S&P 500 performance is driven by only a few stocks 

As we all know, the sunken part of an iceberg can be very different from what is visible on the surface. This allegory seems particularly apt for what we have seen in the U.S. equity markets since the beginning of 2023.
While the S&P 500 and Nasdaq indices have performed remarkably well since the beginning of the year (8% and 17% respectively), only a very limited number of stocks have contributed to the indices' advance.
Indeed, the 10 largest market capitalizations are responsible for 86% of the overall performance of the S&P 500 Index since the beginning of the year. As shown in the chart below, an index made up of the 5 largest market capitalizations (Apple, Microsoft, Amazon, Nvidia and Google/Alphabet) is up +29.8% since the beginning of the year. The index made up of the other 495 stocks in the S&P 500 is up "only" 3.7% over the same period.

Performance of the 5 largest contributors to the S&P 500, the S&P 500 Index and the index that represents the remaining 495 stocks

Source: TME, Jefferies, Factset

Artificial intelligence and LLM, the new 
bubble? 

For JP Morgan strategists, a new thematic bubble may be forming, of companies exposed to artificial intelligence and Large Language Model (LLM). As shown in the chart below, the 6 large caps related to this theme (Microsoft, Alphabet/Google, Amazon, Meta/Facebook, Nvidia and Salesforce. com) have recorded a cumulative market capitalization increase of $1.4 trillion since the beginning of the year, a weighted performance of +45%. 

A breakdown of the S&P 500 Index performance in 2023

Source: JPM

The current corporate earnings season reveals the importance given to this theme by companies and investors. It is no coincidence that Alphabet/Google executives mentioned the term artificial intelligence no less than 58 times during the press conference following the release of Q1 results.    

Amazon's real revenue growth (%)

Source: Crescat Capital 

Yet, the valuation multiples paid by investors show that GAFAs are still considered growth stocks by the market. Based on the last 12 months' earnings, Alphabet/Google's price per earning (P/E) is 22x, Meta/Facebook's is 24x, Apple 28x, Microsoft 32x, Amazon 115x and Nvidia 156x.. Concerning the latter, which is often perceived as the most emblematic large tech capitalization of the AI bubble, the price/revenue ratio is 24x which also seems very generous (even if the revenue could continue to grow by 15-20% in the coming years).  
If we take into consideration the valuation of the entire Nasdaq 100, we see that the "earnings yield" (i.e. the inverse of the P/E) is equivalent to the 10-year yield on US Treasury bonds. This indicator, which compares the valuation of an equity market to that of bonds, seems to show that the Nasdaq 100 index (heavily weighted in technology stocks) has rarely been so expensive compared to bonds.  

Conclusion

Too often, investors tend to comment on how the market should behave, rather than trying to decipher the message it is giving. Clearly, the behavior of the S&P 500 index has been ambivalent and difficult to decipher since the start of the year. On the surface, the S&P 500's increase of more than 8% (i.e. an annualized performance of nearly 26%) gives a rather reassuring picture of the US economy and investor sentiment. On the other hand, the strong dichotomy in performance between the 5 or 10 largest capitalizations and the rest of the index gives a completely different picture,  suggesting that the economy is far from being in good shape. In fact, the Russell 2000 index of small and mid-cap stocks - the one that best represents domestic economic activity - is virtually unchanged since the beginning of the year.