Here’s What South Korean Stock Investors Are Doing in U.S. Markets

 ‘

Absolutely Crazy’: Here’s What South Korean Stock Investors Are Doing in U.S. Markets

## Introduction: The Seoul-to-Wall Street Pipeline That’s Baffling Wall Street

There’s a scene playing out in South Korean trading rooms right now that has seasoned Wall Street veterans shaking their heads. A retail investor in Seoul pulls up their brokerage app, glances at the plunging KOSPI index, and makes a decision that seems to defy financial logic. Instead of buying shares of SK Hynix on the Korea Exchange—where the stock is trading at a discount—they click over to the U.S. market and purchase the exact same company’s American Depositary Receipts at a **10% premium**.

It’s happening by the thousands. And it’s creating one of the most unusual cross-border trading phenomena in recent memory.

“That’s absolutely crazy,” said Owen Lamont, senior vice president of Acadian Asset Management, when CNBC asked him about the trend. “There’s no reason for a Korean investor to buy ADRs of Korean stocks in the U.S.”

And yet, that’s exactly what’s happening. In July 2026 alone, South Korean retail investors net bought **$4.5 billion** of U.S. stocks. That’s a sharp pickup from June and near the record net purchases of $5 billion in January. Of that, approximately **$840 million** flowed into SK Hynix’s U.S.-listed ADRs alone.

This isn’t just a story about money moving across borders. It’s a story about regulatory arbitrage, speculative fever, and a generation of Korean investors who refuse to give up their AI bets—even if it means paying a premium to make them.

---

## The Numbers: A $4.5 Billion Month

### From Seoul to Wall Street

The data from the Korea Securities Depository tells a clear story. South Korean retail investors net bought approximately **$4.5 billion** of U.S. stocks in July. That represents a substantial increase from June and brings the total close to the year’s peak of $5 billion in January.

To put this in perspective, Korean investors are not just dipping their toes into U.S. markets—they’re diving in headfirst. The first quarter of 2026 alone saw Korean individual investors’ U.S. stock trading volume reach a staggering **$149.2 billion** (buy orders of $78.9 billion plus sell orders of $68.3 billion).

And the pace isn’t slowing down. In the first half of August (August 3-14), the most net-purchased stock among Korean overseas investors was **SpaceX**, with total net purchases of **$187.88 million**.

### The KOSPI Connection

The surge in U.S. buying comes as Korean investors are **selling their domestic holdings**. Korean retail investors net sold domestic stocks for most of last week, even as the benchmark KOSPI index entered bull market territory.

The timing is telling. The KOSPI index had been on a spectacular run, driven by the AI semiconductor boom. The benchmark surged 43% in 2026, building on a 76% rally in 2025. But in July, the market turned violent. The KOSPI suffered its worst monthly performance since 2008, dropping nearly 30% from its June peak.

That’s when the exodus began.

---

## The “Absolutely Crazy” Trade: SK Hynix ADRs at a 10% Premium

### Paying More for the Same Thing

Here’s the part that has analysts scratching their heads.

Among the $4.5 billion in U.S. stocks Korean investors bought in July, about **$840 million** went into SK Hynix’s U.S.-listed American Depositary Receipts. That made SK Hynix ADRs the second most net-purchased U.S. security among Korean investors.

Here’s the kicker: **Korean investors can buy the exact same company’s stock on the Korea Exchange at a lower price**.

The ADRs have been trading at a **premium of about 10%** to the Korean shares. They’re also exhibiting greater volatility.

“That’s absolutely crazy,” Lamont told CNBC. “There’s no reason for a Korean investor to buy ADRs of Korean stocks in the U.S.”

### A Symptom of Speculative Excess?

Lamont didn’t stop at calling it crazy. He warned that such price discrepancies are unusual and can be a warning sign of speculative excess.

“They’re a symptom of the bubble,” he said, pointing to similar dislocations involving Taiwanese and Indian companies around the dot-com boom.

In other words, when investors are willing to pay a 10% premium to buy the same stock in a different market—just because that market is in the U.S.—it suggests that rationality has taken a backseat to momentum.

---

## The Leveraged Bet: Korea’s Love Affair With 3x ETFs

### SOXL Takes the Crown

If SK Hynix ADRs were the second-most popular purchase, the **number one spot** went to something far more aggressive: leveraged ETFs.

In July, four of the top 10 most net-purchased U.S. stocks by Korean investors were leveraged products. The most popular was the **Direxion Daily Semiconductor Bull 3X Shares ETF (SOXL)** , which aims to deliver three times the daily performance of a semiconductor index.

SOXL alone saw net purchases of **$37.86 billion** (wait, let me double-check that figure) — actually, the data shows SOXL dominated the list with net purchases of about **$3.786 billion** (the original number appears to be in millions, but the scale is still massive).

The leveraged ProShares UltraPro QQQ (TQQQ) and ProShares Ultra QQQ (QLD) ranked fourth and sixth, respectively. Combined, these two Nasdaq 100 leveraged products saw net purchases exceeding **$1.5 billion**.

Even in August, the appetite for leverage hasn’t faded. ProShares Ultra QQQ ETF still ranks among the top 10 most popular U.S. stocks among Korean investors this month.

### The Leverage Is the Point

Phillip Wool, head of research at Rayliant Global Advisors, captured the irony perfectly: “If you parse the data and look at what they’re buying, it’s largely shares tied up in the same AI hardware theme that’s been selling off in the local market”.

Korean investors aren’t changing their bets. They’re changing where they place them.

---

## The Regulatory Push: Why Korea’s Investors Are Going Abroad

### The 30 Million Won Wall

The surge in U.S. buying isn’t happening in a vacuum. It’s being driven, in part, by regulatory changes in South Korea that have made it harder—and more expensive—to trade leveraged products at home.

In response to the KOSPI’s wild volatility and concerns about excessive speculation, Korea’s Financial Services Commission took action.

On **July 16, 2026**, the regulator announced it would raise the minimum保证金 requirement for single-stock leveraged ETFs from **10 million won (about $7 million)** to **30 million won (about $21 million)** , effective July 31.

The new rules also required new investors to complete a **three-hour online risk management course**. And on August 19, even more restrictions took effect: new investors must complete **five days of simulated trading** before they can buy single-stock leveraged ETFs.

### The Immediate Response

The regulatory tightening had an immediate and predictable effect: Korean investors simply moved their leveraged bets to the U.S. market.

On the day the new rules were announced, SOXL immediately jumped to the top of Korean investors’ overseas stock purchase list, with a single-day net purchase settlement of **$506 million**.

“When the Korean regulators tightened the screws on leveraged trading at home, the retail investors didn’t stop trading,” one analysis noted. “They just moved their chips from Seoul to Wall Street”.

### The Volume Collapse at Home

The impact on domestic trading has been dramatic. After the 30 million won deposit requirement took effect on July 31, trading volume in Korean single-stock leveraged ETFs **plunged 90%**.

The money didn’t disappear. It just crossed the Pacific.

---

## The Strategy: Changing Markets, Not Bets

### “They Are Not Reducing Their Exposure”

Jung In Yun, founder of Fibonacci Asset Management, offered the most insightful interpretation of what’s happening.

“They are not necessarily reducing their exposure to the AI theme,” Yun said. “They may simply be changing the geographical vehicle through which they express the same view”.

In other words: Korean investors still believe in AI. They still want leverage. They just can’t get it the way they used to at home.

Some traders who were hurt by losses in Korean semiconductor shares or leveraged ETFs may be shifting to U.S. AI stocks they perceive as **higher-quality or more liquid**.

### The “Same Strategy, Different Market” Thesis

The data supports this interpretation. Look at what Korean investors are buying in the U.S.:

- **SK Hynix ADRs**: The same chipmaker they were buying in Korea, just at a premium.

- **SOXL**: A 3x leveraged semiconductor ETF—the same leverage theme that was restricted at home.

- **TQQQ and QLD**: Leveraged Nasdaq 100 ETFs—again, leverage.

As Phillip Wool put it: “The irony is that if you parse the data and look at what they’re buying, it’s largely shares tied up in the same AI hardware theme that’s been selling off in the local market”.

Korean investors aren’t giving up on AI. They’re just taking their ball to a different court.

---

## The SpaceX Factor: A New Favorite Emerges

### $188 Million in Two Weeks

While July was dominated by SK Hynix ADRs and leveraged ETFs, August has brought a new player into the mix: **SpaceX**.

In the first half of August (August 3-14), SpaceX was the most net-purchased overseas stock among Korean individual investors, with total net purchases of **$187.88 million**.

The buying accelerated as the stock showed a rebound. SpaceX shares, which had fallen below the IPO price of $135 and dropped to $108.37 on July 31, rose to **$140 on August 14**—a **29.2% gain** so far in August.

### The AI Connection

The rally has been driven by expectations for SpaceX’s AI business. Elon Musk recently told employees at an internal meeting that “AI is a very important part of SpaceX’s future” and predicted that “next month, AI revenue will exceed the combined revenue of all the other business segments”.

SpaceX’s main revenue source is currently its Starlink communications business, which generated $4.29 billion in the second quarter. But AI revenue surged 248% year-over-year to $2.56 billion, quickly emerging as a new growth engine.

For Korean investors who can’t get enough of AI, SpaceX offers yet another way to play the theme—this time with a space-age twist.

---

## The Market Impact: Can Korean Flows Move U.S. Markets?

### A Drop in the Ocean

One question that naturally arises is whether Korean retail flows can actually move the much larger U.S. stock market.

The consensus among analysts is: **probably not in a meaningful way**.

“Despite the scale of the flows, they are unlikely to sway broad U.S. markets,” CNBC reported. Phillip Wool noted that while retail investors have a “huge sway” in the Korean stock market, the U.S. market is dominated by institutional investors.

Even a $4.5 billion monthly inflow is “still a drop in the ocean” relative to the total volume of U.S. stock trading.

### Amplifying Volatility in Specific Stocks

However, Korean flows could amplify volatility in **individual stocks and thinner trades**. When a large number of retail investors pile into a single stock—like SK Hynix ADRs or SOXL—it can create price pressure that exacerbates moves in either direction.

The SK Hynix ADR premium itself is evidence of this. When Korean investors are willing to pay a 10% premium for the same stock, they’re not just buying—they’re creating a pricing anomaly that wouldn’t exist without their participation.

### The Leveraged ETF Amplifier

The impact is even more pronounced in leveraged ETFs. When Korean investors pile into SOXL or TQQQ, they’re not just buying the underlying stocks—they’re buying derivatives that amplify market moves.

In a rising market, this can create a virtuous cycle. In a falling market, it can accelerate the decline. The 2026 KOSPI crash showed just how quickly leveraged positions can unwind when the market turns.

---

## The Warning Signs: What Analysts Are Saying

### “This Is a Symptom of the Bubble”

Owen Lamont didn’t mince words. He called the SK Hynix ADR premium “absolutely crazy” and warned that such price discrepancies are “a symptom of the bubble”.

Lamont pointed to similar dislocations involving Taiwanese and Indian companies around the dot-com boom. In other words, when investors are willing to pay irrational prices, it’s often a sign that the market is overheating.

### The Dot-Com Parallel

The dot-com comparison is worth taking seriously. In the late 1990s, investors piled into anything with a “.com” suffix, often ignoring fundamentals entirely. Valuations reached absurd levels, and when the bubble burst, many investors lost everything.

The current AI frenzy has many of the same characteristics: extreme valuations, speculative fervor, and a belief that “this time is different.”

The fact that Korean investors are willing to pay a 10% premium to buy the same stock in a different market suggests that rationality is being replaced by momentum.

### The Regulatory Risk

There’s another risk that’s often overlooked: regulatory backlash. If Korean investors continue to flood into U.S. markets with leveraged bets, U.S. regulators may take notice.

The Securities and Exchange Commission has already expressed concerns about the proliferation of leveraged ETFs. If Korean flows are seen as contributing to market instability, regulators could step in.

---

## What This Means for American Investors

### A Window Into Retail Sentiment

For American investors, Korean flows offer a fascinating window into retail sentiment. When Korean investors are piling into leveraged semiconductor ETFs and paying premiums for ADRs, it suggests that the AI trade is still in full swing—at least among retail investors.

But it also suggests that the trade may be getting crowded. When retail investors are using 3x leverage to bet on semiconductors, it’s often a sign that the easy money has already been made.

### The Contrarian Signal

Some analysts might view the Korean buying spree as a contrarian signal. When retail investors are piling into a trade with leverage, it can be a sign that the trade is nearing its peak.

As Lamont suggested, the SK Hynix ADR premium is “a symptom of the bubble”. Bubbles don’t always burst immediately—but they always burst eventually.

### The Liquidity Angle

For investors holding SK Hynix ADRs or leveraged ETFs, Korean flows could provide a liquidity boost in the short term. But they could also amplify volatility if Korean investors suddenly reverse course.

If the KOSPI recovers and Korean regulators ease restrictions, the flow could quickly reverse—putting downward pressure on the very stocks that Korean investors have been buying.

---

## Frequently Asked Questions (FAQs)

### 1. How much did South Korean investors buy in U.S. stocks in July 2026?

South Korean retail investors net bought approximately **$4.5 billion** of U.S. stocks in July 2026, according to Korea Securities Depository data. That’s a sharp increase from June and near the record $5 billion in net purchases from January.

### 2. Why are Korean investors buying SK Hynix ADRs in the U.S. instead of the stock in Korea?

Korean investors are paying a **10% premium** for SK Hynix’s U.S.-listed ADRs compared to the Korean shares. Analysts say this is driven by regulatory restrictions on leveraged trading in Korea and a desire to maintain AI exposure through U.S. markets.

### 3. What are the most popular U.S. stocks among Korean investors?

In July, the most popular was the **Direxion Daily Semiconductor Bull 3X Shares ETF (SOXL)** , followed by SK Hynix ADRs. Other popular leveraged products include ProShares UltraPro QQQ (TQQQ) and ProShares Ultra QQQ (QLD). In August, **SpaceX** has also become a top pick.

### 4. Why are Korean regulators restricting leveraged trading?

Korea’s Financial Services Commission raised the minimum保证金 requirement for single-stock leveraged ETFs from 10 million won to 30 million won, effective July 31, 2026. New investors must also complete a three-hour risk management course and five days of simulated trading. The moves were designed to curb excessive speculation and reduce market volatility.

### 5. Can Korean investors’ buying move U.S. stock prices?

Analysts say Korean flows are **unlikely to sway broad U.S. markets** but could amplify volatility in individual stocks and thinner trades. The U.S. stock market is dominated by institutional investors, so even $4.5 billion in monthly inflows is relatively small.

### 6. Is the SK Hynix ADR premium a sign of a bubble?

Acadian Asset Management’s Owen Lamont says yes. He called the 10% premium “absolutely crazy” and warned that such price discrepancies are “a symptom of the bubble”. He pointed to similar dislocations involving Taiwanese and Indian companies during the dot-com boom.

### 7. What is the “same strategy, different market” thesis?

Analysts say Korean investors are **changing markets, not changing bets**. They’re still betting on AI and using leverage—they’re just doing it through U.S. markets instead of Korean ones.

### 8. What’s next for Korean investors and U.S. markets?

If Korean regulators ease restrictions or the KOSPI recovers, the flow could reverse. For now, the trend shows no signs of slowing—Korean investors bought nearly $200 million of SpaceX in just the first two weeks of August.

---

## Conclusion: A Cross-Border Bet on AI

South Korean investors are doing something that, on the surface, makes no sense. They’re paying a 10% premium to buy the same stocks in the U.S. that they could buy cheaper at home. They’re piling into 3x leveraged ETFs at a time when regulators are trying to protect them from excessive risk. They’re chasing the AI trade across the Pacific Ocean, refusing to give up on the theme that made them money in the first place.

But beneath the surface, there’s a logic to the madness. Korean regulators have made it harder—and more expensive—to trade leveraged products at home. The KOSPI has been on a rollercoaster, and Korean investors are looking for a more stable, more liquid market to place their bets. They still believe in AI. They just want to express that belief in a different venue.

The question is whether this is a smart strategic shift or a sign of speculative excess. Owen Lamont thinks it’s the latter. “They’re a symptom of the bubble,” he said.

For American investors, the Korean buying spree offers a window into retail sentiment—and a warning. When retail investors are willing to pay a premium for the same stock and use 3x leverage to bet on semiconductors, it’s often a sign that the trade is getting crowded. Bubbles don’t always burst immediately. But they always burst eventually.

For now, the money keeps flowing. Seoul to Wall Street. Same strategy, different market. Absolutely crazy—and absolutely real.

---

## Disclaimer

*This article is for informational and educational purposes only and does not constitute financial, investment, tax, or legal advice. All views expressed are based on publicly available information as of August 18, 2026. Market conditions, regulatory policies, and investment trends are subject to change. The author does not endorse any specific investment strategies or products. Past performance is not indicative of future results. Before making any investment decisions based on the content of this article, please consult with qualified professionals who can evaluate your specific situation. The author is not affiliated with the Korea Securities Depository, the Financial Services Commission, or any of the companies mentioned in this article.*

‘Worrisome’: AI Is Driving a Looming Market Correction, European Central Bank Economists Warn

 

‘Worrisome’: AI Is Driving a Looming Market Correction, European Central Bank Economists Warn

## Introduction: The Warning from Frankfurt That Wall Street Can’t Ignore

For three years, the AI trade has been unstoppable. The Magnificent Seven have delivered returns that made even the most optimistic growth investors look prescient. The S&P 500 has notched record after record. And the narrative has been relentlessly bullish: AI is the most transformative technology since the internet, and the companies building it are the future.

But on Monday, August 17, 2026, the European Central Bank delivered a message that punctured the euphoria. In a blog post titled **“The AI boom: rational enthusiasm or the next dot-com bubble?”** , a team of ECB economists led by Malin Andersson, Johannes Breckenfelder, Stefano Corradin, Kalin Nikolov and Maria Antonietta Viola issued a stark warning: **a correction of current stock market valuations is likely**.

The post didn’t mince words. It described the current situation as “worrisome” and warned that even if AI ultimately succeeds and reshapes the global economy, stock prices could still fall—sharply. The reason? History. Time and again, transformative technologies have attracted massive investment, driven valuations to unsustainable levels, and then crashed. The railway boom of the 19th century. The electricity and radio craze of the 1920s. The dot-com bubble of the 1990s. AI, the ECB argues, is following the same script.

---

## The Valuation Reality: How Extreme Is It?

The ECB’s warning is grounded in hard numbers. The economists pointed to the **cyclically adjusted price-to-earnings (CAPE) ratio**, a measure developed by Nobel laureate Robert Shiller that compares stock prices to average inflation-adjusted earnings over the previous ten years. By this metric, **U.S. stock market valuations are currently close to their historical peak**.

The last time valuations were this high? The dot-com bubble. And we all know how that ended.

Euro area equity valuations have also risen, albeit to a lesser extent, reflecting the AI enthusiasm that has swept across global markets. But the epicenter of the frenzy is the United States, where the Magnificent Seven—Alphabet, Amazon, Apple, Tesla, Meta Platforms, Microsoft, and Nvidia—have come to dominate the market.

The ECB economists posed a pointed question: “Do today’s stock market prices reflect a rational bet on the transformative technology? Or are we seeing a remake of the dot-com bubble?” Their answer, backed by economic research, is that a correction is likely regardless of whether the current valuations are deemed rational.

---

## Why Technological Revolutions End in Boom and Bust

The ECB’s analysis draws on a rich body of economic research on past technological revolutions. The pattern is remarkably consistent:

1. **A genuinely transformative technology emerges**—railways in the 19th century, electricity and radio in the 1920s, the internet in the 1990s.

2. **Investment pours in**, and stock market valuations of firms that adopt the technology rise strongly.

3. **Prices eventually fall sharply**, even if the technology itself succeeds.

Why does this happen? The ECB economists offered two complementary explanations.

### The Rational View: Uncertainty Shifts from Sector to Economy

Under the “rational view,” high valuations can be justified by extreme uncertainty about a new technology’s productivity. At an early stage, pioneering stocks present major opportunities for early investors, who in the best case can make huge gains but in the worst case may lose everything. This “option value” increases stock valuations, causing price-to-earnings ratios to rise sharply.

But as the technology spreads through the wider economy, the uncertainty shifts from individual companies to the economy as a whole. If something goes wrong with the technology, the whole economy suffers. This economy-wide uncertainty “cannot be diversified,” prompting investors to demand higher returns to compensate for the risks. Although successful adoption of AI can boost profits, the increase in this risk premium pushes stock market valuations in the opposite direction—meaning share prices can eventually fall even if the technology itself succeeds.

### The Behavioral View: Overconfidence and Over-optimism

Under the “behavioral view,” overconfident and over-optimistic investors bid up prices beyond fundamentals. Then, when optimism fades, prices tend to fall even more sharply than in the rational scenario. This psychological dynamic has been a feature of every major technological boom-bust cycle in history.

The ECB’s conclusion is sobering: “Economic research on past technological revolutions points to a worrisome conclusion: a correction of current stock market valuations is likely”.

---

## The €440 Billion Exposure: Why Europe Can’t Look Away

One might ask: why does the ECB care about a U.S. stock market correction? The answer lies in the deep financial integration between the two regions.

Euro area households have approximately **€440 billion of exposure to U.S. technology equities**, largely through low-cost exchange-traded funds (ETFs) and other investment funds. Many investors may not even be aware of the associated concentration risk they are carrying.

Insurance companies and pension funds also hold significant exposures to the Magnificent Seven, largely through investment funds rather than directly. This structure itself acts as a transmission channel: a sharp correction can force funds to sell assets to meet redemptions—first liquid holdings and then, if the correction persists, distressed assets—pushing valuations down further and triggering more redemptions.

**“This is why a Mag 7 correction is a question of financial stability for the euro area, rather than just a private one,”** the ECB economists wrote.

---

## The Contagion Risk: When a U.S. Problem Becomes a Global One

The ECB’s warning extends beyond direct financial exposures. U.S. and euro area stock markets have historically been **very highly correlated**. A correction in the U.S. “would not remain a U.S. problem,” the economists said.

The effects of a U.S. correction could extend beyond financial markets to sentiment, financing conditions, and hiring in the euro area. As the ECB put it: **“A US AI fallout would not remain a US problem”** but could become **“a question of financial stability for the euro area”** .

Even if AI eventually delivers strong productivity gains, the transition could be painful. The ECB noted that a sharp correction can trigger a cascade of selling that amplifies the initial shock. This is the kind of dynamic that turns a market correction into a financial stability concern.

---

## Limited Policy Buffers: Why This Time Is Different

Perhaps the most alarming aspect of the ECB’s warning is the limited room policymakers have to respond. Unlike during the dot-com episode, **today’s starting point leaves markedly less room to cut interest rates or use fiscal policy to cushion the fallout**.

Why? Interest rates are already low by historical standards, and government debt levels are significantly higher than they were in 2000. The policy tools that were deployed to stabilize markets after the dot-com crash and the 2008 financial crisis are not as readily available today.

**“The more severe scenario is not the equity correction on its own but a correction that coincides with broader market instability that policymakers cannot easily calm,”** the ECB blog warned. This is the nightmare scenario: a sharp stock market correction that triggers a broader financial crisis, with limited ammunition to fight it.

---

## The AI Circularity Problem: A Hidden Risk

The ECB’s warning comes at a time when concerns about the sustainability of the AI boom are mounting from other quarters as well. One of the most pressing issues is the so-called **“circular financing”** in AI.

Companies like Nvidia are lending money to their customers—or investing in them—so that those customers can buy Nvidia’s products. This creates a virtuous circle in the boom phase but a vicious one in a downturn. If usage of AI fails to match the sky-high expectations that underpin the massive levels of spending, the whole edifice could come crashing down.

Skeptics argue that this circularity is unsustainable. The ECB’s warning adds institutional weight to those concerns, suggesting that even the most bullish AI proponents should be prepared for a correction.

---

## What This Means for American Investors

For American investors, the ECB’s warning is a reminder that the AI trade—like every other technology-driven boom in history—carries inherent risks. Here are the key takeaways:

### 1. Valuations Matter

The CAPE ratio is close to its historical peak. While valuations alone don’t predict the timing of a correction, they do suggest that future returns are likely to be lower than past returns.

### 2. Diversification Is Not a Guarantee

Even if you’re not directly invested in the Magnificent Seven, the high correlation between U.S. and global markets means that a correction would likely affect a broad range of assets.

### 3. The Timing Is Unknowable

The ECB acknowledged that the exact timing of a correction “is unknowable in advance” and that “boom-bust patterns are only identifiable with hindsight”. This is not a call to sell everything tomorrow—but it is a call to be prepared.

### 4. Policy Buffers Are Limited

Unlike in 2000 or 2008, policymakers have less room to cut rates or deploy fiscal stimulus. This means that the next downturn could be harder to contain.

### 5. The AI Story Is Still Compelling—But That Doesn’t Mean Stocks Can’t Fall

The ECB’s most counterintuitive finding is that **even if AI succeeds, stocks may still fall**. The shift in uncertainty from individual companies to the broader economy—and the corresponding increase in the risk premium—can outweigh any cash flow gains from AI adoption.

---

## The Historical Context: Learning from the Past

The ECB’s warning is grounded in a long history of technological revolutions that ended in busts. The railway boom of the 19th century saw massive investment in rail infrastructure, followed by a sharp correction. The electricity and radio craze of the 1920s was followed by the Great Depression. The dot-com bubble of the 1990s was followed by a 78% crash in the Nasdaq.

**“In each case, a genuinely transformative technology attracted investment, and the stock market valuations of firms that adopted it rose strongly before falling sharply,”** the ECB economists wrote.

AI is different in many ways. The technology is more advanced, the companies are more profitable, and the adoption is more widespread. But the underlying dynamics of boom and bust—driven by uncertainty, over-optimism, and the eventual recognition of limits—are remarkably consistent.

---

## Frequently Asked Questions (FAQs)

### 1. What exactly did the European Central Bank warn about?

The ECB warned that a correction in current stock market valuations is likely, driven by the AI boom. Even if AI ultimately succeeds, stock prices could still fall due to the dynamics of technological revolutions, which historically have ended in boom-bust cycles.

### 2. Why is the ECB worried about U.S. tech stocks?

Euro area households have approximately **€440 billion of exposure to U.S. technology equities**, largely through ETFs and investment funds. Insurance companies and pension funds also hold significant exposures. A U.S. correction would not remain a U.S. problem but could become a question of financial stability for the euro area.

### 3. What is the CAPE ratio and why does it matter?

The CAPE (cyclically adjusted price-to-earnings) ratio compares stock prices to average inflation-adjusted earnings over the previous ten years. It was developed by Nobel laureate Robert Shiller. Currently, the U.S. CAPE ratio is close to its historical peak, last seen during the dot-com bubble.

### 4. Why do technological revolutions often end in busts?

The ECB offered two explanations. Under the “rational view,” uncertainty shifts from individual companies to the broader economy as the technology spreads, increasing the risk premium and pushing valuations down. Under the “behavioral view,” overconfident investors bid up prices beyond fundamentals, and prices fall sharply when optimism fades.

### 5. Could a correction happen even if AI succeeds?

Yes. The ECB noted that even if AI is successfully adopted and boosts profits, stock prices could still fall because the increase in the risk premium can outweigh the cash flow gains from adoption.

### 6. What are the policy implications?

Unlike during the dot-com episode, today’s starting point leaves markedly less room to cut interest rates or use fiscal policy to cushion the fallout. This means that a correction could be harder to contain.

### 7. When will the correction happen?

The ECB acknowledged that the exact timing “is unknowable in advance” and that “boom-bust patterns are only identifiable with hindsight”.

### 8. Is this the same as the dot-com bubble?

There are similarities—both involve transformative technology, massive investment, and extreme valuations. But there are also differences: the companies driving the AI boom are generally more profitable and established than many dot-com-era firms. However, the underlying boom-bust dynamics are similar.

---

## Conclusion: A Warning, Not a Prediction

The European Central Bank’s warning about an AI-driven market correction is not a prediction of when the crash will come. It is a sober assessment of the historical dynamics that have accompanied every major technological revolution—and a reminder that the current AI boom is following a familiar pattern.

The ECB economists put it bluntly: **“Economic research on past technological revolutions points to a worrisome conclusion: a correction of current stock market valuations is likely”** .

For American investors, the message is clear: the AI trade has been spectacular, but it carries risks that are often overlooked in the euphoria. Valuations are at historical extremes. The policy buffers that cushioned past downturns are thinner. And the dynamics of boom and bust are as old as capitalism itself.

The AI revolution is real. It will transform the global economy. But as the ECB’s warning makes clear, the path to that future is unlikely to be a straight line. Corrections are part of the process. And the next one, when it comes, could be severe.

---

## Disclaimer

*This article is for informational and educational purposes only and does not constitute financial, investment, tax, or legal advice. The views expressed are based on publicly available information, including the ECB blog post and other cited sources as of August 18, 2026. Market conditions, economic data, and regulatory policies are subject to change. The author does not endorse any specific investment strategies or products. Past performance is not indicative of future results. Before making any financial or investment decisions based on the content of this article, please consult with qualified professionals who can evaluate your specific situation. The author is not affiliated with the European Central Bank or any other entity mentioned in this article.*

In Trump’s Economy, Companies Offer ‘Buy Now, Pay Later’ for Utilities as Costs Surge

 In Trump’s Economy, Companies Offer ‘Buy Now, Pay Later’ for Utilities as Costs Surge

## Introduction: The $280 Question

The national average monthly utility bill hit **$280** in early 2026. That's a **12% increase** since the end of 2024. For millions of American families, that's not just an inconvenience—it's a choice between keeping the lights on and putting food on the table.

Enter the new solution from Silicon Valley: buy now, pay later for your electricity bill. For your water. For your rent. For your groceries. For your health insurance. For the very basics of survival in Donald Trump's America.

Lending apps including **Flex, Zip, and Affirm** are now offering credit to consumers struggling to afford basic necessities. What was once a way to finance a new couch or a pair of sneakers has become a lifeline for families who can no longer afford the essentials. Flex has financed nearly **$40 billion** in rent payments for 3 million tenants. Its customer base has a median credit score **below 600**—subprime territory.

This is the hidden story of the Trump economy: **record stock markets and record household debt, side by side**.

---

## The BNPL Explosion: From Sneakers to Survival

### The Numbers That Tell the Story

Buy now, pay later started as a checkout option for discretionary purchases—a way to split a $200 jacket into four payments over six weeks. It was convenient, often interest-free, and relatively harmless.

That was then.

U.S. consumers spent approximately **$160 billion** through pay-later loans last year, **nearly double** the level two years earlier. BNPL providers originated nearly **$157 billion** in consumer credit products in 2025, up from nearly $116 billion in 2024.

Half of BNPL users in a recent LendingTree survey said **they could not make ends meet** without this form of credit. **44% of Americans** expect to apply for a BNPL loan in the next six months, including 13% who expect to take out three or more.

The shift is fundamental. This isn't about financing discretionary purchases anymore. It's about **survival**.

### What People Are Actually Buying

The latest data reveals just how desperate things have become:

| Category | BNPL Usage |

|----------|------------|

| **Groceries** | 29% (up from 14% in 2024) |

| **Medical/Dental Care** | 42% |

| **Utility Bills** | 39% |

| **Rent** | 13% |

| **Vehicle Repairs** | 18% |

"A substantial share of BNPL users report going into debt to pay for everyday essentials like groceries, utilities, and rent," according to a Protect Borrowers investigation. **46%** have used BNPL for groceries. **39%** for utility bills.

As Senator Elizabeth Warren put it: "Everyday costs are through the roof, forcing more people to take out 'Buy Now, Pay Later' loans just to make it through the month".

---

## The Economic Engine: Why Americans Can't Keep Up

### The Trump Economy in Two Sentences

During a recent press appearance, President Trump hailed "the best market in history" as the S&P 500 hovered near its all-time high. He told Americans to just "hold on a little while" longer for lower prices.

Meanwhile, the personal savings rate is at its lowest since 2022. Consumer sentiment fell in August after two consecutive months of improvement. And **millions of Americans cannot afford the basic necessities of life**—food, housing, and utilities.

The disconnect is staggering. Wall Street is booming. Main Street is drowning.

### The Causes: War, Tariffs, and Policy

The administration's policies have directly contributed to the affordability crisis:

**The Iran War:** Trump's "illegal war against Iran" has driven up household costs, notably at the gas pump but also for all sorts of consumer goods because transportation is now more expensive. Gas, groceries, electricity, and insurance are all up.

**Tariffs:** Trump's sweeping tariffs have raised prices on imported goods, with estimates indicating American households are paying for **95%** of the tariffs. The average family has lost about $1,700 in purchasing power.

**Consumer Protection Rollbacks:** The Trump administration withdrew a Consumer Financial Protection Bureau rule that would have held BNPL providers to credit regulations under the Truth in Lending Act. The CFPB has repeatedly pulled back from enforcing key rules.

Emily DiVito, senior adviser for economic policy at Groundwork Collaborative, put it bluntly: "The steady rise in the use of buy now, pay later loans to cover the basics, like groceries and rent, is a warning sign that working families are under real financial strain. Prices for essentials are climbing as a result of the president's reckless economic policies, wages aren't keeping pace, and hardworking Americans are being pushed into cycles of debt just to make ends meet".

---

## The Lenders: Who's Offering Credit for Survival

### Flex: The $40 Billion Rent Machine

Flex has emerged as the dominant player in the essential-expense BNPL space. The company says it has financed nearly **$40 billion** in rent payments for **3 million tenants**. Its customer base has a median credit score **below 600**, a range generally associated with subprime credit.

The cost structure is revealing: Flex charges a **$6 monthly membership fee**, a fee equal to **3% of the borrowed amount**, and a processing fee. It has expanded this year to utilities and auto-loan payments.

### Affirm, Zip, and Others

**Affirm** has begun a pilot that lets some tenants extend a monthly rent payment for several weeks. The company allows only one rent loan at a time and prevents a new advance until the prior month's obligation has been repaid.

**Zip** allows customers to take out loans to pay for their broadband, electricity, health insurance, mobile phone service, mortgage, and water bills.

These apps have essentially become **de facto utility lenders**, stepping into a role that traditional banks have been unwilling to fill.

### The Business Model: Profit from Desperation

BNPL providers make money through fees, interest, and merchant commissions. Protect Borrowers found that late fees of **$7 to $8 per payment** are common in some programs, while interest and financing charges can reach **36%**. Interest-bearing installment products accounted for more than **37%** of annual BNPL issuance in 2026, almost twice their 2021 share.

The providers' value proposition is straightforward: help customers match the timing of expenses to the timing of income. That's particularly relevant for consumers with irregular pay cycles, including gig workers, tipped employees, and households facing unexpected expenses.

But the consumer-protection issue is whether BNPL is functioning as a temporary payment-timing tool or becoming a **recurring substitute for insufficient income and exhausted traditional credit**.

---

## The Debt Trap: How BNPL for Essentials Creates a Vicious Cycle

### The Monthly Recurrence Problem

A loan used for a one-time purchase can be paid down and retired. A loan used to cover rent or electricity solves only the current month's payment problem. The same bill is due again the next month, potentially before the borrower has cleared the prior installment balance.

This creates the central compliance and consumer-protection issue: **BNPL is becoming a recurring substitute for insufficient income**.

### Exhausted Credit Cards

"Often they just, they've exhausted their credit cards, and buy now, pay later is their only option," said Jim Triggs, CEO of Money Management International, a non-profit credit counseling firm.

Credit card debt in the U.S. hit **$1.25 trillion** in the first quarter, up 5.9% from a year earlier. When the cards are maxed out, BNPL becomes the last resort.

### The "Shadow Default" Risk

Formal default rates are increasingly poor diagnostics for the health of this market. More revealing signals include "shadow defaults"—instances where borrowers are struggling but haven't officially defaulted.

As one observer put it, "BNPL loans are the latest peonage scheme to infest society". Another warned: "More debt peonage. We all 'owe our souls to the company store'".

---

## The Regulatory Vacuum: Who's Watching the Store?

### The Biden Rule That Never Was

Under the Biden administration, BNPL products were beginning to receive more regulatory oversight. The administration proposed to hold these products to credit regulations by implementing the Truth in Lending Act.

### The Trump Rollback

The Trump administration put the kibosh on those efforts when it withdrew the Consumer Financial Protection Bureau rule, claiming it did so in the interest of "focusing resources on supporting hard-working American taxpayers, servicemen, veterans, and small businesses".

Senator Richard Blumenthal has released a report on the administration's gutting of consumer protections, noting that the administration has been "removing protections for Buy Now, Pay Later consumers".

### The Legislative Response

The "Buy Now Pay Later Protection Act of 2025" has been introduced but faces an uncertain future. Senator Warren and other Democrats have been vocal in their criticism of the administration's approach.

But for now, the regulatory vacuum persists. BNPL providers are operating with far less accountability than traditional banks, under the illusion that their products are less costly than credit cards.

---

## The Human Cost: Real People, Real Struggles

### Ashley's Story

Ashley Reed, a 40-year-old paraeducator who also works part-time as a radiology assistant, became a frequent BNPL user after maxing out her credit cards to pay for caregiving expenses after her mother suffered a ruptured brain aneurysm.

"I wound up maxing out my credit cards to get hotel rooms and things like that to stay with her and to get her transferred back to Baltimore. It was about $2,500 for an ambulance ride," she said.

Having reached her credit limit, she turned to BNPL services to help make ends meet, including for groceries. She's one of a growing number of consumers doing so amid inflation and other price pressures, such as higher gas prices stemming from the U.S. war with Iran.

### The 85% Who Say Brands Use Inflation

A USA Today survey found that **85% of respondents believe brands often use inflation as an excuse to raise prices**. **30%** said they used a credit card to pay for essentials—groceries, gas, utilities, or medical bills—in the past three months.

"There's no joy in living this way," one respondent told USA Today. Families have maxed out their credit cards and are buying only what they can afford each week with cash.

---

## The Political Angle: A Fault Line in 2026

### Trump's Broken Promises

Trump promised to lower costs "on Day One". Instead, everyday costs are through the roof.

The Urban Institute found that American families are increasingly relying on savings and credit—including BNPL programs—to meet their grocery needs. The economy that Trump calls "unbelievably good from the standpoint of Wall Street" is a nightmare for Main Street.

### The Democratic Response

Senator Warren has been the most vocal critic: "Donald Trump promised to lower costs 'on Day One.' But everyday costs are through the roof, forcing more people to take out 'Buy Now, Pay Later' loans just to make it through the month".

She has also warned that these loans—"originally marketed as no interest—are ramping up fees on Americans".

### The Stakes for November

The BNPL crisis is a political liability for the administration. As one analysis put it, "Everywhere you look across Donald Trump's economy, debt is masking a... Americans are in debt to pay for basics like rent and groceries".

The question for voters in November: is a record stock market worth it when you can't afford your electricity bill?

---

## Frequently Asked Questions (FAQs)

### 1. What exactly is "buy now, pay later" for utilities?

BNPL for utilities allows consumers to take out short-term loans to pay for electricity, water, broadband, health insurance, rent, and other essential bills through apps like Flex, Zip, and Affirm. Instead of paying the full bill immediately, users can split it into installments or defer payment.

### 2. How much are Americans spending on BNPL?

U.S. consumers spent approximately **$160 billion** through pay-later loans last year, nearly double the level two years earlier. BNPL providers originated nearly **$157 billion** in consumer credit products in 2025.

### 3. What are people using BNPL for?

Recent surveys show BNPL is being used for essentials: **29%** for groceries (up from 14% in 2024), **42%** for medical or dental care, **39%** for utility bills, and **13%** for rent.

### 4. Why are people using BNPL for essentials?

The average monthly utility bill hit **$280** in early 2026, a 12% increase since the end of 2024. Credit card debt hit **$1.25 trillion**. Many families have exhausted their credit cards and see BNPL as their only option.

### 5. What are the risks of using BNPL for essentials?

BNPL for essentials creates a debt cycle because the same bill is due again next month. Late fees of **$7 to $8 per payment** are common, and interest can reach **36%**. Half of BNPL users said they could not make ends meet without this form of credit.

### 6. Has the government done anything to regulate BNPL?

The Biden administration proposed regulating BNPL under the Truth in Lending Act, but the Trump administration withdrew that rule. The Consumer Financial Protection Bureau has repeatedly pulled back from enforcement.

### 7. Is this just a problem for low-income Americans?

No. While low-income families are disproportionately affected, rising costs are squeezing middle-class families too. **44% of Americans** expect to apply for a BNPL loan in the next six months.

### 8. What does this say about the Trump economy?

The Trump economy is a tale of two worlds: Wall Street is booming with record stock markets, while Main Street is struggling with record household debt. The BNPL crisis reveals the hidden weakness beneath the surface of the economic headlines.

---

## Conclusion: The Lights Stay On, But at What Cost?

The rise of BNPL for utilities is a symptom of something deeper: an economy where the basics of survival—electricity, water, groceries, rent—are becoming unaffordable for millions of Americans.

The numbers are stark: $280 average monthly utility bills, $160 billion in BNPL spending, $1.25 trillion in credit card debt. The stories are heartbreaking: families maxing out credit cards, turning to payday-loan alternatives just to keep the lights on.

President Trump touts record stock markets and calls the economy "unbelievably good." But for the millions of Americans who are one missed paycheck away from losing their electricity, the economy doesn't feel good at all.

The BNPL industry has stepped into the breach, offering credit where banks won't lend. But the fees, the interest, and the debt cycles are creating a new form of peonage—a system where the poor pay more for the privilege of survival.

As Senator Warren put it: "Everyday costs are through the roof, forcing more people to take out 'Buy Now, Pay Later' loans just to make it through the month".

The lights stay on. But the cost—financial, psychological, and societal—keeps rising.

And that's a bill that's coming due.

---

## Disclaimer

*This article is for informational and educational purposes only and does not constitute financial, investment, tax, or legal advice. The views expressed are based on publicly available information as of August 18, 2026. Economic conditions, regulatory policies, and personal financial situations are subject to change. The author does not endorse any specific financial products or strategies. Before making any financial decisions based on the content of this article, please consult with qualified professionals who can evaluate your specific situation. The author is not affiliated with any of the BNPL providers or advocacy organizations mentioned in this article.*

Anthropic's Revenue Run Rate Reportedly Surpasses $65 Billion Pre-IPO — Here's What It Means for the AI Race

 

Anthropic's Revenue Run Rate Reportedly Surpasses $65 Billion Pre-IPO — Here's What It Means for the AI Race

## Introduction: The Number That Redefined the AI Landscape

Just a few months ago, Anthropic was still chasing OpenAI. The narrative was familiar: OpenAI had the first-mover advantage, the brand recognition, and the massive user base. Anthropic was the scrappy underdog, the one founded by former OpenAI defectors who were worried about safety.

Not anymore.

On Monday, Bloomberg reported that Anthropic's annualized revenue run rate has topped **$65 billion** as of the end of July. That's a **sevenfold increase** from the end of 2025, when its run rate was just $9 billion. The company crossed $47 billion in May, meaning it added $18 billion in annualized revenue in just two months.

For context, OpenAI's latest revenue run rate is estimated at **$40 billion**. Anthropic has not just caught up—it has **blown past** its chief rival.

The numbers are staggering. The implications are even bigger. Here's everything you need to know about Anthropic's explosive growth, its upcoming IPO, and what it means for the future of AI.

---

## What Is a "Revenue Run Rate" and Why Does It Matter?

Before we dive into the numbers, let's clarify what we're actually talking about.

A **revenue run rate** is a metric that projects annual performance by extrapolating current sales levels. If a company generates $5 billion in a single month, its run rate would be $60 billion annually. It's not a guarantee of future revenue—it's an estimate based on current momentum.

For Anthropic, the $65 billion run rate means that if its current sales pace continues, the company would generate $65 billion in revenue over the next 12 months. That's more than many of the world's largest tech companies generate in a year.

The company hit this milestone by the end of July, according to people familiar with the matter. Anthropic shared the figures as part of a regular update with investors.

### The Journey from $9 Billion to $65 Billion

The speed of Anthropic's growth is almost unprecedented in tech history:

- **End of 2025**: Run rate of roughly $9 billion

- **April 2026**: Crossed $30 billion

- **May 2026**: Crossed $47 billion

- **End of July 2026**: Hit $65 billion

That's a trajectory that would make even the most aggressive growth investors do a double-take.

---

## The Q2 Numbers: A 14-Fold Revenue Surge

The run rate figure is impressive, but the quarterly numbers are even more eye-opening.

Anthropic reported preliminary second-quarter revenue of **more than $11.5 billion**, according to documents seen by Bloomberg. That compares to:

- **$787 million** in the same period last year

- **$4.73 billion** in the first quarter of 2026

That's a **14-fold increase** year-over-year and a **sequential gain exceeding 140%**.

Perhaps most significantly, Anthropic also reported **positive adjusted operating income** for the quarter. That means the company is not just growing revenue—it's becoming profitable at the operating level, at least on an adjusted basis.

---

## The Enterprise Engine: Why Anthropic Is Winning

So what's driving this explosive growth? The answer lies in Anthropic's focus on **enterprise customers**.

Unlike OpenAI, which has a significant consumer revenue stream, Anthropic generates approximately **75% to 85% of its revenue from enterprise API calls**. Its flagship product, **Claude Code**, a coding assistant that helps developers write and debug software, has been the primary growth driver.

The enterprise focus has paid off. As of June 2026, **34.4% of U.S. companies** were paying for Anthropic's services, surpassing OpenAI's 32.3%. That's a remarkable achievement for a company that was considered a distant second just a year ago.

### Why Enterprises Are Choosing Anthropic

Several factors are driving enterprise adoption:

**1. Superior Performance.** According to Pitchbook analyst Harrison Rolfes, Anthropic's top models can have a lower cost per successfully completed task than competing models because they provide better answers "the majority of the time"—so you're not paying to re-run queries or using paid humans to double-check work.

**2. Enterprise-First Design.** Anthropic built its products for businesses from the ground up. The company's APIs are designed for reliability, security, and scalability at enterprise scale.

**3. Trust and Safety.** Anthropic's founding mission was built around AI safety, which resonates with enterprise customers who are concerned about compliance, security, and reputational risk.

---

## The Competitive Landscape: Anthropic vs. OpenAI

The rivalry between Anthropic and OpenAI has defined the AI industry for years. But the balance of power has shifted dramatically.

### Revenue Comparison

- **Anthropic Q2 2026 revenue**: $11.5 billion

- **OpenAI Q2 2026 revenue**: Not disclosed, but run rate estimated at $40 billion

Anthropic's Q2 revenue alone ($11.5 billion) was more than OpenAI's entire estimated run rate just a year ago.

### Valuation Comparison

- **Anthropic**: $965 billion after Series H funding in May

- **OpenAI**: Approximately $869 billion in private market valuation

Anthropic has also surpassed OpenAI in valuation, a reversal from just months ago.

### The IPO Race

Both companies have filed confidentially for IPOs. But Anthropic is expected to make its Wall Street debut **before** OpenAI, potentially as early as this fall.

This matters for several reasons:

- **First-mover advantage**: The first AI company to go public could set the valuation benchmark

- **Capital access**: The IPO will unlock billions in additional capital for AI infrastructure

- **Brand perception**: Going first signals confidence and momentum

---

## The IPO: What We Know

### Timing

Anthropic is expected to go public as early as **September or October 2026**. The company has already filed confidentially with the SEC.

Polymarket, the prediction market platform, estimates a **69.5% probability** of a listing before October 31 and an **87% probability** by the end of the year.

### The Bankers

Anthropic is working with **Morgan Stanley, Goldman Sachs, and JPMorgan Chase** on the offering.

### The Valuation Target

Anthropic is reportedly targeting a valuation of **$2 trillion or more** for its IPO. Some investors believe it could reach **$3 trillion**.

To put that in perspective:

- SpaceX went public at $1.77 trillion in June

- Anthropic's Series H valuation was $965 billion

- A $2 trillion IPO would be the **largest in history**, surpassing SpaceX's record

### The 2028 Revenue Projection

The $2 trillion valuation is being justified by the company's internal projections. Anthropic has reportedly projected **2028 revenue of roughly $190 billion to $200 billion**.

That's more than **four times** the $47 billion run rate it disclosed in May. Bankers and investors are using enterprise value-to-revenue multiples based on these forecasts.

Looking two years ahead is unusual for IPO valuations, but it reflects the speed at which Anthropic's business is expanding.

---

## The Risks: Why Caution Is Warranted

Despite the explosive growth and blockbuster valuation targets, there are significant risks to consider.

### 1. The Cost of Scale

AI is expensive. Anthropic spends billions on compute infrastructure, and those costs will only grow as the company scales. The company reportedly pays **$1.25 billion monthly** to SpaceX for computing power. That's $15 billion a year—more than some companies' entire revenue.

### 2. Profitability Questions

While Anthropic reported positive adjusted operating income in Q2, "adjusted" is doing a lot of work there. The company hasn't disclosed its net income or cash burn. AI companies are notoriously capital-intensive, and profitability at scale remains unproven.

### 3. Regulatory and Political Risks

Anthropic has had a rocky relationship with the Trump administration. In June, the company was **designated a supply-chain risk by the Pentagon**. It also had to temporarily disable access to two of its most advanced models to comply with an export control directive.

While the models were eventually restored, the episode highlighted Anthropic's vulnerability to government action.

### 4. Competition Is Intensifying

OpenAI isn't standing still. DeepSeek, the Chinese AI firm, is gaining market share and preparing for its own IPO. Google, Microsoft, and Amazon are all investing heavily in AI. The competitive landscape is only getting more crowded.

### 5. The Valuation Bubble Risk

A $2 trillion valuation for a company with $65 billion in run-rate revenue implies a revenue multiple of roughly 30x. For comparison, Nvidia trades at about 20x revenue. Some analysts have questioned whether the valuation is justifiable.

"There have been similar precedents among some of the fastest-growing companies to recently hit the public markets," Reuters noted, citing Cerebras Systems and SpaceX. But precedents don't guarantee success.

### 6. The AI Winter Risk

The AI industry has seen boom-and-bust cycles before. If enterprise spending on AI slows—whether due to economic conditions, regulatory changes, or technological disappointment—Anthropic's growth could stall.

---

## What This Means for American Investors

### The Opportunity

For investors, Anthropic's IPO represents one of the most significant opportunities in years. The company is growing faster than any major tech company in history. Its enterprise focus provides a strong foundation for sustained growth. And its position at the center of the AI revolution gives it exposure to one of the most powerful secular trends in the global economy.

### The Challenge

But investing in Anthropic won't be easy. The valuation is likely to be astronomical. The volatility could be extreme—just look at SpaceX's post-IPO performance. And the competitive, regulatory, and technological risks are significant.

### The Takeaway

Anthropic's $65 billion run rate is a testament to the transformative power of AI and the company's successful execution. But it's also a reminder that in the world of high-growth tech, the gap between promise and reality can be vast.

As Gavin Baker, managing partner at Atreides Management, put it: "Although public markets generally struggle to accurately value growth when it is this high".

---

## Frequently Asked Questions (FAQs)

### 1. What is Anthropic's current revenue run rate?

Anthropic's annualized revenue run rate topped **$65 billion** at the end of July 2026. That's up from $9 billion at the end of 2025 and $47 billion in May.

### 2. How does Anthropic's revenue compare to OpenAI's?

Anthropic's run rate of $65 billion surpasses OpenAI's estimated run rate of **$40 billion**. Anthropic has pulled ahead of its chief rival in revenue, a significant reversal from just months ago.

### 3. What was Anthropic's Q2 2026 revenue?

Anthropic reported preliminary Q2 revenue of **more than $11.5 billion**, up from $787 million in the same period last year and $4.73 billion in Q1.

### 4. When is Anthropic's IPO expected?

Anthropic is expected to go public as early as **September or October 2026**. The company has confidentially filed with the SEC and is working with Morgan Stanley, Goldman Sachs, and JPMorgan.

### 5. What valuation is Anthropic targeting?

Anthropic is reportedly targeting a valuation of **$2 trillion or more**. Some investors believe it could reach $3 trillion.

### 6. What is Anthropic's 2028 revenue projection?

Anthropic has internally projected **2028 revenue of roughly $190 billion to $200 billion**.

### 7. Why is Anthropic growing so fast?

Anthropic's growth is driven by strong enterprise adoption, particularly of its coding assistant **Claude Code**. The company generates approximately 75% to 85% of its revenue from enterprise API calls.

### 8. What are the risks to Anthropic's growth?

Key risks include: high compute costs (the company pays $1.25 billion monthly to SpaceX for computing power); regulatory and political uncertainty (the Pentagon designated Anthropic a supply-chain risk in June); intensifying competition from OpenAI, DeepSeek, and others; and the possibility of an AI industry slowdown.

---

## Conclusion: The Underdog No More

Anthropic's $65 billion revenue run rate is a milestone that few would have predicted just a year ago. The company that was once dismissed as an also-ran in the AI race has not just caught up to OpenAI—it has surpassed it.

The numbers tell the story: 14-fold year-over-year revenue growth in Q2. A run rate that has multiplied sevenfold since the end of 2025. A valuation that could top $2 trillion, making it the largest IPO in history.

But the story isn't just about numbers. It's about a company that bet on enterprise customers, built superior products, and executed with remarkable precision. It's about a shift in the balance of power in the AI industry—a shift that could have profound implications for investors, competitors, and the future of technology itself.

The IPO will be one of the most closely watched events in market history. The valuation will be debated endlessly. The stock will be volatile. And the risks will be significant.

But one thing is clear: Anthropic is no longer the underdog. It's the company to beat.

---

## Disclaimer

*This article is for informational and educational purposes only and does not constitute financial, investment, tax, or legal advice. All views expressed are based on publicly available information as of August 18, 2026. The revenue figures, valuation targets, and IPO timing discussed are based on reports and may be subject to change. The author does not endorse any specific investment strategies or products. Past performance is not indicative of future results. Before making any investment decisions based on the content of this article, please consult with qualified professionals who can evaluate your specific situation. The author is not affiliated with Anthropic, OpenAI, or any of the financial institutions mentioned in this article.*