Higher for longer: Revisiting our rates thesis from last year

Now that the convulsions in the bond markets have subsided and we can breathe again, it is time to revisit our “higher-for-longer” thesis.

G7 bond yields have risen sharply in recent months, with 10-year US Treasuries approaching their 2023 highs. Despite the Fed's recent decision to pause rate hikes, most market participants anticipate at least two rate cuts this year.

??We maintain that even with one or two potential rate cuts, Fed rates are likely to remain elevated for an extended period. In a worst-case scenario, rates might even need to increase by 2026.

In a post on Feb 13, 2024, we posited the theory that the Fed was unlikely to be able to cut rates by too much. The Fed funds rate was likely to stay higher for longer, with the new lows now much higher than those seen over the past decade or so.

??Our original thesis was predicated on the strong underlying fundamentals of the U.S. economy; more specifically the number of jobs that are likely to be created from increased capex on new manufacturing capacities in the U.S.

Below, we'll discuss this thesis with updated data and examine additional factors influencing interest rates.

Updated data holds up: Factory capex still running at 3X the usual level

??Analysis of monthly data on U.S. manufacturing construction spending (Exhibit below) indicates a continued upward trend since our last update in Feb 2024. The peak spending was recorded in June 2024, reaching an annualized monthly rate of US$238.4bn. This monthly spending has remained similarly high in subsequent months, with the most recent data for November 2024, showing a spending of US$236bn. ?

So, it is still well over three times of the usual capex levels before the IIJA, the IRA and the CHIPS Acts, etc caused the U.S. factory expansion to take off in the first place (Exhibit below).

Perhaps this manufacturing capex is peaking? The monthly figures do suggest this as the figures rose to a high in June 2024 and have plateaued at that level ever since.

The monthly jobs data is not showing great gains for manufacturing. Yet.

The latest jobs data out on Jan 10 were strong enough to trigger a rise in yields again, although the subsequent CPI data tempered inflation expectations and brought some calm to the markets. The jobs data, however, showed a strong growth in services while manufacturing hiring was somewhat subdued.

There are several factors behind this:

• Manufacturing job lag: Large factories take at least 2-3 years to construct, and even after they are ready, they may take up to 3 years again to reach full 100% utilization rates. Only by then will they have hired all the workers they need to hire. Consequently, manufacturing jobs growth may only become evident 3 to 6 years after the initial capex growth. Given that manufacturing capex growth accelerated in late 2021 (see first Exhibit), we may begin to observe manufacturing job gains in 2025.
• Interim services sector growth: While these factories are under construction, the jobs they are creating are primarily in services sectors, such as construction, transportation and logistics. As these are sectors that are classified under “services” sectors rather than under “manufacturing”, it may appear as if services jobs are growing, but not manufacturing jobs. Thus it may be misleading to think that all this manufacturing capex is not leading to any growth in manufacturing jobs.
• Job multiplier effect: In the earlier Feb 2024 note, we had mentioned the Job Multiplier for manufacturing jobs as 7.4: that is, each new manufacturing job tends to create, on average, more than 7 additional jobs. It is worth noting that these additional jobs may not necessarily be in manufacturing itself, with many in ‘services” – in businesses that provide services to these new manufacturers (such as vendors, transportation and logistics companies).
• Economic impact of services jobs: Many of these new service sector jobs may be in lower-productivity areas like transportation and retail, rather than high-productivity sectors like financial services. Workers in these jobs tend to have a higher propensity to consume, rather than save, their marginal earnings. This could potentially boost consumption growth and, by extension, inflation.

Nevertheless the basic thesis remains intact: new manufacturing jobs will be created as the factories currently under construction come online. These, in turn, will create additional manufacturing and service-sector jobs. Collectively, this job creation will keep consumption humming. All in all, this should keep interest rates elevated for longer.

There is also the specter of rising US Treasury issuances

Not only has the pace of issuance risen along with the deficits, the US Treasury has been relying more and more on short-term debt funding to finance the deficit. As of December 2024, Treasury bills constituted approximately 22% of the government's debt stock, up from 15% two years before.

During this period of rising Treasury supply and concerns over the deficit, there has also been a drop in demand from many traditional buyers such sovereign states and banks. This imbalance itself could see rates ticking higher to entice buyers back to the market. (There is, of course, another possibility: the overlord of the Department of Government Efficiency manages to cut so much spending so quickly that there is no deficit left to be financed). ?

Gross quarterly U.S. Treasury issuance (US$ tn):

What could go wrong with this thesis?

• Policy shift: The new Trump administration appears to be taking a different “stick” approach to bring manufacturing back to the US, as opposed to the “carrot” approach of the Biden administration. It is very likely that some of the signature incentives being provided to manufacturers (the IIJA Act, the IRA Act) referenced in our Feb 2024 note will be scaled back by President Trump. In its place will come tariffs, which may or may not create manufacturing jobs in the short run.
• Targeted industries may create fewer jobs. Some of the incentives were for the energy transition industries - something the new administration has announced it will no longer support. Though some of these factories are already under construction, it is unclear if they will have funds available to become operational and thus create as many jobs as was expected.
• Impact of immigration. One of the reasons to expect a manufacturing-driven jobs growth was that there had been a drop in US labour force during the 2020-21 covid period. There was also a drop in both authorised and unauthorised immigrants. The immigration numbers have picked up since then (chart below), and should they continue to grow, despite the efforts of the new Trump administration, the resulting growth in the labour pool could keep wages in check. (If, on the other hand, the new administration does manage to lower immigration levels, it could create an environment where jobs and wages are both growing. This would create a goldilocks scenario for our thesis.)

• Productivity gains could undo a part of our thesis. The U.S. labour force has been unusually productive. Productivity growth has been strong (exhibit below) and higher than in most other OECD countries. And all this was achieved even before potential productivity gains kick in from wider adoption in new AI technologies, the effects of which could further boost U.S. labour productivity in the coming years. Should growth in productivity continue, it could lead to lower inflation and hence, lower rates. ?

Labour Productivity: Output per Hour (% change from quarter a year ago):?

?

?