对冲基金经理Russell Clark：美债是比AI更大的投机泡沫，AI巨头烧钱是为「防住马斯克」

Original Author: Zhao Ying

Source: Wall Street News

A hedge fund manager bluntly stated: The biggest speculative bubble in the market isn't AI, it's US Treasuries. He predicts the 10-year Treasury yield will rise to 10% and believes tech giants' massive spending on AI is essentially a defensive move to "fend off Musk."

Recently, hedge fund manager Russell Clark appeared on the podcast "Other People's Money" for an interview with host Max Wiethe, sharing a series of disruptive views on several hot market topics, including the US Treasury market, the rationale behind AI capital expenditure, trends in the semiconductor industry, and risks in private credit.

Clark, currently based in London, manages a hedge fund and continues to write market commentary on Substack. His series of judgments have attracted attention in the market – the core logic being: we are transitioning from an era of "capital surplus and low interest rates" into a new political and economic cycle characterized by "high wages, high inflation, and high interest rates."

US Treasuries are the Biggest Bubble: 10-Year Yield Target at 10%

Amid the heated debate over the AI bubble, Clark turned his sights on a much larger market.

"I still have a 10% yield as my target for Treasuries this year," he stated this striking figure directly.

His logical chain is clear: If the US political goal is to make housing affordable again for people under 40, wages need to grow by about 7% annually, doubling within 10 years. Meanwhile, nominal housing prices should remain stable, leading to a continuous decline in real prices. To achieve this, real interest rates need to be around 3% – with nominal rates plus inflation, this means rates need to rise to approximately 10%.

"If real interest rates stay around 3%, people will put their money in the bank instead of investing in real assets."

Clark places this judgment within a longer historical perspective. Citing the leading indicator significance of Japanese government bonds (JGBs) – "I've always believed JGBs are an excellent leading indicator for US Treasuries" – he notes that JGBs were once called the "widowmaker trade," with the market discussing their debt unsustainability for nearly 30 years "until it actually collapsed."

He also points to the freezing of Russia's foreign exchange reserves in 2022 as a significant signal: "If a country holds foreign exchange reserves, why keep them there?" He believes global reserve assets will naturally migrate from Treasuries to gold – a trend quietly occurring, albeit slowly.

More broadly, Clark believes the current political transformation is the fundamental driver. Since the Reagan Revolution in 1980, capital accumulation has suppressed wages and interest rates. Now, the political pendulum is swinging back, with demands for "full employment and high wages" re-dominating the policy agenda, implying continued inflation and upward pressure on interest rates.

"The AI Giants Aren't Burning Cash for AI – It's to Keep Musk at Bay"

On the issue of AI capital expenditure, Clark offers an interpretation starkly different from the mainstream narrative.

"The real problem is that Elon Musk, through SpaceX, is essentially signaling: I want in on AI too. I manufacture computing equipment, and I have a way to make cheaper computing devices."

He believes this is the real motive behind the massive spending spree by tech giants like Google, Microsoft, and Amazon – not betting on AI's future, but defensively protecting their existing business moats.

"Companies like Google, Microsoft, and even Amazon have very profitable businesses. They are all trying to stay ahead and prevent Elon Musk from getting a foothold. That's my understanding of this situation."

Clark draws an analogy with the rise of Tesla: traditional automakers struggled to launch competitive electric vehicles because they were trying to protect their legacy internal combustion engine businesses. The result was "Tesla's market cap is several times that of traditional internal combustion engine manufacturers." He argues the tech industry is undergoing the same logic – "If we don't invest, we're finished."

For this reason, he is highly skeptical of views predicting a sharp cutback in AI capex: "I am very skeptical that tomorrow we'll see Microsoft, Meta, Google, and Amazon announce a 50% cut in AI capital expenditure. I think the ones who cut spending first are often the companies that are about to lose money."

As for whether AI will disrupt the labor market, potentially undermining his macro narrative of a high-wage era, Clark is dismissive. He believes AI's impact is primarily concentrated among professional white-collar workers – lawyers, accountants, fund managers, senior doctors – "whose salaries have always been tied to asset prices" – rather than impacting the bottom end of the labor market. He cites the post-WWII era as an example: major technological breakthroughs like nuclear energy and jet engines followed one after another, yet wages rose by 1000%. "Technological change and the wage issue are actually two different topics."

Semiconductors are the New Oil; Supply Constraints Will Support Prices

Clark proposes an imaginative analogy: semiconductors today are like oil in the 1970s.

"If you look at the 70s, holding both oil and gold was actually a pretty good strategy. Oil was key to economic growth everywhere, and its supply was constrained... Modern economic growth is actually driven by semiconductors or computer technology, so the price of semiconductors remains as high as oil was in the 70s."

He notes that Nvidia's chip prices have remained high for the past five or six years. Traditionally, semiconductor prices fall with capacity expansion, but this hasn't happened this time. Additionally, supply-side hard constraints exist, highly analogous to the logic of oil in the 70s.

Private Credit: A Severely Underestimated Time Bomb

Clark specifically calls out the private credit and private equity sectors – where he sees the market's most overlooked risk.

"Why would I want to hold this illiquid private credit fund? I have no idea about the value of these assets, and their condition is quite poor."

He notes that "redemption gates" have appeared in the market – redemptions have exceeded new subscriptions for the first time, forcing funds to implement redemption restrictions. He states bluntly that once money market fund yields reach 7% or 8%, rational investors will start questioning the need to hold highly illiquid private credit funds.

"Businesses like private equity and private credit emerged in the 1980s, when we had stopped following pro-labor policies. To me, these businesses are actually relics of that era."

He believes problems in this asset class have been apparent for a year and a half, but the market has been slow to acknowledge them – while credit spreads remain extremely low and the stock market is at highs, masking the underlying real risk. "The problem will still continue to affect the market, slowly but surely."

Transcript

Russell Clark: 00:00

If you look at people 40 and under, let's say the 20 to 30 age group, the biggest problem they face is they can't afford housing. If you want to bring housing costs back to a more reasonable level, then wages need to grow by about 7% per year, which would double them in 10 years. At the same time, nominal housing values should stay stable, and real values should fall. That requires real interest rates to be around 3%, so people put money in the bank instead of investing in real assets. That puts rates at around 10%. That's still my target for this year: Treasury yields at 10%. So the question is: how high can wages actually go?

Russell Clark: 00:49

This episode of Other People's Money is sponsored by the Tocurium Soybean Fund, ticker symbol Soy B. Welcome to Other People's Money. I'm Max Wiethe, and joining us today from London is hedge fund manager Russell Clark.

Max Wiethe: 01:02

Russell, thanks for joining us. You're not just running a blog, you're also managing a hedge fund. I recently read your work and found your article last week on AI investing very interesting. Many people think this might be the end of a massive speculative bubble. In the AI investing space, you pointed out there's another asset class that you think is much larger and more speculative. Can you tell me why you think this larger market has a lot of risk right now? I assume you mean the Treasury market. Two questions: first, is the AI market speculative?

Russell Clark: 01:44

So why do I think there's speculative behavior in the US Treasury market? Usually, when I look back at any major selloff event in my investment career, there were always clear signs that things weren't right. But people chose to ignore them, partly due to human psychology – when there's a problem and action is needed, people often prefer to ignore it because it's easier. That's probably human nature.

Russell Clark: 02:25

For example, during the 2008 financial crisis, people saw the housing problem coming 3 to 4 years before it happened. The issue was starting to show. Everyone thought it was just something we could handle, because we'd dealt with similar things before. Of course, others said the bank balance sheets were in terrible shape, so this housing crisis would cause bigger problems. Eventually, everyone accepted it.

Russell Clark: 02:58

What I'm saying is, especially regarding the US Treasury, but government bonds in general. In recent years, voters and politicians have gradually realized that the government will do whatever it takes to maintain economic growth, so it seems more willing to spend. Therefore, if any problem arises, the government steps in – that's the Trump administration's case. They even took it further: willing to spend what's necessary while not taxing anyone, especially large corporations.

So, the government spends but doesn't try to increase taxes. If you look closely at government financial statements, current revenue can barely cover necessary expenses like social security and interest payments. I think that accounts for about 90% of spending. Of course, that doesn't include other areas like defense, education, and infrastructure. So overall, the government's spending and tax mechanism is pretty well-established. This applies not just to the US, but also to Japan.

Russell Clark: 04:23

What's interesting to me is that in 2022, I was briefly bearish on Treasuries. There were other reasons, but mainly because when Russia's foreign exchange reserves were frozen, the Russian government couldn't use them after invading Ukraine. I thought, if a country has foreign exchange reserves, and it's the Russian government, why keep that money there? Thinking further, why would any country choose to hold Treasuries as foreign exchange reserves?

Russell Clark: 05:20 So, I expected to see a natural transfer from the Treasury market to the gold market. To me, that seems likely. However, I also suspect that investors seeking fixed income, especially those investing in sovereign government bonds, are gradually disappearing. Actually, that is the case. The Treasury market has performed relatively well.

But if you look at more peripheral sovereign bond markets, like Japan, it's different. Japan is one of the world's largest sovereign bond markets, but yields there have risen sharply. The UK situation is more complex, and the market is still very unstable. In the long term, investors are still selling Treasuries. I think the US Treasury market has performed fairly well, but genuinely willing investors are disappearing.

Russell Clark: 06:13

This is what I've been emphasizing when discussing this issue. I'm 52 now, getting older.

Russell Clark: 06:24 The idea of setting up large sovereign wealth funds and accumulating massive foreign exchange reserves is actually relatively new. Before 1980, people didn't really know how to hold another country's fixed income as foreign exchange reserves. That makes sense because all reserves were basically gold. Later, Japan started buying a lot of Treasuries because they didn't want their currency to appreciate.

Max Wiethe: 06:55

So, when you look at those 500-year charts, you see that reserve currencies used to be the British pound, and before that, another European currency. We can trace it back to the Portuguese era, where the currency was thought to be closely tied to the world's strongest navy. But actually, that wasn't the case. Unlike now, we didn't hold other countries' bonds or currencies.

Russell Clark: 07:17 So, reserve currencies are a relatively new concept. Historically, gold has always been the only form of foreign exchange reserves. Typically, countries with strong armies held the most gold for various reasons – basically, they would get it from other places or countries that had gold. Therefore, if a country lost a war, its gold reserves would be used to compensate the victors.

So when people talk about foreign exchange reserves, they often conflate them with the main trading currency or the currency used for transactions. And these currencies were often backed by gold. In fact, the US dollar was backed by gold until the 1970s. You know, after World War I, the British Empire began to disintegrate.

Max Wiethe: 08:11

You saw the British pound depreciate because their calculations didn't accurately reflect the real situation. So, do you think we are now returning to a historical period – a period where hard assets, especially gold, become the main component of foreign exchange reserves, or perhaps the concept of foreign exchange reserves has changed.

Russell Clark: 08:32

They will really disappear. Yes, I do think so, because I feel this is all just a political debate, not reasoning based on facts. So people often use empirical data to refute me, claiming that's how it was in the past. My response to them is that we are in a constantly changing political environment, so this change is an inevitable outcome.

Russell Clark: 08:59

I believe that after 1980, with the Reagan Revolution, people gradually stopped emphasizing full employment and rising wages. Instead, they preferred to let prices float freely and let wages adjust according to market conditions. Wages can adjust in two ways: either cut wages or devalue the currency, thereby lowering wages, improving competitiveness. Therefore, I think from the 1980s onwards, when many countries faced fiscal, financial, or current account crises, they often chose to devalue their currencies. This lowered the wages of domestic workers, allowing them to drive economic growth through exports.

This model was further developed in Japan, which kept the yen weak by buying Treasuries, trying to create inflation and economic growth this way. Part of these arguments also involved free trade, i.e., lowering trade barriers.

Russell Clark: 10:14

We are gradually moving away from that government-led model of industrial organization. When I was a kid, all the major airlines were state-owned. Later, the government sold them off, and the unions disappeared. In those countries, there were three major automakers – General Motors, Ford, Chrysler – which were strictly regulated and protected by the government. After 1980, Japanese auto companies entered these industries, disrupting the union organizations in these companies. Therefore, the entire environment became very unfavorable for wage growth.

For highly competitive countries like Switzerland, Japan, or even Germany, they would strive to appreciate their currencies, then offset this appreciation by buying US dollars. So, this capital-driven growth model was actually designed to maintain low wages in some way.

Russell Clark: 11:13

But now I feel the political environment is gradually shifting towards a post-WWII state – achieving high