How Much Government Debt is Too Much?

The “virus recession” and the policy response to it is going to bring record deficits in dollar terms AND near-record levels in % of GDP as well. For many countries, this comes on top of an already-high government debt. Many people seem to wonder if there is some magic and we can just go on like this no problem. Here is a text to think clearly on this topic.

What does NOT matter is the dollar amount of debt. This must stop. If a person earning 10 million per year and without any other debts has a debt of 1 million dollars, it is honestly not that much of a big deal. The same dollar amount of 1 million dollars of debt for a person earning 100k per year is another story.

What matters is debt relative to income or potential income / sources of financing. Lots of people compare government debt to a household debt, and in SOME ways, it is OK: you can’t just spend to infinity without thinking at all. That is obvious. But there are 3 differences for the government relative to a household, and they matter a great deal:

1. The government can ultimately “print money” to pay its own debt. Although this is not officially allowed in most industrialized countries, that is what is happening in reality: central banks are buying government debt by the truck loads (and this is equivalent to printing money to pay for debt). The government then owes the debt held by the central bank to… the central bank… and the central bank pays the immense majority of its profits to… the government! So, debt held by the central bank is the government owing money to itself: it doesn’t really “count”! That is why the simplistic measure of “total debt / GDP” is misleading, because it counts the debt held by the central bank and it should not. A much better measure of “debt pressure” would be “debt outside the central bank / GDP”.
2. The government can ultimately increase taxes to pay it’s debt obligations, which a household can’t do. This is politically explosive, but if push comes to shove, it is possible. Countries with already-high tax burdens (France, Germany, Italy) can’t afford this luxury as much as those who currently have low tax pressure in general (USA, Australia, Switzerland), while those with mid-way tax pressure also have mid-way space to cover the burden of debt with tax increases (Canada, UK, Japan).
3. The government can push back forever the actual payment, because the government can “re-borrow” (debt rollover) when payments are due and does not have the money. This is not an option for individuals, who must ultimately pay debts at one point.

Of course, once a government gets into “debt trouble” and the heat is on and increasing debt more becomes impossible and even “rolling over the debt” becomes hard and / or too expensive, the choices boil down to:

1. Tax more.
2. Cut government spending in a major way.
3. Print money to pay the debt.

The first 2 options are politically sensitive and even socially sensitive: who and how and how much will you tax more? Where is the bulk of government spending that could be cut to get some extra breathing room? If military spending cuts are not an option due to geopolitical strategies, then the only big ones are “everything social”: social security, healthcare, education, etc. This is rough. So all options become explosive… but you can also print!

When is debt too much? What really matters in the end is the DEBT SERVICE relative to GDP, because GDP is the size of the “taxable cake” and also gives an order of magnitude of how much money could be printed (bigger GDP = more space to print money).

Here is my calculation-approximation of total government debt service (interest payments on the debt) as % of GDP. It is an OVER estimation, but it is a good approximation:

The debt service of the US government has almost almost never been as low as it currently is in the last 60 years! Obviously, there is no current problem, as I explained in “Is There a Government Debt Problem”?

So what is the metric for the limit? What is “too much”? There are 3 variables to this puzzle that are really just “one” variable, because they are so intertwined:

1. Interest rates.
2. Inflation.
3. The value of the currency.

I don’t want to give a full course on the bond market, so I’ll just make it short and oversimplified for brevity: the government “borrows” by selling “bonds”, which is “debt”. These are sort of “contracts” that say something like this: “I am selling you this piece of paper for 98 000$ now and I’ll give you 100 000$ back next year.” This is a simplification, but it is essentially like that. As long as people WANT to buy the bonds of a government, then the prices of these bonds remain high and the interest rate that the government pays on its debt remains low (prices of bonds and interest rates go in opposite direction).

If people start to NOT want the bonds of the government for some reason, then demand drops, prices drop, and interest rates can increase rapidly, as experienced by Greece a while back:

But… when the public decreases its demand for the bonds of the government, the central bank can buy then instead, and that’s easy for the central bank, because they go on the “ central bank computer”, they type in the amount they want, and they can use that newly and freely created money to buy those bonds! MAGIC! By buying government bonds, the central bank removes those bonds from circulation, thus reducing supply, increasing prices, and decreasing interest rates (remember that prices of bonds and interest rates go in opposite direction).

The limit of this magic is inflation. As long as inflation is in check, the central bank can buy all the bonds in the world, and as previously explained, this is money the government pays to the central bank and gets back anyways, so it’s not “real” debt. Now the party comes to a grinding halt when inflation starts to rise uncomfortably and for a prolonged period. Not one or 2 months or even a year… I mean more than a year of clearly rising inflation… 3%… 3,5%… 4%… 5%… at one point, the central bank is forced to stop the printing party and increase interest rates… and that could make an easy-to-manage debt service become hell-on-Earth very quickly, because 1% of a large debt could be OK, and 2% on a large debt could become hard to tolerate… and 3% could be a catastrophe. So inflation is the ultimate “limit” of government debt. No inflation? Borrow and print baby!

Finally, and this really is related to the inflation and “bond demand” variables, the value of the currency: if a considerable chunk of the government debt is held by foreigners, they fear that the value of the USD will fall (lets say we are speaking of US government debt), because the USD payments they receive from the US government would then be worth less and less in THEIR currencies… this is a form of “implicit default”: I pay to you the amounts, but what I pay is worth less than before and less than what was expected!

Some governments could have not that much of a debt towards foreigners, but a “domino effect” could start if those few foreigners decide they no longer want the bonds of your government: others see this and don’t want to hold those bonds due to a potential price drop… and prices drop… and interest rates rise… and the currency crashes because the rest of the world no longer wants the bonds and other assets of your country, and the dominos start falling.

Is this likely for the USA? From what I can see, no. The only potential danger I see is that all this “stimulus” will still be floating around once this virus passes and the economy could overheat and bring inflation, hence rising interest rates on the back of a much bigger debt… but inflation in rich countries is simply nowhere to be seen and has been quite in control for a LONG time, so it is hard to imagine an “inflation spike” and that is why I am not overly worried. But you never know.

So what should we look for in terms of potential signs of problems? Here are the variables I see as major flashing points:

1. Inflation: steadily rising above 3% and 4% would become risky.
2. The 10-year government bond yield: convincing increases could signal some problems ahead.
3. The value of the currency: a steady and prolonged drop would signal falling asset demand from the rest of the world, and that could also be a sign of things to come.

Pascal Bedard

Walking towards a new climbing area to explore new crags.