It is the worst day since the 2004 IPO and the second-largest daily loss in American history, behind Nvidia’s in January 2025. The Nasdaq Composite fell roughly 1.4% to 1.7%, and the Nasdaq-100 ended October down more than 6%, with estimates of over $2 trillion in market capitalization wiped out since the month’s highs.
This is not an isolated accident. It is the moment when markets begin to demand that massive investments in artificial intelligence translate into visible cash flow, and not merely into revenue growth or promises.
Spending that outruns incoming cash
The five main players – Amazon, Alphabet, Meta, Microsoft and Oracle – have guided for capital expenditure (capex) of around $720 to $735 billion in 2026, according to aggregates of their own statements. Amazon is aiming for about $220 billion, Alphabet $195 to $205 billion, Microsoft around $170 billion and Meta $135 to $145 billion. These amounts mainly fund data centers, chips and the electricity to power them.
The problem is that this spending now absorbs a growing share – sometimes more than all – of operating cash flow. In the third quarter, Alphabet posted negative free cash flow of about $9.1 billion, with capex of $51.4 billion. Amazon reported negative free cash flow of $42.4 billion over the trailing twelve months, compared with +$14.8 billion a year earlier. The company explicitly attributes this reversal to increased investment in AI-related equipment.
S&P Global pointed out that hyperscalers are likely carrying more AI-related debt or commitments than clearly appears on public balance sheets, through interconnected financing structures, leases and guarantees. Rating agencies have not yet downgraded en masse, but they note the shrinking room for maneuver.
The Alphabet case and monetization
Alphabet’s results served as the trigger. Revenue reached $127.6 billion (+25% year over year) but slightly missed expectations. Earnings per share fell to $2.12, well below forecasts of around $3. Search advertising grew, but more slowly than expected, and executives acknowledged that AI search mode, which answers questions directly, monetizes less well than traditional ad clicks.
Google Cloud, by contrast, accelerated sharply, with growth above 100% by some measures, confirming that demand for infrastructure exists. That is where the tension lies: the cloud pays, but the historic core – advertising – is under pressure just as infrastructure spending explodes. Investors did not wait to see whether the return on investment would arrive in three or five years.
Amazon went through a similar dynamic the next day. AWS grew 39.3% to about $46 billion, its fastest pace in several years. Yet overall results missed expectations and negative free cash flow weighed on the stock, which fell several points. Intel dropped nearly 30% on October 30 after disappointing fourth-quarter guidance, despite strong growth in its data center business.
China as a leading indicator
In China, the deflation is already more advanced. The STAR50 index, concentrated in AI and semiconductor stocks, has lost about 40% since its July high, erasing nearly $480 billion in market capitalization, according to Reuters in late October. Valuations remain high (around 100 times trailing earnings), compared with about 30 times for the Nasdaq-100. The causes cited include overly optimistic demand expectations, export restrictions and a more fragile retail investor base.
This precedent does not prove that Wall Street will follow the same path, but it shows that high valuations combined with massive investment can correct quickly once confidence in monetization weakens.
The crypto parallel and financing
On the same day as Alphabet’s results, Binance suffered a theft of about $6.3 to $6.4 billion, the largest in cryptocurrency history. Analysts at TRM Labs, Elliptic and Chainalysis attribute it with high confidence to the North Korean group TraderTraitor, linked to Lazarus. Entry reportedly came through a compromised laptop belonging to a DevOps engineer. Part of the funds (mostly stablecoins) was frozen; about $2.4 billion in Bitcoin and Ethereum remains harder to block. Binance announced it would cover customer losses.
Bitcoin fell to around $76,000-$77,000 in the hours that followed. This is not the heart of the AI story, but it illustrates how confidence shocks spread across assets that are already under pressure. The Magnificent Seven account for roughly 34% to 35% of the S&P 500. Alphabet and Meta were already in bear market territory (more than 20% below their highs). A correction concentrated on these stocks mechanically hits the broad index and passive portfolios.
Here are the verified key figures as of late October 2026:
- 2026 guided capex for the main US hyperscalers: around $720 to $735 billion (Amazon ~$220B, Alphabet $195-205B).
- Amazon 12-month free cash flow: –$42.4 billion.
- Anthropic net loss in the first half of 2026: $80.7 billion, of which about $72 billion in non-cash accounting charges tied to the valuation of convertible securities; operating loss much smaller (~$693 million).
- Chinese STAR50: about –40% since July, ~$480 billion erased.
- Binance hack: $6.3-6.4 billion on October 29, likely attributed to North Korean actors.
What this really means
AI is not a pure speculative bubble in the sense that the technology doesn’t exist or produces nothing. Cloud revenue is accelerating at several players, and real use cases are being deployed. The main risk is an investment cycle that is too far ahead of monetizable returns, exactly as with the railway mania of the 19th century or the dot-com bubble of 2000. The infrastructure changed the world; a lot of capital was destroyed along the way because returns took longer to appear than valuations allowed.
Today, markets are testing whether free cash flow will turn positive again within a reasonable horizon, or whether part of the capacity being built will sit underused. More opaque financing structures add a layer of credit risk that was not as visible in previous tech cycles. These companies are not short of cash in absolute terms – Alphabet and others hold substantial reserves – but they are converting it into concrete, chips and energy at a pace that forces investors to reassess multiples.
The ongoing correction is not yet a systemic collapse. Balance sheets remain solid for most of the giants, and demand for cloud infrastructure is real. But the market has clearly shifted from a phase in which spending growth was celebrated to one in which the return on that spending is the central question. This is precisely the moment when expectations meet the accounts.
FAQ
Are 2026 AI investments already profitable?
Not in aggregate, as visible in the free cash flow of several hyperscalers. Cloud growth is strong, but infrastructure spending still far exceeds incoming cash for some major players.
Is Anthropic’s $80 billion loss a cash loss?
No. The vast majority is a non-cash accounting charge tied to the increased valuation of convertible securities. The operating loss is much smaller.
Is the Binance hack directly linked to the drop in AI stocks?
No. It is a parallel confidence shock in another market. It amplified general nervousness, but the main trigger for tech stocks remains earnings and guidance on AI spending.
