In brief
- A soft landing in 2024 would mean no recession but also slower economic growth. While concerns would recede about price inflation, they would also likely shift from idiosyncratic drivers, like supply chains and housing supply, to the more typically cyclical tightening of economic slack, and keeping economic slack from becoming too tight means keeping economic growth at or below potential.
- After a soft landing the economic expansion could last for some time—possibly for some years. Apart from inflation, there are none of the usual historical recession culprits looming—such as private sector exuberance, high leverage and excess investment.
- Growth could still surprise to the upside, in the event of a productivity boost—including one coming from artificial intelligence. Boosting productivity could also ease supply constraints, helping dampen inflation and further extending the cycle.
- Cyclical extension would benefit our equities overweight. We would expect a broadening from initial beneficiaries—technology and other quality growth sectors—to cyclicals. AI-driven productivity gains could ultimately drive broad margin expansion.
A soft landing, and comfortably avoiding recession, is now our base case for the U.S. economy in 2024. We see U.S. growth dipping a little below trend and the unemployment rate rising only modestly, which should help core CPI inflation fade meaningfully and sustainably below 3%.
The economy is already well on its way: Underlying growth, though modestly above trend—as reflected by Q4 GDP data this week—continues to decelerate toward trend at a controlled pace, and inflationary pressures are fading significantly. To be sure, the economy hasn’t landed quite yet, and we still expect some bumps on the way there. But it’s a good time to start asking: If the economy makes a soft landing, what happens next?
A soft landing means slower growth
First, a U.S. economy that manages a soft landing wouldn’t reaccelerate into full-throttle liftoff. To the contrary, sustaining the expansion—and moderate inflation—would require real growth to decelerate from the above-potential pace of the last three years.
That above-potential pace largely owed to the rapid (re)creation of lots of jobs. Today, nonfarm payroll expansion is trending around 160,000 jobs per month, while growth in the working age population might support just 100,000 jobs. Each month that employment grows faster than this “breakeven” pace further reduces economic slack, which may already be a bit too tight: The current unemployment rate of 3.7% has tended historically to coincide with inflation creeping above the Federal Reserve (Fed) target—even during the prior expansion, which was marked by notably low average inflation.
Creeping above target doesn’t mean surging. Although pro-worker actions in the U.S. last year made headlines, the number of work stoppage days was comparable to 2019, while union membership in 2023 hit a record low of just 10.0% of workers. The labor actions, moreover, mostly sought to protect worker livelihoods from soaring inflation. Since early 2020, U.S. workers’ real wage gains have significantly decelerated from their pre-pandemic trend—lagging inflation by some measures.
Our base case this year therefore expects that a steady expansion of the working-age population (with some help from immigration), and increased productivity, will allow the economy to eke out just under 2.0% real GDP growth, slightly below trend—down from 2.5% last year.
More years of expansion?
If the bad news is slower growth, the good news is fading near-term odds of recession and the possibility of a significantly longer economic cycle. If the U.S. avoids a recession in 2024, the downturn doesn’t automatically get bumped to 2025. We see a long runway—of years, potentially—for a continued expansion.
Most U.S. postwar recessions were caused by some form of overheating: market price bubbles, rapid buildups of private sector leverage, unsustainable patterns of investment and consumption or simply excess inflation. For the last several years, our Business Cycle Scorecard has tracked the economic and market indicators that tend to coincide with cyclical progression and ultimately lead to recession (Exhibit 1).
In the current cycle, inflation has been the only potential source of overheating. If inflation can be managed, we foresee no other ultimate cause for a recession. The private sector has meaningfully deleveraged in this cycle, thanks in large part to fiscal transfers in 2020–21, on top of deleveraging after the global financial crisis. Household and business sentiment and major spending plans remain relatively subdued.
That lack of exuberance has so far kept consumption and investment patterns more sustainable, relative to the end of past cycles. Consider: Real residential fixed investment has already contracted nearly 20% since Q1 2021 (its share of GDP is now half what it was in 2005). Homebuilding activity now sits below our estimate of its cycle-neutral level, leading us to expect some reacceleration in the months ahead, as rates decline.
One caveat regarding a potentially extended cycle: In addition to slower growth, a low rate of personal saving means household spending would be less buffered against any disruption in real incomes. Thus, even as we see no clear recession catalysts on the horizon, the economy could be more susceptible than usual to unforeseeable shocks.
Squeezing out more upside
When businesses can’t find new workers, they’re likely to focus on boosting real output per worker. Now is an opportune time, given the emerging potential for investments in artificial intelligence (AI) to boost labor productivity. Over the next 10 years, we estimate AI could potentially add up to 1.4%–2.7% to real growth.1
Although the timing of AI’s meaningful effects is highly uncertain, anecdotal evidence of generative AI’s real-world implications already includes studies showing everyday workers can reduce time spent on common writing tasks by as much as 40% while improving quality, and that nurse practitioners can take on some advanced clinical processes and decision-making from primary care physicians. Early data also suggests labor productivity picked up in 2023. In addition to boosting productivity, AI automation might help extend the cycle by loosening supply constraints—echoing the internet-driven extended late-cycle period of the late 1990s.
The sectoral composition of the workforce also drives changes in average labor productivity. The economic value added per hour by employees in the information sector, for instance, is many times greater than for most other service workers.
In the coming years, output might pivot from sectors with lower labor productivity toward those with higher labor productivity, such as manufacturing, finance and information. This pivot could be fueled, in part, by the buildup of economic excesses that have been unusually absent from this cycle; and by secular drivers, including AI, the reshoring of some manufacturing and the greening of the economy.
What would an extended cycle mean for markets?
An extension of the economic cycle should be positive for risk assets such as equities, but after the landing, the underlying drivers of returns may differ from the current landing phase.
During the landing phase, we expect U.S. equity returns to be driven by earnings growth, rather than expanding valuations. While equity valuations are not at historical extremes, they are already elevated relative to history and unlikely to expand, particularly as economic growth slows. In addition, we expect the quality growth segments of the market, such as technology, to lead. As the economy takes a potentially bumpy path to further disinflation, accompanied by monetary policy uncertainty, investors are likely to favor companies that provide both fast earnings growth and strong balance sheets.
After the landing, cyclical extension would be characterized by a rise in capex, so we would expect performance to broaden from technology and other quality growth to those companies most geared to the economic cycle. Beneficiaries by sector would likely include industrials and by region, the eurozone. Select cyclical equities also trade at relatively low valuations. Their rerating could lift index-level valuations.
An extension of the cycle characterized by accelerating productivity because of AI should benefit both AI providers (such as the Magnificent 7) and in the medium term, AI adopters. In this scenario, margins would expand across the board, particularly for companies benefitting most from AI automation.
Asset class implications
Our current expectation of moderate growth, cooling inflation and less restrictive policy should generally support asset markets in 2024. We are modestly overweight equities. Along with U.S. equities, Japanese stocks are especially attractive, given rerating potential driven by corporate governance reforms. We maintain a constructive view on duration and continue to find pockets of opportunity in credit—although spreads are historically tight, carry is still attractive and we see opportunities in securitized credit, shorter-dated crossover credits and high yield. Asynchronous rolling slowdowns should produce tradable differences among sectors and regions as well as a good outlook for stock selection alpha.