The slippery slope

 

In a recent blog post, the White House reflected on the topic of recession, stating that “it is unlikely that the decline in GDP in the first quarter of this year – even if followed by another GDP decline in the second quarter – indicates a recession". When public authorities start managing market expectations in the same way as central banks do, one could be forgiven for raising a conspicuous eyebrow. Sure enough, less than a week later a negative US GDP print for Q2 was released. Indeed, most of the macro data now signal a significant slowdown, not only in Europe but also in the US. This holds true in the case of housing data, leading indicators and even jobless claims (which are still very low but are showing signs of bottoming. The consistency of data is enough to avoid doubt in anybody’s mind: the world economy is slowing down, at least in real terms. While the Q2 earnings season is not yet showing many signs of this slowdown, as sales growth remains elevated, markets know better. Investors have started looking past this to ask the question of whether “we are knocking on the door of a recession?” In such a context, we believe it is essential to pay attention to the fixed income world, especially to the 2-10 slope; however, this does not necessarily reflect the typical ‘inverted yield curve predicts a recession’ framework. Here is why.

The 2-10 slope is the difference between the 10-year and 2-year nominal yields attached to government bonds. Its variations are usually regarded as an advanced indicator of economic health: a high spread (‘steep’ curve) is indicative of strong growth conditions whereas a low spread (‘flat’ curve)is consistent with a more challenging environment. This can be explained by the expectation element implied in yields: the 10-year yield is seen to represent an average of all future overnight rates over the next ten years. When a recession is expected, markets start pricing the fact that central banks will lower rates to a point that eventually puts the 10-year yield into decline versus the 2-year yield. Fixed income portfolio managers know these cycles all too well: before a recession, the central bank fights against inflation. By doing so, short term rates rise while longer term rates do not rise as much – a “bear flattening” period. This may result in an inverted yield curve, prompting a wave of forecasts that a recession is coming in ‘x’ number of months. However, once the recession starts, the opposite happens: the central bank cuts rates, often at a faster pace than expected, leading short-term rates to decline faster than longer-term rates – a “bull steepening” episode. These episodes are then followed by two other periods:

• a “bear steepening” period: 10-year rates rise faster than 2-year rates as investors cut their long-term bonds in favour of risky assets.
• a “bull flattening” period: investors start to add long bonds again, causing 10-year yields to decline faster than 2-year rates.

Chart 1 illustrates how the 2-10 slope has evolved during previous recessions (based on US data). Two years before a recession, the 2-10 slope typically declines (a bear-flattening episode) – this is similar to today’s situation. Yet, investors are currently overlooking the fact that upon entering a recession, the 2-10 slope starts to steepen (a bull steepening episode kicks in). This indicator may very well be one of the best anchor-points we should look at during such times: as the economy enters a recession, the slope of the term structure of interest rates steepens violently. Over the past six recessions, as the US economy was entering the recession, the 2-10 slope rose by an average 123 bps during the three months that followed. The sharpest steepening episodes occurred during the first three months of the late 70s / early 80s recessions (incidentally these were the only recessions which began with an inverted 2-10 curve).  This tells us that a coming recession is announced by a flattening curve, but it is nowcasted by a steepening curve – and these days nowcasting is key to gauging recession risk.

 

Chart 1: 2-10 slope development through US recessionary periods (1977-2022)

 

Source: Bloomberg, LOIM. Moves are relative to the starting ‘spot’ point of the 2-10 slope for each individual recession.

 

By this measure, we are not in a recession yet as we have not seen this steepening move. This raises the following question: what could cause such steepening to happen? Or, more specifically, what type of data? Most economists know that the 2-10 slope is generally linked to job market data. Chart 2 illustrates that when jobless claims start to pick up (a growing number of US workers are seeking unemployment benefits) the slope for the term structure of interest rates rises. This can be explained by the Federal Reserve’s (Fed) dual mandate to achieve maximum employment and price stability: once unemployment ticks up, it becomes much harder for the Fed to focus solely on the price component of its mandate. This worked remarkably well in 1990, 2001 and 2008. But Chart 2 actually shows much more than this simple message. It depicts two essential pieces of information relevant to today’s situation:

• First, when observing the previously mentioned three recessions (1990, 2001 and 2008), irrespective of the rise in jobless claims, the increase in the slope of the term structure was more or less the same and quite orderly. The depth of the crisis did not fully explain the level of the slope. Note, we have intentionally left the 2020 crisis out of this chart, as the Fed’s more activist stance capped the steepening of the curve efficiently as public spending was used to counteract the impact of lockdowns.
• Second, while the relationship holds well for these three recessions, those that took place in the 70s and 80s tell a somewhat different story. Persistently high inflation pushed the Fed into maintaining high rates, after a brief pivot, even though the US economy was re-entering a recession. When faced with the choice of supporting the economy or fighting inflation, the Fed chose to go with the latter. Sound familiar?. The 2-10 slope rose sharply during the first three months of the first recession, before retreating substantially and flattening. The steepening episode then restarted in 1981 and remained volatile. In fact, the existence of Quantitative Tightening (QT) this time around is likely to add further volatility to the mix. While we expect QT to support a steepening narrative – it is a buying force that has been driving long-end rates lower for the best part of a decade and is now working in reverse – but how it might run during a recession is currently unknown.

This historical perspective reinforces our call for caution: during periods of controlled inflation rises (1990, 2001 and 2008), steepening happened relatively smoothly and with comparable scale. In the 70s-80s, the need to maintain high rates even as the recession hit the economy prompted 2-10 variations to be much more volatile. If history provides a lesson, it is that persistent inflation can supersede recessionary concerns and therefore exploiting moves in the curve will require more vigilance and nimbleness. Fed chair Powell is not a monetarist, but he is showing increasing dogmatism on the inflation question, which likely reflects political pressure. This is an element that must not be overlooked by investors.

 

Chart 2: Jobless claims vs. the US 2-10 slope (1976-2019)

Simply put, the 2-10 slope historically rises when a recession hits the US economy. Yet, if history is any guide, stubbornly high inflation and a dogmatic, inflation-fighting Federal Reserve may cap and complicate its progression.

 

---

Macro/Nowcasting Corner

The most recent evolution of our proprietary nowcasting indicators for world growth, world inflation surprises and world monetary policy surprises are designed to keep track of the latest macro drivers making markets tick. Along with it, we wrap up the macro news of the week.

In the US, most news flow points to slower growth momentum: the CFNAI index declined from 0.01 to -0.19, having been expected at 0. The Dallas Fed manufacturing declined from -22 to -18, expected at -18. New homes sales declined from 696K to 590K, expected at 655K. Only the Richmond Fed manufacturing survey rose from -11 to 0, having been expected at -14. Jerome Powell mentioned these mounting recession risks, without making a call for the US to be in a recession at the time of the meeting. The advance GDP numbers added to this overall “declining growth” picture: in annualised terms, US GDP declined by 0.9% over Q2 after its -1.6% Q1 technical decline. Consumption remains solid so far, but its growth rate has moderated.

In the Eurozone, European Commission surveys are showing further signs of a macro deceleration – but not of a recession yet. Consumer confidence shows the most worrisome reading, while it failed to decline this time, it remains at -27 (a lower level than in 2020). Industrial confidence declined from 7.4 to 3.5 (expected at 5.4). Service confidence declined from 14.8 to 10.7, expected at 13.3. These numbers will have to fall below 0 to show the real tremors of a recession, but these surveys show how the deceleration is at our gates. 

In China, industrial profits showed a first sign of recovery. In year-over-year terms they rose by 0.8% versus a contraction of 6.5% last month. This improvement is one of the many signs that explain the recovery of our China Growth nowcaster.

Factoring these data points into our indicators, our nowcasting indicators currently point to:

•    Worldwide growth is clearly declining. The US and Eurozone have entered a declining growth period. In China, growth momentum remains subdued, but more than 50% of data are now improving.
•    Inflation surprises should remain positive for the Eurozone but are declining elsewhere.
•    Monetary policy is set to remain on the hawkish side: central bankers are likely to be more hawkish than expected.

World Growth Nowcaster: Long-Term (left) and Recent Evolution (right)

 

World Inflation Nowcaster: Long-Term (left) and Recent Evolution (right)

 

World Monetary Policy Nowcaster: Long-Term (left) and Recent Evolution (right)

Reading note: LOIM’s nowcasting indicator gather economic indicators in a point-in-time manner in order to measure the likelihood of a given macro risk – growth, inflation surprises and monetary policy surprises. The Nowcaster varies between 0% (low growth, low inflation surprises and dovish monetary policy) and 100% (the high growth, high inflation surprises and hawkish monetary policy).