UBS: US companies' willingness to pay Capex has bottomed out and rebounded; further interest rate hikes may have a limited impact on the economy.

U.S. businesses' willingness to spend capital is rebounding from a low point, so the marginal drag on economic growth from further interest rate hikes may therefore be less than historical experience suggests.

In a recent report, UBS pointed out that US corporate investment has been weak for an extended period, with current capital expenditures starting at a significantly lower level than in previous cycles. Against this backdrop, even if interest rates continue to rise, their ability to suppress further corporate investment may be limited.

Meanwhile, AI investment remains a major pillar of capital expenditure growth in the United States. AI-related investment has increased by 26% year-on-year over the past eight quarters; more noteworthy is that, excluding AI factors, corporate capital expenditure intentions have also begun to rebound, indicating that signs of investment stabilization may be spreading from the technology sector to a broader range of corporate sectors.

UBS believes that pent-up maintenance and upgrade demand accumulated during a prolonged period of sluggish growth, ample operating cash flow from businesses, and the potential for AI demand to spread to other industries could all support a recovery in capital expenditure. This means that assessing the impact of tightening policies requires not only looking at interest rate levels but also observing marginal changes in corporate investment.

Traditional investment has remained weak, and the potential for interest rate shocks may be narrowing.

According to the rule of thumb of the Federal Reserve's FRBUS model, every 100 basis point increase in interest rates will roughly reduce economic growth by 60 basis points in the first year, and the cumulative drag could reach up to 150 basis points over two years. This is also an important basis for the market to assess the impact of further tightening.

However, UBS points out that the difference in this cycle is that corporate investment is not entering a contraction phase from a high level.

Over the past eight quarters, U.S. private non-residential fixed investment has grown by an average of only 0.1% year-on-year, far below the long-term average of 3.2%; residential investment has declined by an average of 2.0% year-on-year, also significantly lower than the long-term average of 2.2%. Apart from the technology sector, traditional U.S. investment has remained weak for at least two years.

Therefore, with investment already at a low level, further increases in interest rates will continue to constrain the market, but the room for further reducing corporate spending may have narrowed. The interest rate shock reflected in the FRBUS model has not failed; it's just that the historically healthy investment starting point is significantly different from the current situation.

Non-AI capital expenditure willingness rebounds, investment shows early turning point

AI remains the strongest driver of growth in US capital expenditures. Over the past eight quarters, AI-related investments have increased by 26% year-over-year, indicating that US corporate investment is not contracting across the board, but rather highly concentrated in the AI sector.

However, UBS observed that the willingness to spend capital in non-AI sectors has also begun to improve.

UBS compiled the capital expenditure willingness indicator from 14 regional Fed manufacturing and service sector surveys and took the median. This indicator fell to the third lowest percentile in history after April 2025, but rebounded to the 36th percentile by August 2026.

This level is still not high, but the direction has changed. Historical data shows that every 1-point increase in this indicator, roughly equivalent to 1.2 standard deviations, typically corresponds to an increase of about 5.5 percentage points in capital expenditure growth. While intentions do not equate to actual spending, they usually lead changes in capital expenditure.

UBS also excluded AI factors from its fixed capital investments, removing five categories of AI-related investments. However, the "non-AI" category may still include some AI spending, so this change is more accurately interpreted as a sign of stabilization in non-technology investments rather than a complete elimination of the AI impact.

Update demand and ample cash flow support Capex recovery

The rebound in capital expenditure willingness may be driven by the release of previously postponed maintenance and upgrade needs. After at least two years of persistently weak non-technology investment, some companies may find it difficult to continue postponing equipment and facility upgrades. Therefore, the recent rebound in Capex may not necessarily indicate a full-scale expansion, but could primarily stem from the restart of essential spending.

Corporate cash flow provides the financial foundation. Operating cash flow increased by $825 billion year-on-year, a growth of 25%; meanwhile, except for hyperscale cloud service providers, capital expenditure as a percentage of operating cash flow remains near historical lows.

This means that companies are not lacking funds, but rather that they haven't previously converted more of their cash flow into investment expenditures. If the willingness to invest in capital continues to recover, ample internal cash flow is expected to reduce companies' reliance on external financing and weaken the direct constraint of high interest rates on corporate investment.

Furthermore, the recovery in capital expenditure could expand further if demand for AI infrastructure, software, and automation continues to spread to non-tech sectors. Current data is insufficient to prove this trend is fully established, but the rebound in non-AI investment intentions from historically low levels is a noteworthy early sign.

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.