Interest Rates to 10%: Why the Treasury Market is the Real Speculative Bubble (Not AI) | Russell Clark

Russell Clark argues that the US Treasury market, not AI stocks, is the defining speculative bubble of the current era, and he targets a 10-year Treasury yield of 10 percent as his long-term destination. He derives that figure by combining roughly 7 percent annual wage inflation, which he considers necessary to restore housing affordability for people under 40, with a real interest rate of approximately 3 percent required to keep money on deposit rather than flowing into real assets like housing.

Clark frames the bond bear case as a political argument rather than a debt-sustainability argument. He notes that during the 1970s, when US yields reached 15 to 20 percent, federal debt to GDP was below 20 percent, demonstrating that high yields and low debt levels can coexist and that fiscal ratios are not the primary driver. The political shift he identifies is a return to the post-World War Two model of full employment and rising wages, reversing the Thatcher-Reagan era that suppressed wages, broke unions, and generated deflation through competitive imports. He points to Japanese bond investors beginning to sell JGBs in March 2020 as a leading indicator, reading the political response of wage protection and massive spending as a signal the world had structurally changed.

The freezing of Russian foreign reserves after the Ukraine invasion was a turning point for Clark's bearish Treasury view. He argues it raised the question of why any country that could disagree with the Trump administration would continue holding reserves in assets subject to seizure, and he predicts foreign reserve flows will shift from Treasuries into gold, noting that until 1980 all foreign reserves were held in gold rather than other countries' fixed income. He also observes that competitive economies like Japan, Switzerland, and Germany, which historically bought Treasuries to suppress their currencies and sustain export-led growth, are losing the political will to continue as a weakening yen causes real Japanese wages to fall and turns domestic politics against the policy.

On the fiscal side, Clark notes that US government revenue now covers only about 90 percent of mandated expenses such as social security and interest payments before accounting for defense, education, or infrastructure, and that the Trump administration plans to maintain spending without raising taxes on large corporations. He argues austerity is dead as a political policy, meaning growth will remain strong, inflation will remain elevated, and fiscal pressure will compound the structural rate argument.

Clark draws an analogy between Nvidia chips today and oil in the 1970s, arguing both represent the key input to economic growth with restricted supply. He notes Nvidia chips have remained expensive for five to six years without the price declines typical of semiconductors, and that generic chips like DRAM and NAND are now priced similarly to Nvidia products. He argues that stripping out the semiconductor sector, much of the broader market already resembles the 1970s pattern of high volatility with no real returns, and that higher interest rates are producing profound weakness in parts of the economy that semiconductor performance is masking in aggregate index data.

Clark argues that big tech AI spending is primarily defensive rather than profit-seeking. Google's core advertising business is directly threatened by LLMs because users increasingly search via ChatGPT, and Elon Musk entering the data center business has pushed incumbent CEOs to accelerate spending to make entry as expensive as possible. He considers it very unlikely that Microsoft, Meta, Google, and Amazon will cut AI capex by 50 percent in the near term, and he notes that cheaper and more efficient AI models do not reduce capex pressure because demand immediately absorbs efficiency gains. AI spending is also politically insulated because neither left-wing nor right-wing politicians would want to appear weak on China in the AI race.

Clark sees serious structural problems in private equity and private credit, which were built on the assumption of ever-lower interest rates. The Cliffwater private credit fund was forced to gate redemptions after outflows overwhelmed subscriptions. He attributes ongoing stress in these sectors to poor underlying asset quality and notes that money market funds offering seven to eight percent returns made illiquid private credit funds less attractive by comparison. He warns that problems in markets often persist for two to three years before they metastasize into something worse, and that shrinking pools of capital combined with a rising cost of capital compound each other negatively for these asset managers.