Interest rate peak in sight?

The rise in inflation rates led to a sharp increase in short and long-term interest rates in many countries from the start of 2022. Now that the ECB has raised interest rates for the last time and the US Federal Reserve is likely to raise rates for the last time in November, have interest rates in the industrialized countries reached their peak?

The rise in interest rates from the beginning of 2022 does not look particularly dramatic for 30-year US government bonds (chart 1, orange line), but holders of these securities have nevertheless suffered a 50% loss since the first lockdown in April 2020 due to subsequent interest rate rises, although they were able to increase their wealth by a factor of 36.88 between 1979 and 2020.  

The interest rate level for 30-year mortgage loans in the USA has developed in a very similar way (chart 2). This rise in interest rates does not look dramatic either, but it has severely impaired the feasibility of house purchases in the USA (chart 3). In addition to mortgage interest rates, which were still around 10% at the end of the 1980s (currently 7.8%, chart 2), the surprising fact that house prices have actually risen up to the second quarter of 2023 despite the rise in interest rates (chart 4) is also responsible for the dwindling attractiveness of US residential real estate.

The reason for this is that many Americans took advantage of the low interest rate period until 2021 to finance their homes with 30-year fixed-rate mortgage loans. If they want to buy a new home now, they will have to take out a new, more expensive mortgage loan, as they are not allowed to transfer the favorable loans to the new home. As a result, the supply of "second-hand" houses has fallen sharply, so that even low demand has led to further increases in prices (source: The Economist, Sept. 2023). The orange line in chart 2 shows that the average interest rate on mortgage loans in the USA has barely reacted to the rise in interest rates to date due to the proportion of contracts from the period of low interest rates. It is therefore likely to be several years before the US residential real estate market comes under pressure due to interest rates.

Stricter bank lending conditions will have a much faster impact, causing a decline in credit growth with a five-quarter delay or even a contraction in the credit volume, as is already the case (chart 5). This means that the US economy is likely to be burdened by a shrinking credit volume until at least the second half of 2024.

Clear recession signals for the US economy also come from the inverse interest rate structure, which has only sent a false signal once since 1953, namely in the mid-1960s (chart 6). Otherwise, all 10 US recessions since 1953 have been correctly predicted by this forecasting model. An inverse yield curve exists when short-term interest rates (here: the yield on 1-year government bonds) are higher than long-term interest rates (here: the yield on 10-year government bonds).

To date, an increase in the unemployment rate of at least 0.33 percentage points after the respective cyclical low (chart 7, red dots) has predicted a US recession even more reliably.

This was reached in the current cycle in April 2023 at 3.4%. The low was exceeded by 0.4 percentage points with the unemployment rate in August 2023 (3.8%). This forecast model has not produced a single false signal since 1953.

Finally, a forecast model calculated by the New York branch of the US central bank itself on the basis of the interest rate structure also indicates a high probability of recession (chart 8), which was only reached in the first oil crisis from 1973 and exceeded in the second oil crisis from 1979. All other recessions occurred with a recession probability of less than 50%.

The economic situation in the eurozone is no better. The sharpest interest rate hike since the European Central Bank was established (chart 9) has contributed to a contraction in the money supply (chart 10), the likes of which had not been seen for decades before the creation of the eurozone. With the exception of the coronavirus crisis 12 months later, even a fall in the money supply growth rate to below 5% since the start of the eurozone in 1999 has resulted in stagnation (2001) or recessions (2009 and 2011 to 2012) (chart 11).    

Interest rates should actually have peaked.

However, the clear warning signals for economic development in the USA and the eurozone have not yet been confirmed by corporate bonds with low credit ratings (chart 12). These so-called high-yield bonds, which are very sensitive to the economy, show no signs of weakness in either the eurozone or the USA. In times of crisis, these bonds fall significantly, so that the yield differential to safe government bonds rises sharply. At present, however, there are clearly no concerns about the financial stability of companies, either in the USA or in the eurozone.

However, the picture on the stock markets is less uniform. While the US stock market is valued far above average in terms of company earnings (dividends and cash earnings) and equity (book value) (chart 13), the German stock market, for example, is almost as cheap as it was during the various crises of the last 50 years (chart 14). Accordingly, from a stock market perspective, Germany is in crisis, while the USA can enjoy a rosy present and future.

However, the difference in valuation could also be interpreted to mean that Germany's major problems are generally known, while in the US, investors are more interested in future topics such as artificial intelligence than any problems. However, should economic difficulties also arise in the US, the US equity market would be more at risk than the German market, which is also reflected in earnings expectations.

In the US equity market, these are only just under 2% p.a. (chart 15), whereas in Germany they are 13% p.a. (chart 16), derived from the current valuations of the equity markets (charts 13, 14).

The other major equity markets have also become much cheaper since the internet and telecoms bubble burst in 2000 and are currently valued significantly lower than the US equity market, whose valuation has only clearly set itself apart from the other equity markets since the coronavirus crisis in early 2020 (chart 17). As a result, earnings expectations on the equity market are at least 6 percentage points higher than on the respective bond market everywhere except in the US (chart 18, gray bars). In the USA, however, government bonds have become a very attractive investment alternative with yields of 4.5%.

A recent study by the respected US investment bank Goldman Sachs examines the impact of the use of artificial intelligence on productivity growth in various economies (chart 19). The result definitely does not justify the high valuation difference between US equities and the other major stock markets; according to the study, annual productivity growth in the US would be 0.07 percentage points p.a. higher than in the eurozone, the same as in the UK and slightly lower than in Japan.

We now return to the 30-year US government bonds and their performance (interest income + price changes). Chart 20 shows the perhaps somewhat surprising fact that the price gains of the US equity market from 1979 to 2020 were no higher than the performance of a bond portfolio that was always invested in 30-year US government bonds. The bond portfolio was much more stable and continued to rise with little fluctuation until 2020. In contrast, share prices had risen 97% more than the bond portfolio in the years before 2000 (technology bubble), only to lose the entire lead again by mid-2002 and follow the bonds until the Lehman bankruptcy in the fall of 2008. The stock market then broke out to the downside, but after a strong recovery it reached the blue line again, which it then followed until 2020. Since then, the gap to the bond portfolio has widened even further than at the peak of the technology bubble in 2000, making bonds quite attractive after the sharp price losses and increased yields, while the equity valuation reflects more risks than opportunities (see also chart 18).

The attentive reader will rightly object that bond performance should not be compared with share prices, but with share performance, i.e. price changes + dividend yield. However, as in 2000, this hardly matters for future performance, as dividend yields of just over 1% in 2000 and currently 1.5% have not been able to compensate for the risks of the high valuation of US equities, because their contribution to performance in times of overvaluation is very low.

If share prices fall, the US central bank in particular will deduce that the economy is becoming increasingly weak and quickly loosen monetary policy. The interest rate peak is therefore not far off, and we may already be there.

Finally, our key statements from the Capital Market Outlook from September 2020, which you can find here:

The main topic was the tendencies towards increasingly anti-growth and inflationary policies that were already in place three years ago and which we could also be complaining about in 2023. Among other things, it was about the increasing armament, which has now led to a war in Eastern Europe, the turning away from free world trade and the hostility towards education - most recently in Germany, where politicians have recently called for the abolition of school grades and homework. From this, the danger of soon rising inflation rates was derived, which then materialized from 2021.

‍

You can also download the capital market outlook here.