The Biden Administration, backed by influential economists, claims that borrowing more trillions on top of the trillions added to counter the pandemic, is a low-risk road to prosperity. That’s a fantasy. The recent, sudden jump in interest rates exposes the “don’t sweat the numbers, just go big” offensive as a potentially perilous gamble, based on sunny forecasts that are already looking highly questionable.
The $1.9 trillion COVID relief package passed by the House on February 27 marks the only the first step in President Biden’s plans for giant increases in deficit spending this year. He’s also pledged to propose a $2 billion infrastructure and green energy measure this spring. In her Senate confirmation hearing, Treasury Secretary Janet Yellen said a new round of “stimulus” was essential to restoring growth to pre-pandemic levels. “Neither President Biden, nor I, propose this relief package without an appreciation for the country’s debt burden,” stated Yellen. “But now, with interest rates at historic lows, the smartest thing we can do is act big.”
The case for big spending
The Administration’s position that trillions more debt isn’t especially worrisome dovetails with the theory advanced in a recent paper by Larry Summers, Treasury Secretary under President Clinton, and Jason Furman, formerly top economic advisor to President Obama. Summers and Furman argue that what matters in gauging how much the U.S. can safely borrow isn’t the traditional yardstick of debt to national income, but rather the share of interest outlays to GDP. As Furman stated in a blog post, “Rather than focus on the size of the U.S. debt, policymakers should assess the fiscal capacity of ‘real’ interest payments, ensuring they remain comfortably below 2% of GDP. By this measure, there is room for substantial additional fiscal support.”
The Biden-Yellen stance, endorsed as sound economics in the Summers-Furman manifesto, suffers from a big weakness. It’s based on the assumption that the interest rates pulled to historic lows by the ravages of the pandemic will stay abnormally tame for years to come. They may be right. But big spending now leaves the U.S. far less margin for error if they’re wrong. “What if we have another financial crisis, or a second pandemic hits, or Italy goes bankrupt, or a trade war with China intensifies?” says John Cochrane, an economist at the Hoover Institution. “Then you have no growth or a recession, and rates could soar. The U.S. would hit an iceberg.” During mid-to-late 2008, when the U.S. was mired in the Great Recession, the 10-year Treasury (long bond), 2.5 points above its current level.
Put frankly, Biden, Yellen and brainiacs Summers and Furman and the seasoned forecasters at the Congressional Budget Office can only make their best guess at where rates are headed. The CBOs estimates change substantially from forecast to forecast. The 10-year treasury yield’s surprise jump from 0.93% at the start of the year to 1.56% on March 4, underscores how quickly and unpredictably rates can move. That quicksilver shift already means that U.S. will be forced to roll over debt at rates much higher than the Congressional Budget Office was predicting at the start of February. It may also signal that the U.S. in en route to the much higher yields that generally accompany a robust recovery.
What the “go big” crowd underplays is the risks imposed by the sheer volume of U.S. debt. That load increased by one-quarter or $4.3 trillion to $21.09 trillion trillion at the end of the nation’s September fiscal year––that’s 102% of GDP, equalling France’s. (It was $21.7 trillion at the end of January.) Besides the proposed packages that could add another $4 trillion in “emergency” spending over the next few years, the U.S. is running chronic, unsustainable deficits of around 6% of GDP that extend to the horizon. The burden is already so big and rising so fast that a spike in rates levels that prevailed as recently as late 2018 would swamp the budget with added interest costs.
Policymakers are on the verge of putting America in a predicament where the cost of being wrong about interest rates has never been greater. And if Yellen, Summers et al are wrong, we’ll be a bind that can only be remedied by draconian spending cuts or epic tax increases.
The problem is not just that we’re borrowing so much, but how. The U.S. has been bulking up ultra-short term borrorowings, leaving us highly vulnerable to rising rates, just the trend that hammered folks relying on adjustable rate mortgages in the housing frenzy of 2006 and 2007.
At the close of FY 2019, of America’s $16.8 trillion in debt held by the public, $4.5 trillion were maturing in a year or less. That segment consisted mostly of short-term Treasury bills issued for up to 12 months. Another $5.3 trillion were coming due in years two through five. By September 30 of last year, the amount due in one year on the much higher total load of $21.09 trillion had soared from $4.5 trillion to $7 trillion, and the number maturing within five years zoomed from $9.8 trillion to $14.2 trillion. The U.S. found itself with 85% as much debt payable by the fall of 2026 as it held in total borrowings twelve months earlier.
Through February, the U.S. has borrowed an additional $553 billion, and the portion due in 12 months has declined a bit as the Treasury attempts issue longer-term bonds to fund new spending and refinance debt coming due. But the high-risk profile hasn’t changed much. In fiscal 2020, the U.S. paid an average of just 0.2% on those year-or-less Treasury bills, down from 2.1% in 2019. The U.S. relied much less on bonds at longer maturities of one to ten years, but still benefited from the incredible drop in yields from 1.9% at the start of 2020 to .between 0.6% and 0.7% from early April through September. Those falling rates lowered average payments on these Treasury notes from 2.2% to 1.9%.
In early February, the CBO issued its “The Budget and Economic Outlook: 2021 to 2031” report, forecasting a continuation of ultra-low rates in the years ahead. The CBO predicted that the long bond (10-year Treasury) yield would average 1.1% this year, 1.3% in 2022 and 1.5% in 2023. The agency posited that the U.S. could keep borrowing new trillions and refinancing maturing debt on such favorable terms that interest expense would fall sharply while debt soared. The CBO forecast that as borrowings increased from $21 trillion in 2020 to $27.6 trillion by 2026, interest costs would first fall sharply then reach $365 billion at the end of the five year window, less than in 2019, and just a sliver more than last year.
Sounds great. The U.S. could borrow trillions more through 2026, and pay lower interest expense than before the big ramp-up even started.
These numbers raised jubilation among some House Democrats, who claimed that the savings from super-low rates would effectively pay for the cost of the proposed infrastructure bill.
The CBO’s projections now look unrealistic, for two reasons. First, debt will be much higher than the CBO predicted. In part, that’s because it’s required to include only spending under current law, and hence can’t incorporate an estimate of the annual costs of the proposed COVID relief and infrastructure measures. Brian Riedl of the conservative Manhattan Institute reckons that the expenditures in the cards will lift total debt by the end of 2021 to almost $24 trillion, $1.2 trillion more than the CBO’s “under current law” estimate of $22.5 trillion. In the years that follow, the CBO’s numbers include the expiration of middle class tax cuts that probably won’t happen. As a result, Riedl forecasts that by the close of the CBO’s forecast period in 2031, total U.S. debt will reach $40 trillion, or 125% of GDP.
Second, the recent spike in rates to 1.5% means the CBO’s estimate of 1.1% for this year is already far off the mark, and that the long bond yield now already exceeds its 1.3% forecast for 2022, and equals its number for 2023. Instead of refinancing debt at 10-year maturities at the 0.93% rate on January 1, the Treasury will now need to pay between 0.4% and 0.6% more. It’s only other option to shouldering higher than projected costs is to do what it did in 2020: Keep doing the ARM thing by borrowing heavily in Treasury bills that roll over in a year or less.
If rates flattened at 1.5% for the next five years, the CBO projections would be far less favorable because first, the debt will be higher than it’s able to bake into its official estimates, and second, rates will be loftier as well. If rates stay that low, the U.S could could forestall a calamity for a few years. But disaster would still lurk towards the end of the decade. Employing conservative assumptions, the CBO foresees interest expense of $800 billion by 2031, when it expects the U.S. to be paying average rates of 2.3%. At a more reasonable $40 trillion in debt, that burden rises to $920 billion. That’s the equivalent of half the predicted outlays for all discretionary spending.
What if the CBO’s forecast is wrong? As recently as January of 2020, it was predicting that the 10-year would yield would average 2.2% in 2021, double its February forecast, rising to 2.6% in 2022 (double again), and well over 3% by the end of the decade. Don’t discount the possibility we could get back on that path. In that scenario, interest on $40 trillion in debt by 2031 would reach $1.3 trillion, equalling over one-fifth of all revenues.
The February rate scare raises the probability that debt picture will present a big problem well before 2031. The U.S. will be forced to roll over the trillions its borrowed on ARM-like terms in longer-term bonds whose yield recently doubled, and if they doubled again, still wouldn’t nearly reach the long-term averages.
It’s a big mistake to assume that a “known,” a galloping run-up in borrowing, isn’t a looming peril when what you’ll pay on that known depends on the unknowable. We just got a shocking reminder that the future of interest rates is one of the economy’s great unknowables.