2023-09-27 01:51:49 ET
Summary
- Historically, stocks have outperformed bonds, but there may be times when it makes sense to reallocate capital from stocks to bonds.
- The yield curve, particularly an inverted yield curve, can provide insight into future economic conditions and stock performance.
- During periods of yield curve inversion, it may be wise to prioritize allocating capital towards bonds rather than stocks.
We normally favor being overweight stocks over bonds for the simple reason that they have historically outperformed bond returns by several percentage points. For example, if we compared the total returns of the S&P 500 index ETF ( SPY ) with the Vanguard Total Bond Market ETF ( BND ), the outperformance is quite clear.
How much to allocate to each asset type depends of course on the investor, their investment objectives and investment horizon. However, we were wondering if there are times where it makes sense to re-allocate some capital from stocks to bonds.
It is well known that the yield curve can give a lot of insight related to expected economic conditions in the future. In particular an inverted yield curve, where short-term rates exceed long-term rates, is often seen as a recession warning. One of the most popular ones is the 10 year treasury rate minus the 3 month treasury interest rate. This can be easily found in the FRED website for free. As can be seen below, there have been several instances where an inversion was soon followed by an economic recession, which is shown in grey.
Yield Curve Inversion Research
There has been interesting research done on the application of the yield curve to gain insights into future stock performance. For example, in "Application of the Yield Curve Inversion Indicator to Determine the Current Phase of the Stock Market" by O. Benenson, he finds that:
...the difference of the yield curve of 2-year and 10-year US Treasury bonds is a fairly reliable tool for determining of the approaching recession in the economy, but at the same time it is not possible to determine the exact time of the recession. It is shown that this indicator is expedient to use for early warning about a possible fall of international stock markets. At the same time, it was found that not every inversion of the yield curve is followed by a fall in the stock market, but every fall is preceded by an inversion. It was noted that the current dynamics of the yield curve is signaling a possible significant drop in the US stock market in the near future.
Given that he focuses on the 10Y-2Y term spread, let's take a look at its current spread. As can be seen, that yield curve is also currently inverted, and in a very significant way.
Another interesting study , this one focusing on the use of machine learning to evaluate different term spreads to predict recessions is "Predicting Recession Probabilities Using Term Spreads: New Evidence from a Machine Learning Approach" by Jaehyuk Choi, Desheng Ge, K. Kang, S. Sohn. One important conclusion they reach is that there is good support for the conventional use of the 10Y -3M Treasury yield spread, and that the use of machine learning to evaluate other pairs is not necessarily that beneficial.
Practical Implications
While research on this subject is interesting, we are mostly looking for practical strategies that can help with investment decisions. For example, a yield curve inversion might indeed be a very good predictor that a recession might be approaching, but what is the impact on stock and bond returns.
To answer this question we ran a couple of analyses that we'll share with you, for a simple strategy that has historically helped decide when to move more towards bonds, and when to move more towards stocks. We calculated the average monthly return for the S&P 500 index during periods when the 10Y-3M yield curve was inverted. We show these periods shaded in grey, with the average monthly return during these periods since 1962 of -0.02%. This might not be a huge average loss, but the opportunity cost was significant considering that average annualized, 3 months Treasury rate during these periods was 7.75%. Another interesting finding was that this relatively anomalous situation of having long-term rates higher than short-term rates was rare, but not extremely rare. According to our calculations the 10Y-3M spread was negative ~11% of the time.
Analyzing the 10Y-5Y spread we obtained similar results, but what was very relevant is that this is a much more common occurrence. The 10Y-5Y term spread has been negative about 19% of the time. The average monthly return for the S&P 500 during negative 10Y - 5Y spread periods since 1962 is -0.47%.
Strategy to Consider
During periods where the 10Y-5Y treasury yield curve is negative, it is probably a good idea to prioritize allocating capital towards bonds, instead of adding to ETFs like SPY, or stocks in general. Of course, this might be different for different investors, and a good stock picker might still outperform bond returns during these inverted yield periods.
Right now the 10Y-5Y curve is inverted by -0.13%, which means we are currently during one of those periods where it makes sense to favor bonds over stocks. Given the extremely high valuation metrics for the S&P 500 index, and with 3 months Treasuries yielding almost 5.5%, it makes even more sense to reconsider favoring bonds at this time.
Improving the strategy
Since the yield curve inversion is supposed to signal an impending recession, and that is usually accompanied by stock market declines, it might further improve the strategy to wait some additional time after the yield curve stops being inverted before returning to stocks.
In a future article we plan on analyzing what is usually the ideal time to wait, once the yield curve is no longer inverted, before reallocating capital towards stocks.
S&P 500 Current Issues
One of the best valuation metrics to determine whether the S&P 500 index is overvalued is the cyclically-adjusted price to earnings ratio. This helps reduce the effect that an anomalous year might have when companies experienced extraordinary profitability, or had below average earnings because of a recession. This valuation metric has been shown to have good predictive power in the long term (e.g. expected returns for the next ten years). The current Shiller PE ratio is ~29.5x, while the index PE ratio is ~22x. It is clear from the chart below, which can be found on the Multpl website, that we are currently at an above average valuation.
The Multpl website also shows the S&P 500 index earnings yield, which is significantly below the long-term average of the past few decades.
Another critical issue with the S&P 500 index at the moment is its high concentration in technology stocks that look overvalued. According to Morningstar, the index has ~30% of its assets in its top 10 holdings. While some of these top holdings are characterized as 'Consumer Cyclical' or 'Communication Services', many investors would actually consider them technology stocks.
Looking at just the top 5 holdings, Apple ( AAPL ), Microsoft ( MSFT ), Amazon ( AMZN ), NVIDIA ( NVDA ), and Alphabet ( GOOGL ), it is clear that several of these companies are trading above their average historical valuation of the last ten years.
Two that look particularly overvalued in our opinion are NVIDIA and Microsoft, both with EV/Revenue multiples significantly higher than their ten year averages. It appears they are both currently benefiting from the significant interest that investors are showing on AI related companies.
What can investors expect from the S&P 500 index for the coming ten years? One of the best analysis we have seen trying to answer this question comes from Michael Finke, who wrote an article in Advisor Perspectives titled "The Remarkable Accuracy of CAPE as a Predictor of Returns". It is clear that there is a tight correlation between the CAPE multiple and the average ten year future S&P 500 returns. Based on where the CAPE multiple currently stands, investors can expect a ~5% average annual return in the coming ten years. This is very close to the average interest rate that high-quality bonds are currently offering. It makes it harder to justify taking the additional risk that comes with stocks.
Risks
This strategy is based on historical averages, and is not guaranteed to always work. In fact, we have had the yield curve inverted for much of the past few years, with the stock market performing well. Still, we think information coming out of the bond market should not be easily dismissed. There is a reason many bond investors are willing to take a lower interest rate for a longer term commitment. They wouldn't be doing so if they didn't see a good chance that the FED will have to lower rates in the future.
There are other factors to consider, and each investor is responsible for evaluating what their optimal stock-bond allocation should be.
Conclusion
From the data we have analyzed, it seems a good idea to favor bonds over stocks during periods of yield curve inversion. We are currently in one of those periods, and given the stretched S&P 500 valuation, and relatively high interest rates, it is probably a good time to favor bonds over stocks. This strategy might not suit everyone and isn't guaranteed to always work. Also, there are some sectors in the market such as REITs that are looking relatively undervalued. Based on current conditions and the S&P 500 valuation, we are rating SPY as a 'Sell' and BND as a 'Buy'.
For further details see:
SPY Vs. BND: The Untold Strategy For Timing Bonds Over Stocks