Students of inversions in the bond market – when short bonds pay more than long bonds – worry that each dip in yield heralds a recession. Historically, these inversions are road signs to potential recessions, but not every inversion is followed by a contraction in the overall economy. Furthermore, not every contraction has been preceded by an inversion.
There were nine inversions in the U.S. bond market between Sept. 1, 1966, and Dec. 31, 2018, but only seven recessions. An astute investor might say an error rate of 22% isn’t reason enough to flee the market. Yet an oracle who’s right 78% of the time deserves respect.
On April 22, 2019, Government of Canada bonds offered 1.62% for two years, 1.61% for five years, 1.79% for 10 years and 2.08% for long-term. The curve for U.S. bonds offered 2.38% for two years, 2.36% for three, 2.59% for 10 and 2.99% for 30.
The two inversions are just one or two basis points; they are close to the front of each curve and neither qualifies as a macroeconomic crisis. With a very flat curve, these inversions can appear and disappear daily. What is significant is that 30-year bonds barely cover anticipated inflation.
Furthermore, the present inversions are less about changing macroeconomic trends and more about new institutional trading and investing rules.
Bond yields, which move inversely to bond prices, are down because long government bonds are up in price. The price trend reflects hefty buying of the longs by banks that are required by Basel III, the international regulatory framework, to hold their reserves in government bonds. Longs pay better than shorts now, so that’s where banks like to put a lot of reserve capital. That drives up prices and pushes down yields. For banks, the flatness of the curve is a minor curse, since they cannot charge as much for mortgages over GIC rates as they could with a steeper yield curve.
Life insurers also are required to hold reserves in government bonds. For them, flatness merely affects pricing when setting premiums.
“The increase in regulatory demand for safe assets is the issue,” says Avery Shenfeld, managing director and chief economist at CIBC Capital Markets in Toronto. The long end of the curve is held down by institutional buying, he says, so short rates don’t have to rise as much as they did in the past. Thus, inversions are easier to come by, he says. “In the ’70s, ’80s and ’90s, there was an extra reward for locking money away for 10 to 30 years. With prevailing low rates, the premium has been wiped away and inversions happen more easily.”
Pension funds and authorities running infrastructure projects also buy long bonds. For example, a bridge or highway authority that wants to have funds in 20 or 30 years for major maintenance can use longs to build reserves.
Demand for longs is such that if a central government decides to issue less long-term debt, the provinces will issue their own debt to fill the space, says Edward Jong, vice president for business development at T.I.P. Wealth Manager Inc. in Toronto.
There is little risk of long debt flooding the government bond market, however. “Governments are issuing at the short end and moving debt from the long end. That process tends to flatten the curve,” says Charles Marleau, chief investment officer at Palos Management Inc. in Montreal.
One could have said 30 years ago that inversions indicate reduced confidence in the future – perhaps stagnation from poor expectations of growth, for example, Shenfeld notes. The curve isn’t such a pure predictor anymore because central banks injected a great deal of cash into the banking system by buying bonds in the 2008 rescue. That buying spree drove down rates, and yield curves are just normalizing now.
Nonetheless, Chris Kresic, head of fixed-income and asset allocation at Montreal-based Jarislowsky Fraser Ltd., says the long end of the curve is pricing in slowing global growth. That adds to government pressures to hold long rates down – if only to make paying the debt piled up during the rescue mission after 2008 easier.