economist

Global bond markets have been sending three overlapping warning signals about the health of the world economy. Bond yields have fallen to record lows. August saw yields on Australian 10-year government bonds fall below one per cent for the first time while the yield on 30-year US government treasuries slumped to a new low.

The value of bonds trading with a negative nominal yield rose to almost US$17 trillion by the end of August (more than a quarter of the global bond market).

Key segments of the US Treasury yield curve have inverted. The term spread between 10-year Treasury bonds and three-month Treasury bills has been negative since May, while the spread between 10-year and two-year Treasury bonds turned negative briefly in mid-August.

Start with record low yields. Remember, the nominal yield on long-term debt can be split into two components: the average of the short-term interest rates expected to prevail over the bond’s life and a term (or risk) premium. The former comprises projections of real short-term rates and inflation while the latter is the extra return investors require as compensation for the increased risk of holding longer-term debt. Today’s record low yields therefore imply a combination of very low central bank policy rates, very low inflation, and no significant risks to those expectations for a decade or more.

Moreover, not only are yields historically low, for more than a quarter of all bonds they are negative. As of 5 August, that means if an investor were to buy and hold to maturity, say, a German government bond of any maturity out to 30 years, they would expect to lose money on the transaction. The buyer is paying the German government to take their money while Berlin is being paid to borrow. Negative nominal yields first appeared in post-bubble Japan a couple of decades ago, but have since spread to the Eurozone, Sweden, Switzerland and Denmark, some European emerging markets and even in parts of the corporate junk bond market.

Curve ball

Finally, there’s the recession predictor that is the slope of the US yield curve. The slope reflects the difference between the yield investors require to buy debt with a short-term maturity and debt with a longer-term maturity. Since expected short-term rates should on average be equal to the current short-term rate, while the term premium is assumed to be positive to compensate for the greater risk involved in longer-term investments, a “normal” yield curve slopes up. But if the yield on long-term debt falls below that on short-term debt, the slope turns negative and the curve “inverts”. That pattern of yields can manifest when investors expect the central bank to cut rates in the future, for example to combat a looming recession.

Historically, US evidence for the yield curve’s forecasting ability is compelling (although that relationship is weaker elsewhere, including Australia). The spread between 10-year and three-month US Treasuries has been negative before each of the past seven US recessions — a recession typically following a persistent inversion with a lag of 12–18 months. That past predictive power makes markets very sensitive to Treasury yield curve inversions. Together, these three indicators would usually imply bond markets foresee a future global economy mired in low growth and lowflation for years to come, along with a significant risk of a US and likely global recession in the next one to two years.

There is one major complicating factor in the form of the market distortions created by unconventional monetary policy. This has brought negative policy rates to economies comprising nearly 25 per cent of world GDP at market prices, including Japan and the Eurozone, along with massive asset purchase programs under Quantitative Easing (QE). That may have compromised traditional market indicators. QE and a decline in the inflation risk premium, for example, have likely served to compress the term premium and even turn it negative, flattening the yield curve, thereby making inversions more common even without any change in recession probabilities. This time might be different, but don’t count on it.