On Dec. 11, Oracle lost about $80 billion in market capitalization due to lower-than-expected sales, and management raised AI-related capital spending from $35 billion to about $50 billion, partially funded by increased debt.
Stocks fell as much as 16%, dragging Nvidia, AMD, and the entire Nasdaq lower.
Reports have described the move as fueling fears of an “AI bubble,” with investors questioning whether the rewards from building out large-scale data center capacity will come fast enough to justify the costs.
On the same tape, Bitcoin fell below $90,000, likely due to concerns about the AI sector’s declining risk appetite.
This one-day episode encapsulates Bitcoin’s new structural vulnerabilities. Bitcoin has a high beta tail in AI trading, moves in lockstep with tech stock sentiment, and bleeds even harder when AI stocks break.
According to analysis by 24/7 Wall St., the correlation between Bitcoin and Nvidia reached approximately 0.96 in the three months leading up to Nvidia’s November earnings.
As for the Nasdaq, the 30-day aggregate Pearson correlation coefficient was 0.53 as of Dec. 10, according to data from The Block.
Furthermore, since the Fed began cutting interest rates on September 17th, Bitcoin is down about 20%, while the Nasdaq is up 6%. This suggests that when tech stocks crash, Bitcoin’s price will fall sharply.
The AI bubble story has matured rapidly in recent weeks.
Reuters reported in late November that while macro indicators such as AI-related valuations and Buffett Indices have pushed overall U.S. stock valuations beyond the extreme levels of the dot-com era, AI-heavy indexes are showing sharp pullbacks and increased volatility, even as enthusiasm remains high.
Moreover, big tech companies have raised hundreds of billions of dollars in corporate bonds this year to fund data centers and hardware. Morgan Stanley estimates there is a roughly $1.5 trillion funding gap for AI infrastructure, and Moody’s chief economist Mark Zandi warned that AI-related borrowing now exceeds the run-up of tech companies before the dot-com crash.
The Bulletin of the Atomic Scientists and an essay in The Atlantic both state that AI spending this year will be about $400 billion, while revenue will only be about $60 billion.
This calculation suggests that most companies are running deep losses and that the entire economy is now partially reliant on an AI investment boom that will not last forever.
Liquidity mechanisms that exacerbate Bitcoin’s AI failure
If the AI bubble bursts, the damage to Bitcoin will go beyond simple correlation, as AI capex becomes increasingly a confidence story.
According to estimates, AI-related data center and infrastructure financing transactions will jump from about $15 billion in 2024 to about $125 billion in 2025, driven by corporate debt issuance, private credit, and asset-backed securities.
In a Reuters article, analysts compare some of the structures and opacity to pre-2008 patterns and warn of “untested risks” if tenants or cash flows fail to meet expectations.
Central banks are now treating this as a financial stability issue. The Bank of England’s recent stability update clearly highlights the inflated valuations of AI-focused companies. It also warns that a sharp correction in AI stocks could threaten the broader market through leveraged players and private credit exposure.
A similar point was made in the ECB’s November 2025 Financial Stability Review. The AI investment boom is increasingly being financed through bond markets and private capital, making it more exposed to fluctuations in risk sentiment and credit spreads.
Oracle is a typical example. The $50 billion capital spending plan for AI data centers, along with a nearly 45% jump in long-term debt and record credit default swap spreads, represents exactly the kind of oversized balance sheet that regulators are concerned about.
When the AI bubble bursts, spreads will widen, refinancing costs will soar, and funds that have long been leveraged in AI-themed bonds and stocks will be forced to reduce their total exposure. Bitcoin is at the end of that chain.
An analysis of Bitcoin versus global liquidity by Chinese researchers found a strong positive relationship between Bitcoin price and global M2 or broad liquidity index. Their paper calls BTC a “liquidity barometer” that performs well when global liquidity is high and performs poorly when it contracts.
The liquidity story is simple. If the AI bubble bursts and a credit crunch is forced, the first-order effect will be global risk reduction and reduced liquidity.
Bitcoin is one of the first things macro and growth funds sell when margin calls occur, and over-sensitivity to liquidity exacerbates drawdowns.
Act 2: How policy responses will fuel Bitcoin’s next bullish cycle
The other half of the story is what happens after the first wave of deleveraging.
The same agencies concerned about AI-driven remediation are also implicitly pointing to possible responses. If AI and credit market overleverage are sufficiently shaken to threaten growth, central banks will ease financial conditions again.
The IMF’s latest Global Financial Stability Report warns that AI-driven equity concentration and the expansion of risk asset valuations are increasing the likelihood of a “disorderly adjustment” and highlights the need for careful but ultimately supportive monetary policy to avoid escalating shocks.
History provides a template. After the coronavirus shock in March 2020, the market capitalization of cryptocurrencies significantly increased from approximately $150 billion in early 2020 to approximately $3 trillion by late 2021 due to aggressive quantitative easing and liquidity provision.
A recent Seeking Alpha report mapping Bitcoin against global liquidity and dollar indices shows that once easing begins in earnest and the dollar weakens, Bitcoin tends to rally significantly over the following quarters.
The rotation of the story is also important. If AI stocks experience the classic post-bubble hangover of lower multiples, negative headlines, and political backlash against wasteful capital spending, some of the speculative and macro capital could turn into bets on alternative “futures of money” and “anti-establishment.”
Bitcoin is the cleanest non-corporate candidate.
Recent market stress has already focused capital on Bitcoin rather than alt currencies. Amid recent declines in liquidity and increased volatility, Bitcoin’s dominance has risen to around 57%, with ETFs serving as the institutional gateway.
Furthermore, although Bitcoin has recently shown correlation with tech stocks, decentralization and scarcity are still at the core of the “hedging” narrative.
Bitcoin trade-offs cannot be avoided
Bitcoin’s structural problem is that it cannot be separated from AI trading in the short term, but its upside potential in the medium term depends on policy responses to AI failure.
Immediately after the AI credit crunch, Bitcoin will bleed as it has a high beta tail of macro risk and global liquidity contracts faster than most assets can adjust.
In the ensuing months, Bitcoin has historically reaped significant gains as central banks respond with new easing measures and the dollar weakens, as liquidity returns to risk assets and the speculative narrative is reset.
The question for allocators is whether Bitcoin can survive the initial blow enough to benefit from a second wave.
The answer will depend on how sharp the AI adjustment is, how fast the policy pivot is, and whether institutional capital flows through ETFs and other instruments hold up or collapse under stress.
Oracle’s December 11 earnings error is a sign of foreboding. Bitcoin fell below $90,000 on the same tape that wiped out Oracle’s $80 billion market cap, showing that the correlation is alive and the sensitivity is real.
When the AI bubble fully unwinds, Bitcoin will be the first to be attacked. Whether it becomes even stronger depends on what central banks do next.
However, one positive short-term indicator emerged late in yesterday’s trading session. Nvidia has recovered 1.5% from its intraday low, with Bitcoin following suit, rising more than 3% to regain $92,000.