Share prices at record highs - are there no more risks of recession?

Share prices have reached new all-time highs not only in the US, but also in Japan, France and even in Germany, which is burdened by economic weakness. Investors seem to be convinced that the central banks will implement several interest rate cuts over the course of 2024 and that a recession can therefore be avoided, particularly in the US.

In the following, we examine how realistic these widespread assumptions are.

First, we examined the phases of interest rate hikes over the last 70 years in the USA, whose economy and capital market are still dominant worldwide, and their impact on the economy. Chart 1 shows the interest rate hikes by the US central bank up to the respective peak in percentage points. The largest interest rate hike of 10.7 percentage points took place at the end of the 1970s, when money market rates were raised from 4.5% in April 1977 to 15.2% in March 1980. All other interest rate rises were aligned with this peak, ending at 10.7 percentage points but not starting at zero. This shows that the current phase of interest rate rises, which presumably ended in October 2023 (chart 1, red line), was above average. The two weak rate hikes up to August 1984 and February 1995 were not followed by a recession, but all other rate hike phases caused recessions (chart 2). Therefore, the absence of a recession after the most recent rate hike would be rather unusual.

The US Federal Reserve itself currently estimates the probability of a recession in the US in the next 12 months at 61.5% (chart 3). In January 2023, the model had already seen an above-average probability of a recession at 57%. In all previous recessions, the recession occurred either exactly at the predicted time (1990, 2007) or a few months before. This time the expected recession has been delayed because the US government has built up enormous government deficits to support the economy by spending over 9% of national income p.a. on average over the last 5 years (chart 4) - last year it was also over 7%. The eurozone managed a modest 3.4% p.a. in the same period. The US economy will have to make do with significantly less government support in the foreseeable future; there is no such risk in the eurozone.

The developments in the unemployment rate and money market interest rate also match the coming recession (chart 5). Before every recession in the last 70 years (golden bars), the unemployment rate stagnated at a low level compared to previous years, only to shoot up as soon as the recession began. It was precisely at this point that the US central bank began to cut interest rates, which continued until at least the end of the recession. Obviously, the US central bank always waited for unemployment to start rising before starting the interest rate cut phase. It never saved the economy from a recession by starting to cut interest rates beforehand, but started to cut interest rates after a recession had begun and ended the rate cut phase when the recession had also ended. This is definitely not in line with current expectations on the US capital market. The current picture of an unemployment rate that has been stagnating at a very low level for around two years and a simultaneous sharp rise in money market interest rates has also been seen in the past before every recession.

The US central bank is likely to have figured out that the recession had just begun whenever the unemployment rate rose by 1/3 of a percentage point compared to the average of the previous 12 months (red dots in chart 6). Since the US central bank, unlike the ECB, not only has the task of ensuring monetary stability, but also has to support the economy, it has a very good argument for lowering money market interest rates due to the onset of a recession even if the unemployment rate rises slightly.

The renewed significant decline in wage increases (chart 7, blue line) is an indication of a gradually weakening labor market; this was also the case in the recessions of 2001 and 2008. The brief coronavirus recession was triggered by political measures to contain the pandemic, not by economic factors, and is therefore only of limited significance for analytical purposes.

Equally typical for the period before a recession are the rising default rates on consumer loans, long before unemployment rates soar at the start of the recession (chart 8: there is a blue circle in front of each orange circle). The reason for this lies in the rising money market interest rates before each recession (chart 5), which make consumer loans more expensive and thus trigger more bankruptcies. If unemployment then also rises, the fear of saving begins, consumption falls and the recession is here.

The commercial real estate market is also likely to weaken the US economy. High vacancy rates in office buildings and shopping malls are causing rental income to fall, but at the same time lending rates are rising sharply when loans from the low-interest phase mature and the currently much higher interest rates have to be paid in future. As a result, defaults on debt secured by commercial real estate have been rising sharply for a year now (chart 9).

The probability of a recession in the US economy is therefore high enough to examine the consequences of a recession for the capital markets.

A German investor should not assume that the consequences of a recession for US investments will be mitigated by a rising dollar exchange rate. The dollar has weakened against the euro five times in the last eight recessions, albeit only slightly three times, and only appreciated three times (chart 10).

In contrast, 10-year US government bonds can certainly be expected to make price gains during a recession. With the exception of 1974 and 2020, interest rates have always fallen (chart 11), meaning that interest income has mostly been enriched by price gains.

The consequences of a US recession are likely to be less favorable for US equities. In the 12 months before the start of a recession, US share prices have fluctuated between +20% and -20% over the last 70 years (average -0.6%, chart 12). However, share prices have always fallen after the start of a recession, although interest rate cuts also began at the start of the recession (chart 5). The average price loss up to the low point of share prices after the start of the recession was 18%, with only a moderate 2% in 1980 and 50% in the recession of 2008/2009 (financial crisis with the collapse of Lehman Brothers). Only when the interest rate cuts were already at an advanced stage did share prices begin to recover from their respective lows by an average of 19%.

The question now arises as to whether high or only moderate price losses are to be expected in the likely coming recession. The respective dividend yield of the US equity market (chart 13) at the start of the interest rate cuts (not always exactly congruent with the start of the recession) provides a fairly reliable indication of this.

If the stock market is valued very low, i.e. has a high dividend yield, there are no or only minor price losses (chart 14). In 1980, the interest rate cuts began in March. However, the stock market immediately rose by 12% due to the low valuation (dividend yield at an unusually high 6.1%) (blue dot at top right in chart 9). However, with dividend yields below 3.5% (the three blue dots on the left in chart 14: 1973, 2000 and 2007), the price losses reached 39%, 38% and 48%. At 1.36%, the current dividend yield is almost at the historic low reached in the two years of the technology and telecoms bubble in 1999 and 2000 (chart 13).

Although there are only 9 data points for the correlation between interest rate cuts at the start of recessions and subsequent share price losses (chart 14), this correlation is large enough with an R² of 0.77 (R² can fluctuate between 0 (no correlation) and 1 (exact correlation)) for caution to be exercised given the current extremely low dividend yield and the considerable probability of recession on the US stock market.

To summarize, we can say that 9 of the 11 phases of interest rate hikes over the last 70 years were followed by a recession. Only 2 phases with interest rate hikes of less than 3 percentage points were not followed by a recession. The current rate hike is 5.2 percentage points until October 2023 and was above average. The narrative that the US central bank would prevent a recession by cutting interest rates soon is unrealistic, as interest rates have not been cut before a recession in the last 70 years, but always after it has begun. In addition, the US central bank has estimated the probability of a recession at over 50% since January 2023, i.e. several months before the end of the interest rate hikes. It may therefore have carried out the last interest rate hikes in the full knowledge that they would probably trigger a recession in order to combat inflation. As the start of interest rate cuts was always initially associated with price losses on the stock market, the size of which can be easily explained by the dividend yield on the stock market, US equities currently pose considerable short-term risks as the dividend yield is very low.

The long-term outlook for US equities is not particularly good either. Valuations are currently close to historical highs (1999 and 2021, chart 15). This results in a return expectation of -2% p.a. for the next 10 years (chart 16); unfortunately, the forecast model has a very high reliability with an R² of 0.87.

An examination of recessions and share price trends in Germany since 1969 (chart 17) confirms the results of this analysis for the USA. Share prices also rose or fell before a recession (average value +6.4%), but from the beginning of a recession there were only price losses (average 21% up to the respective low point, USA: 18%). Share prices in Germany then recovered by 17% by the end of the recession (USA: 19%).

The influence of the dividend yield on the extent of price losses from the beginning of the interest rate cuts to the low point of share prices also corresponds to the behavior of the US stock market, whereby the dividend yield in Germany is almost 2 percentage points above the American level and is therefore not particularly low, unlike in the USA (chart 18). A high dividend yield of over 6%, as at the beginning of the interest rate cuts in 1980 and 1981, was even followed by price gains in 1980 and only a moderate price loss in 1981, while the interest rate cut that began in 2001 started with a dividend yield of a very low 2.1% and the price loss thereafter reached 36% (chart 19, blue dot on the far left). This means that the recession-related price risks on the German equity market are currently significantly lower than in the USA (charts 14 and 19).                                            

The long-term prospects on the German equity market are also better than in the USA. Valuations are favorable (chart 20) and the earnings outlook for the next 10 years is over 10% p.a. (chart 21). Investors may be overlooking the fact that the current uninspiring state of the German economy has little impact on large German companies, which generate more than 80% of their sales outside the country's borders (source: onvista, January 2024).

This could once again confirm the old rule that the less popular investments - in this case German equities, but this also applies to European and Asian equities and emerging markets - will outperform technology companies, which are currently much more popular with investors, both in the next recession and over the next 10 years.

Finally, our key statements from the capital market outlook from February 2021, which you can find here can be found here:

After the equity markets had recovered strongly following the coronavirus trough, we showed that the equity markets were no longer particularly cheap just under a year later and would therefore be exposed to significant price losses again in the event of a significant rise in interest rates. We estimated the risks in the USA to be higher than in Europe, but this proved to be incorrect in 2022. With the rise in interest rates starting at the beginning of 2022, US equities fell slightly less than European equities, by around 10%.

‍

You can also download the capital market outlook here.