Debt Sustainability: The Interest Bill Comes Due
INTRODUCTION
This brief paper explains the key issues around government debt sustainability. Government debt has risen substantially, and interest rates are no longer low. The result is that debt interest is becoming a real burden on government budgets that are already stretched after years of profligacy.
It also sets out which data to track, which commonly quoted statistics mislead, and why, for a government that borrows in its own currency, the real question is never whether it can fund itself, but at what price.
HIGH DEBT LEVELS AND WHY THEY ROSE
The absolute level of government debt matters because its cost, i.e. interest, competes with everything else a government spends money on (social security, infrastructure, pensions etc.).
The simplest measure is the Notional amount of debt outstanding. In the US, for example, total marketable debt is $32.4trn.
To judge whether that is large, it is compared to the size of the economy (GDP), which gives the Leverage. In the US, Debt/GDP is 117% (Source: BIS). Note that the Debt/GDP ratios in this paper use the BIS measure of government debt, which is broader than marketable debt, so the two figures are not directly comparable.
Both have risen significantly over the last 20 years. Notional debt rises whenever a government runs a deficit; Leverage rises when debt grows faster than nominal GDP. (Inflation, by lifting nominal GDP, actually works to reduce Leverage.) The charts below show both measures for the US. The same trends can be seen across most Advanced Economies (AEs).
Before the Great Financial Crisis (GFC) of 2008, Leverage rose and fell with the economic cycle. The GFC and then the Pandemic each produced a large jump (green circles), and in neither case did Leverage come back down in the recovery that followed. Outside those two events it has been broadly stable, if drifting higher. The result is a ratchet: debt jumps in crises and is not paid down in the good times.
Today Leverage is high almost everywhere. Only a handful of countries have Debt/GDP below 50%, and many are above 100%. For the G20 as a whole, Debt/GDP is close to 100%, up almost 40 percentage points since before the GFC.
THE COST OF HIGH DEBT
Governments pay interest on their debt mainly through the fixed coupons on the bonds they issue. If the US government borrows $100bn for 10 years at 2%, it pays $2bn a year for 10 years and then repays the $100bn. That $2bn is a fixed annual expense for the life of the bond. (The bond will yield 2% to an investor who buys it at issue and holds it to maturity.)
A government's annual interest cost therefore depends on two things:
The size of its debt
The rate of interest that it pays on its debt
After the GFC and during the pandemic, when debt rose most, interest rates were very low. Central banks held policy rates near zero and bought huge quantities of government bonds through Quantitative Easing (QE) programs, suppressing bond yields. This is commonly known as Zero Interest Rate Policy (ZIRP). The chart below shows policy rates across Advanced Economies since 1970: compare the ZIRP period with the decades before it and the years since.
So as debt rose, the annual interest bill barely moved: governments issued vast amounts of bonds yielding 0-1%. The US interest bill, for example, rose from
$186bn in 2007, when debt totalled $4.5trn (58%/GDP), to
$221bn in 2013, when debt totalled $11.9trn (96%/GDP).
Thus, a mere $35bn increase in annual interest costs, while the debt increased by $7.4trn.
That world has gone. As the chart of central bank rates shows, ZIRP is over and governments are now issuing debt at much higher yields. High debt AND high rates have together driven a sharp rise in interest costs around the world.
BUT there is a delayed effect, and it means interest costs will keep rising even if rates go no higher.
Remember the 2% bond? Its cost stays fixed at 2% until it matures. Governments running deficits cannot repay maturing bonds out of revenue, so they must refinance them (and borrow more on top) at whatever the market rate is at the time. The average rate on the debt stock therefore lags market yields: when yields rise, it climbs only gradually, as old, low interest paying bonds are replaced with ones with higher interest rates.
With the US 10-year yield currently at 5.25%, the $100bn that was costing $2bn a year will cost $5.25bn a year once refinanced: more than double, for the next 10 years.
The charts below show the US Government's annual interest cost (left) and the Weighted Average Interest Rate (WAIR) on its outstanding debt (right). The interest bill has ballooned as the WAIR has more than doubled, from around 1.5% in 2021 to around 3.45% today. Crucially, the WAIR is still nearly 2 percentage points below current market yields, so the refinancing effect has much further to run.
MEASURING DEBT SUSTAINABILITY
Debt/GDP is a reasonable measure of Leverage and allows comparison across countries. Interest Cost/GDP, however, is a poor guide to sustainability, even though it is widely quoted (Financial Times, The Economist etc.).
GDP does not pay the government's interest bill; tax revenues do. GDP is only the base from which taxes are raised, and governments differ enormously in how much of it they collect: France's government takes in around half of GDP in revenue, the US federal government less than a fifth. The same Interest Cost/GDP can therefore mean a very different burden from one country to the next.
Taxes pay for interest as well as pensions, healthcare, social security, education and other public services, so every extra dollar spent on interest is a dollar not available for anything else.
So, what is “Debt Sustainability”? It is a government's ability to pay its interest costs without having to change fiscal policy, i.e. without cutting other spending, raising taxes, or borrowing more simply to pay the interest. This is a different question from whether a government can fund itself, which is covered in the final section.
The best gauge is therefore Interest Cost as a share of Budget Tax Revenue. It shows directly how much interest is crowding out everything else, and whether it is likely to force a change in fiscal policy.
The table and chart below show this ratio for five major economies. It stayed low and stable through the ZIRP years even as debt rose, then turned higher as central banks raised rates.
The table also projects the ratio forward, assuming interest rates follow current market forward rates, debt and tax revenues keep growing at their recent pace, and maturing bonds are refinanced across similar maturities to current issuance. Note that revenue growth, the channel through which nominal GDP growth helps a government, is already built in, and the ratios still rise. The common argument that debt is safe as long as growth exceeds the interest rate does not offset the cost of refinancing cheap bonds at today's yields.
GOVERNMENT POLICY AND DEBT IMPLICATIONS
Many governments already face difficult budget decisions, with large deficits under pressure for different reasons:
In the US, the DOGE initiative has not delivered the hoped-for cuts in spending, while tax cuts have kept the deficit at its highest peacetime, non-recessionary level.
In Europe (including the UK), higher spending has pushed up deficits, and large tax rises have not been enough to bring them meaningfully down. Defense spending also has to rise significantly in the coming years.
In Japan, and in Canada (not shown in the table), governments have made pledges that will put further pressure on deficits.
If rates stay at current levels, which are elevated relative to the ZIRP years, interest costs will rise faster than tax revenues. Unless tax or non-interest spending policy changes, deficits will widen. Governments then face four options, or a mix of them:
Cut other spending by the amount of the extra interest cost
Raise taxes by the amount of the extra interest cost
Run a larger deficit, borrowing more to pay the extra interest
Hold borrowing costs down: central bank bond buying, yield caps, or tolerating higher inflation to erode the debt (financial repression)
Option 1 looks very unlikely. Governments have found it extremely hard to cut spending: over the past year (financial year to date versus the same period last year), government expenditures have risen by 2.2% in the US, 3.5% in Japan, 7.1% in the UK, 4% in France and 2.4% in Italy, and the political pressure (defense, pensions, healthcare) is to spend more, not less. Option 2 is all but ruled out in the US under current Trump policy and elsewhere runs into the commitments of governing parties or recent tax rises that are already weighing on growth. The market should therefore brace for Option 3: wider deficits as the interest bill grows. And the larger those deficits become, the greater the temptation to reach for Option 4.
These projections are not a worst case. They assume economies keep growing and tax receipts keep rising. In a slowdown, tax revenue growth falls away and the ratios would deteriorate meaningfully further.
GOVERNMENT DEBT CAN ALWAYS BE FUNDED – THE QUESTION IS THE PRICE
If an individual or company runs out of money and cannot raise new finance, it goes bankrupt. A government that borrows in its own currency is different: with the help of its central bank, which can create the money to buy its bonds, it will always be able to raise funds, however large its deficit or debt. The question is simply at what price, and what the consequences will be for the economy and other asset prices. Euro-area members such as France and Italy are in a different position: they borrow in a currency they do not control and depend on the ECB choosing to act.
Bloated debt, large deficits and rising interest rates are not new. The combination has occurred many times in history, including in Italy and Belgium in the 1980s and 1990s. What is new is that it is happening at the same time across the major Advanced Economies, including the reserve-currency issuers. Comparisons with the stagflationary 1970s are misleading: when Volcker raised rates to tame inflation, US debt was far lower (Debt/GDP stood at 43% in 1973, against 117% today). Back then it took a decade of high rates and deficits for the interest burden to become a serious problem; today the debt is already there, and it is far more sensitive to rates from day one.
So the point of the sustainability ratios is not to decide whether debt will be funded. It will be. It is to judge the price of that funding, and how that price will feed through to the broader economy and other asset prices.
If, as projected, US interest costs consume 27% of tax revenues by 2028 (and 30% by 2030), investors will demand significantly higher yields. They will anticipate that, to keep paying, the government will eventually lean on Option 4, artificially depressing yields or debasing the currency, and they will want to be compensated for that risk.
Three factors determine how exposed a country's bond market is to sustainability concerns, and therefore to higher yields:
Who owns the debt. Foreign investors can be fickle sellers, while central banks rarely sell except in calm conditions. A market with heavy foreign ownership (e.g. France) is more at risk than one with little foreign ownership and large central bank holdings (e.g. Japan).
Weighted Average Maturity (WAM). A government that has issued a lot of short-term debt (e.g. the US) refinances more often than one with a long WAM (e.g. the UK), so its interest bill responds faster to higher rates (and, equally, to lower ones).
Weighted Average Interest Rate (WAIR). If the WAIR is close to market yields, new bonds cost about the same as the ones they replace. Where maturing bonds carry very low coupons (e.g. the US and Japan), every refinancing adds substantially to the interest bill, which is why the US and Japanese ratios above are rising so quickly.
The 2010-12 euro peripheral crisis is a useful warning. Italy's debt was ~118%/GDP and the Berlusconi government was running a deficit of 4.5%/GDP, even though its primary (non-interest) budget was roughly balanced: the deficit was essentially the interest bill. Investors grew worried about sustainability, bond auctions cleared only at sharply higher yields, and the spread over Germany widened. The economy slowed abruptly, the stock market fell sharply, and political upheaval and market volatility followed.
Calm returned only in stages: the ECB began buying Italian bonds in August 2011, a new government under Mario Monti committed to fiscal consolidation in November, the ECB's long-term loans (LTROs) supported the banks that winter, and Mario Draghi's “whatever it takes” in July 2012 finally brought yields and spreads back towards normal. Italy is the clearest recent example of how a debt shock unfolds, and of what happens without a national central bank: the market set the price until the ECB chose to act.
The next paper will look at what happens around a bond market shock: its impact on other assets and on the wider economy, and the fiscal and monetary responses it is likely to force.
For now, the key point is the impending burden of rising interest costs on already stretched budgets. Short of fiscal consolidation that looks politically out of reach, the only thing that halts it is a meaningful fall in interest rates. But how that fall comes about matters:
If it has to be engineered by central banks while inflation is above target, rather than delivered by the market, it is Option 4 by another name.
If it is the result of an economic slowdown, the ratios are likely to deteriorate further, as tax receipts fall and the automatic fiscal stabilisers push up spending.
Comments