Is the Era of High Interest Rates Driven by AI Investment Here?

This sharp rise in Treasury yields is creating significant ripple effects in global financial markets and is acting as a factor that increases future economic uncertainty.

The prevailing optimistic view was that AI would increase overall economic efficiency, alleviating inflationary pressures, which would eventually lead to central banks cutting interest rates.

Currently, large-scale investments in the AI industry are causing a short-term scramble for capital, acting as a major factor in rising market interest rates.

As AI-related big tech companies competitively issue corporate bonds to expand their businesses, this leads to overall upward pressure on interest rates.

The projected issuance of AI-related corporate bonds by big tech companies is expected to reach an astonishing $420 billion by 2027, a 60% increase from this year.

Furthermore, by 2030, the scale of AI-related debt issuance is predicted to hit $4.1 trillion, meaning that AI investment as a percentage of GDP will increase to 2.5% in 2027 and 2.7% in 2028.

This astronomical scale of AI investment creates a stronger demand for capital than any previous technological revolution, impacting the market.

The Federal Reserve's projected neutral interest rate has risen by 0.7 percentage points, from 2.5% to 3.2%, in just three years.

This suggests a structural shift where the value of money is increasing due to insufficient supply compared to growing demand. An analysis indicates that rising prices are necessarily accompanied by interest rate hikes, confirming a structural characteristic and suggesting that high interest rates could become the new norm.

Of Microsoft's $72 billion in capital expenditures this year, $25 billion was purely due to rising component prices.

Additionally, a surge in power demand from data centers is leading to increased electricity costs in some regions.

The US House of Representatives passed a bill by an overwhelming margin of 417 to 3, requiring companies to bear the cost of expanding the data center power grid, suggesting that the burden of power infrastructure costs may increase.

This indicates that the market's perception of risk regarding AI companies is expanding, and investors are demanding a higher risk premium.

As the US 10-year Treasury yield reaches 5.2%, it is nearing the profit margins of S&P 500 companies.

This causes confusion in investment decisions as returns on risky assets like stocks become equal to safe assets like Treasuries, effectively eliminating additional compensation for holding stocks.

Considering capital expenditures of $44.9 billion, it recorded a cash flow deficit of $5.9 billion, leading to an increasing trend of issuing long-term corporate bonds to cover this.

Companies like Oracle also face growing doubts about cash flow generation, with only 13% of contract amounts actually reflected in revenue within a year.

This is attributed to pressure for power expansion due to local opposition and midterm elections, amplifying market doubts about Oracle's ability to fulfill sales contracts and secure cash flow.

The US 10-year Treasury yield, which was only 0.99% five years ago, has surged to 5.2%, increasing debt repayment pressure on companies.

Companies face a situation where maturing corporate bonds issued during low-interest periods must be refinanced at much higher rates, raising doubts about their ability to monetize revenue.

They are caught in a vicious cycle where they cannot stop as long as competitors continue to invest, leading to rising financing costs and a 'chicken game' that increases the likelihood of restructuring.

Ultimately, a winner-take-all structure is being reinforced in the AI industry, where companies that run out of capital first are eliminated.

Productivity improvements lead to a virtuous cycle of lower corporate costs, stable prices, and falling interest rates.

Similar to the US productivity growth of 2.5% annually from 1995 to 2000, only if AI investment translates into actual productivity improvements can the current vicious cycle of rising interest rates be broken.

True productivity improvement requires creative destruction and the establishment of new systems.

Considering that productivity gains during the past PC revolution were limited to a 1.1 percentage point increase, and given the current scale of US debt and corporate bond issuance, at least 4-5% productivity improvement is needed to lower interest rates.

Bessen argued that no further interest rate hikes are necessary based on AI productivity improvements, citing the 1996 case where Chairman Greenspan predicted productivity growth from technological innovation and did not raise rates.

Just as the IT revolution in 1996 doubled productivity and led to a boom without interest rate hikes, if AI follows a similar path, productivity improvements could break the vicious cycle of rising interest rates.

Oil shocks from the Strait of Hormuz and the persistence of high oil prices are identified as key variables for interest rate policy.

The current situation of persistent price instability and oil shocks in 2026 implies a potential similarity to the 1847 crisis.

Oil shocks from the Strait of Hormuz and the persistence of high oil prices are analyzed as major variables for future interest rate policy.

High interest rates in the US worsen the financial situation of AI infrastructure investment companies, directly affecting Korean HBM-related stocks.

Because companies borrow money to invest, they are inevitably sensitive to interest rates. Fear that contract fulfillment will become uncertain if someone is pushed out of the chicken game means Korean stock prices undergo greater adjustments than US stocks when interest rates rise.

Particularly, with Samsung Electronics and SK Hynix's share buybacks scheduled to end in mid-October, this period will be a test for future stock price maintenance.