Growth in the United States is expected to come close to its potential pace in 2026. This resilience would mask “K-shaped growth”, supported by AI-optimism related investment and consumption by the wealthiest. Investment in other areas of the economy is not as dynamic, while most Americans face persistent inflation and a deteriorating labour market. At the end of Q1 2026, the Fed is expected to end its cycle of monetary easing, due to an emphasis on the employment component of its dual mandate. The fiscal impulse is expected to remain slightly negative in 2026 due to tariffs, with their scope still a key issue.
Growth supported by optimism around AI
Growth experienced unusual variability in 2025 due to temporary factors: volatility in goods imports (H1 2025) and the federal government shutdown (Q4 2025). In our view, the latter reduced Q4 2025 growth to +0.0% q/q (a loss that will be recouped later) against +0.8% q/q in Q3[1]. GDP growth is expected to be more stable in 2026, at around +0.4% q/q per quarter (from Q2), with an annual average of +1.9%, as in 2025. Therefore, growth would be close to its potential pace, but characterised by a change in its drivers and a moderation in employment.
GROWTH AND INFLATION (YEARLY AVERAGE)The dynamism of investment against a backdrop of AI-optimism is the distinctive feature of growth. At the same time, other components of investment have declined (Chart 2). In H1 2025, AI-related categories accounted for 9% of GDP, but drove 40% of its growth. However, this was accompanied by an increase in related imports, putting the previous result into perspective and highlighting the United States' external dependence in this area. In addition, the valuation of the tech sector's "Magnificent Seven"[2] has contributed significantly to the recent increase in household financial wealth (+30ppsince Q3 2023, to 420% of GDP in Q2 2025), which supports consumption among those households that hold shares.
UNITED STATES - GDP: A K-SHAPED FIXED INVESTMENT (100 = Q1 2023)US growth is therefore “K-shaped”. The divergence in investment is likely to continue in 2026–27, and household consumption should continue to be supported by the top end of the income and wealth distribution (see our EcoInsight on Consumption). Household sentiment reflects these contrasts, as it remains particularly weak at an aggregate level (at its lowest since 2015), with a widening gap between the top income segment (USD > 120,000/year) and almost the entire rest of the distribution in 2025, according to the Conference Board. Growth is therefore vulnerable to a sharp correction in stock market valuations, which are historically high.
Decline in household and corporate debt ratios
Household debt fell to 93.6% of GDP in Q2 2025, its lowest level in over twenty years. Corporate debt also declined, to 46.1%, but remained close to its historic highs (excluding COVID). Growth in outstanding household mortgage loans remained fairly strong (+2.9% y/y in H1 2025), but property purchasing power remained weak (high interest rates and prices). Although 30-year fixed rates (90% of new lending) have fallen since May (-65 bp to 6.3% in November), this trend is unlikely to continue, given our scenario of rising long-term sovereign rates (see below). Refinancing demand will also remain modest, as the vast majority of loans were taken out at fixed rates below 6% (78% of outstanding loans in Q2 2025). Demand for consumer credit, which was virtually stable y/y in September, remains depressed. Conversely, corporate financing flows (new credit production, and debt securities issues) increased in H1 2025, helping to support investment. This trend could continue, driven by lower short-term rates.
Employment: less demand, less supply
In 2026, the unemployment rate is expected to be around 4.4%, close to full employment. Job growth has declined (non-farm payrolls averaged +59k in September 2025, compared with +133k a year earlier), and the annual revision confirms this trend (-911k jobs between April 2024 and March 2025). The available labour force is also declining with the administration's "net zero immigration" goal. This is resulting in a paradoxical situation where the downturn in hiring is not leading to higher unemployment, but where potential growth is being reduced by immigration policy.
UNITED STATES: AI-RELATED EMPLOYMENT IS NOT SURGINGThe support provided by AI-optimism to growth is not having a significant impact on employment, either directly or indirectly. Payroll employment has even declined in the relevant sectors since January 2023 (Chart 3). In addition, the labour market is mirroring the "K-shaped" growth pattern. As a fact, in Q3 2025, the wage-growth differential between the first and fourth quartiles reached its highest level since 2013 (Atlanta Fed). The associated challenges in terms of household consumption are more significant due to stable aggregate wage growth (+3.8% y/y in September, stable in 2025), the rebound in inflation and increased household fears about employment.
Inflation: no return to target before 2028
Inflation (CPI) is expected to exceed +2% y/y until at least 2028, according to our forecasts, peaking at +3.3% y/y in mid-2026 for both headline and core indices. The rise began in April 2025, with the CPI rising from +2.3% to +3.0% y/y in September (+2.8% to +3.0% for core). Additional customs duties are the main factor, due to the increase in the average effective tariff on imports (15.9%, compared to 2.3% in 2024, although the actual revenue collected currently stands at 11.5%). However, the pass-through to consumer prices has so far been weaker than expected. The deterioration in household prospects may have led to a greater absorption of tariff increases by businesses. Nevertheless, disinflation is expected to be slow after the peak in mid-2026, with the CPI reaching +2.5% y/y in Q4 2027, above the "pre-tariff" level of April 2025.
The Fed faces the challenge of neutrality
In September, the Federal Reserve (Fed) resumed its monetary easing policy due to the rebalancing of risks between employment and inflation, cutting rates three times (-25 bp) and reducing its restrictive bias. We anticipate a final cut in March 2026, with a target rate of 3.25% - 3.50% (slightly below our estimate of the neutral rate of 3.75%). The risks are skewed towards further easing, given the bias towards employment among the most influential members of the FOMC. Furthermore, the committee is very divided on rate cuts against a backdrop of persistent inflation. At the same time, the USD is expected to weaken moderately against the euro (1.20 at the end of 2026, compared with 1.16 today) against a backdrop of stronger-than-expected growth in the Eurozone (see chapter) and a narrowing of the policy rate differential.
The Supreme Court (SCOTUS) ruling in Trump v. Cook and the replacement of Chair Jerome Powell are major issues for the future of monetary policy. A ruling in favour of the administration would grant the US President broad discretionary power over the composition of the FOMC, while the future Fed Chair could be chosen for their bias in favour of easing. Another risk is the possibility of resignations from the board, including by Jerome Powell, which could lead to a more profound reshuffle, opening the door to a change in institutional structure and a significant loss of independence. This could result in an erosion of the Fed's credibility and its commitment to price stability. However, this is not our scenario (see our Editorial, Fed: Powell's legacy should endure): decisions would continue to be guided by economic fundamentals, despite a persistent imbalance in favour of the "employment" part of the dual mandate.
A moderately negative fiscal impulse
The total fiscal deficit is expected to reach 6.5% of GDP in 2025 and 6.3% in 2026, which are particularly high levels, but down from 7.0% in 2024, due to the student-loan reform and higher customs revenues. The latter should help to reduce the primary deficit, which is expected to fluctuate between 1.0% and 1.5% of GDP between 2025 and 2027, but remains volatile (recent exemptions on food) and uncertain (see below). At the same time, interest payments are weighing on the total deficit (5.0% of GDP in 2026, rising steadily since 2020). The fiscal stance has negative distributional effects on growth and contributes to the "K-shaped" economy: customs tariffs are borne by the lowest-income households, and the One Big Beautiful Bill Act (OBBBA) essentially extends favourable taxation for high incomes. As a result, the public debt ratio is expected to continue rising (126.4% of GDP in 2027, compared to 124.7% in 2025), but will be limited by continued high nominal growth. Furthermore, there is a high risk of another government shutdown.
Deficit reduction and FOMC rate cuts are unlikely to stem the rise in bond pressure in 2026. According to our projections, 10-year and 30-year rates will reach 4.5% and 5.0%, respectively, in Q4 2026, compared with 4.2% and 4.8% on 9 December. Persistent inflation and deficits, combined with better-than-expected growth, have contributed to this forecast. Furthermore, Q3 saw a moderation in the absorption of US debt by foreign investors, whose demand increased in volume but moderated as a proportion of net Treasury issuances, which were particularly significant after the debt ceiling was raised this summer.
Foreign trade: uncertainty has receded, but remains significant
Data available for Q3 (July and August) put the trade balance at an average monthly USD -68.9 billion, compared with USD -75.2 billion in 2024. Imports are in line with Q2, which was marked by a sharp backlash following frontloading in Q1, and exports have been generally stable since January. A relative balance appears to have been reached on the issue of customs tariffs, with a one-year truce (until 10 November 2026) with China. Trade uncertainty has therefore declined significantly since its peak in spring 2025, but remains well above pre-tariff levels. However, there is a strong possibility that reciprocal tariffs will be invalidated by the WTO, opening the door to more sectoral countervailing tariffs. Finally, other sectoral tariffs are possible, particularly on semiconductors and pharmaceuticals.
Article completed on 11 December 2025
