In June 2026, global capital markets are facing an unprecedented dual squeeze.
On one side, the new Federal Reserve Chair, Kevin Warsh, delivered a hawkish debut. The June FOMC meeting kept the federal funds rate unchanged at 3.50% to 3.75%, but the dot plot reversed dramatically in just three months—from "12 members supporting rate cuts" to "9 members supporting hikes." The 10-year US Treasury yield climbed to around 4.5%, hitting a new high since the onset of geopolitical tensions.
On the other side, the AI arms race continues to heat up. Amazon, Microsoft, Alphabet, and Meta are projected to spend a combined $725 billion in capital expenditures in 2026, up 77% from $410 billion in 2025. This marks the largest corporate investment initiative in peacetime history, now colliding head-on with an increasingly restrictive monetary environment.
These two trends intersected by mid-2026: rising interest rates have increased the cost of capital for all risk assets, while massive AI-related capital expenditures are draining the tech giants’ unprecedented cash flows. Which side will buckle first? The answer hinges on a core variable: as financing costs rise and capital expenditures remain inflexible, how will the market reprice risk?
March to June: A 180-Degree Turn in the Dot Plot
On June 17, 2026, Warsh chaired his first FOMC meeting as Fed Chair. The rate decision itself was uneventful—unanimous approval to hold the federal funds rate steady at 3.50% to 3.75% for the fourth consecutive time. The real shock came from the dot plot.
In March, not a single one of the 19 Fed officials expected a rate hike in 2026, with a median rate forecast of 3.4% and as many as 12 projecting a rate cut within the year. By June, the situation had completely flipped. Warsh himself did not submit a rate forecast—consistent with his long-standing reservations about the dot plot. But among the 18 officials who did, 9 anticipated rate hikes in 2026—3 expected one hike, 5 expected two, and 1 expected three hikes. Only one official still projected a rate cut for the year.
The median year-end 2026 federal funds rate forecast rose from 3.4% in March to 3.8% in June. Median projections for 2027 and 2028 were also revised upward to 3.6% and 3.4%, respectively. Bank of America took an even more aggressive stance, forecasting the Fed would raise rates by 25 basis points each in September, October, and December—a total increase of 75 basis points.
This shift was backed by two sets of data: US CPI rose to 4.2% year-over-year in May, driven mainly by a rebound in energy prices; and 172,000 new nonfarm jobs were added in May, far exceeding expectations, with the unemployment rate holding at a low 4.3%. With high inflation and steady employment, the Fed had no reason to cut rates.
Warsh himself reshaped market expectations on three fronts. First, the policy statement was sharply condensed from 341 words to about 130, stripping out all forward guidance hinting at possible future rate cuts. Second, at the press conference, he placed strong emphasis on inflation risks, making it clear that the Fed would not revisit its inflation target until inflation returned to 2%. Third, he announced the creation of five independent working groups, covering Fed communications, balance sheet management, data sources, productivity and employment, and the inflation framework.
How Bond Yields Squeeze Risk Assets
The shift in the dot plot immediately rippled through the bond market. On June 17, the 2-year Treasury yield surged about 16 basis points to 4.21%, and the 10-year yield climbed 6 basis points to 4.49%. As of June 24, the 10-year yield was hovering near 4.48%.
Rising Treasury yields exert direct valuation pressure on risk assets. When the 10-year yield climbs above 4.5%, the higher risk-free rate means discount rates for all risk assets rise. For equities, a higher discount rate directly reduces the present value of future cash flows. For cryptocurrencies, the opportunity cost of holding non-yielding assets like Bitcoin increases.
On June 23, all three major US stock indices fell sharply. The Nasdaq dropped 579.56 points, or 2.21%, to 25,587.04; the S&P 500 fell 1.44% to 7,365.48. The tech sell-off spread further—Nvidia fell 4.15%, the VanEck Semiconductor ETF dropped 7.01%, Micron closed down 13.18%, and SanDisk fell 13.64%.
The pressure on the cryptocurrency market was even more pronounced. On June 24, Bitcoin dropped 5% to $59,018, breaking below the $60,000 mark and hitting a new year-to-date low. Since the start of the year, Bitcoin has fallen more than 30%. Total crypto market capitalization dropped to $2.15 trillion, the lowest since February 2024. This decline triggered $237 million in long position liquidations within four hours, with total crypto market liquidations reaching $486 million over the same period.
Ethereum’s drop was even steeper. As of June 24, ETH was trading at $1,662, down 3.7% over 24 hours and extending its weekly loss to 7.2%. The ETH/BTC ratio fell to 0.027, a near two-year low. This ratio has dropped sharply from 0.038 at the start of the year, reflecting Ethereum’s weakening position in capital allocation. Over the past 24 hours, total network liquidations reached $2.544 billion, with long liquidations accounting for $2.404 billion, or 94%.
Zach Pandl, Head of Research at Grayscale, pointed out that if the Fed holds rates steady for the rest of 2026, Bitcoin prices could keep pace with stock market gains. But the reality is that the market is now pricing in rate hikes—CME data shows the probability of a September 2026 hike has surged from negligible to over 50%.
The $725 Billion AI Bet
In sharp contrast to rising rates is the tech giants’ aggressive push into AI.
According to company guidance: Amazon plans to invest about $200 billion in 2026; Microsoft expects around $190 billion; Alphabet has guided for $180–190 billion; and Meta has raised its full-year guidance to $125–145 billion. Combined, the four companies are expected to spend a staggering $725 billion in capital expenditures. Morgan Stanley estimates this is about 2.2% of US GDP.
This investment frenzy is transforming the financial structure of tech companies. For years, light capital investment was a key attraction for investors—these companies boasted extremely high free cash flow and steady stock buyback programs. Now, they’ve abruptly become capital-intensive businesses.
Bloomberg economists note that current capex by tech giants far exceeds expectations and is crowding out buyback budgets. Both Microsoft and Meta have redirected more than 100% of their operating cash flow into AI’s "bottomless pit," forcing the industry to issue record amounts of debt to shore up finances. Alphabet is considering its first new stock offering in 20 years to raise about $85 billion. Meta is also reportedly weighing a new share issuance to raise tens of billions.
Stock buybacks were once a key pillar supporting the rally in large-cap US tech stocks. Yet in Q1 2026, after spending a combined $27.9 billion on buybacks in Q1 2025, Meta and Alphabet did not repurchase any shares. Goldman Sachs projects that these four companies alone will spend over $5.3 trillion on capex from 2025 to 2030.
At the same time, debate over AI’s return on investment is intensifying. Nvidia CEO Jensen Huang told shareholders that the company’s fiscal 2026 revenue will grow 65% to $216 billion, with operating cash flow reaching $103 billion. But whether downstream cloud providers can turn these compute investments into sustainable profit growth remains uncertain. Morgan Stanley forecasts that hyperscale cloud capex as a share of sales will reach 36% in 2026 and climb to 44% in 2027—far surpassing the 32% peak for telecom services during the dot-com bubble.
Where the Two Curves Meet
Rising rates and the AI arms race are not parallel storylines. They intertwine on three levels, together shaping the market’s trajectory.
Rising financing costs are eroding the financial viability of AI investments. Tech giants’ massive capex relies on debt and market financing. As the 10-year Treasury yield rises from 4.2% to 4.5%, corporate borrowing costs rise in tandem. For companies needing to raise hundreds of billions each year, every 100-basis-point increase in financing costs translates into billions in extra expenses. Alphabet’s planned $85 billion stock offering will face higher dilution and lower pricing in a pressured market.
AI spending is making inflation more persistent. One of the main drivers behind the Fed’s dot plot shift is inflation—May’s CPI was up 4.2% year-over-year. The AI infrastructure boom is itself an inflationary force: data center construction is boosting demand for building materials, electricity, and chips, while competition for engineers and tech talent is driving up wages in the service sector. In other words, the more intense the AI arms race, the harder it is for the Fed to cut rates—creating a self-reinforcing cycle.
Risk asset liquidity is being squeezed from both ends. On one hand, higher Treasury yields are drawing capital away from risk assets and into safer havens. On the other, reduced stock buybacks by tech giants mean the "passive bid" that supported risk asset rallies in recent years is disappearing. Together, these trends are putting sustained liquidity pressure on both crypto and tech stocks.
Conclusion: Who Will Blink First?
Back to the original question: which will give way first—rising rates or the AI arms race?
This isn’t a binary choice. Both lines of pressure are tightening simultaneously, and the market will find a new equilibrium at their intersection.
If inflation cools gradually amid falling energy prices and easing geopolitical risks, the Fed may hold rates steady or even pivot to cuts—giving risk assets and AI investments some breathing room. Grayscale’s Zach Pandl bases his outlook on this scenario.
But if inflation remains sticky and the Fed is forced to hike rates in the second half of 2026—as Bank of America predicts with 25-basis-point hikes in September, October, and December—then financing costs for AI giants will climb further, and risk asset valuations will face even steeper compression. At that point, the $725 billion capex plans will have to be reevaluated.
For the crypto market, Bitcoin’s drop below $60,000 on June 24, 2026, may be just the beginning. On-chain data shows that if Bitcoin falls below $58,000, over $1.6 billion in leveraged long positions could be liquidated. Market participants are closely watching the June 30, 2026, window. On a broader level, crypto asset pricing is shifting from a "rate-cut trade" to a "rate-hike narrative"—a fundamental paradigm shift.
The race between rising rates and AI’s cash burn will ultimately be decided by inflation data. And the finish line for this contest may arrive sooner than anyone expects.
FAQ
What did the Fed decide at the June 2026 FOMC meeting?
On June 17, the Fed kept the federal funds rate unchanged at 3.50% to 3.75% for the fourth straight meeting. However, the dot plot showed that nine officials expect at least one rate hike in 2026, and the median year-end rate forecast was raised from 3.4% in March to 3.8%.
Why is the 10-year US Treasury yield so important for risk assets?
The 10-year Treasury yield anchors global asset pricing. When yields rise, the risk-free rate increases, raising the opportunity cost of holding risk assets like stocks and cryptocurrencies. The 10-year yield is currently near 4.48%.
How large is tech giants’ AI capex in 2026?
Amazon is planning about $200 billion, Microsoft about $190 billion, Alphabet $180–190 billion, and Meta $125–145 billion. Combined, that’s around $725 billion.
Why did Bitcoin fall below $60,000 on June 24, 2026?
On the macro front, the Fed’s dot plot turned hawkish, and rising rate hike expectations dampened risk appetite. On the market side, Bitcoin slid from a June 23 high above $65,500 to $59,018 on June 24, down over 30% year-to-date.
How will the AI arms race affect the Fed’s rate decisions?
AI infrastructure investment is boosting demand for related goods and services, potentially making inflation more persistent. The more aggressively capital is spent on AI, the harder it becomes for the Fed to cut rates—creating a self-reinforcing policy constraint.
