Big shots gave a warning: U.S. Treasury bonds may collapse first next year, the market has shifted from "rewarding AI spending" to "rewarding sell-offs," and the appreciation of Asian currencies will end the gold bull market

Renowned investor and CEO of Gavekal Group, Louis-Vincent Gave, made a shocking prediction during an interview on December 4th: the convergence of the Federal Reserve and the Treasury has been established, and its long-term impact will be the collapse of the U.S. bond market.

Gave cited the deterioration of the Japanese government bond market as a precursor and described this pattern as the "Turkey scenario"—sacrificing bond and local currency value in exchange for nominal GDP growth. In such an environment, savers will abandon bonds and flock to tangible assets that can generate cash flow (such as stocks and precious metals).

In terms of asset allocation, Gave pointed out that the erosion of capital value in a zero-interest-rate environment is the main driving force behind investors chasing risk assets. He believes the outlook for gold depends on the movements of Asian currencies, particularly the severely undervalued yen. He speculated that Asian currencies might appreciate by 2026, which would prompt capital to flow back to Asia, weakening the demand for gold and thereby changing the effective trading logic for gold in recent years.

Additionally, Gave raised sharp questions about the current AI giants like Anthropic preparing for IPOs, suggesting that this might signal a turning point for capital-intensive bull markets and signs of a bubble. He observed that the market sentiment is shifting from "rewarding cash burning" to "rewarding asset sell-offs," doubting that the AI market has missed the optimal IPO window.

Here are the key points from the interview:

• The reality is that whether the Federal Reserve cuts rates in December or makes more cuts in 2026, the overall situation is set: we are witnessing the convergence of the Federal Reserve and the U.S. Treasury.
• What will collapse first? The stock market, the dollar, or the bond market? I believe the most likely candidate is the bond market.
• Again, using the Japanese government bond market as a reference: it has been declining all year, and the decline is accelerating, while the stock market remains indifferent.
• Most people view gold and silver as inflation hedges, but they are actually tools for hedging against zero interest rates.
• Therefore, I think the key question for 2026 is: will Asian currencies appreciate? If so, many effective trades in recent years may no longer work.
• So I wonder if we are currently transitioning from a phase where "the market rewards spending" to a phase where "the market begins to reward asset sell-offs." If so, IPOs for Anthropic or OpenAI will be very difficult to achieve.
• I believe there is an inherent contradiction in AI: at this point, the numbers are so large that to rationalize them, you must believe we are undergoing a massive societal transformation.

Here is the transcript of the interview:

The "most vulnerable link" in the global market? The U.S. bond market may follow Japan's footsteps

Louis-Vincent Gave:

Regarding the Federal Reserve, I think there was a wave of selling in the market in November due to concerns that the Federal Reserve would maintain a tight policy. The reality is that whether the Federal Reserve cuts rates in December or makes more cuts in 2026, the overall situation is set: we are witnessing the convergence of the Federal Reserve and the U.S. Treasury. **

What kind of long-term impact will this have on the US dollar, the bond market, and the stock market? I believe the results are slowly becoming apparent. In terms of the bond market, Japan's current situation can be seen as a precursor. In Japan, the Bank of Japan (BOJ) has kept interest rates at levels far below what they should be, and the long-term bond market has begun to experience sell-offs, even though the Japanese government has not actually issued bonds at the long end. The situation at the long end of the Japanese Government Bond (JGB) market is starting to become quite dire, which may be a rather ominous sign for the US Treasury market itself. So back to your point, what will collapse first? The stock market, the dollar, or the bond market? I think the most likely candidate is the bond market.

So what can the Federal Reserve do? Treasury put options, Fed put options, Trump put options have all played a role in some way at different points in 2025. If the bond market starts to awaken in 2026, it could shake up other global markets and assets. At some point, yes, but again using the Japanese government bond market as a reference: it has been declining all year, and the decline is accelerating, while the stock market seems indifferent to this. The yen has been weakening, but overall the Japanese economy is doing well. This is precisely the pattern we refer to in our research as the "Turkey scenario," which mimics the monetary policy that Turkey once adopted: sacrificing bondholders and the value of the domestic currency in exchange for substantial nominal GDP growth, thereby maintaining capital inflows and economic operation. In such an environment, savers will seek tangible assets, which can be precious metals, real estate, or stocks, as stocks do indeed generate cash flow.

So back to your question, if bonds are sold off, will it crush the stock market? I'm not sure. Within a certain range, like the 4%-4.1% bond yield you mentioned earlier, if we rise to 4.5% or 4.75%, will that crush the stock market? Probably not, because while the increase in financing costs is not a good thing, how much actual punishment does it impose on most companies? Perhaps not that much.

Global Asset Choices Under the "Zero Interest Rate Trap"

Host:

Do you think this dynamic we see in 2025—one South Korean market commentator described the investor psychology as being unable to tolerate holding cash in their portfolios, while investing in risk assets is indeed the right strategy in 2025—will be different next year?

Louis-Vincent Gave:

It depends on what kind of risk it is. One of the best-performing assets this year has been all precious metals, such as silver and gold. I believe silver has risen 100% this year, and gold has risen 55%. This goes back to one of our long-term themes, most people view gold and silver as inflation hedges, but they are actually tools for hedging against zero interest rates.

Regarding the South Korean savers you mentioned, interestingly, when you look at where global gold purchases are occurring, you find they are concentrated in Japan and South Korea, which are zero interest rate regions. In the US and Europe, you do not see large-scale gold purchases because they have positive interest rates. So when people look at their bank accounts and see interest rates at zero, especially in a situation where inflation is simultaneously reaching 3% in places like Japan or the US, you feel that the value of your capital is being eroded So you have to do something about it. The next question is, how should I cope with the current erosion of capital? Should I invest in the stock market, precious metals, or cryptocurrencies? Here, different people will have different risk tolerances. I do believe this has been a significant driving force in stock markets around the world.

Host:

Do you think this situation will continue? Do you recommend gold now?

Louis-Vincent Gave:

Regarding gold, in recent years, the major buyers of gold have been central banks around the world, which are basically insensitive to price; secondly, retail investors in Japan and South Korea, who have made substantial purchases. The Western world has not really participated in this. So the risk lies in these buyers stopping their purchases. Why would they stop buying? I believe the only reason is that their currencies start to appreciate.

Now, when you take a step back and examine today's world, what is the biggest anomaly? Regarding your question about bear markets, what is the most mispriced asset in the world today? I think the most absurdly priced assets in the world right now are all Asian currencies. Because when the yen is at 155, Japan is also ridiculously cheap. So we live in a world where Asian currencies have become a huge outlier; they are unbelievably cheap.

Therefore, I think the key question for 2026 is: Will Asian currencies appreciate? If so, then many effective trades in recent years may no longer work, because the Japanese and Koreans will start to repatriate capital, buying domestic assets and seeking returns in their local currencies. If the yen starts to appreciate, Japanese real estate investment trusts (REITs) will rise. So my answer to your question is a bit like a Jesuit priest, answering with another question. But regarding gold, you must have a view on Asian currencies because today's buyers are Asians. If you, like me, believe that 2026 will be a year of appreciation for Asian currencies, then you may need to start slightly reducing your bets on gold.

From "Rewarding Cash Burn" to "Rewarding Sell-offs," AI Giants Face Valuation Challenges for IPOs

Host:

Regarding artificial intelligence, what we have recently seen is that large AI company Anthropic is preparing for a possible IPO next year. Is this a sign of an AI bubble? Because it seems that this is how things work: early investors lock in profits and cash out. Or am I being too cynical?

Louis-Vincent Gave:

When it comes to financial markets, I think you can't be too cynical, especially as a journalist reporting on financial markets. When you observe a bull market, particularly a capital-intensive bull market—let's not kid ourselves, AI is extremely capital-intensive—the historical pattern is that it first goes through a phase where the market rewards you for spending money. Remember Chesapeake Energy during the U.S. shale boom, or Lucent during the fiber optic craze of 1999? The more they spent, the higher their stock prices were revalued Then, at some point in time, for some reason, the market sentiment will shift. You are no longer rewarded for spending money; instead, you are rewarded for shedding assets that were acquired at high prices during the boom. I wonder if we are currently experiencing this shift and whether Anthropic has already missed the IPO window.

Let me explain. Remember Oracle a few weeks ago? Oracle announced it would spend $300 billion to build data centers, and the market reacted enthusiastically, with Oracle's market value immediately increasing by $300 billion, and everyone was happy. But then the market began to question, "How are they going to pay for this?" Oracle's credit default swap (CDS) spreads widened sharply, and everyone realized they couldn't finance it. Within two weeks, its stock price fell back to the level it was before the announcement of this capital expenditure, or even lower. Another example is SoftBank, which dropped 40% in two weeks, primarily because SoftBank indicated it would sell its shares in NVIDIA and invest all the funds into OpenAI. The market began to think this was simply illogical.

So I wonder if we are currently transitioning from a phase where the market rewards spending money to a phase where the market starts rewarding the shedding of assets. If so, an IPO for Anthropic or OpenAI will be very difficult to achieve because these companies have already reached valuations in the hundreds of billions in the private equity market. To make the last round of investors profitable, you must IPO them at a very high valuation. Whether the market has an appetite for this is, I believe, an open question.

Host:

Yes, this has always been confusing. If we zoom out and look at the entire AI narrative, the market once disliked companies with unimpressive capital expenditure figures, and then if they spend too much, the market worries again; this cycle is perplexing. But many sell-side analysts say, "No, we are just getting started; we are in the early stages of the AI cycle." So what will the narrative around AI be in 2026? Do you think we are still in the early stages, just that the financial markets are moving too fast?

Louis-Vincent Gave:

You have to understand that when sell-side analysts say we are in the early stages, don't forget their job is to make money by underwriting securities. So of course, they will say, "No, no, you will love these securities; please buy more." That is their business model. So perhaps we shouldn't be too surprised by this.

I think when you look at the numbers surrounding AI, they are already staggering. A recent research report from Bain Capital pointed out that to justify the current capital expenditure figures, AI needs to rapidly increase annual revenues to $2 trillion. For reference, the entire global advertising industry—which is essentially paying for all of this, including ads on Manchester United jerseys and online advertising—has total annual revenues of $1 trillion. For AI investments to make sense, it needs to quickly generate twice the revenue of the global advertising industry So these numbers are extremely challenging. You might say this will happen because AI will replace 10%, 20%, or even 30% of the global workforce. Well, if we get to that point, we will face a whole host of other issues. For example, our welfare state system is not built for a 20% unemployment rate, nor is our society. So I believe there is an inherent contradiction in AI: at this point, the numbers are so large that to rationalize these numbers, you must believe we are undergoing a massive social transformation.