The US yield curve: Risks of inversion

The 10-year bond has been a good proxy for the natural or neutral rate for many years. This may not be so now

Recently most market observers have been very worried about a possible inversion of the yield curve in the US. The yield curve plots the rates and maturities of debt instruments on a curve and is conventionally used as a measure of the difference between short-term interest rates and long-term interest rates. The two classical definitions of the Treasury yield curve are the gap between 10-year bond yields and three-month treasury bills (10-year/3-month) and the gap between 10-year yields and two-year bond yields (10-year/2-year). The yield curve is tracked closely as it has the best long-term forecasting record of any macroeconomic variable. Once the yield curve inverts (short-term rates move higher than long-term rates) within 12 to 15 months, the US economy moves into a recession and stocks enter a bear market. The indicator has a great track record, having given only one false signal in the last 50 years (September 1966-March 1967). Even the path of flattening of the curve usually indicates a slowing economy as policymakers and markets try to rein in economic growth.

While the risks of inversion on the (10-year/3-month) curve are still some time away, the curve has been flattening and the spread is now a little over a 100 basis points. It is on the alternative definition of the (10-year/ 2-year) curve that the chances of inversion are much higher. Two weeks ago, this yield spread had dropped to less than 35 basis points. A little more tightening on the short end, and we could be close. 

Yields across the curve have come off in the last week as the Federal Reserve in the US has signalled a willingness to be more tolerant of any short-term spikes in inflation.  The use of the word symmetry in their commentary has led market observers to believe that we will see the Fed move slower than anticipated in hiking rates and be more tolerant of allowing an inflation print of over 2 per cent. The pressure seems to be off the bond market, at least temporarily, with yields dropping back below 3 per cent on the 10-year yield, but the concerns on inflation and the yield curve are bound to resurface, it is only a matter of time.

Why is this indicator important? What does it really signify? 

Illustration: Binay Sinha

At one level, the yield curve is simply a shortcut to assess monetary policy tightness. It measures the gap between the policy rate and the neutral rate (a theoretical construct, the rate at which monetary policy is neither a headwind nor tailwind to the economy). The three-month Treasury bill or two-year bond is good proxies for policy rates, while the 10-year bond is a proxy for the neutral rate (imperfect proxy but the best historically). When the gap between the policy rate and the neutral rate is significant, monetary policy is loose and economic conditions tend to be robust. If the policy rate is less than the neutral rate (inverted curve), the policy is very tight and will hurt economic growth. Given the leads and lags in the effectiveness of monetary policy, the yield curve is a leading indicator. It gives everyone enough time to react.

An alternative explanation is put forward by the followers of the Swedish economist Knut Wicksell. They argue that there are two interest rates in the economy. A natural rate, which is meant to track the rate of corporate earnings growth in the economy. The best proxy for corporate earnings is nominal GDP growth, and the best proxy for nominal GDP growth is nominal rates on long bonds.

The second rate is the market rate, cost at which companies can borrow. A good proxy for this is short rates. Followers of this school of thought believe that business cycles are based on the interaction between these two rates.

Whenever the natural rate is higher than the market rate, economic actors will borrow as they can earn a return better than their cost of borrowing. Whenever the market rate exceeds the natural rate (yield curve has inverted), companies should pay back debt as they cannot cover their cost of capital. As companies deleverage, they cut capital spending, lay off workers and debt service for weak companies suffers. This will tip the economy into a recession.

Both approaches implicitly use proxies to determine what is the neutral rate in the economy, or the natural rate at which earnings and the economy grow. Both approaches assume that the 10-year bond is a good proxy for this purpose. 

Herein lies the catch. In most circumstances, over the last 50 years, the 10-year bond has been a good proxy for the natural or neutral rate. Today, however, given the extraordinary monetary accommodation over the last decade, these rates may not be accurate. We have seen massive quantitative easing, zero interest rates and other measures being taken globally. There is a huge demand for safe assets.

Many market bulls argue, as an example of policy driven aberrations, that the term premium on 10-year bonds is artificially low today. The term premium is the incremental yield you will receive for holding duration. Today it is negative, highly unusual, over the last 50 years, term premium has averaged 150 basis points. To the extent the term premium is artificially suppressed, even the 10-year yield is lower than its normalised level for this stage of the economic cycle.

The only time the inverted yield curve has failed as a signal was in late 1966 when also the term premium was unusually low at 25 basis points. 

Market bulls argue that because the term premium is artificially low, an inversion of the yield curve cannot be relied upon as a recessionary signal. Had the term premium been at more normalised levels, 10-year yields would be higher, and we would be far from yield curve inversion.

The yield curve is an important macroeconomic indicator of monetary policy, with an enviable track record of predicting economic recessions. While we are witnessing a flattening of the yield curve and may even be close to seeing an inversion of the curve on some indicators, we need to understand the economic context. 

The unprecedented interventions in monetary policy over the last ten years, should at the least force us to normalise variables before assessing their economic impact. We cannot look at the yield curve without making these adjustments. 

The writer is with Amansa Capital