US markets in 2049

Good morning. Walmart reports this morning, and Nvidia tomorrow, a perfect odd couple to bring earnings season to an end. Both should do fine — the digital and analogue economies both appear robust — but a whisper of bad news from either will echo through markets. If you foresee trouble, email us: [email protected] and [email protected].

America’s next quarter century

We have written several times in this space about American exceptionalism. This is the idea that US assets are valued more highly — relative to their current fundamentals — than other kinds of assets, and that they deserve it. The reasons given for paying more for a dollar of US profit than a dollar of international profit include America’s huge internal market, its abundant natural resources, its legal and regulatory regime, its deep and open capital markets, its solid demographic profile, and so on.

What is slightly odd about all these explanations is that they have all been true for a long time, but the American premium has been rising without interruption for almost 15 years. Consider, for example, the ratio of US stock valuations to valuations in the rest of the world:

I was reminded repeatedly of the American exceptionalism debate when I read the long-term asset return report written by Jim Reid and his team at Deutsche Bank. One point they make is that, in the quarter century since the turn of the millennium, US stocks’ real return is a very good, but not spectacular, 4.9 per cent a year. Australia, India, and South Africa have all done better. The reason for this is right there in the chart above: stocks were at a wild peak in valuation in 1999-2000. US stocks are almost as expensive now as they were then. The implication is obvious: the next 25 years of returns probably won’t be stellar.

A more subtle point concerns demographics. Below is Reid’s chart plotting real GDP growth against growth in the working-age population. The correlation is notable.

Over the next 25 years, the US working-age population is not expected to grow anywhere near the 0.8 per cent a year of the last 25, so US real GDP is unlikely to grow at the 2.2 per cent rate. Does this mean the end to US exceptionalism? It depends who you compare to. Working-age population growth will be much worse in Europe, Japan, China, Taiwan, Korea, as well as a surprising number of emerging economies.

GDP is only weakly correlated to risk asset performance, of course. There are lots of other contributing factors. But one of the most important is valuation, and we’ve just seen where we are on that front.

As a side note, while it is true per capita GDP is what matters personally and politically, even per capita GDP is associated with a growing working-age population. The developed countries with per capita real growth rates of less than 1 per cent or below (Japan, Italy, France and the UK) also tended to have slow growth in the working population. It may be that having a growing working population supports productivity.

Another very important point from the study (relevant to the next section of this letter) is that poor real returns for bonds are associated with high debt burdens. Italy, Japan and France have had negative real returns for bonds over the past 10 years, and over the past 100. With the US debt-to-GDP ratio a hair short of 100 per cent, one might wonder about returns from the bond component of a US equity-bond portfolio. The real return from a US 60/40 portfolio over the past 25 years is 4.1 per cent. Don’t pencil that in for 2049. But, again, the fiscal situation looks more dire in many other countries. The US may remain one of the cleaner dirty shirts.

But how dirty, exactly? The only thing that can offset the drags from high valuations, sluggish demographics, and high debt is high productivity. No wonder, then, there is so much hype around the artificial intelligence revolution. If returns over the next 25 years are to remotely resemble those of the past 25, we need AI, or something like it, to work.

The Doge dodge

Elon Musk and Vivek Ramaswamy have been made the co-heads of the Department of Government Efficiency (Doge, named after Musk’s preferred cryptocurrency), a two-year initiative aimed at shrinking the US government. Doge differs from the many past Federal efficiency initiatives in structure and ambition. The pair will sit outside the government, and Musk has claimed Doge will cut $2tn, or about a third, of the federal budget. This won’t happen. Not even close.

The US’s fiscal trajectory is unsustainable, and that is something investors in US Treasuries absolutely need to worry about. The Congressional Budget Office projects the national debt will increase from nearly 100 per cent of GDP in 2024 to 172 per cent in 2054 — with the annual federal deficit growing from $1.7tn in 2023 past $2tn year over year.

Mandatory spending, or payments for benefit programmes already enshrined in the law, accounts for 60 per cent of the budget, or $3.8tn in 2023. Trump’s campaign pledged to not touch Social Security and Medicare, worth $2.1tn in 2023. Government pensions, veterans programmes, and bank deposit insurance, at $500bn, are almost certainly untouchable, too. The remaining $1tn is help for the poor: primarily Medicaid and income supplements such as food stamps. Cuts here may well be politically acceptable to Trump’s constituency. But this would likely require an act of Congress — and with an almost evenly divided House of Representatives, a few Republican defectors could scuttle the plan.

A tenth of the budget goes to repaying the government’s debt obligations. Slashing that would precipitate a default, making it a no-go. Another 20 per cent goes to defence spending. Despite Trump’s isolationist pledges, we did not see any big cuts to defence spending in his first term — we don’t expect to see any in this term, either.

The rest, about $917bn in 2023, is divided among agencies such as housing and urban development, state, and transportation. This is “the administrative state” where Doge will focus most, and where they might get buy-in from Congress. But as pointed out to us by Jennifer Selin, a law professor at Arizona State University, even if Doge completely cuts an agency, or downsizes its staff, “the agencies [would] still have statutory responsibilities, or those responsibilities [would] just get shifted to another agency”. Who will voters blame when, say, highways and mass transit start to fall apart?

Think of it this way. If every civilian employee of the federal government, including all non-military contractors, were fired without severance and replaced with a magically costless artificial intelligence, and every penny of direct federal healthcare and income support for the poor were eliminated, Doge would still be about $300bn short of Musk’s $2tn target.

It is possible that the Doge team finds more savings than we expect, perhaps using novel legal strategies to get around Congress. But taking the knife to “the deep state” will not be enough to meaningfully change the trajectory of the deficit. That will require cuts to Social Security, Medicare, and the military, as well.

(Reiter and Armstrong)

One good read

“The revolutionary technology meant to bypass the establishment has become another product it controls.”

FT Unhedged podcast

Can’t get enough of Unhedged? Listen to our new podcast, for a 15-minute dive into the latest markets news and financial headlines, twice a week. Catch up on past editions of the newsletter here.