The Daily Spark

Some loans let the borrower add interest to the principal rather than pay it in cash, an arrangement known as payment-in-kind (PIK). The yield on PIK debt is the rate lenders earn on those loans, and it normally sits above the yield on cash-pay loans because borrowers who need the PIK option tend to be weaker credits. PIK debt yields in listed BDCs have started to rise, with software sector headwinds driving the repricing, see chart below.

The risks are rising that long rates six months from now could be a lot lower than where they are today.

Long rates are high today because of inflation and fiscal problems. But these forces could end up being dominated in early 2027 by what happens to AI, see the first chart below.

If AI succeeds and tech companies generate trillions in revenue, AI will be massively deflationary and push rates lower.

If AI does not work out, the bubble bursts and the Nasdaq is down 50% as investors rotate out of equities into Treasuries and long rates fall dramatically.

Over the next six months, the market will make up its mind about which AI scenario is playing out.

The bottom line is that financial markets are driven by narratives. The narrative in rates today is all about inflation and fiscal problems. But the narrative going into 2027 is going to be all about either the success or failure of AI. And in both scenarios, long rates are going to be lower.

For 30 years, the 10-year Treasury yield fell almost entirely on the handful of days around Fed meetings, and did essentially nothing the rest of the time, see chart below.

That stopped in 2022 after the Fed began to raise interest rates. The 10-year has since risen roughly four percentage points, but almost none of that came on Fed days. The reason is that each hike was largely priced by the time the Committee met, so the FOMC announcement and press conference carried little news for the long end, and the move higher instead came from higher-than-expected CPI prints, stronger payrolls, increasing Treasury supply and a rising term premium, none of which sit on the FOMC calendar.

This raises the question of whether FOMC meetings are still the most important drivers of long rates. With Fed Chair Kevin Warsh having dropped forward guidance, the most likely scenario is that rates become more market-driven and move more outside of FOMC meetings. In other words, this looks less like a post-hiking-cycle anomaly and more like the new regime.

The bottom line is that we should not expect forward guidance from Warsh today at the Jackson Hole Economic Symposium, but anything he says about framework and the economy, in particular where he sees inflation, unemployment and the neutral rate, could move the long end anyway.

Today's Spark comes with a welcome. Huw van Steenis has joined Apollo in London. He brings decades of experience across European banking and markets, and will be writing here on Europe, the UK and the ongoing evolution of private markets.

We start by looking at how the UK's corporate funding model has changed, see charts below.

Twenty years ago, banks provided 57% of UK corporate debt. Today, investors are the largest source, at 56%.

All the real growth has come from capital markets. Adjusted for inflation, bank lending to UK companies has fallen 10% since 2005, while market-based finance has grown around 50%.

The menu is also much broader. Twenty years ago, market finance meant bonds. Today, public bonds are 40% of UK corporate debt, with private bonds, direct and broadly syndicated loans, and other non-bank lending adding a further 16%. Most of it is investment grade.

Bank lending has gone global. Overseas banks have lifted their share of UK corporate bank lending from 17% before the 2008 financial crisis to 27% today.

The bottom line is that UK companies can now tap a deeper and more diverse pool of capital than at any point in two decades. More sources of funding mean more resilience, better access and less dependence on any single part of the financial system.

Yesterday, Stanley Druckenmiller argued in the Wall Street Journal that the Treasury should stop trying to hold down long-end yields with buybacks and address the primary deficit instead. The fiscal backdrop to his argument is stark.

Since 2006, US gross federal debt has increased by $32 trillion while the annual level of nominal GDP has increased by $19 trillion, see the first chart below. Debt is up nearly fivefold over that period. The economy is up less than 2.5x. Debt is compounding faster than the economy that has to service it, which is why federal debt held by the public has gone from below 40% of GDP to roughly 100% over that period.

The outlook offers no relief. The CBO projects that under current policies, debt held by the public will keep climbing from 100% toward 175% of GDP, see the second chart below. The OMB forecasts budget deficits near 5% of GDP over the coming years, on top of a current run rate closer to 6%, see the third chart below.

Deficits that size are normal in a recession. These are forecasts for a full-employment economy.

The fiscal outlook, a Fed considering a rate hike, and hyperscaler issuance crowding out demand for Treasuries all point the same way.

The bottom line for investors is that interest rates are going to stay higher for longer. Or, as Druckenmiller puts it, the long-term Treasury yield is the only fiscal disciplinarian the US has left.

The chart below shows US employment in each Federal Reserve district. San Francisco accounts for 20% of all jobs in the United States, and Atlanta 14%.

Average recoveries have dropped across the debt stack over the past three years, as out-of-court distressed exchanges reshuffled priority without real deleveraging and a growing share of asset-light software borrowers left creditors with little tangible collateral to seize, see chart below.

If AI were displacing white-collar work at scale, you would expect to see it first in the Philippines and India, where business process outsourcing (call centers, IT support, back-office processing) accounts for a large share of employment.

Instead, the unemployment rate in both countries has continued to trend lower, with the Philippines near 5% and India near 6%, both well below their 2021 levels.

The bottom line is that the hard data still show no signs that AI is generating job losses in the economies and industries most exposed to it.

Application software is the part of software most exposed to AI disruption, and it is also where lenders have the most money at risk: $146 billion of loans outstanding, mostly rated B- or lower.

These are companies that charge based on how many people use their software to run tasks AI can most plausibly automate.

Infrastructure, data and security providers, by contrast, get paid more as AI adds computing workloads and more systems to defend.

China Shock 1.0 flooded global markets with cheap consumer goods in the 2000s, but China Shock 2.0 represents a sharper acceleration in manufacturing exports where the country now dominates advanced sectors like EVs and semiconductors through industrial policy and overcapacity, see the first chart below.

Brad Setser warns that, unlike the first shock, there's nowhere left to move production when China controls the cutting edge, especially as domestic demand is increasingly met by domestic production, leaving massive export surpluses to flood global markets, see the second chart below.

For more discussion, see: The Second China Shock - How This Time Is Different with Brad Setser | Markus' Academy | Ep. 165-1