🔍 Read the full analysis: Six Clues To The AI Boom’s Health Before Big Tech Reports on ThorstenMeyerAI.com
Get audio and creator gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
SK hynix reported preliminary third-quarter operating profit of 83.3 trillion won, while Microsoft and Alphabet were due to report later on October 28 and Amazon on October 29. The figures point to real demand and strong chip profits, but also a concentrated earnings outlook, rising data-center spending and unanswered questions about cash-flow payback.
SK hynix reported preliminary third-quarter operating profit of 83.3 trillion won on October 28, about 7.3 times its result a year earlier, as investors awaited quarterly results from Microsoft, Alphabet and Amazon for further evidence of whether AI infrastructure spending is translating into returns. The memory maker’s record profit signals strong demand at one part of the AI supply chain, but does not settle whether the technology investments made by cloud companies will pay off.
SK hynix said revenue reached 104.6 trillion won, up 328% year over year, and operating margin was 80%. Its net profit was 39.8 trillion won, down 58% from the previous quarter, which the company attributed to non-operating items. The results are preliminary and unaudited. SK hynix also announced a 14 trillion won dividend and a 20 trillion won share buyback this quarter; half of the buyback is for employee compensation and is not counted by the company as a shareholder return.
The company said long-term agreements with 14 major customers cover more than half of its planned 2027 and 2028 capacity. Its expectation that demand will exceed supply through 2030 is a company forecast, not an established outcome. The source report said SK hynix shares rose during the Seoul session and that the company halted a review of funding or a potential listing for its US unit Solidigm, citing market volatility and a correction in semiconductor stocks.
Other indicators show both demand and risk. FactSet estimated on October 23 that S&P 500 earnings would grow 29.5% in the third quarter. Its October 19 analysis found that removing Micron and Nvidia would lower the estimated growth rate from 29% to 20%. The Census Bureau reported that US shipments of core capital goods rose 1.2% in September, while orders for computers and related products were up 20.3% this year. These figures point to activity, but do not by themselves show how much AI investment will generate in future revenue.
Infographic 01 · The boom in numbers
Record profits, narrow growth, rising bills.
Six numbers from company releases, FactSet and the Census Bureau, before Microsoft, Alphabet and Amazon report.
Who carries the earnings boom
Q3 2026 earnings growth, year over year. Blended: reported results plus estimates.
The Wall Street Journal (Dow Jones Market Data) puts the two chipmakers at 34% of the expected growth. Top five contributors per FactSet: Micron, Nvidia, Meta, EchoStar, Boeing.
What the buyers plan to spend in 2026
Capital spending guidance, US dollars. Bars show the top of each range.
Q3 consensus: Alphabet free cash flow −$2.3B on $55.2B capex; Amazon −$9.9B on $58.7B. Definitions differ slightly (Microsoft includes finance leases).
Shovel seller
- Revenue 104.6 trillion won, +328%
- Long-term deals cover >50% of 2027–28 capacity
- About 100 trillion won back to shareholders in 2026
Real economy
- Core capital goods orders +10.9% this year
- Computer orders +20.3% this year
- Census data do not label AI; semiconductors excluded from orders
Debt nerves
- Cost to insure Meta’s debt +177% this year, Amazon’s +129%
- Oracle: $117B long-term debt, rated BBB-
- Off-balance-sheet leases don’t show up as debt (S&P)
Sources: SK hynix 3Q26 release, 28 Oct 2026 (preliminary); FactSet, 19 Oct 2026; Alphabet Q2 call, 22 Jul 2026; Amazon via CNBC, 30 Jul 2026; Microsoft via CNBC, 29 Jul 2026; Meta Q3 release, 21 Oct 2026; FactSet consensus via CNBC, 27 Oct 2026; US Census Bureau, advance durable goods report, 27 Oct 2026. Capex sum and bar scaling: Thorsten Meyer.
Powered by Thorsten Meyer AI
Testing AI Spending Against Returns
The results matter because AI infrastructure requires large, continuing investments in chips, memory, servers and data centers. Strong chipmaker earnings indicate that suppliers are benefiting from demand now. The larger test for investors is whether cloud providers can convert their spending into durable AI revenue and cash flow.
The source report estimates that Alphabet, Amazon, Microsoft and Meta plan a combined $725 billion to $745 billion in capital spending in 2026, based on the writer’s sum of company guidance. It also says analysts expect Alphabet and Amazon to post negative free cash flow again for the third quarter. Those projections are not reported results, and the spending total is an estimate based on guidance. Still, they frame the question facing investors: how quickly can AI-related business support the cost of building capacity?
Market strength is also uneven. The report says Micron and Nvidia account for a substantial share of expected US earnings growth, while China’s STAR50 index has fallen sharply since July. A healthy result from one supplier or one quarter would not establish that the wider AI investment cycle is sustainable.
Products Worth Considering
DJI Osmo Mobile 8 Gimbal with Mic Mini Transmitter, Infinity Black
Bose QuietComfort Wireless Noise-Cancelling Over-Ear Headphones, Black
Sony WH-1000XM6 Wireless Noise-Canceling Over-Ear Headphones, Blue, Bundle
As an affiliate, we earn on qualifying purchases.
Supply Gains, Markets Diverge
SK hynix makes high-bandwidth memory, stacked memory used alongside AI processors in data centers. Its results provide a view of demand for an important component, rather than a complete measure of the economics of AI services. The company’s operating profit rose sharply year over year, while its net profit fell from the previous quarter, underscoring that different measures can tell different stories.
In the United States, the earnings figures cited are estimates issued before all major companies had reported. FactSet’s analysis says the projected S&P 500 growth rate remains 20% even after excluding Micron and Nvidia, which is still substantial growth. The result also shows how headline growth can depend on a small number of companies. Some contributors, including Meta, EchoStar and Boeing, had one-time charges in the year-earlier quarter, according to FactSet, making comparisons less straightforward.
China offers a contrasting market signal. The source report, citing Reuters on October 26, says the STAR50 index had lost 40% since July and about $480 billion in value, while trading at roughly 100 times past earnings. These market figures describe investor pricing and expectations; they do not establish the performance of every Chinese AI or chip company.
“The company expects demand to exceed supply “through 2030.””
— SK hynix
The Payback Is Still Unproven
It is not yet clear whether AI revenue will cover the cost of new capacity across the cloud companies. Their quarterly reports may provide updated spending plans, cloud growth and cash-flow figures, but one quarter cannot establish the long-term return on investment. The combined 2026 spending estimate and expectations for negative free cash flow are drawn from the source report; they are not final annual spending totals or confirmed results for the quarter.
Several earnings-growth figures are forecasts because companies had not reported when the preview was written. SK hynix’s results are preliminary and unaudited, and its forecast of demand through 2030 may change. The source material also does not provide the full quarterly disclosures from Microsoft, Alphabet or Amazon, so their latest reported AI revenue, capacity plans and cash-flow figures are not available here.
The Chinese market decline is another uncertain signal: it records a fall in the STAR50 index, not proof that AI demand in China has collapsed. Likewise, a high valuation multiple does not alone determine future performance. The six indicators offer clues about the boom’s condition, but they cannot resolve whether current investment levels will prove sustainable.
Cloud Results Set the Next Test
Microsoft and Alphabet are scheduled to report on October 28, followed by Amazon on October 29, according to the source report. Investors will be watching company guidance on capital spending, cloud growth relative to that spending, free cash flow and remarks about data-center capacity. These disclosures should add current company-specific evidence to the broader indicators already available.
The source report’s preview calls out four measures to watch: capex guidance, cloud growth, free cash flow and capacity plans. Comparing those figures with prior guidance and reported results will help distinguish planned investment from spending already incurred, and revenue growth from cash returns. No later results or dated update are included in the supplied material, so the outcome of the scheduled reports remains unknown here.
Key Questions
What did SK hynix report?
SK hynix reported preliminary, unaudited third-quarter operating profit of 83.3 trillion won, up about 7.3 times year over year. Revenue was 104.6 trillion won and operating margin was 80%.
Why are Microsoft, Alphabet and Amazon’s results important?
The companies are among the major cloud providers investing heavily in AI infrastructure. Their reports may show whether cloud growth and cash flow are keeping pace with capital spending.
Does SK hynix’s result prove the AI boom is sustainable?
No. It shows strong results at a key memory supplier, but does not establish that cloud companies will earn enough from AI to recoup their infrastructure investments. The company’s supply outlook through 2030 is a forecast.
How concentrated is expected US earnings growth?
FactSet’s October 19 analysis said removing Micron and Nvidia would reduce estimated third-quarter S&P 500 earnings growth from 29% to 20%. Those figures were estimates before all companies had reported.
Source: ThorstenMeyerAI.com
