
Big Tech earnings take centre stage this week as Alphabet, Amazon and Apple report results within 48 hours, forcing Wall Street to judge whether the largest artificial intelligence spending programme in corporate history is actually paying off. The verdict arrives in the same 48-hour window as the Federal Reserve's interest-rate decision on Wednesday, October 28, creating a two-day test of both company fundamentals and the wider market mood.
Alphabet reports third-quarter results after the market closes on Wednesday, October 28, with a conference call at 4:30 p.m. ET. Amazon and Apple follow on Thursday, October 29: Amazon holds its call at 2 p.m. PT, and Apple reports its fiscal fourth quarter after the close that evening. Together, the three companies will show whether cloud growth is still strong enough to justify capital spending now running into the hundreds of billions of dollars a year.
Wall Street heads into the busiest stretch of the earnings season with a single question dominating trading desks: can cloud growth keep pace with the enormous sums being poured into AI infrastructure? Previews published ahead of the reports frame the week as a patience test. Investors are less interested in last quarter's sales in isolation than in what executives say about future spending, and what return that spending is producing.
The schedule itself adds to the pressure. Alphabet's report lands on the same day as the Fed's decision, with the central bank announcing at 2 p.m. ET and holding a press conference at 2:30 p.m. ET. Amazon and Apple then report the following day, alongside the September Personal Consumption Expenditures report, the Fed's preferred inflation gauge. A Dow Jones estimate compiled by MarketWatch puts headline PCE up 0.5% in September on a seasonally adjusted basis and core PCE up 0.3%, with annual rates forecast at 3.7% and 3.1% respectively. Weekly jobless claims, released the same morning, stood at 194,000 in the latest week, near a 57-year low.
Other major names report in the same week, including Robinhood, Reddit, Boeing, T-Mobile and Verizon, as well as Eli Lilly, Chevron, Exxon Mobil, PayPal, Corning and Rivian. But it is the three megacap reports that will set the tone, because their spending plans now move bond markets as well as stock markets.
The scale of the buildout explains why this earnings round matters more than a routine quarterly update. Hyperscaler capital expenditure — the money spent building data centres and buying the chips and servers inside them — is expected to reach roughly $800 billion this year, according to a Reuters preview cited in market coverage this week. That spending is a deliberate shift from pursuing short-term profitability to building long-term infrastructure, and it has already changed how investors judge these companies.
Microsoft and Meta provided the cautionary example. Microsoft beat Wall Street's estimates in its recent report, posting $4.74 in earnings per share against a $4.24 forecast, yet the stock fell as much as 5%. The reason was a single number: Microsoft now expects $190 billion in capital expenditures for 2026, a 61% jump from 2025, after guiding to closer to $150 billion earlier in the year. Its AI revenue run rate is now $37 billion, up 123% year over year, but investors sold anyway.
Meta set the tone days earlier. Its shares fell roughly 6% to 8% after it warned that its own 2026 capital spending would be a good deal larger than in 2025, without giving investors a figure to hold onto. It also booked a one-time $15.9 billion non-cash charge tied to a deferred tax adjustment. Neither stock dropped because the AI business is failing. Both dropped because investors could not yet draw a straight line from the spending to a return big enough to justify it.
The debt market is telling a similar story. Investment-grade debt issued by AI-focused Big Tech companies in September and October alone came in at more than double the sector's average annual issuance between 2015 and 2024, according to Reuters data cited in market reporting. Meta sold $30 billion of bonds, Oracle added $18 billion, and a separate $38 billion high-grade loan was arranged around Oracle's Vantage data centre projects. Five-year credit default swaps on hyperscaler debt are sitting near their widest levels of the year, a sign bond investors are pricing in real doubt about how this spending gets paid back.
Alphabet enters its report with the cleanest defence in the group. The company has lifted its 2026 capital spending guidance twice this year, first to a $175 billion to $185 billion range in February, then up again in July to $195 billion to $205 billion. Its answer to the spending question is Google Cloud: revenue grew 82% year over year in the second quarter, to $24.77 billion. That is the kind of growth that at least gestures toward paying for the servers. Consensus estimates cited in previews call for total revenue of about $127.4 billion to $131.8 billion and earnings per share of $3.04 to $3.10, while Fiscal.ai data cited elsewhere put the figures at $127.83 billion and $3.00. Whether Alphabet holds up on Wednesday is the first real data point in this round.
Amazon's case looks the strongest of the three going in, on cloud growth alone. Amazon Web Services grew 37% in the second quarter, its fastest pace in 18 quarters, pushing cloud revenue to $42.2 billion. Chief executive Andy Jassy raised 2026 capital spending guidance by $20 billion to roughly $220 billion, with most of it earmarked for AI infrastructure. The market rewarded that July report, with shares rising more than 8% afterwards. But the same quarter produced negative free cash flow of $7.6 billion — a reminder of just how much cash this buildout is burning, even at a company with Amazon's scale and retail cushion. Consensus estimates for the new quarter put revenue at about $202 billion and earnings per share at $1.96 to $2.01.
Apple is the outlier of the three. It has far smaller AI capital spending exposure and a cloud business that barely registers next to the other two, so its report on Thursday is more about iPhone demand and services growth than about feeding the same capital spending narrative. That difference could make Apple a relative haven if Alphabet or Amazon disappoint — or simply a separate story in a week dominated by data centre economics.
The market reaction to Microsoft and Meta is the backdrop every analyst note cites this week. Both companies reported demand that, on its own, would normally support their shares. Both fell because the spending outlook rose faster than the comfort level of investors who must now fund it. The Nasdaq posted its worst week since June in mid-October amid an AI selloff, and sentiment toward broad market funds turned extremely bearish in retail surveys late last week, even as the SPDR S&P 500 ETF gained 0.15% in Friday's after-hours session.
The bond market reaction may matter as much as the equity reaction. If Alphabet's cloud growth holds near its second-quarter pace and Amazon's AWS growth holds near 37%, the spending looks earned rather than reckless, and credit spreads may ease. If either number slips even modestly from where it stood in the second quarter, the week changes character: it stops being the moment that resolves the AI spending debate and becomes the moment that deepens it. Bond markets, already primed, are likely to notice first.
For the Federal Reserve, the backdrop is a market that wants clarity on rates without a shock. Previews expect the central bank to keep its target range at 3.75% to 4.00%, with CME FedWatch data cited in market coverage showing 90% odds that rates stay on hold. Interest rates and tech valuations have a well-worn relationship: companies valued on future earnings look less attractive when money costs more today, since those future dollars are discounted more heavily. A hold, with steady language on the path of rates, would remove one source of pressure from a week already full of company-specific risk.
This week matters beyond the three companies reporting because their spending now shapes the whole economy's investment cycle. Data centre construction, chip orders, power contracts and debt issuance all flow from their capital plans. If cloud revenue keeps growing at the pace seen in the second quarter, the buildout looks like infrastructure for a new computing era. If growth slows while spending keeps rising, it looks like a race with no finish line — and the repricing will not be confined to technology shares.
It also matters for ordinary investors. Retirement funds, index funds and pension portfolios hold these companies in large weights, so a sharp move in Alphabet, Amazon or Apple moves the market as a whole. The same logic applies to the Fed decision: a surprise in either direction would land on top of earnings reactions that are already hard to untangle, especially with Alphabet reporting hours after the central bank speaks.
The sequence to watch is clear. Wednesday brings the Fed decision at 2 p.m. ET, followed by Alphabet after the close. Thursday brings Amazon and Apple, the PCE inflation report and jobless claims. Investors should watch three things closely: cloud growth rates against the second-quarter benchmarks, any changes to capital spending outlooks, and the Fed's language on the path of rates.
If Alphabet and Amazon both confirm that cloud demand is still accelerating, the AI spending debate eases — at least until the next quarter's guidance. If either disappoints, expect the focus to shift quickly to balance sheets, free cash flow and debt, and to whether the $800 billion annual buildout pace can survive a more sceptical bond market. Either way, by Thursday night Wall Street will have a much clearer answer to the question it has been asking all autumn: is the AI boom paying for itself yet?
Alphabet reports third-quarter results after the market closes on Wednesday, October 28, with a call at 4:30 p.m. ET. Amazon and Apple both report on Thursday, October 29: Amazon's call is at 2 p.m. PT, and Apple reports its fiscal fourth quarter after the close.
Previews cite consensus estimates of roughly $127.4 billion to $131.8 billion in total revenue and $3.04 to $3.10 in earnings per share, with Fiscal.ai data putting the figures at $127.83 billion and $3.00. Google Cloud's second-quarter growth of 82%, to $24.77 billion, is the benchmark investors will compare against.
Both companies reported strong demand but raised capital spending expectations sharply. Microsoft guided to $190 billion in 2026 capital spending, up 61% from 2025, and its shares fell as much as 5% despite beating earnings forecasts. Meta warned its 2026 spending would be much larger than in 2025, and its shares fell roughly 6% to 8%.
Market previews widely expect the Federal Reserve to hold its target range at 3.75% to 4.00%, with CME FedWatch data cited in coverage showing 90% odds of a hold. The decision is announced at 2 p.m. ET, followed by a press conference at 2:30 p.m. ET.
Hyperscaler capital expenditure is expected to reach roughly $800 billion this year, according to a Reuters preview cited in market coverage. Alphabet has guided to $195 billion to $205 billion for 2026, and Amazon to about $220 billion.
Related coverage on US News Pulse: Meta's $95 billion AI raise leads a $198 billion day of AI spending, Fed rate decision October 2026: why a pause looks likely despite high inflation, and Anthropic IPO filing: Claude maker seeks Nasdaq listing in $2 trillion test.