It appears to me that the post-WWII order which shaped global trade, capital flows, and monetary policy for decades is beginning to unravel. Not because one election changed the world overnight, but because the incentives that sustained the old system have gradually eroded and are now fraying.

There are plenty of ways to tell the history of the postwar order. I am not interested in its complete history end to end, but in one particular feature of it: the alignment of incentives that allowed an enormous global financial and trading system to organize itself around the United States and its currency.

The old system

In the wake of World War II, the United States stood as the globe’s sole solvent superpower, owning much of the world’s industrial capacity and monetary gold. The new monetary system agreed to in 1944 at Bretton Woods placed the U.S. dollar at its center, which laid the groundwork for what has been called Pax Americana: the period of relative peace among the major powers of the U.S.-led order that emerged after World War II.

This relative peace was in some parts an illusion, as much of American statecraft and military policy was preoccupied by the Cold War from the late 1940s until the Soviet Union fell in 1991. But for nations outside the Soviet sphere of influence, Pax Americana saw the U.S. providing security guarantees, mediation of geopolitical disputes, and monitoring of sea lanes backed by the world’s most powerful military. Under that security umbrella, trade barriers fell, capital flowed increasingly freely across borders, supply chains prioritized efficiency over resilience, and many export-oriented economies increasingly sold into U.S. demand while recycling part of their accumulated dollar surpluses into U.S. assets.

The mechanics of this system evolved quite a bit over the decades. Bretton Woods initially linked the dollar to gold at a fixed exchange rate, but rising overseas balances increasingly strained America’s ability to simultaneously maintain dollar-gold convertibility and run independent fiscal/monetary policies, leading to Nixon’s closure of the gold window in August 1971.1 What came afterward was an evolution of the dollar-based system: first through the recycling of oil exporter surpluses starting in the 1970s, then later through the gigantic reserve accumulation of East Asian economies (especially China) following the Asian Financial Crisis of the late 1990s. The sources of surplus capital changed, as did the plumbing (gold convertibility ended), but the broader pattern endured: foreign dollar surpluses continued finding their way back into American financial assets, most importantly government bonds.

The economic order that followed Bretton Woods didn’t develop separately from this American-led security architecture. American military and economic dominance helped keep trade routes relatively secure, allowed allies to devote fewer resources to their own defense, and gave countries room to organize their economies around international, rather than purely domestic, considerations. In that sense, the globalization that followed was partly a product of the political order beneath it.

This post-war arrangement, which participating nations had powerful incentives to sustain, made the dollar - and the financial, cultural, and economic system surrounding it - one of America’s most consequential exports. In practice, it meant the U.S. could run persistent trade and fiscal deficits with fewer immediate constraints than other countries faced, sending dollars abroad and absorbing excess savings in the form of foreign appetite for its bonds and equities.

It worked because each participant in the system solved someone else’s problem: America received financing; exporters received demand and security; and investors received deep, liquid markets.

And for a time, it was good.

Why other nations participated

A monetary system gains its strength from network effects. More participants trading in a currency means more demand for that currency, a rising share of trade and liabilities get denominated in it, and the system begins to feed on its own momentum. With the U.S. providing security over shipping lanes and encouraging trade barriers to fall between non-Soviet aligned countries, trade cooperation rose and dollar demand increasingly became self-reinforcing. It went something like this:

The system endured not because it was perfectly fair, but because every participant had a reason to keep playing. The arrangement gave both sides powerful reasons to sustain it: export-driven economies gained a safe, liquid home for their growing dollar reserves and secure passage for their goods abroad, while the United States received stronger structural demand for its currency and assets, unusually deep financial markets, and a greater capacity to run persistent external deficits at lower financing costs than otherwise.

The system also grew far beyond dollars physically leaving the United States through trade. Beginning in the postwar decades, banks outside the U.S. increasingly accepted dollar deposits and extended dollar-denominated loans: the somewhat oddly named Eurodollar market. That mattered because the dollar could now be created and lent outside the American banking system. A borrower in London, Tokyo, or elsewhere who owed dollars also needed to acquire future dollars to service that debt. This created a self-reinforcing process: the more the world borrowed, traded, and saved in dollars, the greater its need for dollar liquidity.2

Foreign governments were not buying U.S. assets to expand American hegemony. They were acting in their own interests. Exporters wanted American demand and competitive exchange rates. Banks wanted dollar funding. Companies and investors wanted access to the world’s deepest capital markets. Other countries benefited from an increasingly open trading system and, in many cases, lower security burdens under the American umbrella. Americans got cheaper imports, American corporations got efficient global supply chains and higher margins, and Wall Street sat at the gooey center of an enormous global financial system. That these separate incentives also financed American deficits and reinforced U.S. power was a consequence of the system, not its primary objective.

What the system has wrought

No system is without tradeoffs. So far we have focused mostly on the financial benefits of the arrangement for the United States. However, the system also affected America’s productive economy.

Some erosion of America's extraordinary postwar industrial dominance was probably inevitable. The United States emerged from World War II in an almost unnaturally advantageous position: much of its competitors' productive capacity had been destroyed while its own industrial base had expanded. As Europe rebuilt and Japan and other Asian economies industrialized, that gap narrowed. American manufacturers increasingly faced capable foreign competitors.

Throughout this process, the U.S. increasingly accepted persistent trade deficits while embracing open trade and globally optimized supply chains. Corporations had strong incentives to move production toward wherever goods could be produced most efficiently. American consumers benefited enormously from cheaper imports, while American firms benefited from lower production costs and access to increasingly sophisticated foreign supply chains.

In that environment, the location of productive capacity increasingly became an economic optimization problem rather than a strategic one. The question was less Where should this be made? and more Where can this be made most efficiently? Over time, however, that meant accepting that some domestic productive capacity would migrate elsewhere.

Then the scale changed. The collapse of communism integrated an enormous new pool of workers into the global trading system, most consequentially in China. A 2015 BIS paper calculated that the combined labor forces of the former Soviet states and China increased the supply of labor available for global production by roughly 120%.3 China’s 2001 entry into the World Trade Organization accelerated the process, connecting its enormous labor force more deeply with Western capital and consumer demand.

China was not merely a source of inexpensive labor. It pursued industrialization as a national development strategy, building infrastructure and output capacity at extraordinary speed. For Western corporations already incentivized to optimize production globally, China offered a combination of labor cost, scale, infrastructure, and manufacturing depth that became difficult to ignore. The addition of so much relatively inexpensive labor put enormous competitive pressure on manufacturing workers in developed economies, while production and investment increasingly flowed toward China.

But China was not simply absorbing Western manufacturing capacity. It was also becoming one of the largest participants in the dollar flywheel described above. American demand for Chinese goods helped generate large Chinese trade surpluses and growing dollar reserves, some of which were recycled into U.S. financial assets. That recycling supported the financial side of the same arrangement: Americans continued consuming inexpensive imports while China continued building productive capacity around Western demand. Each side was responding to its own incentives. All of this worked not because Beijing was trying to finance American power out of generosity, but because the arrangement served China’s own development strategy.

The bargain was functioning extremely well.

It worked so well that it ultimately changed the balance of power upon which the bargain rested.

The system is unraveling

That is the point where I think the old bargain begins to come apart. Once China grew from being a participant in the U.S.-led system to being its principal economic and strategic rival, questions that had been primarily economic started taking on a national security tone. Where semiconductors are fabricated matters. Who builds ships matters. Where critical supply chains are located matters. Productive capacity itself starts to look less like an output of comparative advantage and more like a form of national power. The ethos of Pax Americana isn’t necessarily gone, but the assumptions beneath it are changing along with the old bargain it supported.

What happens when the country at the center of that system decides that some of the things it spent decades encouraging—open trade, globally optimized supply chains, and relatively free movement of capital—are now weakening its own strategic position?

And this is not only a problem for the United States and China. America’s allies built their own economic models around many of the same assumptions.

More in Part 2…

Author’s Note: This essay is the first in a series exploring how the post-war economic status quo is changing. Subsequent parts of the series will be spread among future issues. If you’d like to follow the series or stay in the loop about any of my research, please consider subscribing.

Subscribe now

A lot more debt, but not a lot more duration

Simply put, the government needs a shitload of money. The latest Treasury estimates of borrowing for Q3 and Q4 of 2026 are $739 billion and $628 billion respectively. That is $1.37 trillion in six months. And the Q3 figure is more than 10% higher than was projected back in May.1

The Treasury Department’s latest Monthly Treasury Statement showed that the federal government, in just the first 10 months of fiscal year 2026 (through July), borrowed more than the entirety of fiscal year 2025. $432 billion of borrowing in July alone.

The government is borrowing dramatically more money, but the scheduled size of coupon auctions2 is exactly where it was more than two years ago. Something else has to absorb the financing requirement. For the past few years, that something has been bills.

This policy began as a necessity with Janet Yellen in summer 2023 after the debt ceiling was suspended and Treasury needed to rebuild its cash balance. But by the November refunding of that year, the strategy began to look less temporary. With long-term yields rising and investors demanding greater compensation for duration, Treasury slowed increases at the long end while continuing to use bills as the marginal financing valve. This was arguably the first modern episode in which Treasury was explicitly confronted with the problem of enormous financing needs + increasingly price-sensitive duration buyers. The bill-heavy policy has continued with Scott Bessent at the helm.

An interesting thing happened this week

Two events hit the news wire this week that together made me go ‘hmm’. First, the 30-year yield hit 5.33%, the highest level since June 2007. One day later, the Treasury announced they were doubling the size of their buyback operations at the long end of the yield curve.3

Higher yields are the market’s mechanism for finding buyers. There is no literal shortage of buyers. At a sufficiently high yield, buyers will appear. The question for Treasury is how high that yield needs to go before the amount of duration they need to sell clears the market, and what those higher yields do to government financing costs and the economy. They can tweak the amount of duration that needs to find a clearing yield in the market by financing relatively more (or less) at the short end, as they’ve been doing.

Both bills and long-term coupons finance the same dollar of government spending, but they place very different risks into private hands. A bill is a very short-term cash-like instrument. A 30-year bond, on the other hand, asks someone to warehouse decades of duration risk and inflation risk.

All of this matters because the long end of the Treasury market does not exist in a vacuum. Government bonds provide the benchmark against which mortgages, corporate debt, and countless other assets are priced. Higher long-term Treasury yields raise financing costs across the economy while gradually raising the government’s own interest burden. And, if they’re too high, they can crowd out meaningful investment. With Washington borrowing trillions of dollars, how much long-term debt Treasury asks markets to absorb (and at what price) is becoming increasingly important.

What are buybacks?

Treasury buybacks are first and foremost a market management tool. They allow the Treasury to manage pockets of the bond market which are less liquid and more volatile. There are some nuances, but the stated purpose is basically to allow holders of older, less-liquid “off-the-run” bonds to exchange them for cash. Treasury then continues financing itself according to its broader issuance schedule.4

That’s what makes the maturity composition of marginal issuance relevant. Buybacks don’t eliminate Treasury’s need to borrow, but Treasury also does not mechanically replace every long bond it retires with another long bond. If coupon auction sizes remain unchanged (as we noted they have been), while additional marginal financing continues to come from bills, the composition of privately held Treasury debt can still shift shorter.

Zoom out and it looks less isolated

The timing of Wednesday’s buyback announcement, a day after fresh 19-year highs in the 30-year yield, naturally invited speculation that Treasury was growing worried about the long end. But the scale of the operation is worth keeping in perspective. Even after doubling the maximum size, Treasury can buy back up to $32 billion of 10-to-30-year debt per quarter, assuming they continue the established schedule of four operations in each long duration bucket, compared with $231 billion of scheduled 10-to-30-year issuance in Q3. In other words, maximum buybacks amount to only about 14% of new issuance. The amounts are too small to constitute yield curve control, but the direction of travel may still be worth paying attention to. Despite their relatively small size, the evolution of buybacks suggests Treasury is taking conditions at the long end, and the market’s capacity to absorb duration, more seriously.

The Treasury Borrowing Advisory Committee (TBAC) has specifically noted that there is still strong market appetite to sell off-the-run Treasuries into the buyback operations, which supports the idea of expanding their sizes. However, this strategy is running into a limit. TBAC projections suggest the current auction structure eventually won’t be sufficient to meet projected funding needs, and coupon issuance probably has to rise in fiscal year 2027 to fill that gap.5

Treasury could lean more on bills and retire some old long bonds, but this exposes the government to greater rollover and interest-rate risk down the road. Not ideal if interest rates keep climbing. Unless deficits shrink, either bill share keeps climbing, coupon auctions eventually rise, or some combination of both occurs. There isn't a magic third option that makes the financing requirement disappear.

The Fed complicates the picture

Complicating the picture is recent discussion at the Federal Reserve about shortening the maturity of its securities holdings. The Fed is already concentrating its reinvestments and reserve management operations at the short end. The Fed has expressed interest in reducing its balance sheet maturity to give it more breathing room to respond to the next crisis, as those events are typically ones that see central bank balance sheets growing longer (in duration terms) rather than shorter. Before the Great Financial Crisis when the world changed forever, the Fed’s holdings were about 38% bills. Since that period they’ve been single digits and currently stand at 8%. It has room and precedent to go much higher.

For the Fed to achieve this, it would need some combination of balance sheet roll-off, reinvestment policy, and open market operations to shift its holdings. The exact mechanics aren’t really important for my point here. If the Fed wants to own less duration, somebody else has to own more of it. All else equal, that puts upward pressure on long-term yields. But Treasury can offset some of that pressure through its own issuance choices. By leaning heavily on bills, limiting growth in coupon auction sizes, and expanding its capacity to buy back less-liquid long-term debt, Treasury may be creating more room for the Fed to eventually shorten its portfolio without overwhelming private markets. There is no evidence yet that this is an explicitly coordinated strategy, but it is worth monitoring. As my Twitter friend Adam Butler said, “the Fed and Treasury are partners in a dance, and neither can dance alone.”6

With financing needs this large, Treasury increasingly appears sensitive to how much duration the private sector must absorb. If the Fed eventually shortens its own portfolio, Treasury's issuance and buyback decisions could determine how disruptive that transition becomes.

Businesses aren’t hiring, but they aren’t firing either. The result is a strange labor market equilibrium: historically difficult for someone looking for work, yet remarkably stable for someone who already has a job.

Cold job demand

For starters, job growth is very weak. You can see the two labor surveys below showing month/month changes and a six-month moving average. The household survey (top, orange) is negative, averaging -260K jobs lost per month over the past half year. The establishment survey (bottom, green), also known as Nonfarm Payrolls, recently bounced off zero but remains very low around 61K jobs created monthly. While the surveys measure employment differently, both point toward the same conclusion: the number of people working is barely growing—and may currently be shrinking.

To smooth the noise from the individual series, I like to average the two into a single dataset, which I’ll call employment composite. This results in a rolling average of about -100K jobs lost per month. Some of the weakest job growth we’ve ever seen (below).

To put the current environment in perspective, today’s job growth in the context of past business cycles is the lowest run rate (-1% annualized) of any non-recessionary period in American history. The chart below shows current job growth in black relative to when the Fed first raised interest rates in 2022. Other colored dashed lines represent past recessionary cycles. The vertical bars at the bottom of the chart indicate when recessions began.

In some ways the trend of current job growth looks similar to the Great Financial Crisis period (red dashed line), but the broader economy looks quite different from those years. We do not appear as though we’re repeating 2008. But it does raise an eyebrow.

With a typical recessionary outcome, negative jobs tend to be the harbinger of recession. It is the shedding of payrolls by businesses that tends to ring in the downturn. By the time job growth flips negative, the economy is 2-3 months into what is later dated an official recession by the NBER.

In our current cycle, job growth using the same measurement has been negative for six months, and yet we do not see a big spike in layoffs nor initial jobless claims.

Not only are we missing any observable impact to hiring, firing, or jobless claims, the official unemployment rate has actually declined in recent prints from a post-pandemic high of 4.5% to 4.1%.

All measures of explicit stress in the labor market are muted. Meanwhile, job openings have finally returned to their pre-pandemic baseline, indicating a full reset of labor demand.

So how can job growth be this weak without producing the unemployment we normally associate with it?

The answer may lie on the other side of the equation: labor supply.

Americans are getting older

A good explanation for much of the weirdness in the labor market absent a bona fide recession is that our demographics are changing. Americans are aging as a group, living longer, and having fewer children. This means fewer new wage earners entering the labor force relative to the entire pool of consumers in the population. This trend is likely with us for the foreseeable future, and it begs us to evolve our expectations.

At first, the shift toward longer lifespans and fewer children did not have much impact on the data. However once we reached the early 2000s and the first Boomers became eligible for early retirement, the nation underwent a shift in labor participation that has endured ever since. The all-time peak of the labor force participation rate was January 2000, but it would not last long. The oldest Boomers reached 62 years of age in 2007, and by 2010 retirements were fully underway.

Looking deeper into labor participation, the aging dynamic becomes clear. If we split out the prime aged working cohort, meaning workers aged 25-54 years, we can see that their labor participation rate has remained very near its year 2000 high even as the headline figure, which counts the total population including retired people, has declined steadily.

The fact that 25-54 year labor participation is near the series high is another notch in the cap of a labor market that is not entering a recession. The vast majority of prime aged workers are employed. And since we already noted that job openings have returned to their baseline, the reasonable triangulation of all these data points is that weak growth in labor demand is currently being matched by unusually weak growth in labor supply.

With an aging population having fewer children, the natural result is a long-term slowing of growth in the labor force, and thus a slowing in the growth of the pool of available workers. That is precisely what we have observed.

While the labor force is still growing, that rate of expansion has slowed dramatically to less than 1% per year. This is a big change from what the U.S. has enjoyed historically. Couple that long-term trend with the current administration’s priority of large scale deportations of illegal immigrants (the White House estimates 2.5M exits), and recent figures of labor force growth may be flat or even negative.

A strange equilibrium

When labor force growth is this low, it necessitates adjusting what is known as the “breakeven rate” of job growth- that is, the rate of job growth necessary to maintain a stable unemployment rate. If the labor force is effectively not growing and prime aged workers are as close to fully employed as we’ve ever seen, then it’s reasonable to expect little to no job growth (even occasional negative job growth) would be “normal” in that environment.

Imagine an economy adding 200,000 potential workers each month. Businesses would need to create roughly that many jobs to keep unemployment from rising.

Now imagine essentially no one new is entering the labor force. Suddenly the economy can create zero new jobs without unemployment rising at all.

That is increasingly the labor market we have.

The Federal Reserve wrote about this earlier in 2026 (emphasis mine):

Labor force growth has been slowing and could be near-zero starting this year, driven by weak population growth reflecting low net immigration and by declining labor force participation reflecting population aging. Such weak growth in the labor force is unprecedented in the United States’ recent history. In this note we highlight two significant implications of near-zero labor force growth: First, near-zero labor force growth implies that breakeven employment growth (i.e. the pace needed to maintain a steady unemployment rate) would also be near-zero—making negative job growth almost as likely as positive job growth in any given month. Second, it implies that any growth in potential GDP will need to come entirely from productivity growth. These features would represent a significant shift in U.S. labor market dynamics and the composition of economic growth.

This explains much of the current oddness with the labor market. Labor participation among prime-aged workers is at highs, labor force growth is flat to negative, and thus breakeven job growth all else equal should be flat to negative.

And what a strange environment this must be for workers. Businesses are not hiring much, but neither are they firing much. There are not enough new workers entering the labor force to produce a lift in the unemployment rate, so it declines. Churn between jobs is minimal. The most effective avenue for workers to increase their incomes, by job switching, feels closed off because of low demand. Someone who already has a job can see the low unemployment rate and think everything is okay. Someone trying to get a job can experience total stagnation and a frozen market.

This is a historically bad labor market for finding a job while simultaneously being an okay market for keeping one.

What it means for the future

The important distinction going forward is whether this remains a low-hire, low-fire labor market or it becomes a low-hire, high-fire one. With labor supply barely growing, weak payroll prints alone may no longer tell us what they once did. The recession signal would be businesses moving from simply refusing to hire toward actively shedding the workers they already have. In this new equilibrium, monthly job growth may need to take a backseat. Layoffs and unemployment are where I’ll be looking for clearer signs that something has actually broken.