Macro Moments Weekly: Debt, Hyperscalers and Inflation—Three Rumblings Before the Storm

Key Takeaways:

• Rising debt costs are becoming a major challenge for governments: Higher interest rates and elevated debt levels mean countries will face increasing debt-servicing burdens, putting pressure on future spending and taxation decisions.
• Hyperscalers are affecting bond markets through long-term issuance: The concern is less about the volume of borrowing and more about the unusually long maturities being issued, which are disrupting long-duration bond markets.
• Inflation may prove more persistent than expected: Risks from European gas supplies, food prices, and weather-related disruptions could keep inflation elevated and delay further declines in core inflation.

In pretty much every article you read on the recent bond market sell-off, three things will be said:
1) Governments need to start to rein in their spending, because interest costs as a percentage of gross domestic product (GDP) are spiralling and will continue to do so if they don't. 2) The hyperscalers have gone way over the top and ruined the market. 3) Inflation is not a worry.

What most fail to do is put any numbers to their commentary, so it’s time to have a go, firstly with debt servicing costs as a percentage of GDP. The starting point is normally pre-COVID-19, when base rates were close to zero, and long-bond rates were not much higher. Of course, debt servicing was lower, and as it was pre-COVID, debt-to-GDP was often 20% or 30% lower than today. So the shock of debt servicing costs rising is no different than anyone with a mortgage whose five-year fixed deal done in 2021 at 2% is now suddenly facing a 4.5% rate. The mortgage holder is forced to cut spending, forgo a holiday or a new car. Governments don't do that, and new issues, such as defence, mean they want to spend even more. The bond market is acting as their bank manager and forcing a change of behaviour. It just takes more time, or more taxes. The United Kingdom currently spends close to 6% of GDP on interest costs, and as existing debt is rolled, that will rise over time to 10%. France is in a similar position, although starting at 3.4%. The United States is similar. But China has been able to tap its huge domestic savings market akin to Japan 30 years ago, and has thus kept its costs down, even as the debt level rises to close to 100%. So the markets are right—it is a problem for tomorrow, that needs addressing today.

Secondly, we have the hyperscaler issuance flooding the market, so they say. And with it, there’s an implication that this is a boom and all that capex will go up in smoke. The numbers suggest something rather different. Yes, this has been a year of high issuance from non-financial investment-grade companies. And yes, no one expected SpaceX to join in. But the overall level of issuance will match the previous record year of 2020 of around US$650 billion net. Where it gets interesting is the tenor, the debt length. Issuers have gone for very long-term debt, and as a result have been responsible for 28% of all 7-to-10-year issuance, 16% of all 10-30 year and a whopping 64% of all 30-year plus. And by doing so they have planted their debt tanks on the governments’ lawns, causing upset all around the market. The size of issuance is large, but not unmanageable. The premium they are paying is 25 to 40 basis points; as an investment-grade borrower, it’s the illiquidity of the long-term markets that has caused the problem. Next year, when it is likely they will return for similar levels of funding, the tenor will no doubt be shorter in parts of the market that are much more liquid.

Thirdly, is Inflation a worry—or not? Last week, I asked should we be surprised if inflation hangs around when the economies are doing quite well. The evidence this week from the Purchasing Managers’ Index data has backed that up. But there are a couple of issues that may make this much more pressing—one being gas prices in Europe. The restocking process for winter is way behind where it should be, especially in Germany. Prices are beginning to rise sharply, although not to 2022 levels. And unless supply changes with the reopening of the Gulf, prices may stay high. With low stocks there is always the risk of a cold snap, forcing Europeans into the spot market, bidding up prices further. Some countries like the United Kingdom have little or no storage and thus are particularly exposed. Then, there’s a risk from food prices. As a result of high prices in 2025, food prices have not risen as much as energy costs have. The drought has further complicated the situation, bizarrely reducing the price of milk, where output has risen because cows are not eating pasture, but nutrient-rich processed food. This will end, and we will likely see prices of soft commodities rising sharply, and urea price increases would impact fertilizer, which may reduce land planted this winter/spring. What impact El Niño will have on Pacific and Brazilian crops is another risk. Food prices could start to steadily rise and may well prevent core inflation from falling at all next year.

So does this all amount to the perfect storm? The peculiarities of governments realising that the party is over, the heavy long duration issuance of the hyperscalers and the probability that inflation has a couple more tricks up its sleeve could very well be rumblings. Looked at in another way, one factor is structural, one factor is cyclical (compounded by poor management) and one is a combination of geopolitics and bad luck.

Priced in? Two are probably in the price. One isn't. Is two out of three "ain't bad"?