing in the same week.
But volatility doesn’t mean the trend has reversed.
The long-term AI capex cycle isn’t over; we’ve just entered the phase where companies have to show me the numbers. Three reports come out on the 28th.
The amount of information that night could determine where things go over the next month.$IBMB
For more than a year, cloud providers and AI labs have kept ramping up capital spending.
GPUs, HBM, advanced packaging, optical modules, data center power—the whole supply chain has been lifted by this spending.
It no longer matters how compelling the story sounds.
Now we need to see how much of that money comes back after it’s spent.
Earnings reports start coming thick and fast in the second half of October.
➢ Around the 21st and 22nd: IBM, Tesla, and Intel
➢ On the 28th: Microsoft, Google, and Meta all report on the same day
➢ On the 29th: Apple and Amazon
I’m most interested in the group reporting on the 28th.
If Microsoft, Google, and Meta keep raising their capex guidance without squeezing cloud margins, the case for the hardware and power supply chains will remain solid.
On the other hand, if spending growth slows, or free cash flow comes under significant pressure,
the first to be sold will be the high-valuation stocks propped up by a narrative.
//
As it happens, FOMC is also that same week, on the 27th and 28th.
Long-term yields are already at multi-year highs.
Oil prices have climbed back toward $100 on geopolitical concerns.
If the meeting minutes or dot plot remain hawkish, and oil prices stay elevated, real interest rates will be hard to bring down.
Tech stocks are already stretched on valuation; add this on top and volatility will be high.
━━━━━ • ━━━━━
The direction of money has actually become clear lately:
it’s moving toward areas with real supply-and-demand bottlenecks.
Memory, especially HBM and scarce DRAM; data center power; independent power producers and nuclear power; optical modules; advanced packaging. These areas have long order visibility and strong pricing power. They won’t be afraid of questions about where the money went this earnings season.
Companies relying purely on the AI story, but without capacity or locked-in customers, will be most vulnerable to repricing when earnings pressure and interest-rate pressure hit at the same time.
——
My view is this:
this earnings season will separate the winners from the rest.
Companies that can deliver orders and cash flow will keep attracting capital.
Those that can’t will give back some of the gains they made in recent months on the back of a narrative.
Tech stocks as a whole will be more volatile in the short term, especially with FOMC and earnings landing in the same week.
But volatility doesn’t mean the trend has reversed.
The long-term AI capex cycle isn’t over; we’ve just entered the phase where companies have to show me the numbers. Three reports come out on the 28th.
The amount of information that night could determine where things go over the next month.$IBMB