Please read the Disclaimer before reading this article.
Introduction
If you would have read independent investment research in the beginning of the year, the general consensus was that the first half of the year would continue to be rather difficult for the market after which a better second half would follow.
Consensus was wrong and the first half of the year has shown strong market returns.
In this article I would like to continue to explain my strategic framework for the equity market. Please note that these are no tactical views but they are a framework to invest through a complete market cycle.
I don't think we've seen the lows in this market cycle but this doesn't mean I'm always bearish. I update my tactical views on twitter. I haven't been long or short this market since the beginning of the year for the simple reason that I don't like to trade against my core strategic fundamental view. I'm big picture still bearish so I'll go short the market but won't go long on bear market rallies. I'll go long when I think we've seen a sustainable market bottom and the beginning of a new bull market. So this year, I've been sitting on the sidelines.
With the technology sector being up more than 35% since its low at the end of last year, the question to ask is if my big picture bearish view is wrong or not.
Please note that several technical indicators signaled buying opportunities throughout the year. 2023 offered some great opportunities if you invested in the right sectors. However, the analysis in this article is about the framework I use to analyse a complete market cycle and therefore I won't discuss any tactical tools or views.
Stages of a market cycle
The chart below shows the average evolution of a bear market.
First, inflation starts to increase (red line) and the Federal Reserve reacts to this by increasing short-term interest rates (grey line). Long-term interest (dark blue line) rates started to rise before the pick up in inflation and continue to do so during the Fed hiking cycle.
This rise in the discount rate puts pressure on valuations and the stock market (purple line) declines. (Please note that the market can perfectly rise during Fed rate hikes when the economy is strong and earnings are rising, I showed this here.)
When the Fed rate hikes start to have an impact on the economy and on earnings, we see that earnings start to decline (light blue line) and the stock market (purple line) also continues to decline.
Note how the stock market (purple line) bottoms way before earnings (light blue line) do.
Chart 1
Source: Jurrien Timmer, Fidelity
The first part of this story is what we saw during the first 6 to 9 months of 2022. The stock market declined because the discount rate rose sharply and valuations declined. This while earnings remained strong. (You can see this on the chart below: the stock market declined and this was caused by lower valuations, the grey line, even though earnings, the green line, remained strong.)
Chart 2
Source: Michael Kantrowitz, Piper Sandler
Since the market bottomed in October of last year, we see that the driving force behind the rally has been rising valuations (grey line on the chart above) because earnings (green line on the chart above) have been declining.
Before we continue to analyze the two charts above, I would like to talk about earnings. As we saw on the first chart, the market bottoms well before earnings do. The chart below shows that historically this time gap is about 3 to 6 months.
Chart 3
Source: Warren Pies, 3Fourteen Research
So if October 2022 was the market low, we should have seen the bottom in earnings for this cycle.
There is however one problem with this, I don't think we have. I've shown you my own leading indicator on earnings before and it's not looking good. Below I share two other models that are showing the same negative outlook as my own models are.
Chart 4
Source: Beowulf's Treasury
Chart 5
Source: Morgan Stanley
So if earnings haven't bottomed, chances are very low that October 2022 was the ultimate market low.
It's interesting to note that some independent investment research providers are bullish on the market even though they think earnings will continue to decline. On the face of it this doesn't need to be so strange because the market frontruns these earnings. If you look at the performance of the stock market over the period of 1 year, valuations have a much bigger impact than earnings do on its performance.
Their reasoning goes as follows: interest rates will decline and therefore valuations will increase and so will the market despite lower earnings. I've shown in detail before (see here) why this line of thought is wrong. In short, valuations are driven by anything that alters the perception of financial and/or economic risk. If interest rates decline but credit spreads rise, valuations decline.
The chart below shows this. The highlighted areas are periods when interest rates and valuations both declined together because the perception of risk, credit spreads, increased.
Chart 6
Source: Michael Kantrowitz, Piper Sandler
You can see this on the first chart from Fidelity as well. The stock market (purple line) doesn't necessarily bottom when interest rates peak (grey and dark blue line). The market continues to decline because the green line, which represents the interest rate on BAA-rated bonds, starts to increase when the yield on government bonds stabilizes/declines, thereby leading to a rise in credit spreads. This rise in credit spreads shows a higher perception of risk by investors and thus leads to lower equity valuations.
If you now look back at the second chart in this report, you can see the same process in action again. The higher valuations (grey line) since the bottom in October last year were accompanied by lower credit spreads.
Is the risk priced in?
The chart below from Bridgewater shows a similar decomposition of the drivers of a bear market as the first chart did. Here they focus on inflationary bear markets in particular. We see the same pattern that we discussed earlier. First the discount rate rises which puts pressure on the stock market (effect shown by red line). Cash flows only start to struggle later on when the rate hikes start to impact the economy. The negative impact of these declining cash flows (represented by the blue line) only ends well after the stock market bottoms. The green line is the line I want to focus on. It shows that the final push lower in the stock market is caused by a rise in risk premiums. This is exactly the same thing that the first chart from Fidelity showed by the rise in credit spreads.
Chart 7
Source: Bridgewater
We'll now focus on two risk premiums. Credit spreads and the equity risk premium (ERP).
I've shown before in detail why I don't think credit spreads can decline sustainably as long as the economy is weakening. You can read about it here. (Of course credit spreads have declined the last few months causing valuations to rise but this is a tactical move. I'm interested in knowing whether or not we've seen the absolute peak in credit spreads or not for this whole cycle. If I know this, I know whether or not a sustainable new bull market has begun.)
Next I want to show some other reasons why I think credit spreads are too low.
The chart below shows that banks have been tightening lending standards (dark line) quite hard. A tightening of this magnitude has historically always led to a recession (grey areas). Because there has been a lot of talk about soft and hard landings, a lot of people seem to think that a recession is priced in. This is wrong. As the chart below shows, credit spreads normally move together with these lending standards. This is because there is a high correlation between lending standards and corporate default rates. A decline in the availability of credit leads to more defaults and higher credit spreads. We've seen the banks tightening their willingness to make loans but credit spreads haven't reacted to this yet. It's only a matter om time before they will. To me this is a very clear indication that recession risk is not priced into the market.
Chart 8
Source: Ian Harnett, Absolute Strategy Research
The following chart shows credit spreads together with a model that estimates the correct level based on several inputs. Historically there is a very close fit between the model and actual credit spreads. Again, according to this model, credit spreads are too low.
Chart 9
Source: Simplify Asset Management, shared by Tier1 Alpha
There also seems to be this idea with investors that contained credit spreads show that risks are contained. This is wrong. Warren Pies from 3Fourteen Research notes the following: "In general, credit spreads do not blow out until there is equity market stress. Much more reactive than predictive...Credit spreads rarely offer an advanced signal."
The chart below shows this. The horizontal axis shows 3 month equity returns. The vertical axis shows the 3 month change in credit spreads. In the top left panel you have rising credit spreads and a declining stock market. This is just as expected but the interesting thing to note is that the relationship is not linear. Just as Warren Pies noted, credit spreads are more reactive than predictive.
Chart 10
Source: Warren Pies, 3Fourteen Research
Next, we'll have a look at the equity risk premium (ERP). The ERP is the excess return earned by an investor when they invest in the stock market over a risk-free rate.
Looking at valuations in isolation, some analysts state that equities currently offer good value. While this might be true, the problem is that you can't look at valuations in isolation. With 10-year US government bonds currently yielding 3,8%, the alternative to equities has become much more interesting than before. It's not possible to compare the current S&P500 forward PE ratio of 18 to periods in 2017 or 2019 when the valuation was similar because interest rates were way lower back then. It's also not possible to compare the current 'low' valuation to those seemingly high valuations of 2021 because again, interest rates were way lower back then. To correct for this, we can use the ERP because this measure is an estimate of the excess return you get from investing in stocks over bonds.
The chart below shows an estimate of the ERP. It uses a combination of several models to arrive at a single estimate. Since there is no way to get an exact measure of the ERP, it's not the exact level that is of interest to me but its movement.
Normally when you have a recession, there is an increase in the ERP. Investors are afraid and they demand a higher return on equities over the risk-free rate to compensate for the increased risk. You can see this on the chart by the sharp increase in the ERP around the shaded grey areas which indicate a recession. All bear market bottoms are also accompanied by a sharp increase in the ERP. Markets bottom when equities become interesting again compared to its alternative, bonds. Over the last few months we have seen a decrease in the ERP. This shows that equities are not as cheap as a look at valuations would indicate. It also shows that a recession isn't being priced in at all by the equity market. As Michael Arnett from Absolute Strategy Research notes: "This chart of the US Equity Risk Premium shows that there is almost zero probability of recession priced into US markets. ERPs RISE in recessions - the last six months have seen the ERP FALLING on a consistent basis. Investors are betting on recovery, not recession…"
(Please note that a low ERP isn't necessarily bad for forward stock returns. There was a low ERP during most of the 1980's and the 1990's but stock returns were great. That's why I primarily focus on the movement of the ERP in relationship to the economic environment that I'm seeing/expecting.)
Chart 11
Source: Ian Harnett, Absolute Strategy Research
Not only the movement in the ERP is interesting to analyze, it's also interesting to compare the ERP to where fundamentals indicate it should be in theory. The chart below from Nordea does exactly that. In red we can see a measure of the ERP. In blue we have a fundamental model that shows where the ERP should be based on several economic inputs. As you can see, the fit is very close historically. Over the last few months however the ERP has been declining whereas fundamentals indicate that it should be rising. If the model is right, it implies that valuations are currently way too high.
It's not the first time there has been a divergence between the ERP and the model. We saw the same thing during the 1998 LTCM stock market decline. The market declined around 20% and the ERP rose. The model indicated there were no fundamental problems and thus a divergence between the model and ERP developed. In the end the model was right and the ERP declined sharply again thereby creating a rising stock market.
We saw another divergence just before the Great Financial Crisis. The ERP was declining but the model was rising. Again, the model was right and what followed was a big decline in the stock market. We have a similar setup today.
Chart 12
Source: Mikael Sarwe, Nordea
What is the economy telling us?
Investors are talking about soft and hard landings, about future economic weakness being priced in, about a Fed pivot being bullish for the market,..
A lot of these basic ideas are wrong.
A Fed pivot is only bullish for the market when there is a soft landing. When we have a hard landing (economic recession), a Fed pivot is not bullish for the market. (I showed this in detail here.)
Most of the talk about soft and hard landings is also not helpful because most people don't seem to know that there is a very clear relationship between how markets behave around both type of landings. The interaction between monetary policy, the economy and asset markets is crucial and has repeated itself throughout history. I showed this in detail here.
One of the elements that differentiates a soft landing from a hard landing is the way the unemployment rates behaves. Simply put, a hard landing is when the Fed raises interest rates, the economy weakens, and the unemployment rate starts to rise thereby creating a recession. A soft landing is when the Fed raises interest rates, the economy weakens, but the unemployment rate continues to decline. The way asset markets behaves once the Fed stops raising interest rates is completely different.
I've shown before in detail why I think that we'll see a hard landing and why leading indicators are showing an increase in the unemployment rate is coming (see here and here).
Some might argue that a rise in the unemployment rate is a general consensus view. If it happens, it won't matter for the stock market because it's expected. In my opinion, this is wrong. History clearly shows that when there is a hard landing and the unemployment rate rises, the market never bottoms before the rise.
The vertical lines on the chart below indicate stock market bottoms around hard landings. As you can see, they always took place during the rising phase of the unemployment rate. Why would this time be different? I also added initial unemployment claims (inverted on this chart) which are a leading indicator of the unemployment rate. Once initial unemployment claims start to rise (on this chart it means that the blue line is declining since they are inverted) around hard landings (the black lines), the stock market declines with them.
This clearly shows that the market doesn't price these kind of things in.
Chart 13
The following chart supports this. On the left we see the evolution of the unemployment rate from its lows. A recession starts once it rises by 0,5% from its lows. The right part shows what happens to the stock market once this happens. As you can see, the market didn't price this in and it declines to its final low.
As Variant Perception concludes: "The final crash in a recession starts when the unemployment rate picks up 0,5% off the lows. This is the most obvious confirmation that recession has started."
Chart 14
Source: Variant Perception
So what does the market price in?
The stock market prices in a lot of things. However, just assuming that everything you can think about is being priced in is wrong. The market is much more reactive than you think. The stock market is viewed as a leading indicator. That is why it's very closely correlated with leading economic indicators like the ISM PMI. It's also why all stock market bottoms that went together with a hard landing took place when the ISM PMI bottomed. I showed this in detail in a previous article. If you look back at chart 1 from this article, you'll see this once again. The market (purple line) bottomed way before earnings (light blue line) but it bottomed together with the ISM PMI. Why? Because both are leading indicators.
When will I be wrong?
This is perhaps the most important question to answer. Below are the three most important elements that can make my conclusion wrong:
1) There are several leading indicators on the PMI's that are showing we've reached the bottom. You can see 4 of them in the chart below. If these indicators are correct, and given the interplay between PMI's and the stock market, chances are high I'll be wrong.
One thing I would like to note is that most of these indicators have a lead time of 1 to 9 months. If I look at indicators with a longer lead time of about 18 months (as shown in previous articles), I don't see signs of a bottom. Time will tell if the shorter term models will prove correct or if they will turn down again.
Chart 15
Source: GMI (top-left), Steno Research (top-right), Clocktower Group/Marko Papic (bottom-left), Deutsche Bank (bottom-right)
2) The biggest mistake I've made is underestimating the strength of the labor market. (In a future article I will investigate how we can improve the timing of a softening labor market.) Historically, the labor market dictates if we have a soft or hard landing and it also helps us to interpret what the market has priced in and what not. If the unemployment rate remains low, my thesis will be wrong. However, all leading indicators are indicating the unemployment rate will rise. Some might argue that the current situation in the labor market is completely different or that the markets will react differently. Comparisons to the period after World War II are made but even then, the unemployment rate rose when their was a recession and even then, the interaction between monetary policy, the economy and the stock market was the same as in all the decades that followed: the market always bottomed in the rising part of the unemployment rate. The market rally we saw in 2023 was not abnormal. Early this year, the market was already pricing in several rate cuts for 2024. With an unemployment rate that remained low and the market expecting a Fed pivot, a relief rally followed. This has been a rather consistent pattern throughout history. I'll show this in more detail in a future article. So it's not that the framework didn't work, it's that I thought we were in a different part of the framework. Stated differently, I didn't expect the unemployment rate would stay this low for this long.
Some final thoughts
In a previous article, I wrote about how people expecting a soft or hard landing wasn't really helpful in predicting what actually would happen. Last year for example, we saw a lot of people talking about a soft landing but there were also a lot of people talking about a hard landing. You can see this in the chart below. The orange line indicates the number of stories talking about a recession, the white line shows the same but now for a soft landing. In Q3 last year, both the white and orange line were spiking together. As I said in the previous article, you could have made a contrarian case for both last year. What's more interesting however is the sequencing between both lines. What we can see is that talk about a soft landing (white line) starts to rise sharply when we approach the end of a Fed rate hiking cycle. This optimism recedes afterwards and talk about a recession comes back. Simply stated, in every hard landing we've seen, people were expecting a soft landing when economic data seemed robust. At times when the leading indicators were indicating a recession, this optimism was proven wrong and a recession arrived. Will this time be different?
Chart 16
Source: Michael Kantrowitz, Bloomberg
Another final point I want to touch on is the yield curve. We saw an inversion in the 10Y3M yield curve in late October, early November last year. Generally speaking, this is considered by a lot of people as a recessionary signal. (Please not that in a lot of countries outside the US, it doesn't have such a good track record at predicting recession.) We're +-10 months later and people are wondering where the recession is and why the stock market is up so much. As the chart below shows, there is actually nothing strange about the current price action. The chart shows the evolution in the stock market after several yield curve inversions. As we can see, in a lot of cases, the market has a significant rise after inversion. However, the one important thing to notice is that the market never bottoms before the inversion takes place. If October 2022 was the low, this would be very abnormal.
Chart 17
Source: Cheng-Wei Chin (@ChengWeiChin1)
And this brings us to a general point I want to make. Indicating a market bottom in real time is very difficult. We know that it happens at a time when nothing looks good and that a lot of fundamentals will bottom several months later. However, if we look at where fundamentals stand 6 to 9 months after a stock market bottom, things are very consistent. So this means that if October 2022 was the low for this cycle, we can look at the current market and economic fundamentals and compare them to those that were present 6 to 9 months after market bottoms. If we do that things look very strange. Normally, when there is a hard landing, markets bottom:
-after the Fed started cutting rates (not this time, the Fed is still hiking)
-after yield curve inversion (not this time, the market bottomed before yield curve inversion)
-when earnings bottom a few months later (not this time, expectations of earnings are rising but leading indicators are indicating further weakness)
-and are followed by a very strong initial rally (not this time, we've seen a very strong rally in the technology sector but if we compare the current rally in the s&p to those after bear market bottoms, the current rally is very weak)
-during the rising part of the unemployment rate (not this time, the market bottomed with the unemployment rate being at a multi-decade low)
-when leading economic indicators like the PMI bottom (could be the case this time if the shorter-term models are correct)
Thank you very much for reading my article.
If you want, you can follow me on twitter: click here.