Heavily dependent on AI! The US economy is undergoing a "dangerous boom"

UBS has poured cold water on the US economy.

According to Chasing Wind Trading Desk, UBS points out in its latest research report that beneath seemingly robust growth data, the foundation of US economic expansion is rapidly narrowing: marginal improvements in investment, consumption, and employment are almost entirely tied to the singular theme of artificial intelligence. If AI investment slows or asset prices adjust downward, the current expansion will lose its core support.

UBS models show that the probability of a US recession in the next 12 months is about 50%. If the AI investment boom cools, the US economy will rapidly lose its core backing. Nevertheless, UBS remains cautiously optimistic about the long-term outlook, with potential GDP growth expected to rise to around 2.5%, benefiting from a lessening demographic drag and improved productivity.

Meanwhile, tariff shocks are reshaping the inflation structure as a "slow variable." The effective US tariff rate has jumped from 2.5% at the beginning of the year to over 13%, equivalent to a hidden taxation of about 1.1% of GDP. UBS believes this is not a one-off shock, but a persistent factor that will continue to push up core inflation over the coming years, keeping it above the Federal Reserve’s 2% target for an extended period.

On the policy front, fiscal and monetary policy have entered a typical "hedging—restriction" dynamic. In 2026, the OBBBA ("Outstanding Beautiful Bill Act") is expected to bring about $55 billion in tax rebates, briefly boosting consumption in Q2, but the effect will fade quickly thereafter; meanwhile, although the Federal Reserve plans further rate cuts, it is constrained by cost-driven inflation from tariffs, limiting the scope for monetary easing.

Against this backdrop, UBS’s assessment is not aggressive, but strikingly sharp: the US has not entered a recession, but this is an expansion with extremely low margin for error. The fate of economic growth depends on whether AI can, within the next 18 months, transition from a capital market narrative into real, broad-based productivity gains.

Growth persists, but the engine is highly concentrated

On the investment side, US economic polarization is markedly evident.

In the past four quarters, investment in AI-related equipment grew around 17%, covering computers, information processing devices, and electronic components. By contrast, investment in non-AI equipment fell about 1% in the same period. In non-residential building investment, data center construction contributed roughly 0.7 percentage points of growth, while other areas combined dragged growth down by about 7 percentage points.

UBS estimates reveal contributions to GDP growth:

AI investment: about 0.5 percentage pointsConsumption by high-income groups: about 0.8 percentage pointsAll other economic activities combined: only 1.1 percentage points

Residential investment contracted in four of the past five quarters, and non-residential building investment has shrunk for six consecutive quarters. If AI is excluded, the US economy is closer to "low-speed wandering" than stable expansion.

The "resilience" in consumption comes from wealth concentration, not income improvement

Consumption data is likewise misleading.

While real disposable income increased by only 1.5%, real personal consumption expenditure rose 2.6%. UBS points out that this divergence is not due to wage or employment improvements, but rather the concentrated unleashing of equity wealth effects.

In Q2 2026, stock assets will account for a record 35% of total household wealth, and market performance is primarily driven by AI and tech sectors. The result: high-income household consumption is greatly amplified, while middle- and low-income groups continue to suffer from inflation and tariff erosion.

This means US consumption has become heavily reliant on asset prices, and any stock market volatility will quickly erode the buffer for consumption.

Tariffs: A brewing "hidden tax increase"

UBS offers a judgment on inflation that sharply diverges from prevailing market optimism.

The effective US tariff rate has climbed to 13.2%, approaching levels seen under the Smoot–Hawley Tariff Act of the 1930s. Based on import structure calculations, the actual effective rate is about 12.2%, equivalent to an economic tax hit of about 1.1% of GDP.

Unlike traditional shocks, tariffs have not immediately suppressed imports but instead seep slowly into the price system from the cost side. UBS expects that over the next four years, tariffs will cumulatively drag on real GDP growth by about 0.8 percentage points, with the impact mainly concentrated in 2025–2026.

Inflation already reflects this trend: core PCE rebounded sharply midyear, and UBS forecasts it will peak in Q2 2026, then remain around 3% for an extended period. This is classic cost-push inflation and the root of the Federal Reserve’s policy constraints.

Employment data "maintain appearances," but underestimate underlying weakness

Although the unemployment rate remains relatively low, UBS believes it no longer accurately reflects the labor market’s condition.

Excluding health care and social assistance, nonfarm employment dropped by an average of 41,000 jobs per month over the last four months. The unemployment rate rose to 4.5%, while the U-6 broad unemployment rate reached 8.43%, significantly above pre-pandemic levels.

Multiple survey measures are weakening simultaneously: business hiring expectations are falling, manufacturing ISM dropped below 40, and commercial loans are down 2% year-over-year. UBS notes that this is a demand-driven, chronic employment contraction—more hidden than mass layoffs and harder to reverse quickly.

OBBBA: Short-term support, hard to change the medium-term trend

OBBBA will provide temporary support in 2026. UBS expects roughly $55 billion in tax rebates to be released in Q2, offering a significant lift to consumption that quarter.

However, this stimulus is distinctly "front-loaded." UBS estimates that the positive fiscal policy contribution to GDP will turn negative by 2027, with the deficit ratio still above 6% of GDP. Fiscal policy is more of a stopgap than a new engine of expansion.

Monetary policy walks a tightrope

UBS expects the Fed will cut rates twice in 2026, reducing the policy rate to 3.00%–3.25%. But under tariff-driven inflation, the room for easing is limited and policy differences may widen.

Meanwhile, the Fed has shifted toward expanding its balance sheet, using reinvestment and reserve management purchases to stabilize financial conditions.

Notably, in the context of "AI concentration + cost-push inflation," the rationale for gold allocation has changed. Real interest rates are squeezed, policy visibility is down, and growth is increasingly dependent on a single narrative—transforming gold from a cyclical safe-haven to a structural asset in a high-uncertainty era.

Conclusion

This is not a recession that has already happened, but an expansion period highly dependent on a single engine.

The real test for the US economy over the next 18 months is: Can AI turn from narrative to structure before the next shock arrives?

Risk Disclosure and DisclaimerThe market has risks; investment needs caution. This article does not constitute personal investment advice, nor does it take into account individual users' special investment goals, financial situation, or needs. Users should consider whether any opinions, perspectives, or conclusions herein are suitable for their particular circumstances. Invest accordingly, at one's own risk.