As long-term US Treasury yields remain elevated, Oaktree Capital co-founder Howard Marks has issued a warning: the loss of US fiscal control is the root cause of higher interest rates, and any market intervention that sidesteps this issue is merely a temporary fix that fails to address the underlying problem.

On Tuesday, Marks published his latest memo on the Oaktree Capital website, directly targeting the deep-seated risks in US fiscal policy. He criticized the Treasury's expansion of its long-dated bond buyback program, arguing that such actions can only suppress yields in the short term but cannot resolve the fundamental drivers pushing rates higher. 'Forcibly buying bonds to push down yields is like a doctor putting an ice pack on a feverish patient,' Marks wrote. 'The ice pack may bring the temperature down temporarily, but the patient is unlikely to truly recover until the underlying cause is addressed.'

In Marks' view, the real 'causes' include persistent inflationary pressures, continuously expanding government debt, and the enormous capital demand represented by AI infrastructure construction. He emphasized that the US fiscal deficit currently stands at roughly 6% of GDP, noting this is 'an extremely abnormal level for an economy enjoying prosperity with an unemployment rate of only 4%.' Notably, he also pointed out that under current circumstances, selling US stocks and dollar assets is not the answer, as shifting to non-US assets carries risks that cannot be ignored.

Operation Twist' Fails to Convince the Market

The US Treasury previously announced it would at least expand its long-dated bond buyback scale to $4 billion per operation, with Treasury Secretary Bessent subsequently signaling a 'whatever it takes' stance. After the announcement, long-term yields fell that day but rebounded the very next day. Marks compared such market intervention to holding a ball aloft in the ocean with a water column: the ball floats as long as the water jets upward, but once the pump stops, the ball falls.

He cited investor Druckenmiller's commentary in the Wall Street Journal to support his point: 'Every basis point of artificial yield suppression is a subsidy for delay... Governments defending prices against fundamentals have never won; the only variable is how much money they spend before admitting defeat.'

Deficit, Inflation, and AI Capital Demand: Triple Pressures Pushing Long-Term Rates Higher

In his memo, Marks systematically laid out the structural drivers behind rising long-term interest rates. First is the fiscal deficit issue. He noted that the current ~6% of GDP deficit is a rare high level during an economic expansion phase. Net interest spending is projected to exceed $1 trillion this year, surpassing the defense budget, and as debt continues to expand, this figure will accelerate. He criticized the current government for large-scale borrowing during prosperous times, completely deviating from the Keynesian logic of 'deficits during downturns, repayment during recoveries.'

Second is inflation stickiness. Marks mentioned in the memo that PCE inflation remains above the Fed's 2% long-term target, forcing the Fed to maintain a tighter stance; large deficits themselves also have inflationary effects, as the liquidity injected through government spending exceeds what is withdrawn via taxes, further boosting aggregate demand.

Third is the wave of AI-driven capital demand. Marks cited McKinsey forecast data that global investment in AI-related data center construction will exceed $5 trillion by 2030. This capital demand, combined with the Treasury's approximately $2 trillion in annual net new bond supply, collectively pushes up the cost of capital. 'Increased demand leads to higher prices, which is the simplest economic principle,' he wrote. 'The upward pressure on interest rates from growing capital demand is entirely understandable.'

The Real Solution: Fiscal Discipline, Not Market Manipulation

Marks made clear that pushing down interest rates should not be a policy goal in itself; addressing the fundamental factors driving higher rates is what matters. He outlined what he sees as the only viable long-term solutions: enhancing fiscal responsibility by raising revenue as a share of GDP through higher income tax rates (especially for high-income earners) and cutting tax breaks, while keeping spending growth below GDP growth.

He also noted that increasing GDP growth rates could help improve the deficit situation, with the broad adoption of AI as a productivity tool and pro-business policies playing complementary roles — but only on the condition that new tax revenues are not squandered. Marks concluded by citing Buffett's remarks at the 2025 Berkshire Hathaway annual meeting: 'What worries me is US fiscal policy... The deficit we are currently running is unsustainable over a very long time horizon.'

Diversification Has Value, But Avoid Overcorrection

On the investor-centric question of asset allocation, Marks took a relatively restrained stance. He admitted that this is essentially a political issue but poses a practical challenge for investors. Selling US stocks, however, does not solve the problem — moving funds into dollar-denominated bank deposits, money market funds, or bonds does not eliminate risk; to hedge against dollar depreciation, investors would need to shift into non-dollar-denominated assets, non-financial assets like gold or foreign real estate, or non-US company stocks.

Yet Marks warned that this path is not smooth. Many companies in other developed countries have weaker growth prospects than top US firms and face more regulatory constraints; emerging markets offer growth potential, but realization comes with higher uncertainty. He argued that the US still holds a comprehensive edge in free-market systems, innovation vitality, the rule of law, higher education, and capital market depth, adding that 'no other country possesses these traits to the same extent.'

While Marks did not entirely oppose moderate diversification of dollar assets, he emphasized that timing a large-scale shift is extremely difficult. 'No one knows when the problem will actually explode, and until it does, such a move is likely to look like a mistake for a very long time,' he wrote.