S&P 500: Overvalued or Poised for Growth?

Major tech companies are heavily investing in AI infrastructure, often despite not yet achieving profitability, while rising oil prices, bond yields, and persistent inflation add to market anxieties. Despite these worries, the market continues to hit new all-time highs, sparking a debate on the sustainability of current valuations.

Richard noted that many people are concerned the stock market, especially in the US, is in a bubble.

While individual AI-related stocks may exhibit high valuations, determining if the broader market is in a bubble requires a nuanced perspective.

SpaceX, for example, serves as a high-valued company whose revenue does not place it among the top 200, highlighting how specific company valuations can differ from general market trends. Some analysts argue that current high valuations for AI companies are justified by their long-term growth potential. Richard also indicated that it is difficult to determine if the market as a whole is in a bubble, as it depends on what one looks at.

This index adheres to specific inclusion criteria, notably requiring profitability, which excludes some high-profile companies like SpaceX. Richard noted that it has become a proxy for the US stock market and covers about 83% of total market capitalization, serving as a primary investment instrument for many retail investors.

Richard explained that a common metric used for this purpose is earnings, which results in the widely recognized P/E ratio.

The Shiller P/E (CAPE) ratio currently exceeds 40, a level not observed since the dot-com bubble era, and it is quickly approaching a new all-time high.

This cyclically adjusted P/E ratio utilizes the average inflation-adjusted earnings over the past 10 years to smooth out volatility and mitigate distortions from business cycles. However, the CAPE ratio may understate periods of rapid earnings growth, potentially leading to signals of overvaluation.

This short-term growth is heavily influenced by extensive AI infrastructure build-outs. While trailing 12-month P/E ratios are elevated, they remain substantially below the peak observed during the dot-com bubble, according to Richard.

The forward P/E ratio, calculated by dividing the stock price by the next 12 months of expected earnings, provides a forward-looking perspective on market valuation.

Data from Yardeni Research indicates a current forward P/E of approximately 19 times, which is above the long-term average of 16 times but still well below the 25 times peak seen during the dot-com bubble. Richard noted that this metric is elevated at about 19 times, above the long-term average of 16 times, with consensus estimates representing the average of all analyst forecasts.

Based on this specific metric, Richard stated that the index appears as cheap as it has been in decades, suggesting it is not just cheap, but remarkably so.

Despite this, the S&P 500 holds the lowest PEG ratio at 0.69 among these indices. International markets, as represented by the MSCI ACWI ex USA index, trade at an even lower forward P/E of 12.5x, offering a valuation discount while projecting similar growth rates to the US market, according to Richard.

Richard emphasized that the biggest point is that for future valuations, analysts rely on forecasts, and everyone knows how good people are at that.

Analyst estimates, represented by 'squiggles' over 25 months, frequently show downward trends even as actual earnings rise, indicating their inherent inaccuracy.

Richard stated that analysts tend to be overly optimistic during economic upturns, making these forecasts a limited tool for accurately assessing market value.

These companies command significantly higher valuations compared to the remaining 493 constituents. Richard mentioned that although the index includes 500 different companies, its performance is largely determined by just a handful of firms, and future IPOs of AI leaders like OpenAI or Anthropic could further skew valuation metrics.

Current high earnings growth appears largely driven by cost savings rather than substantial revenue expansion, which is projected at a more modest 15.6%.

Cost savings from AI are expected to face pressure from rising usage costs as AI companies move towards sustainable pricing models. Richard noted that the current high growth is likely not sustainable long-term, and the CAPE ratio was specifically designed to look past such unsustainable growth phases.

Large AI investments, such as chips and data centers, are recorded as capital expenditures on the balance sheet rather than immediate expenses.

Critics express concern that hardware like chips may have shorter lifespans than traditional accounting models assume. Richard pointed out that many of these companies have spent all their profits on AI investments, leading some to question the underlying assumptions of these accounting practices, which can make P/E ratios based on current practices misleading.

The Equity Risk Premium, calculated as the earnings yield minus the 10-year Treasury yield, currently stands at approximately 4%.

This figure, while not as low as during the dot-com bubble, represents its lowest level in recent times, potentially signaling increased market risk. Rising bond yields act as a 'gravity' force on stock valuations, threatening AI companies that depend on external capital for growth, according to Richard.

The equity risk premium quantifies the additional return investors receive for investing in stocks compared to safer government bonds.

It is calculated by subtracting the 10-year Treasury yield from the earnings yield. Richard explained that the forward-looking equity risk premium is currently around 4%, which, while higher than during the dot-com bubble, marks its lowest level in recent memory.

Richard emphasized that higher yields act like gravity on stock valuations, and standard valuation metrics often fail to capture the full potential upside of the AI revolution.

Market timing is a notoriously difficult endeavor that investors should avoid, as assessing current market conditions involves inherent complexity that defies simple narratives.

Richard stated that investors should be cautious of overconfidence, whether in declaring a bubble or believing in an AI-driven paradise, adding that the goal is to provide a nuanced understanding rather than a definitive prediction.