Last week delivered another warning for the U.S. economy. Housing starts fell 12.4% in July, with single-family construction down nearly 10% for the month and 16.6% from a year ago.
This is bigger than housing. Residential construction is one of the economy’s most interest-rate-sensitive sectors, and its weakness spills into building materials, appliances, furniture, mortgage finance and other housing-adjacent industries. It fits the broader pattern we’ve been watching: outside of AI infrastructure and health care, not much is moving forward.
The economy isn’t falling apart. But high borrowing costs are clearly biting, hiring remains weak, and momentum is becoming increasingly concentrated in a handful of sectors.
Tuesday: Can new-home sales surprise?
That makes Tuesday’s new-home sales report particularly interesting.
The consensus expects July sales to fall to roughly 620,000 at an annual rate, from 628,000 in June. That would leave sales about 5% below their July 2025 level. June sales were already 5.6% below a year earlier.
There is a case for an upside surprise.
Mortgage rates have climbed sharply since the beginning of the year, but housing demand has been somewhat more resilient than the increase in monthly payments would suggest. The average mortgage rate rose from around 6% in February to roughly 6.5%-6.7% recently. Yet existing-home sales activity has managed to run modestly above year-ago levels. However, the flow of new resale listings coming on the market has slowed sharply.
That matters for builders. They are still competing against a large stock of homes already on the market, so I would not call resale inventory scarce. But fewer new existing homes are being added to that stock, and builders have another advantage: incentives. Nearly two-thirds of builders are offering some form of sales incentive, and roughly 30% are cutting prices. Rate buydowns and other concessions allow builders to do something most existing homeowners cannot – lower the buyer’s effective financing cost.
So the consensus calls for another decline, but I would not be shocked if new-home sales beat expectations.
Wednesday: A benign PCE print, with an important catch
Next comes the Fed’s preferred inflation gauge.
July’s headline PCE number should still benefit from the drop in energy prices during the month. Consensus estimates point to roughly a 0.1% monthly increase after headline PCE fell 0.1% in June. Core PCE, however, is expected to rise about 0.2% after increasing just 0.1% in June.
And last week’s producer-price report gave us a reason not to get too comfortable with a soft headline number.
Wholesale prices were flat in July. Energy prices fell 3.1%, gasoline fell 5.7% and food prices declined 0.9%. That looks disinflationary. But the components of the producer-price report that feed more directly into PCE were considerably firmer.
Core producer prices excluding food, energy and trade services rose 0.4%. Services excluding trade, transportation and warehousing rose 0.6%. Some of the components that feed into PCE were firm – notably health care and portfolio management, where prices jumped 6.5%. My estimate based on those PCE-relevant components points to roughly a 0.24% increase in core PCE for July, barely above the consensus forecast of 0.2%.
There is another complication. The energy relief is already getting old.
Oil prices have climbed again in August as conflict in the Middle East intensified. Brent crude has moved back above $90 a barrel, erasing much of the energy-price relief that should flatten the July inflation data.
So July PCE could look relatively tame while telling us considerably less about where inflation is headed next.
Friday: Jackson Hole and the problem at the long end
That brings us to the week’s main event.
Fed Chair Kevin Warsh speaks Friday at the Federal Reserve Bank of Kansas City’s Economic Policy Symposium in Jackson Hole. This year’s theme is “Financial Innovation: Implications for Payments and Policy.” The symposium runs August 27-29, with Warsh scheduled to speak Friday morning.
But the most important financial innovation confronting Warsh may have nothing to do with payments.
He arrives in Wyoming with the world’s bond markets under pressure.
The U.S. 30-year Treasury yield climbed to roughly 5.34% last week, its highest level since 2007. Treasury responded by doubling the size of liquidity-support buybacks in the 10- to 30-year portion of the curve, from $2 billion to at least $4 billion per operation. The announcement briefly pulled the 30-year yield down to about 5.19%. By Friday it was back near 5.28%.
The bigger message is that this is not just an American bond selloff.
Long-term borrowing costs have risen across advanced economies. Japan’s 10-year yield has reached its highest level in roughly three decades. German yields have climbed to levels not seen since 2011. French long-term yields are near their highest since 2008, and Britain’s 30-year borrowing cost has been trading around levels last seen in the late 1990s.
Something larger is going on.
Too many borrowers, not enough savings
Start with energy.
War and geopolitical instability in the Middle East have pushed oil prices sharply higher. That does not just raise near-term headline inflation. It increases uncertainty about how quickly central banks can return inflation to target. The Fed’s own July minutes acknowledged that a prolonged conflict could worsen supply-chain problems and keep inflation elevated.
Then there is trade.
The world spent decades building supply chains around the assumption that goods and capital could move increasingly freely across borders. That process helped lower production costs and restrain inflation. First Covid, then tariffs, trade restrictions and increasingly fragmented supply chains now work in the opposite direction. The Fed staff has already attributed part of the recent increase in core-goods inflation to tariffs, even if policymakers generally expect much of that effect eventually to fade.
Fiscal policy may be the bigger structural story.
Global public debt is already around 94% of GDP and the IMF expects it to reach 100% by 2029. Across OECD economies, governments borrowed a record $17 trillion in 2025 and are expected to borrow roughly $18 trillion this year. Outstanding sovereign bond debt is projected to reach about 85% of OECD GDP.
The United States is hardly alone. Governments across the developed world are financing aging populations, defense commitments, industrial policy, energy investment and existing entitlement programs while facing very little political appetite for either higher taxes or substantially lower spending.
Every additional dollar of government borrowing has to find a buyer.
And governments are no longer the only giant borrower in the room.
The AI buildout is turning the world’s largest technology companies into much heavier users of debt markets. Hyperscaler debt issuance has already reached roughly $220 billion this year, compared with just $12.5 billion over the same period last year. The OECD estimates that nine major AI companies raised $122 billion in bonds in 2025 and could spend roughly $4.1 trillion on capital investment between 2026 and 2030.
That investment may raise future productivity. But first it has to be financed.
Sovereigns need more capital. AI firms need more capital. Defense and energy infrastructure need more capital. All of them are competing for the same global pool of savings.
Demographics may slowly make that pool less abundant as well. The aging of the developed world previously helped create a large pool of retirement savings, but as dependency ratios rise, the composition gradually shifts from prime-age workers accumulating assets toward retirees drawing income from them. Life-cycle models imply that this can reduce aggregate saving and put upward pressure on real interest rates. It is a slow-moving force rather than the explanation for last week’s bond selloff – and household saving remains elevated in parts of Europe – but it points in the same direction over time.
This is the new economics of capital: more demand for long-term financing meeting a supply of savings that is no longer expanding as effortlessly as it once did.
Too many borrowers, not enough willing lenders: those supply and demand forces are putting upward pressure on real long-term yields.
What the Warsh Fed is doing – and what it isn’t
That is the backdrop for a Federal Reserve that is changing how it communicates.
The Fed held its policy rate at 3.50%-3.75% in July, but three officials – Beth Hammack, Neel Kashkari and Lorie Logan – wanted a quarter-point hike. It was the largest unified hawkish dissent since 2016. The minutes showed that many participants thought further tightening could become necessary if inflation failed to decline, while some questioned whether financial conditions were restrictive enough to return inflation to 2%.
That tells us what the Warsh Fed is doing: holding rates steady for now, emphasizing the inflation target and keeping the option of another hike very much alive.
It is also giving markets less guidance about what comes next.
Warsh’s first two policy statements averaged just 113.5 words, 55% shorter than Powell’s final eight. All 24 of the recurring Powell-era phrases I identified around the Fed’s reaction function and forward guidance disappeared from both Warsh statements. “Prepared to adjust.” “Assessing the appropriate stance.” The language markets used to reverse-engineer the next move is simply gone.
What the Warsh Fed is not doing may matter even more. It is not promising rate cuts. It is not using the balance sheet to cap long-term Treasury yields. And it is not telling investors that the Fed will insulate them from volatility at the long end of the curve.
The July minutes reaffirmed that changes in the federal-funds rate should remain the primary way monetary policy is adjusted. The Fed may purchase short-dated Treasury securities when needed to keep reserves ample, but those are reserve-management operations – not a program to suppress 10- or 30-year borrowing costs.
That distinction matters because the recent move in bonds has not primarily been an inflation-expectations story. The Fed’s own analysis says the rise in Treasury yields during the intermeeting period was driven largely by real yields, while longer-run inflation expectations remained broadly anchored. Foreign sovereign yields rose alongside U.S. yields.
The Fed can influence the overnight interest rate and the expected path of short-term rates. It cannot manufacture an unlimited supply of global savings.
That is what makes this Jackson Hole different.
Markets will listen carefully for any clue about whether Warsh thinks another rate hike is coming. Even a firm commitment to get inflation back under control may not be enough to make the long end of the yield curve cheap again.
For everyday Americans, that means borrowing costs – on credit cards, auto loans and mortgages – may stay higher for longer, even if the Fed eventually starts cutting rates.
