The Signature of a Slump
Where is the Recession?
In a previous article, titled “A Brief History of the Rumbling Recession”, I argued that the extreme pace of monetary tightening since the beginning of the current US rate hiking cycle has caused enough damage to the financial system to start slowing down the real economy as all the leading indicators have been suggesting.
You can read that article here:
For me (and no doubt many other analysts), the US bank failures in March 2023 were a sign that the chickens were coming home to roost. However, stronger-than-anticipated GDP growth, housing market performance, employment market conditions, and consumer data in recent months suggest that the economy is still performing well and has not yet shown signs of slowing down. So, what does this mean for our recession outlook?
1. What factors does the NBER consider when determining an economic recession?
(4 Points & 4 Charts)
1.1. Industrial Production and GDP
Although there are some false signals on this chart, it’s clear that these metrics dip below zero during recessions. Afterall, a slowdown in the economy, and thus GDP, is the definition of a recession. However, these two metrics are lagging indicators and can’t provide any signal other than after the fact. They have been trending down over the past few quarters but are still in positive territory.
1.2. Corporate & Personal Income
The correlation between these metrics and past recessions seems patchy at best. No reason for alarm here, yet.
1.3. Wholesale & Retail Sales
These metrics seem to correlate well with recessions after the 1980s, providing somewhat of a leading signal in some cases. Currently no reasons for concern here either.
1.4. Unemployment Rate & Nonfarm Payrolls
As in the case of GDP, unemployment has a high correlation to past recessions, but again, it is a lagging indicator and it is currently far from recessionary territory — on the contrary, it is sticky at historically low levels.
2. The NBER has 20/20 vision.
(1 chart)
Of course, the NBER has the advantage of hindsight when declaring a recession. They often do so well after the lagging indicators have confirmed the slowdown. Sometimes, they only declare the recession after it has already ended. See the chart below.
As investors, we are not as lucky. For us to be successful, we must anticipate recessions well in advance. Note that I’m saying “anticipate”, not “predict”. I can maintain a long position in the broad market while anticipating a recession because I recognize the likelihood of its occurrence, although it is not certain, and pinpointing the timing is near impossible anyway. But being aware of the risks and fighting complacency is our job as investors. So, we must look for reliable leading indicators and assign credences (confidence levels) to their signals.
3. The 3-month — 10-year US Treasury Yield Curve
(1 Chart & 1 Table)
This particular yield curve has a 100% success rate in signalling the past eight recessions, spanning over a time period of more than half a century. The chart below provides a visual representation of the curve inversion which happens when the 3-month Treasury yield surpasses the 10-year yield, causing the spread to dip below zero on the chart. Notably, this inversion occurred consistently six to eighteen months prior to the onset of each recession.
As mentioned in point #2 above, It is crucial to emphasize that the declaration of a recession typically occurs retrospectively. At the actual beginning of a recession, it may not yet be officially declared, and therefore one may be unaware of it until much later. Therefore, careful attention to leading indicators, such as the yield curve, becomes paramount for anticipating and preparing for economic downturns.
In the current cycle, the 3-month — 10-year yield curve inverted in October 2022 and has remained deeply negative ever since. It has now been 8 months since the inversion with no signs of a recession in the coincident and lagging data yet. Based on the average and median number of months it takes for a recession to begin after the curve inversion (see table below), we can expect weakness in the economy to appear around October 2023. However, further down in this article, I will make the case for this being pushed out to the first quarter of 2024.
The topic of causal relationships between yield curve inversion and recession is very contentious, and there is no consensus on this among economists. I plan to offer my perspective on this in a future article (I know, I keep promising –Don’t worry, it will happen soon).
4. The Fed-Funds-Rate and Yield Curve
(2 Charts)
Now, let’s see how the curve inversion lines up with the Fed funds rate cycle:
The chart above shows that the peak of each rate hiking cycle (black vertical lines) occurred within a few months of peak curve inversion. We can therefore deduce that the un-inversions were caused by aggressive rate cuts at the short end (Fed funds) rather than a sudden rise in the 10-year yield. Can we assume this is because the tightening cycle starts to bite and slows down the economy too rapidly, and thereby forcing the Fed to ease before a recession hits? I think so. I can even make the argument that this time the Fed might intentionally induce a recession to prevent a potential resurgence of inflation, similar to the situation in the 1970s.
Currently, market participants have reached consensus that the Fed is either at, or very close to the end of its rate hiking cycle. The SOFR curve is currently pricing in a peak Fed funds rate of 5.5%.
So, if historical patterns over the past 57 years continue to hold true, then we should expect the Fed-funds-rate to decrease before we can anticipate the beginning of a recession.
5. Employment and the Yield Curve
(1 Chart)
Given the Fed’s apparent obsession with the labor market’s resilience, let’s examine this indicator through the lens of historical recessions. The chart below illustrates that the 6-month annualized growth rate of employment typically begins to contract (going below zero) around the same time as a recession commences. However, it reaches its lowest point towards the end of the recession.
Currently, the momentum of employment growth has slowed, but it is still positive (jobs are still being added monthly). Only when this metric goes below zero, will alarm bells go off.
6. A sneak-peak into the future for labor
(1 Chart)
Continuing claims
While there are still no signs of weakness in the labor market in terms of the unemployment rate, the pace of job growth is slowing. A reliable leading indicator for the unemployment rate is continuous jobless claims, which counts the cumulative number of people claiming unemployment benefits from the government. Unlike the unemployment rate, which is based on a survey and depends on subjective measures, continuous jobless claims provide hard data based on actual filings.
The chart below depicts the 6-month annualized growth rate of continuing claims for the past eight recessions. It has proven highly reliable in signalling the start of recessions when it goes over 0.26%. Currently, it has crossed this threshold, suggesting that we should already be in a recession. However, the other economic data, including the unemployment rate is still too strong for the NBER to officially declare a recession. One might argue that the impact of Covid lockdowns and government stimulus packages during 2020 and 2021 has distorted the data, leading to amplified swings in this indicator compared to the past. In that case, it would be a classic example of “This time it’s different.”
7. A labor-based Recession Model
(Model by Arturo Estrella)
Now, let’s compare the progression of the current employment cycle with an employment model based on the past 8 recessionary cycles (excluding the 2020 Covid recession, leaving us with 7 recessions). We exclude the 2020 Covid recession as its artificial nature is less relevant to our analysis, and it only distorts the data. Nonetheless, it’s worth noting that the yield curve inversion also forecasted that event.
The chart below illustrates the average progression of the employment cycle following a yield curve inversion for the 7 recessions preceding 2020. The grey line represents the average rate of incremental employment creation after each yield curve inversion. The black line represents the rate of change in employment for the current cycle, which began in October 2022 when the yield curve last inverted.
It is evident that the current cycle is closely aligned with the historical average, suggesting that unemployment should start to decline around 13 months after the curve inversion, or Q4 2023. Consequently, it is premature to declare that a recession has been averted, and it remains reasonable to anticipate economic weakness in the upcoming quarters.
Note that the estimated start according to the median historical lag (12 months) as discussed under point #3 above, was October 2023 and lines up with the unemployment model discussed here. There is a case to be made that the lag of the current cycle could be longer due to the excessive fiscal and monetary stimulus during 2020 and 2021. I tend to agree with this and will be looking for weakness to show up in the lagging data in Q1 2024.
8. The Signature of a Slump
(1 Chart)
Putting all the pieces together, the signature of a recession is as follows:
- #1. The 3m-10y yield curve inverts as the Fed raises rates.
- #2. The Fed-funds-rate peaks.
- #3. The yield curve steepens (and sometimes un-inverts before the recession).
- #4. The labor growth rate goes negative.
- #5. A Recession is declared.
So, let’s throw all of the above onto one chart:
Up until now, #1 has confirmed, #2 seems imminent, and we are waiting for #3,4 & 5 to materialize.
Although this signature occurred in every of the past 8 recessions, we have to be open to the possibility that it might be different this time, even though the odds are firmly in favor of it staying the same.
Finally, the big question:
9. What does all this mean for my portfolio?
(4 Charts)
Follow along on Twitter for real-time updates: https://twitter.com/JohanKirsten1
In February of this year, I had hedged my portfolio to protect it against another leg down. But, when the S&P 500 price action turned bullish, I decided to cut my hedges. The tweet below shows the price-chart-signal that led me to this decision.
Now market sentiment is becoming increasingly bullish again as valuations creep back into very high multiples, but the analysis above still points to a high probability for recession ahead, which the market seems to be disregarding more and more.
Looking at the technical price chart below, there isn’t much to be bearish about, other than the fact that the move seems stretched and due a breather.
So where to from here?
I am comfortable with the discomfort of being long in an increasingly bullish and complacent market. The fading bearish sentiment likely needs a complete reset before an actual economic downturn and corresponding market correction can materialize. The bull in me is happy to be long, while the bear in me is scooping up insurance at lower and lower premiums (put options) as volatility keeps declining. See a chart of the S&P 500 volatility index (VIX) below.
Being able to harness bullish and bearish behavior simultaneously, is perhaps one of the most important skills to master for successful trading and investing. This is a challenge that many people struggle with: How can you be bearish and hold a long position at the same time? Well, I’m long because the market is going up, but I’m bearish as the rally appears increasingly detached from rationality. The bear in me will yield to price action while satisfying himself by scooping up cheap protection along the way. When the reality of the looming recession hits home and markets start to price it in, the bear’s insurance will kick in and the bull will surrender to the price action while looking for signs of a bottom.
If you ostracize the bear completely and become a perma-bull, you will be unprepared and get blindsided when the inevitable reversal comes. If you banish the bull and become a perma-bear, you will miss out on all the upside and probably get hosed if you are positioned net short. Timing the turns perfectly every time is nearly impossible.
Bottom line: I’m not betting big on a crash. I’m only buying insurance against such an event because based on my analysis, it seems to be quite likely, at the same time as insurance (put option) premiums are very reasonably priced. If I’m wrong, I only lose my insurance premium — nothing more. However, if the market goes down, I can sleep well, knowing my portfolio is protected.
10. Summary
The US economy is performing better than anticipated. However, the presence of an inverted yield curve and interest rates reaching their peak (as currently observed) are indicative of a typical late-cycle economic rebound. These two indicators have preceded 8 out of the past 8 recessions.
The next indicators to monitor would be a rise in the unemployment rate and a re-steepening of the yield curve (i.e., the 10-year yield surpassing the 3-month yield). As discussed above, this will most probably occur once the Fed starts cutting rates again. Should this occur, it would be reasonable to start sounding the recession alarm.
Price action is currently bullish and remaining long stocks seems to be the best positioning while accumulating insurance in the form of hedges that become cheaper as bullishness rises and volatility subsides.
My current stance in 8 words: “Cautiously long with one eye on the door”
If you found the content useful, please remember to clap below and share the following link on social media:
https://medium.com/the-investors-handbook/the-signature-of-a-slump-c0dc1b1cc6d
I’m looking forward to your comments — especially to those with opposing views.
Good luck out there.
Follow my medium profile to get notified when I publish, or follow along on twitter:
https://twitter.com/JohanKirsten1
A special shout-out and thanks to those whose work was influential in my research process for this article:
- Arturo Estrella @intheyield
- Jurrien Timmer @TimmerFidelity
- Eric Basmajian @EPBResearch
Disclaimer
The views expressed in this article are the views of Johan Kirsten and are subject to change at any time based on market and other conditions. This is not financial or investment advice, nor a solicitation for investment funds and should not be construed as such. References to specific securities and issuers are for illustrative purposes only and are not intended to be, and should not be interpreted as, recommendations to purchase or sell such securities.
