The interest rate on the US government’s 30-year bond dipped below 2% last week, setting a new record low. It was not that long ago that a three-month bond treasury bill paid that kind of interest.
For most of American history, the longest bond maturity the US government offered was 30 years. But historically low rates have revived a long-simmering discussion about whether the US should issue bonds with even longer maturities, like 50 or 100 years. Austria sold a so-called century bond earlier this year with a mere 1.2% yield (paywall). Even serial defaulter Argentina sold a 100-year bond in 2017, though those buyers may regret it now.
It may seems like a risky bet to lend money for such a long time, but there are reasons why now is a good time to both issue and—crucially—purchase ultra-long dated debt.
The case for borrowers
The case to issue long-dated bonds when rates are low is straightforward. The US government now has the opportunity to lock-in low rates for a very long time. It is a rare opportunity borrow for cheap and take decades to pay it off at a certain, low interest rate. This is ideal for funding infrastructure projects that can take many years to pay off.
No one knows how much longer such low rates will last. Some expect that low rates are here to stay because the population is aging and inflation is low, but some of the biggest buyers of debt may be cutting back. And rates could increase in the future, which means a higher cost of borrowing.
The case for lenders
A better question is: Who would buy an ultra-long bond? Why would anyone commit to such a low yield for so long? Odds are that rates will go up at some point over the next century, making the long-term bond worth less. To make matters worse, long-term bonds are riskier because their prices are more sensitive to interest rates. If a 10-year, zero-coupon bond yield increases from 2% to 4%, the price of that 10 year-bond falls 18%; for a 50-year bond, the price would fall by 62% (assuming you can sell the bond on the secondary market).
But this interest-rate sensitivity is exactly why investors want ultra-long term bonds, assuming they are protected for inflation. Suppose you are managing a pension fund and owe benefits to an employee who is 35 years old. They will retire in 30 years, and then you expect they’ll live another 30. That means in 30 years you’ll need to pay 30 years of benefits. This is the equivalent of a very long maturity bond. The value of this benefit varies with interest rates: $25,000 a year in 30 years is worth $322,000 today if interest rates are 2%. If rates are 4%, it worth only $121,000. Interest-rate uncertainty poses a big source of risk for pension funds and insurance companies, given their long-term obligations. They never know how much they need to save to ensure payouts in the distant future.
They can manage this risk by matching the duration (a weighted average of future payments) of their assets and liabilities. It is similar to owning bonds that pay out at the same time benefit payments are due. The duration of a 35-year-old’s pension benefit is longer than the duration offered by a 30-year bond. Pension funds and insurance companies are usually short duration, meaning the lack of long-dated securities exposes them to interest-rate risk. If you are 35 and have a 401(k) plan you are short duration, too, because you must pay yourself a pension one day using savings with an uncertain value far in the future (stocks don’t have any duration).
So far, bond analysts seem to assume that there will not be much demand for an ultra-long bond. But the Austrian experience and plunging yield on the 30-year Treasury suggests there may be a market for century bonds from a highly rated government like the US. In a world where most savers are short duration, it could be the asset we didn’t know we needed.