Is the yield curve back to the 1950s-1960s?

In recent years, the yield curve has flattened and shifted downwards. The flattening has received a lot of attention because it is viewed as a recessionary signal. In this article, we argue that the fact that the curve has shifted downwards as well as flattening indicates that the yield curve may have changed regime. This regime shift seems to have been caused by lower inflation expectations and lower risk premia. In an environment of lower inflation and lower risk premia, the yield curve may occasionally invert without signalling a recession as it did in 1966.

1. The recessionary signal

Inversions of the yield curve have often been used to predict recessions. The 10-year to 3-month spread between US Treasury yields became briefly negative twice this year in March and the May (figure 1), spurring debate about whether this was signalling a forthcoming recession. The negative spread was mainly the result of a decrease of the 10-year yield – from 2.8 per cent on 1 March to 2.4 per cent on 26 March. A flat or inverted yield curve may be a signal from fixed income markets that short-term rates are too high for the current state of the economy and may risk tipping the economy into recession (Brainard, 2018).

Figure 1: 10y-3m yield spread

Source: US Treasury

Following on our previous analysis (Lenoel, 2018), we have updated our metrics on the probability of a recession. The yield spread averaged 0.17 percentage points in March, which was about the same value as in January 2006, two years before the start of the Great Recession. Based on our estimation, the probability of a recession in 12 months implied by the 10-year to 3-month yield spread (thereafter simply called yield spread) stood at 31 per cent for the average spread during March 2019 (figure 2), significantly higher than the unconditional probability of 13 per cent.

Figure 2: Implied probability of recession

2. Lower term premium

In the following, we propose an alternative interpretation of the current state of the yield curve. The yield curve has not only flattened but also shifted downwards. The low and flat yield curve is the consequence of low inflation expectations and low risk premia. Looking back in history, we notice that the yield curve indicates that the US economy may be back to a situation last experienced in the 1953-1971 period.

An important issue is whether in current economic and financial conditions the yield spread is as informative as it used to be. It might be the case that asset purchases by the Federal Reserve have distorted the signal from the yield curve. Consistent with this, Bonis, Ihrig and Wei (2017) estimated that the three US quantitative easing programmes jointly reduced the 10-year Treasury yield by about 100 basis points by lowering the term premium. However, figure 3 shows that the reduction of the term premium has followed a much longer time trend, with the term premium estimated to have shrunk from about 4 per cent in the early 1980s to a slightly negative value recently. The term premium is estimated to be back to a level similar to what was the case in the 1960s when the term premium was also close to zero.

Figure 3: Implied 10y term premium using 2 models

Source: Adrian, Crump and Moench (2013), Kim, Wright (2005), Federal Reserve

3. Lower inflation expectations

Inflation expectation has followed a similar downward trend since the 80s. Figure 4 shows how 5-year expected inflation declined from close to 10% in the 80s to close to 2% recently, according to the Survey of Consumers published by the University of Michigan. While most of the decline happened in the 80s and 90s, expected inflation continues to be on a declining trend, averaging 2.9% in the 2000s, 2.5% since 2016. The latest reading of 2.4% in June 2019 is the lowest 5-year expected inflation since the survey started in 1979.

Figure 4: Expected Changes in Prices During the Next Five Years

Source: University of Michigan

4. The new regime

Figure 5 characterises the shifts in the yield curve during the past seven decades by separating periods of low or high yields and periods of low or high yield spreads. In the two decades after the Korean War, inflation was low and stable between 0 and 4 per cent. Both the 10-year yield and the yield spread were low; averaging respectively 4.3 per cent and 0.9 percentage points between 1953 and 1971. Investors required hardly any term premium. The 1970s and early 1980s were periods of rising inflation and inflation expectations fuelled in particular by the two oil crises of 1973 and 1979. The risk of persistently high inflation pushed the 10-year yield to a maximum of 15.3 per cent in 1981 and the Federal Reserve responded with an aggressive disinflation policy led by Chairman Paul Volcker (Goodfriend, 1995). While the fight against inflation was successful in bringing inflation below 4 per cent as soon as 1983, the term premium stayed elevated for a remarkably long period, which meant that the yield spread stayed relatively high, averaging 1.9 per cent between 1986 and 2016.

Figure 5: Changes in the yield curve regimes
Source: NIESR

Figure 6 shows graphically how the yield curve has first shifted downwards then flattened since the 1970s-1980s. Remarkably, it has even shifted slightly lower than the 1950s-1960s period.

Figure 6: Changes in the yield curve regimes
Source: Federal Reserve

An inverted yield curve may signal a high risk of recession the following year as it has done successfully on several occasions in the past. But our forecast based on a wider range of economic data does not support this view. An alternative interpretation is that the relationship between yields and spreads may have shifted to a position represented by the lower left quadrant of figure 5 and characterised by ‘low’ yields and ‘low’ spreads, as was the case between 1953 and 1971. By lowering long-term yields, quantitative easing may have contributed to this possible shift. If we are indeed in a ‘low’ yield, ‘low’ spread environment, the yield spread may occasionally turn negative without signalling a recession, as it did in 1966.

References

Bonis, B., Ihrig, J.E. and Wei, M. (2017), ‘The effect of the Federal Reserve’s securities holdings on longer-term interest rates’,GMU Working Paper in Economics, 20 April.
Brainard, L. (2018), ‘Sustaining full employment and inflation around target’, speech on 31 May at the Forecasters Club of New York.
Chinn, M. and Kucko, K. (2015), ‘The predictive power of the yield curve across countries and time’, International Finance, 18(2),pp. 129–56.
Estrella, A. and Mishkin, F. (1998), ‘Predicting U.S. recessions: financial variables as leading indicators’, The Review of Economics and Statistics, MIT Press, 80(1), February, pp. 45–61.
Goodfriend, M. (1995), ‘Acquiring and maintaining credibility in low inflation: the US experience’, Chapter 8 in ‘Inflation targets’,Centre for Economic and Policy Research
Kim, D.H. and Wright, J.H. (2005), ‘An arbitrage-free three-factor term structure model and the recent behavior of long term yields and distant-horizon forward rates’, Finance and Economics Discussion Series, Federal Reserve Board
Lenoel, C. (2018), ‘Predicting recessions in the United States with the yield curve’, National Institute Economic Review, 244 May