On US Debt’s New, Big, Round Number

$30 trillion. That is the big round number US national debt passed for the first time Tuesday, sparking a flurry of headlines—most of them fearful. And political, as is usual wherever federal borrowing and spending are involved—at any time, one party will campaign on fiscal probity while the other talks up “stimulus,” and those roles reverse often. We are going to stay above that fray, as markets don’t care about political personalities. Nor do they prefer any politician or party. But we do think there is a big elephant in the room to address: That headline, $30 trillion figure just isn’t useful for analysis of basically any kind. Understanding why can help investors see through hype and take a more measured view.

The $30 trillion number is what the Treasury calls “gross debt”—total debt outstanding. This includes all the Treasury bonds owned by investors here and abroad, as well as debt owned by the Federal Reserve through its quantitative easing program and normal open market operations. Yet it also includes debt owned by intragovernmental agencies, primarily in trust funds like Social Security. In our view, the most accurate way to measure debt is to exclude these governmental holdings and focus on debt held by the public—investors and the Fed. That figure, also known as net public debt, is just over $23.5 trillion.^[i] If you look up US debt at the Treasury’s Debt to the Penny website, you will see this is their lead debt statistic.

Now, on the surface, it might seem a bit suspicious to exclude nearly $6.5 trillion in intragovernmental holdings.^[ii] We assure you, our preference to focus on net public debt is not simply because we want to make debt look better—rather, we think it is more accurate. One way to see this is by lumping all these government agencies together, along with the Treasury, as “the government.” This $6.5 trillion, which the Treasury borrowed, shows up on the “liabilities” side of the government’s balance sheet. But for the agencies owning this debt, it is an asset—and also shows up on the government’s balance sheet as such. $6.5 trillion in assets cancels out $6.5 trillion in liabilities. Or, if you don’t like balance sheets, you can simplify it as “money the government owes itself effectively cancels.”

The Treasury bonds held in intragovernmental trust funds are also a different breed from your typical interest-paying IOU from Uncle Sam. The Treasury creates the bonds owned by the Social Security trust funds specifically for them to fill a need. By law, the Social Security trusts are required to invest all funds above and beyond what is needed to pay benefits in a given year in securities issued and guaranteed by the government. It is a statutory mandate to set the fund’s market risk at zero (they are focused on default risk here). When bonds owned by the trust mature, the Treasury replaces them with new ones. They are non-marketable—the Social Security trusts can’t just sell them on the open market. They can only ask the Treasury to redeem them if they need cash, which retires the security. Including these securities in a broad tally of funds owed to creditors would therefore be a bit weird.

This same logic explains why we think it is correct to include debt owned by the Federal Reserve in the net public debt tally. Doing so might seem contradictory, as the Fed sends most of the interest earned on its bond portfolio back to the Treasury—sort of falling under the “money the government owes itself” umbrella. But also, not quite. One, the Fed isn’t officially a government body. Two, unlike bonds held in the Social Security trusts, the bonds on the Fed’s balance sheet are marketable. The Fed bought them from banks and, if policymakers chose, can sell them to investors on the open market. Once the bonds mature, the Treasury has to replace them with new marketable securities and sell those to investors. They are walking, talking, outstanding liabilities.

Now, $23.5 trillion is still a big number—a bit bigger than 2021 GDP. We don’t think it would be beneficial for the government to pile on to this exponentially and endlessly. But when considering debt as an economic and market risk, we think it is best to approach the issue like markets do: What is the likelihood that debt becomes problematic in the next 3 – 30 months? In general, “problematic” means being unable to service debt, either because interest payments break the bank or because the Treasury can’t find buyers for new debt to replace maturing securities. That is the predicament Greece was in a decade ago. But the US isn’t anywhere close to that. In fiscal 2021, bond interest payments represented just 8.7% of federal receipts—less than half of 1991’s high water mark, 18.4%.^[iii] It is also down from 2019 and 2020, thanks to low interest payments and rising revenues (a byproduct of economic growth). If interest costs are trending down—and if much higher costs 30 years ago didn’t cause a debt crisis—we rather doubt a debt doom spiral is in the offing now, shocking $30 trillion headline figures or no.