Updated Oct. 25: Company investor-relations calendars continue to place Microsoft and Alphabet on Oct. 28 and Amazon on Oct. 29, keeping the AI-capex read immediately adjacent to the Fed decision.
Next week compresses two of the market’s biggest questions into little more than a day: what will the Federal Reserve do with rates, and how much more capital will the largest AI platforms commit?
The Federal Reserve is scheduled to release its policy decision at 2 p.m. ET Wednesday, Oct. 28, followed by a 2:30 p.m. press conference. Alphabet and Microsoft both report later that day. Amazon follows after the close on Thursday, Oct. 29.
That sequence matters because the market’s AI debate is changing. The question is no longer simply whether demand for compute is strong. It is whether the companies supplying that capacity can keep expanding while preserving cash flow, returns on capital and acceptable financing costs.
The Oct. 28–29 catalyst calendar
| Time (ET) | Catalyst | What investors will be testing |
|---|---|---|
| Oct. 28 · 2:00 p.m. | Federal Reserve decision | The policy-rate path and the cost-of-capital backdrop for long-duration growth assets. |
| Oct. 28 · 2:30 p.m. | Fed press conference | Whether Chair Kevin Warsh reinforces a hold now while keeping further tightening in play. |
| Oct. 28 · 4:30 p.m. | Alphabet Q3 call | Cloud demand, AI infrastructure spending, depreciation and the pace of capacity additions. |
| Oct. 28 · after U.S. close | Microsoft FY27 Q1 call | Azure demand, AI capacity, capex and whether revenue growth is keeping pace with infrastructure investment. |
| Oct. 29 · 5:00 p.m. | Amazon Q3 call | AWS growth, data-center investment, free cash flow and the capital intensity of the next capacity wave. |
Meta changed the tone of the capex debate this week
Meta’s third-quarter results provide the clearest recent example of why investors are becoming more selective. The company reported $34.91 billion of capital expenditures in the quarter and negative $2.12 billion of free cash flow, while narrowing its 2026 capital-expenditure outlook to $135 billion–$145 billion.
Those figures do not imply weak AI demand. Meta said its AI investments are already producing incremental advertising revenue and it plans to keep investing aggressively. But they demonstrate the market’s new constraint: strong demand can coexist with deteriorating near-term cash generation.
The balance sheet is becoming part of the AI product
For the hyperscalers, access to capital is a competitive advantage. A company that can finance servers, networking equipment, power and data centers out of operating cash flow has more room to keep building through a credit-market wobble than a highly leveraged infrastructure provider.
That creates a bifurcation. The largest platforms can still justify huge spending if cloud growth, utilization and monetization keep rising. Smaller or more leveraged providers face a much higher hurdle because widening spreads or expensive secured debt can make otherwise attractive projects uneconomic.
Amazon’s second-quarter release said AWS sales grew 37% year over year to $42.2 billion. At the same time, purchases of property and equipment contributed to trailing-12-month free cash flow of negative $7.6 billion. That combination — very strong cloud growth alongside heavy infrastructure spending — is the central tradeoff investors will revisit when Amazon reports.
Amazon’s cash-flow baseline shows why capex matters
What to compare across Microsoft, Alphabet and Amazon
| Question | Why it matters now |
|---|---|
| Does cloud/AI revenue growth accelerate? | Higher utilization is the strongest answer to concerns that infrastructure is being built ahead of monetizable demand. |
| Does capex guidance rise again? | Another increase would reinforce the scale of the buildout, but also increase depreciation and future cash demands. |
| What happens to free cash flow? | The market is increasingly distinguishing accounting earnings from cash left after infrastructure spending. |
| Are leases, debt or external financing becoming more important? | Funding mix determines how exposed the buildout is to credit spreads and refinancing conditions. |
| Do management teams describe capacity as scarce? | Persistent capacity constraints would support the view that spending is responding to real demand rather than speculative overbuilding. |
Why the Fed comes first
The earnings numbers will land into a rate environment set only hours earlier. A more hawkish Fed can raise the discount rate applied to future AI cash flows and increase the hurdle rate for debt-funded infrastructure. A more benign rates message can ease that pressure, but it does not remove company-specific questions about capital efficiency.
This is why the same earnings report can produce a different market reaction at 4.8% versus 5.3% on the 10-year Treasury. The earnings are the numerator; the cost of capital changes the denominator.
What would calm the market?
The cleanest combination would be strong cloud demand, disciplined or well-explained capex, and evidence that free cash flow can absorb the investment cycle. The riskier combination would be another upward reset in spending with little change in near-term monetization or cash generation.
Markets & News is not treating the outcome as predetermined. The point of the Oct. 28–29 window is that it will provide three large-company tests of the same financing question immediately after the Fed resets the macro cost-of-capital backdrop.
Primary calendars: Microsoft Investor Relations · Alphabet Investor Relations · Amazon Investor Relations