S&P 500 “Waiting for Godot”

The expected return is an estimate of the investment return at a specific maturity, based on all available information at a given point in time. Chart 1 illustrates the expected annual implied return for the S&P 500 over the next ten years, currently at 4.66% (blue line, left axis), while the orange line represents the yield on the 10-year U.S. Treasury, standing at 4.31% (right axis). Using this data, we can calculate the Equity Risk Premium (ERP), which represents the additional return an investor earns by investing in the equity market compared to a risk-free asset like government bonds. As of today, this value stands at 0.35% (4.66% - 4.31%), a very low figure considering the difference in volatility between equities and government bonds.

One of the main reasons the ERP is so low is the current high valuations of the S&P 500. However, these valuations are not a justification for a market crash but rather serve to amplify the impact of a future downturn. A crash typically occurs due to a catalyst that disrupts the market and triggers a sell-off. In fact, when analyzing the S&P 500's returns since the 1950s, we can conclude that valuations have not had a significant impact on the index's performance in the following year. Charts 2 and 3 illustrate this point. In the first analysis, we constructed a distribution of historical S&P 500 P/E ratios from 1950 to the present, dividing them into quintiles. As shown, the five samples do not display significant differences in terms of performance recorded in the following year. The second analysis confirms this: the scatter plot highlights that there is no significant correlation between the initial P/E ratio of the S&P 500 and the subsequent year’s performance.

Chart 4 further illustrates this concept, showing the evolution of the Trailing 12-Month P/E ratio of the S&P 500. The historical average from 1954 until the Global Financial Crisis was 16.70, while the average over the past fifteen years has been closer to 18. 

The situation changes when we extend the time horizon, as starting valuations do impact future performance over the long term. Chart 5 presents the same analysis as before but over different time horizons: 3 years, 5 years, and 10 years. We can observe that, in the long run, starting valuations are negatively correlated with future positive returns. Furthermore, this effect becomes more pronounced as initial valuations increase.

Chart 6 highlights this trend. The histogram illustrates the probability of achieving a negative return over the next ten years when starting from the highest historical P/E multiples of the S&P 500, specifically those in the top decile. This probability stands at 40%.

Recently, Goldman Sachs published a study highlighting that, over the next ten years, the annualized return on the S&P 500 is expected to range between -1% and +7%, with a baseline scenario of +3% (Chart 7), not very different from our estimate.

Vanguard conducted a similar analysis, estimating the expected return over the next ten years for different segments of the equity markets. Table 1 shows that U.S. high-growth companies are the least attractive, while the most appealing opportunities lie in developed non-U.S. markets and emerging market equities.

In our view, the main reason behind muted expected returns for the following decade is that the unprecedented rise in corporate margins, experienced over the last 15 years, is going to be hard to replicate (Chart 8). It’s actually probable that part of the margin increase will mean-revert.

Table 2 was created using our internal quantitative model and allows us to estimate the future growth of the S&P 500 over the next 10 years, assuming different P/E multiples for the S&P 500 (vertical axis) and earnings growth rates (horizontal axis). By varying these metrics, we can observe the sensitivity of the outcome. If we assume the index’s return over the next decade will match its historical average, we should expect earnings growth of 10% and a P/E multiple of 20/22. However, if we assume that multiples will revert to the historical average of 15/17 and earnings growth will be 8%, the result will be an annualized index growth over the next 10 years in the range of 2.8%-4.15%, which aligns with Goldman Sachs' estimate, corollary of that will be risk free like returns but with much more volatility.

Housing and Construction Employment

In the second section, we analyzed the state of the real estate market and the potential consequences that changes in this sector could have on the labor market. The real estate market is characterized by a particular situation: the number of new permits relative to the population is lower than its historical average. This means that the number of homes under construction is lagging natural demand. As a result, there is an unexpressed demand in the market that needs to be addressed. This demand should not pose a problem in the coming years, as the long-term trend remains positive. This slight decline impacts the labor market through a causality effect: the slowdown in new permit approvals leads to a reduction in the construction of new homes, which in turn results in fewer workers, both on construction sites and in the broader economy. Compared to the first part of this relationship, we can say that new permits are a leading indicator of units under construction, as shown in Chart 11. The primary cause of this slowdown are interest rates, which have risen to high levels, leading those who were considering buying or building a home to delay their purchases.

The second part of the sequence highlights the relationship between units under construction and construction employment. The correlation between these two variables is weaker here, as observed in Charts 12 and 13, but still significant. Currently, despite the decline in units under construction, employment in the sector has remained stable. This is because homes must be completed before they can be sold, meaning there is a lag before the dependent variable follows the other. In the coming months, we will begin to observe this dynamic, which is expected to impact about 4% of construction employment, equivalent to a 0.20% increase in the country’s unemployment rate.

A 0.20% increase in the unemployment rate may not have a significant immediate impact, but it will likely generate a secondary effect, leading to a further rise in unemployment. This is because the real estate sector is a key driver of the U.S. economy, and a slowdown in this sector has historically led to broader economic deceleration, which in turn affects the labor market. For this reason, the unemployment rate could rise to 5%, raising further concerns about the resilience of the U.S. economy. This relationship is highlighted in Charts 14 and 15.

S&P 500 Free Cashflow representation

In the final part, we analyzed the sector's free cash flow relative to its weight (sector's market capitalization within the S&P 500) to assess which sectors are overvalued and which are undervalued according to this metric. Table 3 and Chart 16 illustrate this study. The second column of the table represents the sector's weight, with the total equaling 86.63% as Financials are not included. The third column shows the corresponding FCF, the fourth column represents the difference between the sector’s FCF and its weight in the index, and the fifth column reflects the implied growth of each sector. We can observe that there are few sectors that are massively underrepresented in terms of FCF relative to their weight in the index, such as Consumer Discretionary, Utilities and Energy, while one sector in particular, Health Care, has a significantly lower weight compared to the FCF it generates, the other sectors are relatively fairly valued.

Chart 6 highlights this trend. The histogram illustrates the probability of achieving a negative return over the next ten years when starting from the highest historical P/E multiples of the S&P 500, specifically those in the top decile. This probability stands at 40%.