U.S. Government Bonds are Junk Bonds

The risk of high inflation

When you buy a bond with a rating of BBB or higher (Baa using Moody’s), the probability of default is supposed to be less than .1 percent per year. Over a period of ten years, you should have a 99 percent chance of being paid back in full. These are called investment-grade bonds, as opposed to junk bonds.

If you buy a ten-year U.S. government bond, what are the chances that you will be repaid in full? Are they 99 percent or more?

If we are talking about being paid back in dollars, I would say that the answer is “yes.” But what about the purchasing power of those dollars?

To keep the arithmetic simple, suppose that you invest $100 in a bond, and you get no interest until the bond matures. Interest is, say, $5 per year, so that in ten years, when the bond matures, you get $150.

Suppose that $150 ten years from now will buy the equivalent of $100 today. Then you just break even on the bond. That corresponds to inflation of roughly 5 percent per year, not compounded. If inflation is less than that, you do better than break-even, and if it is more than that, you do worse.

I think that it is fair to say that if inflation averages more than, say, 8 percent over the next ten years, then long-term government bonds that you buy today will have effectively defaulted. So if you are going to rate long-term government bonds as being investment grade, you have to assign a very low probability to such a high-inflation scenario.

As you may have heard, one of the credit rating agencies recently downgraded U.S. debt from AAA to AA, which is nothing. Government debt is still considered high quality. That is because its rating only concerns the risk of hard default, not the soft default of rapid inflation. I agree that the risk of hard default is low, but I think that soft default is something to be worried about.

One way to avoid soft default is for there to be a dramatic pickup in the rate of economic growth. That would boost tax revenue and enable the U.S. government to run a balanced budget, or even to enjoy surpluses. This scenario assumes spectacular gains from the application of innovation in biotech and AI. Call this scenario A.

Another way to avoid soft default is for a bipartisan effort to reduce growth in government spending. This almost certainly would require cuts in Social Security, Medicare, and Medicaid. Call this scenario B.

If we have neither scenario A nor scenario B, then government will have to issue more and more debt to finance its spending, including spending to pay the interest and principal on past, present, and future debt. That is what analysts at the Congressional Budget Office and elsewhere worry about when they describe the U.S. fiscal position as unsustainable.

This ever-increasing debt issuance will flood the economy with paper wealth that is not derived from economic activity. This flood of government bonds seems certain to fuel inflation. As inflation goes up, government interest cost goes up, the cost of entitlements and other forms of government spending goes up, and you get into a vicious spiral.

The inflationary spiral does not have to go so far as it did in Weimar Germany, where there was hyperinflation. But inflation could run at 10 or 20 percent, far above the Fed’s inflation target and far above the expectations that are embedded in the market prices of long-term U.S. government bonds right now.

What is your estimate of the chances for scenario A, where we grow our way out of debt? Or scenario B, where we put our fiscal house in order? What does that imply for the chances of an inflation spiral? I think that the chances of the latter are high enough that if it were up to me, I would rate long-term U.S. bonds as junk bonds.

P.S. I am not sure how to rate inflation-indexed bonds. They should be safer than regular bonds, but there are still some risks. They did not do well in the inflation surge last year, when the inflation-adjusted interest rate rose, with bond prices moving inversely with real yields.