EconVerse

Stacker_Stocks

Since October 6, thematic baskets—covering areas such as memory, AI semiconductors, data centers, and optical networks—have fallen by between 20% and 35%, according to Bloomberg data.

These are precisely the assets that had seen the largest gains since the launch of ChatGPT in November 2022.

The memory basket had surged by nearly 700%, while the AI semiconductor basket had climbed more than 500%.

Meanwhile, the Nasdaq has fallen by nearly 10%, the S&P 500 by a much smaller margin, and defensive sectors—such as healthcare and food—have risen by more than 10%.

The broader index masks this correction.

In my view, this rotation aligns with the consensus on AI-related capital expenditure (capex), which suggests that investment growth will effectively drop to zero from 2028 onwards.

![](https://m.stacker.news/165036)

The biggest winners of the AI ​​era are already in a bear market.

According to an analysis by J.P. Morgan, there are approximately $4.5 trillion in off-balance-sheet commitments—consisting of lease agreements that have yet to commence and purchase obligations, primarily for data centers and chips.

This figure is six times greater than the amount currently recorded on the balance sheet, which stands at roughly $760 billion in debt and lease liabilities.

Google leads the list with $1.08 trillion in off-balance-sheet commitments, followed by Anthropic ($865 billion) and Meta ($810 billion).

Scary!

![](https://m.stacker.news/164975)

Big tech earnings reports show only a portion of their AI-related obligations.

The 5-year CDS has moved from around 50 basis points in October 2025 to nearly 128 now, having reached a high of 130 last week.

No one is pricing in a potential collapse of Meta.

What the market is pricing in is a shift in the company’s profile: a firm that once funded everything with its own cash has increasingly turned to debt to support its AI-related capital expenditures (capex).

![](https://m.stacker.news/164720)

The cost of protecting against a Meta default has more than doubled in a year.

The consensus projects that AI-related capital expenditure (capex) by major tech companies will double by 2026, reaching $918 billion, and climb to $1.5 trillion by 2027.

Beyond that point, the curve flattens: 10% growth in 2028, 3% in 2029, and 2% in 2030.

This detail is more significant than the absolute figure.

For companies that sell chips, energy, and data center infrastructure, revenue depends on continued growth in investment—not merely on investment remaining at a high level.

If capex stabilizes at $1.7 trillion, suppliers will need to find growth elsewhere from 2028 onward.

On the flip side, the burden falls on those making the investments.

Spending $1.7 trillion annually requires a return, which is why the credit market has already begun charging some of these companies higher costs for capital.

![](https://m.stacker.news/164719)

Consensus predicts Big Tech AI spending will double by 2026.

In December 2019, the figure stood at US$ 7.2 trillion—meaning that assets under management (AuM) doubled in less than seven years.

What I find most revealing is the composition.

Equities grew from US$ 3.8 trillion to US$ 9 trillion, representing a compound annual growth rate (CAGR) of 13.5%, while fixed income grew at a rate of 5.6% per year, rising from US$ 2.3 trillion to US$ 3.4 trillion.

Even with high interest rates globally, new capital flowed into the stock market.

It is worth noting that a significant portion of this is due to valuation effects (mark-to-market) rather than new inflows.

![](https://m.stacker.news/161696)

BlackRock ended Q3 2026 with $14.8T in assets under management.

It is now filing for bankruptcy as part of a restructuring deal and going private.

Getty was one of the world's biggest stock photo companies. It charged businesses and media for photos and videos from its huge library. 

In 2022, it went public at a $4.8 billion valuation.

Then AI image generators arrived. They make the same images in seconds, and Getty's stock photo sites, Agency and iStock, came under pressure.

Getty went to court. It sued Stability AI, the maker of Stable Diffusion, over its images being used to train the AI. A UK High Court judge largely ruled against it.

Its planned merger with its rival Shutterstock then collapsed, leaving Getty alone with heavy debt and growing competition from AI.

By August, Getty pulled its financial guidance. Its debt was $2.1 billion, and its market value is now just $20.74 million.

Now the Getty family and Koch Equity Development will own 100% of what is left.

Getty spent decades making billions by selling images.

AI made it possible to create similar images in seconds. And now a company once worth $4.8 billion has lost almost its entire value.

This is how fast AI can destroy an established business.

![](https://m.stacker.news/161453)

https://x.com/BullTheoryio/status/2109959945613607170?s=20

Getty Images is down -99.95% in the last four years as AI kills its business

Aggregate net long positions rose to around $33 billion—near the cycle high—according to CFTC data through October 6.

What stands out to me is the speed of this shift. In early 2026, this positioning was net short by more than $20 billion. Within just a few months, it flipped to the exact opposite extreme.

Positioning stretched to this extent rarely marks the beginning of a trend.

When almost everyone is already on the same side of the trade, any surprise regarding interest rates or capital flows tends to inflict far more pain than might be expected.

![](https://m.stacker.news/161447)

Speculators have once again built up long positions in the dollar.

Technology was the sector most heavily sold by hedge funds last week, with flows running nearly one standard deviation below the year's average.

Hardware and semiconductors led the way, driven by both the liquidation of long positions and the opening of new short positions.

However, calling this an "unwinding" strikes me as an exaggeration.

Technology still accounts for 19.5% of these funds' net exposure in the US—a level higher than 76% of readings over the past year—and the "Magnificent 7" remain near record highs.

Taking profits after a year like this is normal.

What catches my attention is something else: if the thesis truly stalls, there is still a great deal left to unwind.

![](https://m.stacker.news/161413)

Technology was the sector most sold by hedge funds last week.

Brazil, the world’s largest producer, is diverting more sugarcane towards ethanol as its more profitable. El Niño effect pressures global supply further.

![](https://m.stacker.news/160852)

https://x.com/BullTheoryio/status/2108503581720715427?s=20

Sugar prices surged over 40% in just 60 days, its highest level in 18 months.

The spread between the 10-year French bond and the German Bund reached around 135 basis points—the highest level since 2012, at the height of the Eurozone crisis.

Between 2015 and 2021, this premium hovered near 30 basis points.

France is not a peripheral economy.

It is the bloc's second-largest economy, yet the market is demanding a risk premium typical of a country facing fiscal distress.

There is recurring political instability, a budget lacking a parliamentary majority, and rising debt with no credible adjustment plan.

When the Eurozone core begins trading like the periphery, the issue ceases to be about just one country.

It is a repricing of the bloc itself.

![](https://m.stacker.news/160547)

France under pressure: the risk to Europe's core is back on the radar.

The 10-year Treasury yield has touched its highest level since 2002, nearing 5%.

In 2020, that rate was hovering below 0.6%.

The recent rise was rapid and driven by the long end of the curve, rather than by expectations of monetary tightening.

It reflects a return of the term premium to pricing: debt at 100% of GDP, interest and mandatory spending consuming nearly all federal revenue, and increasing issuance of long-term bonds.

An entire generation of valuation models was built with the 10-year yield below 2%.

At this level, the math changes for virtually every risk asset.

![](https://m.stacker.news/160540)

The 10-year Treasury yield has touched its highest level since 2002, nearing 5%.

Anthropic reported a loss of $42 billion in 2025, compared to approximately $8 billion in 2024.

However, $34 billion of that figure represents an accounting charge linked to financing instruments that could convert into equity—not actual cash outflow.

Excluding that item, the operating loss was $8.06 billion, versus $2.98 billion in 2024.

Computing and infrastructure costs totaled $7.33 billion—three times the previous year's figure and more than half of total operating expenses.

On the other hand, revenue grew twelvefold, reaching nearly $4.6 billion.

The most significant figure isn't found on the income statement: it is the $518 billion in cloud and infrastructure commitments contracted for the coming years.

A twelvefold increase in revenue validates the thesis.

Half a trillion dollars in contractual obligations represents operational leverage—and that only works if the revenue trajectory maintains this pace.

![](https://m.stacker.news/159757)


https://x.com/GlobalMktObserv/status/2105716131549815027?s=20

Anthropic’s losses are SKYROCKETING

The 30-year Treasury yield hit 5.59%, the highest level since 2000.

In March 2020, that same rate stood at 0.99%.

What is being repriced is not short-term inflation, but the term premium.

Debt at 100% of GDP, interest payments and mandatory spending consuming virtually all federal revenue, and rising issuance at the long end of the curve.

Investors buying 30-year paper today want to be compensated for fiscal risk, not the economic cycle.

Duration is no longer the safe haven it was for two decades.

![](https://m.stacker.news/159344)

The 30-year Treasury yield reached 5.59%, the highest level since 2000.

The S&P 500 implied volatility term structure reveals where the market anticipates turbulence.

Forward implied volatility spikes to nearly 19.5% on October 28 and November 4, contrasting with a well-behaved term curve hovering around 12% to 13%.

These are event dates: a Fed meeting, the peak of earnings season, and the US election—all concentrated within a two-week span.

The market is paying a premium for protection against specific dates while paying little for continuous protection.

This speaks volumes about positioning.

Those anticipating calendar-based risk are hedged; those caught off guard by risks outside that window are not.

![](https://m.stacker.news/159287)

The market is protected against the event, not against the surprise.

The IMF projects 3.1%.

The gap between the two is greater than the acceleration seen throughout the Industrial Revolution, which raised growth from 0.6% to 3% per year.

ARK’s thesis is that AI, robotics, and automation are driving a productivity leap comparable to that of the steam engine—only compressed into a single decade rather than a century.

It is worth keeping in mind who is making the projection and what it implies.

7.3% growth justifies practically any valuation multiple in the technology sector.

At the IMF’s 3.1%, the math looks very different.

![](https://m.stacker.news/159004)

ARK projects real global GDP to grow by 7.3% annually through 2030.

Nvidia’s 5-year CDS spread has risen from around 38 basis points at the end of 2025 to 82 now.

It has more than doubled in nine months—for a company with a net cash position and the best margins in the industry.

What the credit market is pricing in isn't the balance sheet; it’s commitment.

Supply agreements and cross-investments involving AI customers create exposure to counterparties that finance their chip purchases with debt.

If capex slows down, the risk ripples back through the supply chain. 

![](https://m.stacker.news/158811)

Nvidia’s CDS Doubles as AI Supply-Chain Risk Comes Into Focus

Whenever the 10-year Treasury yield has risen rapidly over the past five years, the Nasdaq 100 has subsequently corrected.

The chart highlights these episodes in 2021, 2022, 2023, 2024, and late 2025.

What matters is not the rate level, but the speed of the move.

Abrupt movements leave no time to reprice long-term cash flows, and the technology index is precisely the one most sensitive to duration.

The yield is back near 5.0%.

Past performance does not guarantee future results, but the relationship is too consistent to ignore.

![](https://m.stacker.news/158388)

The effect of the sharp rise in interest rates on the Nasdaq

The Kospi and Taiex fell in tandem during Tuesday's session, with the Taiwanese index dropping more than 1% intraday before recouping some of those losses.

The two represent the physical heart of the AI supply chain: memory chips from Korea and foundry services from Taiwan.

When an adjustment begins there, it isn't merely sector rotation; it is a reassessment of expectations regarding the capex cycle itself.

Asian markets open earlier and often provide the first signal.

A single day does not establish a trend.

However, with the 10-year Treasury yield hovering near 5%, the sector most sensitive to duration and debt-financed capex is the first to feel the impact.

![](https://m.stacker.news/158382)

The Kospi and Taiex fell together in Tuesday's session.

![](https://m.stacker.news/158113)

![](https://m.stacker.news/158114)

https://x.com/WatcherGuru/status/2101791102768419047?s=20

YouTuber Logan Paul purchased this NFT for $635,000 in 2021 Today it's worth $96

The Fed is at 4.00%, the ECB at 2.50%, and the BOJ at 1.25%—the highest level for Japan in nearly three decades.

The key factor is the synchronization.

In 2024 and 2025, the three were either cutting rates or holding steady.

Now they are tightening in unison, removing the safety valve that had always existed: with no major central bank maintaining loose policy, there is no source for cheap capital.

Coordinated tightening differs from isolated tightening.

This is the backdrop sustaining high long-term interest rates and elevated risk premiums across all assets dependent on liquidity.

![](https://m.stacker.news/157705)

Fed, ECB, and Bank of Japan raised interest rates simultaneously again.

It even surpassed the 2022 peak, when the war in Ukraine caused energy and grain prices to spike.

The difference is that the current rise is broad-based, rather than driven by a concentrated supply shock.

Simultaneous increases in metals, energy, and agricultural commodities typically signal a cycle rather than a one-off event.

For Brazil, this means favorable terms of trade and support for the exchange rate.

For the global economy, it represents an additional variable in the path toward disinflation.

![](https://m.stacker.news/157264)