A major transition is underway in the engine of US dollar liquidity expansion. As inflation becomes more apparent, and with the new Federal Reserve Chair, who advocates for balance sheet reduction, taking office, the era of exogenous money driven by Fed balance sheet expansion and fiscal deficits post-pandemic may be drawing to a close. Concurrently, an endogenous monetary expansion driven by AI-related capital expenditures is taking shape—the liquidity engine is shifting from policy to the real economy.
This endogenous monetary expansion enhances economic resilience and deepens inflation persistence, while also accelerating the flow of capital from traditional sectors like real estate and consumption towards high-return technology frontiers. Looking ahead, asset performance is likely to diverge further: assets reliant solely on exogenous liquidity may face pressure, while those representing advanced productivity are poised to benefit from the expansion of endogenous money. The key points are outlined below.
Exogenous Money Peaks, Marking the End of the Flood Era
Exogenous money refers to currency issuance through fiscal channels, typically involving increased government transfers to the private sector, with the Fed expanding its balance sheet to support fiscal expansion. Following the pandemic shock, the US government and the Federal Reserve jointly implemented such measures, injecting substantial liquidity into global markets.
However, this dynamic is changing with the appointment of the new Fed Chair. The Chair has consistently opposed excessive expansion of the Fed's balance sheet and the normalization of quantitative easing, advocating instead for orderly balance sheet reduction and a return to conventional monetary policy. This stance has garnered support from key officials, including the Treasury Secretary and the Fed Vice Chair for Supervision.
Upon taking office, with inflation risks resurfacing, the Chair's primary task is to swiftly establish policy credibility, likely requiring a demonstration of commitment to combating inflation. This could be shown in two ways: hinting at potential future policy rate hikes or controlling the money supply. Given current political and economic constraints, raising interest rates would conflict with the administration's preference for low rates and lack public support, also contradicting the Chair's own inclination towards lower rates. Therefore, strictly controlling the balance sheet size and signaling control over the total money supply may be the optimal policy choice at this stage.
This suggests that the supply of reserves may shift from previously "unlimited easing" to a "relatively scarce" environment, signaling the end of the narrative of massive balance sheet expansion.
On the fiscal front, the liquidity pulse driven by the US budget deficit is also waning. In the first half of 2026, tax rebates related to the Great American Act provided temporary market support, but this essentially represents the "lagging tail-end effect" of existing policies. In the second half, as the Act's disbursements conclude, new liquidity injections from the fiscal side will be limited.
Furthermore, with US federal debt at historically high levels, constraints from the debt ceiling and rising interest payments indicate very limited room for future fiscal stimulus. In fact, the core fiscal legislation of the current administration has been largely "front-loaded." With control of Congress, the administration swiftly passed the Great American Act using budget reconciliation procedures, making key tax cuts permanent. The administration now primarily awaits the execution and effects of this legislation, with no other major fiscal plans in the near term.
The Rise of Endogenous Money: The Multiplier Effect of the AI Wave
Endogenous money refers to money creation through private sector credit expansion, primarily relying on bank lending, capital market financing, and the resulting multiplier effects. It depends not directly on monetary or fiscal policy, but on whether the real economy has strong investment and financing demand, and whether financial institutions are willing to extend credit.
Currently, the accelerated development of the AI industry is driving a significant expansion in corporate capital expenditures, forming a cycle of endogenous money creation. According to the latest financial reports, the capital expenditure of the top five US hyperscale cloud providers is projected to reach $735 billion in 2026, accounting for approximately 2.5% of US nominal GDP.
At this stage, funding for these capital expenditures primarily comes from exceptionally strong corporate free cash flow. This investment activity functions in a "quasi-fiscal" manner—releasing demand upstream to hardware, data centers, power infrastructure, construction, and software services, boosting activity in related industries.
In terms of scale, the fiscal stimulus from the Great American Act is around $300 billion, while the capital expenditure of cloud providers is more than double that amount. This model is also set to evolve: tech giants are shifting from using internal cash to debt and equity financing. Within the endogenous money framework, this does not signal tighter financing conditions. Instead, leveraging their high credit ratings, these companies are channeling idle funds and credit resources from across the economy more efficiently into the AI sector, which represents future productivity growth.
Thus, the shift from cash flow to debt financing is not inherently negative; it represents a market-driven reallocation of financial resources. Credit data also indicates an expansionary phase for endogenous money. This year, US commercial bank loan growth has shown a "structural recovery," with lending to non-bank financial institutions remaining high, and commercial & industrial loan growth seeing its most significant uptick since 2023. This likely reflects substantial funding demand from the AI supply chain and related industries.
In contrast, growth in real estate loans and consumer credit remains subdued. With 30-year mortgage rates and credit card rates persistently high, the traditional "real estate-consumption" credit channel continues to be constrained. This divergence indicates that the US economy is undergoing a structural shift—financial resources are accelerating their flow from traditional real estate and consumption sectors towards high-return technology sectors, exemplified by AI.
The "endogenization" of money creation is a manifestation of this shift. The ongoing expansion of endogenous money also provides favorable conditions for the aforementioned turn towards contraction in exogenous money, such as balance sheet reduction.
Market Implications: Structure and Divergence
Regarding growth, the multiplier effect from AI capital expenditures enhances economic resilience. While this process may exacerbate "K-shaped" economic divergence, as long as endogenous credit expansion continues, the US economy maintains a solid foundation. Its recovery momentum is likely to continue outpacing other major developed economies, with positive spillover effects on the global economy.
On inflation, demand expansion from capital expenditures, combined with supply constraints from tariffs and geopolitical tensions, will push prices higher from both supply and demand sides, increasing inflation persistence. It is also noted that the greatest uncertainty in the second half of the year stems from inflation. This will not only constrain the Fed's policy space but also mean investors must adapt to a "higher for longer" interest rate environment.
In financial markets, strong corporate financing demand and bank credit expansion will pressure the bond market, potentially pushing the yield curve higher. As interest rates rise, the US stock market may transition from a phase driven by Fed and fiscal easing and valuation expansion to a new stage driven by high profitability and efficient conversion of capital expenditures. The core pricing variable may shift from interest rate expectations to whether tech companies can translate massive capital spending into actual profits.
Optimism about the tech revolution may continue to drive capital rotation from interest-rate-sensitive traditional sectors towards AI-related supply chains that benefit from endogenous credit expansion. In contrast, assets that rely solely on liquidity narratives and benefit from dollar oversupply may face sustained pressure. Non-yielding assets like gold and Bitcoin have already shown signs of adjustment over the past six months, reflecting the market's forward-looking pricing of this liquidity source shift.
Consequently, future asset performance is likely to remain divergent. The key differentiation will be whether an asset represents an advanced productivity direction and can benefit from the expansion of endogenous money.
Comments