What Global Yield Curves Signal About 2020 Growth

Are Israel, the US, eurozone and Mexico heading for sharp economic slowdowns?  Is growth in China, Brazil, India and South Africa about to accelerate?  Those are the seeming verdicts of the countries’ respective yield curves.  But how much confidence should be place in them?

In our previous paper, we focused on the US yield curve as an indicator of US economic growth. Our research found two things:

  1. With the exception of a five-year period at the end of the 1990s, the US yield curve generally does a good job of forecasting slowdowns and accelerations in the pace of growth.
  2. The private sector yield curve does a better job of forecasting than the public sector yield curve.

This paper expands upon that analysis by looking at the relationship between private sector yield curve slopes and subsequent growth in GDP across 23 currencies.  Together these 23 currencies account for over $77 trillion of GDP, equivalent to 88% of total global output.  For each yield curve, we use the difference between 10-year swap rates (or the closest equivalent available) minus 3-month interbank rates (expressed as “3M10Y”) for as far back as we can get data.  Depending on the country, those numbers go back as far as 2006 and in some cases as far back as 1990.

The findings of this study are striking.  Looking three-to-five quarters ahead, every single one of the 23 yield curves correlates positively to future GDP growth.  Normally, we include only two lines (give or take) in each chart.  In Figure 1, we make an exception, including all 23, to illustrate the following point: if the yield curve had a random relationship with future GDP growth, there should be a 50-50 chance of the correlation being positive. The likelihood of all 23 currencies demonstrating a positive relationship simultaneously is 1/(2^23), or one in 8,388,608.

Figure 1: The graph is messy but message is clear; yield curve slope correlates positively with growth.

Cutting through the noise by taking an average of the 23 correlations, on balance, the slope of the yield curve, correlation with future growth peaks three to five quarters in advance.  In other words, a steepening of the yield curve today might translate into stronger growth about six to 15 months later.  A flatter yield curve implies slower growth about half-a-year to a year-and-a-quarter in the future (Figure 2).  At the same time, we don’t wish to overstate the significance of these findings.  In many cases the correlations are not particularly strong.  Moreover, as the late 1990s demonstrated in the US, under certain circumstances, such as a soaring equity market, the economy can defy a flat yield curve and grow anyway. 

Figure 2: Correlation between yield curve and growth peaks 2-5 quarters in the future.

The good news is that yield curves have shown themselves to be useful, if imperfect, indicators of the direction of the various economies in the past.  The not so great news is that of the 23 yield curves that we examined, 18 of them are pointing towards slower growth. 

To reduce a large and complicated data set into a single bite-size piece, we examined, relative to their own respective histories, how steep or flat each of the 23 yield curves have been on average in October and November 2019.  If the curve was the steepest it’s ever been, it would have a value of 100%; if it was at its historical flattest, it would have a value of zero. Only five curves, Brazil (BRL), Chile (CLP), China (CNH), India (INR) and South Africa (ZAR) have been steeper-than-average, pointing towards stronger growth.  In each of these five countries, central banks have been easing policy.  The People’s Bank of China has been slashing its reserve requirement ratio.  The others have been cutting interest rates.  The results have been the same: steeper yield curves and the possibility of stronger growth. 

By contrast, the other 18 yield curves have all been less steep than their historical averages, in some cases dramatically so.  For the eurozone (EUR), Israel (ILS), Japan (JPY), Mexico (MXN), Thailand (THB) and the United States (USD), yield curves are all in the bottom 10% of their historical steepness, pointing towards the likelihood of sharply slower growth ahead.  It’s especially striking that Mexico and the US are in this boat, since their respective central banks have both cut rates three times.  Most of the others, including those in Australia (AUD), Canada (CAD), South Korea (KRW), Norway (NOK), Poland (PLN), Sweden (SEK), Switzerland (CHF), Turkey (TRL) and the UK (GBP) are in the bottom third of their historical steepness, pointing towards the likelihood of somewhat slower growth ahead in 2020 and 2021 (Figure 3).  Russian (RUB) and Colombian (COP) yield curves have a more neutral footing relative to their respective histories.

Figure 3: Global interest rate markets suggest China & India will accelerate, Europe & US will slow.

On a GDP-weighted basis, the five countries that seem most likely to accelerate account for $19.5 trillion in 2019 GDP, according to the IMF, or about 25% of the total GDP represented by these 23 currency areas.  The other 75% ($57.9 trillion), looks set to slow, at least if past correlations hold in the near future.  That said, if global interest rate markets are correct, and China as well as India, see a rebound in growth, that could be great news for commodity prices and emerging market currencies.  China alone accounts for 40-50% of the demand for industrial metals like copper and exerts a strong influence on energy and agricultural markets as well as the fortunes of the world’s commodity exporters.

Before throwing in the towel on global growth, however, its worth asking at least a few questions:

  1. Are flat yield curves in places like Europe, Japan and even the US, a distortion caused by QE?
  2. Do negative deposit rates in Europe and Japan also distort the analysis?
  3. Could soaring equity markets offset tight credit conditions in fixed income markets?
  4. How confident should we be in this analysis from a statistical perspective?  Does it vary by currency?

The idea that the Federal Reserve’s (Fed) QE still distorts the yield curve is a hard case to argue.  For starters, the Fed stopped expanding its balance sheet in late 2014.  Secondly, it began passively shrinking its balance sheet in late 2017, reducing it from 25% to 18% of GDP.  Since September it has provided some liquidity to the repo market but is careful to distinguish these actions from a renewed embrace of QE.  Any remaining distortion from QE is probably small, especially given the explosive increase in Treasury debt issuance as the US budget deficit has spiraled from 2.2% to 4.7% of GDP over the past three years.

By contrast, the European Central Bank (ECB) has inaugurated another round of QE and neither the ECB nor the Bank of Japan (BoJ) ever shrank their balance sheets after previous rounds of QE.  One could argue that the extreme sizes of the QE’s (44% of eurozone GDP and 100% of Japanese GDP) distorted interest rate markets and artificially flattened yield curves – both at home and in markets abroad, including in the US.  This is undoubtedly true.

Likewise, negative rates in Europe and Japan rather than boosting economic growth appear to be holding it back while quantitative easing appeared to have little effect either way.  Negative rates may indeed be distorting yield curves as well.  That said, if negative rates are holding back growth in Europe and Japan, that is consistent with their yield curves being flat.  As such, flat yield curves in Europe, Japan, and even the US, might be accurately gauging the failure of QE and negative rates to stimulate growth, just as they may have accurately evaluated the potentially negative impact of so much Fed tightening.  The difference between the Fed’s 2015-18 tightening cycle and the ECB and BoJ’s experiments with negative rates is that the Fed meant to tighten policy.  The ECB and the BoJ tried to loosen policy but, by all appearances, accidentally tightened policy instead by imposing a negative interest rate tax on banks and savers.  As such, it’s not overwhelmingly clear that flat yield curves are really sending bad signals, at least in the case of Europe.

In terms of how much confidence one should place in the yield curves, the answer varies by country and currency.  Of the 23 currencies, Chile, China, Sweden and South Korea demonstrated the strongest relationships between yield curve shape and subsequent GDP growth, followed by Colombia, Canada, Turkey, Mexico, India, the eurozone, New Zealand, and Norway.  That eurozone growth had an above-average correlation between 3M10Y and subsequent GDP growth is impressive given the degree of the ECB’s QE and the existence of negative rates for one quarter of the eurozone’s history as a currency bloc. 

The weakest relationship, however, was Japan, where the BoJ did a QE that was in order of magnitude bigger than what the Fed or the ECB did.  In addition to buying far more bonds relative to the size of GDP, the BoJ also went down much further in terms of credit quality.  Unlike the Fed it didn’t limit itself to government bonds and AAA-rated mortgages.  It went for corporate debt as well and even equity-ETFs. Japan has also spent longer at zero rates than any other country and, unlike, Europe and US, experienced long-term deflation that has only been corrected during the past few years.

Also, unlike the Fed and the ECB, the BoJ’s QE actually did work.  Japan returned to growth and deflation flipped, becoming inflation for the first time in decades.  Japan’s distorted yield curve signaled none of this.  Perhaps Japan’s currently flat yield curve also isn’t really signaling a slowdown either.

Japan was followed by Israel, Brazil, South Africa, the US, UK, Poland, Australia, Thailand and Russia among the weaker relationships between 3M10Y and future GDP growth.  Switzerland was right in the middle of the pack.  You can’t count on much in life, but you can always count on the Swiss to be neutral.

That India and China, which demonstrated among the strongest relationship between yield curve slope and subsequent GDP growth, have relatively steep curves is encouraging.  Likewise, the fact that the flatter curves are in places like Israel, Japan, eurozone and the USA, which have shown weaker relationships between yield curve slope and subsequent GDP growth, may also be encouraging.  At any rate, the next few years will provide yet another test of whether or not yield curves do an accurate job of forecasting growth.

For the details of each market as well as the exact Bloomberg codes for each currency/country, please see the Appendix. 

Bottom line

  • Yield curves are pointing to stronger growth in China, India, Chile and Brazil.
  • Yield curves are pointing towards sharp slowdowns in the US, eurozone, Japan and Israel.
  • All 23 yield curves examined have a positive correlation with growth 3-5 quarters in the future.
  • There is a wide variance in the forecasting accuracy among the yield curves.


Figure 4: Chile, China, South Korea and Sweden have the strongest yield curve-GDP correlations

Figure 5: Canada, Colombia, Mexico and Turkey have the second strongest batch of correlations

Figure 6: The eurozone, India, New Zealand and Norway have the third strongest set of correlations

Figure 7: Australia, Russia, Switzerland and Thailand have the fourth strongest set of correlations

Figure 8: Poland, South Africa, UK and US have the fifth strongest set of correlations

Figure 9: Brazil, Israel and Japan have the weakest correlations between yield curve slope and GDP

Figure 10: Australia has data since mid-1993

Figure 11: Brazil’s data dates to 2004

Figure 12: Canada’s data goes back to the early 1990s

Figure 13: Switzerland’s data dates back to the early 1990s

Figure 14: Chile’s data begins in late 2004

Figure 15: China has consistent yield curve data since mid-2006

Figure 16: Colombia has consistent yield curve data since mid-2006

Figure 17: The eurozone is unique in that over a dozen countries issue debt in a single currency

Figure 18: UK data begins in the early 1990s

Figure 19: India’s data goes back to the early 2000s

Figure 20: Israeli data goes back to 2006 and has a low correlation to GDP growth

Figure 21: Deflation, Zero/Negative Rate Policy, QE weakened Japan’s yield curve-GDP correlation

Figure 22: South Korean yield curve times series date from the early 2000s

Figure 23: Mexican data goes back to the early 2000s

Figure 24: Norway’s data stretches back a quarter century

Figure 25: New Zealand’s yield curve data goes back to 1997

Figure 26: Poland’s data goes back to 2004

Figure 27: Russia’s yield curve timeseries begin in late 2006

Figure 28: Sweden’s data goes back to the early 1990s

Figure 29: Thailand’s data goes back to the turn of the century

Figure 30: Turkish data goes back to 2006

Figure 31: US private sector data goes back to 1989; public sector data many decades further back

Figure 32: South Africa’s data goes back to the mid-1990s


All examples in this report are hypothetical interpretations of situations and are used for explanation purposes only. The views in this report reflect solely those of the author(s) and not necessarily those of CME Group or its affiliated institutions. This report and the information herein should not be considered investment advice or the results of actual market experience.

About the Author

Erik Norland is Executive Director and Senior Economist of CME Group. He is responsible for generating economic analysis on global financial markets by identifying emerging trends, evaluating economic factors and forecasting their impact on CME Group and the company’s business strategy, and upon those who trade in its various markets. He is also one of CME Group’s spokespeople on global economic, financial and geopolitical conditions.

View more reports from Erik Norland, Executive Director and Senior Economist of CME Group.

Equity Index Options

Our Equity Index options on futures offer around-the-clock liquidity, market depth, and extensive product choice on the world's benchmark indices to suit a variety of trading strategies. Capitalize on potential margin offsets on futures and options strategies, advanced on-screen spreading capabilities, and the certainty of central clearing.

Start trading