US Stocks Stay Strong Despite Prolonged High Rates

According to Park Shin-young, services account for more than 60% of US nominal GDP, while goods make up only about 30%, reflecting a service-centered economic structure.

Because of this, the impact of interest rate hikes spreads more slowly than in a manufacturing-centered economy, and existing service demand does not react immediately to rate changes, which contributes to overall economic stability.

Park Shin-young explained that the net worth of US baby boomers stood at $97.4 trillion as of Q2 2026, accounting for 53% of total household assets.

They hold substantial assets, including $35.2 trillion in stocks and $20.6 trillion in real estate, and their debt share is only 21%, so rising rates increase their interest income and help them maintain spending power.

Thanks to the high share of low fixed-rate loans secured during the pandemic, the US is sustaining today's high Treasury yields.

About 49.9% of mortgages carry rates of 4% or lower, and roughly 78% carry rates below 6%. As a result, the interest burden on existing debt is fixed, and the impact of rate hikes falls mainly on new borrowing.

Park Shin-young noted that when employment holds, household wages keep flowing in, so spending is limited in its decline even in a high-rate environment.

In fact, US nonfarm payrolls rose by 162,000 in August, the unemployment rate held at 4.1%, and real personal consumption expenditures grew 3.4% annualized in Q2, showing solid spending.

The federal fiscal deficit stands at 5.8% of GDP, above its 50-year average, while AI investment is shifting from training to inference, driving demand for CPUs and GPUs. The federal deficit is above the 50-year average at about 5.8% of GDP.

On the private side, AI investment has emerged as a new growth axis. It is shifting from training-centered to inference-centered, driving demand for CPUs and GPUs.

Capital expenditure by hyperscalers such as Amazon, Alphabet, and Microsoft is expected to reach about $800 billion this year and about $1 trillion next year.

Such high real rates mean investors can earn stable returns from US Treasuries. This forces stricter standards on the valuation of AI companies that depend on future earnings and intensifies competition between growth expectations and stable bond yields.

The S&P 500 trades at about 19 to 20 times expected earnings over the next year, reflecting high market expectations.

The Shiller CAPE ratio, which compares prices to long-term average earnings, is around 40, similar to the 44 seen just before the 2000 dot-com bubble. High valuations mean a few large AI stocks drive the index higher, while the median constituent is 16% below its 52-week high, showing market polarization.

The AI segment has yet to prove a clear return on investment (ROI). Right now, cash generated by the existing high-margin businesses of hyperscalers such as Microsoft, Alphabet, Meta, and Amazon is closer to funding investments in data centers, AI chips, and power infrastructure.

Goldman Sachs projects 2027 as an inflection point for verifying AI investment profitability, while Morgan Stanley expects free cash flow to decline through 2027 due to AI spending before recovering in 2030.

Increased Treasury issuance from a widening deficit leads investors to demand higher yields. Rising long-term rates raise government interest costs, which can in turn widen the deficit in a vicious cycle, while oil price increases driven by geopolitical conflict can also fuel inflation.

However, if asset prices adjust, baby boomers' asset income and spending could shrink, and whether the pace of AI company earnings growth can offset the high-rate environment is expected to be the key point to watch for the stock market.