US Economy Is Probably Not Accelerating. Does This Matter for Markets?

Summary

• Bottom-up and top-down analysis suggest the US economy is not accelerating, despite recently better data.
• This makes little difference to markets as equity markets are more driven by AI sentiment than by GDP growth. Even if I am wrong and growth is accelerating, the acceleration is unlikely to be strong enough to fundamentally change the macro picture.

Market Implications

• I still expect two-three 2025 cuts and at most one 2026 cut against markets pricing two and three cuts, respectively.

Bottom Up: Recession Risks Still Rising

Here, I argue the US economy is probably not accelerating but that it matters little for market performance.

Recently positive data including a pickup in the economic surprise index and a strong July consumption print have led to perceptions the US economy is reaccelerating (Charts 1 and 2).

Bottom-up (= data-based) and top-down (= policy shocks-based) analysis show growth is more likely to remain near or below H1 1.4% average than to accelerate over the remainder of 2025.

Bottom-up indicators show rising recession risks but generally remain below the recession triggers.

I think the strongest recession indicator is based on the second derivative of unemployment. In all recessions the unemployment increase is nonlinear, likely due to negative feedback loops, for instance between employment, household income and demand for businesses products (Chart 3). These feedback loops take root once an economic slowdown crosses a threshold.

Measuring the speed of increase in unemployment is the essence of Sahm’s rule, which is based on monthly unemployment from the household survey or of the Richmond Fed’s SOS indicator, which is based on the weekly insured unemployment rate. I prefer the latter that has a stronger track record and is available weekly. It is currently rising but remains well below the recession trigger (Chart 4).

Sahm’s rule and the SOS indicators (like most quantitative indicators) must be put in context. In 2023, both gave false positives due to labour market normalisation from extreme labour shortages and historically low unemployment.

Two more recession indicators indicate rising risks. The University of Michigan survey breaks down consumer perceptions of their current situation by income third (Chart 5). During expansions, upper income consumers are more positive on the economy than lower income consumers. By contrast, at the onset of recession upper income consumers perception tend to converge to that of lower income consumers.

The intuition behind the indicator is that higher wage employment (e.g., for college graduates) tends to be more recession proof than lower wage employment. So, for upper income consumers to become as negative as lower income consumers the economy must be experiencing strong downward momentum.

Upper income consumers perceptions have recently worsened, which could signal recession. This data is confirmed by corporate earnings and media reports showing weaker middle class spending.

Lastly, the Conference Board leading indicators (LI) have been hitting their recession trigger for the past three months (Chart 6). The indicator has a strong track record (dating back to the 1970s), except for 2022-23 when it gave a false positive. The false positive largely reflected that the yield curve, a key constituent of the LI, had inverted. But in 2022-23, I think the inversion did not signal recession but rather a transition to a higher inflation regime. This time the yield curve (10Y/FFR spread) is flat and is therefore unlikely to be a big driver of the LI.

Chart 1: Unclear If Consumption Is Rising	Chart 2: Economy Surprises Are Not GDP
Chart 3: UR Speed Strong Recession Signal	Chart 4: Rising Recession Risks
Chart 5: Gloomier Upper Income Consumers	Chart 6: LI Hit Recession Trigger!

Overall, bottom-up recession indicators signal an economy that is still losing momentum. A top-down discussion of policy shocks shows why that is.

Top Down: Fiscal Consolidation, Consumer Deleveraging Hits Growth

Historically, US recessions have reflected either restrictive fiscal and monetary policies (up to the mid-1990s) or balance sheet fragility (from the mid-1990s). The Trump administration has added tightening through policy uncertainty. Policy uncertainty measures soared after ‘Liberation Day’, with the St Louis Fed concluding such extreme uncertainty could trigger a recession.

However, since then policy uncertainty has decreased and private confidence has stabilised, though it remains well below the post-elections high. But the US economy has been hit by new shocks.

The administration has tightened fiscal policy since January, mainly through rescissions (Chart 7). As a result, the 12m deficit has shrunk to 6.5% of GDP in July from 7.3% in January.

And while the Fed has cut 100bp in 2024, transmission to the real economy has been limited. The 1yr lookback Fed FCI is less supportive of growth now than before the Fed’s cuts (Chart 8).

Also, consumer borrowing rates have fallen by less than the FFR cuts and mortgage rates are now higher than before last year’s Fed cuts (Chart 9). The limited passthrough from the FFR to household lending rates explains why households’ debt service ratio is almost back to pre-pandemic level even though household leverage is significantly lower (Charts 10 and 11).

Household deleveraging is also likely to reflect households’ efforts to strengthen their balance sheets and improve creditworthiness. Lower delinquencies shows this (Chart 12). Long-term, this is a positive development for financial stability but is likely to weigh on growth short term.

Chart 7: Policy Uncertainty Remains High	Chart 8: Private Confidence Remains Weak
Chart 9: De Facto Fiscal Tightening	Chart 10: Poor Transmission of Fed Easing
Chart 11: HH Lending Rates Remain High	Chart 12: HH DSR Rising
Chart 13: HH Are Deleveraging	Chart 14: HH Credit Quality Improving

Bottom-up and top-down analysis suggest the US economy is unlikely to be re-accelerating. That said, this will make little difference to market performance.

Growth to Have Limited Market Impact

Equity market performance is currently more driven by AI sentiment rather than GDP growth (Chart 15). GDP growth would impact equity markets indirectly if it led the Fed to cut less than expected (i.e., two cuts in 2025). This would also lead to curve flattening.

This seems unlikely. Unemployment would have to fall well below the current rate of 4.2% for the Fed not to cut this year. Also, the Fed would have to become concerned that tariffs are triggering a lasting inflation increase. In turn, this would manifest itself through inflation contagion from tariffed to non-tariffed goods and to services, wage growth acceleration, or a steady rise in long-term BEs, exactly the opposite of what we have seen so far.

Such a reversal of current data patterns would require growth to accelerate above trend, which would be unlikely to happen without a new policy shock. For instance, a fiscal stimulus. But that is more of a 2026 risk.

I expect growth in 2025 to remain well below trend of about 2%, and for the Fed to cut 25bp two to three times at the September, December and possibly the October FOMCs.

The Fed would be much more aggressive with cuts if recession risks rise materially (the speed of the unemployment rise is key here), but that is not my base case scenario.

I still expect two-three 2025 cuts and at most one 2026 cut against markets pricing two and three cuts, respectively.

Chart 15: AI Sentiment Not GDP Driving SPX	Chart 16: Weak Economy Helps Disinflation

The commentary contained in the above article does not constitute an offer or a solicitation, or a recommendation to implement or liquidate an investment or to carry out any other transaction. It should not be used as a basis for any investment decision or other decision. Any investment decision should be based on appropriate professional advice specific to your needs. You are not permitted to publish, transmit, or otherwise reproduce this information, in whole or in part, in any format to any third party without the express written consent of Macro Hive. This includes providing or reproducing this information, in whole or in part, as a prompt.)