Henrik Zeberg: Why Bond Yields Are About to Collapse

Zeberg describes a fragile labor market, citing revised payroll growth of about 16,000 jobs per month from April 2025 through March 2026, GDP growth supported by current job creation of roughly 0.5% or zero, and 1.7 million people leaving the labor force since November. He argues that the 4.1% unemployment rate masks this weakness because falling participation reduces the economy’s production capacity.

Zeberg argues that AI investment and government spending have helped keep GDP above water, with the AI boom contributing an estimated 1% to 2% of GDP, but says consumers are too strained to offset a pullback in AI investment. He contrasts roughly $3 trillion in annual AI revenue needed to justify current investment with about $175 billion in current revenue.

Zeberg portrays households and housing as vulnerable: he says pandemic-era savings are gone, the savings rate is 3%, and the median consumer may struggle with an unexpected $2,000 bill. He also says long-term unemployment is rising and existing-home sales are at levels comparable to deep recessions, while home-price increases are slowing.

Zeberg says his business-cycle model reached a structural recession signal in July, which he says has preceded recessions by one to three months in prior cases, though the lag could be four months; he cautions that indicators should be watched before declaring an outright recession. He expects short-duration yields to fall sharply, a move he says would signal the economy is nearing its final phase before rolling backward.

Zeberg links recent higher yields to capital demand from the AI boom, which he estimates contributed 1% to 2% of GDP, and says that upward pressure could fade if that demand does not continue. He also expects weakening job creation to bring inflation and yields lower, arguing that the current supply shock is not demand-driven.

Zeberg argues that weakening growth and demand should bring disinflation, even if a supply shock temporarily lifts prices. He says a different phase could follow a bust: if growth resumes while money circulates faster than production, stagflation could last one to three years and be favorable for gold and commodities.