AI-Driven Stock Market Valuations: S&P 500 Metrics Explained

The current stock market, particularly in the United States, is a subject of intense debate, with many questioning whether it's in an AI-driven bubble. While tech companies dominate major indices and invest heavily in AI infrastructure, concerns about profitability, rising oil prices, treasury yields, and inflation persist. Despite these headwinds, the stock market continues to reach new highs, leading many to believe valuations are stretched.

Are Valuations Truly Stretched?

While some individual AI companies, like SpaceX, exhibit questionable valuations (high market cap with relatively low revenue), determining if the entire market is in a bubble is complex. The answer often depends on the metrics one examines. Some argue that the high valuations of AI stocks are justified by their massive future potential, while others foresee a potential crash. Predicting the future is difficult due to numerous variables.

To simplify the analysis, we can focus on whether the S&P 500 is expensive based on historical standards. Even this question yields varied answers depending on the chosen valuation metric. Understanding these different perspectives is crucial for investors.

Valuing the S&P 500

The S&P 500, comprising 500 of the largest US stocks by market capitalization (with additional inclusion criteria like earnings requirements), serves as a proxy for the broader stock market, covering about 83% of its total market capitalization.

To value an index, its level alone is insufficient. It must be compared to fundamental metrics of the underlying companies' performance or expected future performance. Unlike individual companies where a discounted cash flow (DCF) model can be used, applying DCF to 500 companies would introduce too much error. Instead, a relative valuation is used, comparing the index level to a market-cap-weighted average of a fundamental metric across all 500 companies. Earnings are the most common metric, used to calculate the Price-to-Earnings (PE) ratio.

The Schiller PE (CAPE Ratio)

The Cyclically Adjusted Price-to-Earnings (CAPE) ratio, or Schiller PE, suggests the market is significantly overvalued. Currently above 40 times, a level last seen during the dot-com bubble, it's approaching new all-time highs.

The CAPE ratio is cyclically adjusted, using the average inflation-adjusted earnings over the past 10 years. This smoothing factor reduces year-over-year volatility and provides a long-term view. However, it places less weight on recent earnings, which can be problematic during periods of rapid earnings growth. If earnings are growing quickly, the CAPE ratio might overstate valuations.

For instance, in Q3 of this year, S&P 500 earnings grew by nearly 30%, marking the third consecutive quarter of over 25% year-over-year growth (compared to a 25-year average of 6.4%). If AI is indeed a revolutionary technology driving rapid future growth, the CAPE ratio might be overstating current valuations.

Trailing PE Ratio

A more simplified trailing PE ratio, which uses earnings from the past 12 months, presents a different picture. While more volatile, this measure is currently elevated but well below the peak of the dot-com bubble. This suggests that current company performance is backed by earnings, unlike during the dot-com era.

Forward-Looking PE Ratios

Investors often base decisions on future expectations. Forward-looking PE ratios divide the stock price by estimated earnings for the next 12 months. This introduces the challenge of forecasting future earnings. Consensus estimates, which are the average of analyst forecasts, are often used.

According to Yardi Research, the forward PE for the S&P 500 is around 19 times, above the long-term average of 16 times, but still below the dot-com peak of 25 times. It has also been declining over the past year.

Forward PEG Ratio

The forward PEG ratio (forward PE divided by the 5-year estimated earnings growth rate) offers another perspective. A ratio above one generally indicates expensiveness, while below one suggests a cheaper valuation relative to growth. This metric controversially suggests the S&P 500 is currently cheap, even the cheapest it has been in decades. This implies analysts expect earnings growth to outpace current higher valuations. The consensus 5-year earnings growth estimate has surged to 27.3%.

Comparing the S&P 500 to Other Indices

US Small and Mid-Cap Indices

Comparing the S&P 500 to the S&P 600 (small-cap) and S&P 400 (mid-cap) indices reveals that these smaller companies have lower forward PE ratios (around 14-15 times) compared to the S&P 500's nearly 20 times. However, when looking at the PEG ratio, the S&P 500 is actually on the lower end (0.69), indicating it might be cheaper relative to its expected growth compared to the S&P 400 (0.8) and S&P 600 (0.96).

International Markets

Comparing the S&P 500 to the MSCI All Country World Index ex-US (an international index covering 22 developed and 24 emerging countries) shows that international markets are relatively cheap by historical standards, with a forward PE of 12.5 times. This is partly due to surging earnings estimates for these markets. While a direct PEG ratio comparison is difficult due to data availability, international markets are expected to grow earnings at a similar pace to the S&P 500 despite trading at a significant discount.

Important Caveats and Considerations

1. Reliance on Forecasts: Forward valuations depend on forecasts, which are often overly optimistic, especially during good times. Analyst estimates tend to have downward revisions over time.
2. S&P 500 Concentration: The index's performance is heavily influenced by a few large companies, such as the "Magnificent 7" (Google, Amazon, Apple, Meta, Microsoft, Tesla, Nvidia), which represent a third of the index and have above-average valuations. The eventual IPOs of highly unprofitable AI companies like OpenAI or Anthropic could further skew valuations.
3. Source of Earnings Growth: The massive forecasted earnings growth appears to stem predominantly from cost savings (15.6% revenue growth forecast) rather than rapid revenue expansion. Cost savings have limits and could be diminished by rising AI usage prices as currently unprofitable AI companies seek sustainable pricing models. The CAPE ratio was designed to account for such unsustainable growth periods.
4. Underrepresented Costs: Much of the AI spending is in capital expenditures (chips, data centers), which are treated as investments on the balance sheet rather than immediate expenses on the income statement. This can make earnings figures appear higher. Concerns exist about the lifespan of these assets being shorter than accounting assumptions, leading to potential overvaluation based on PE ratios.

The Equity Risk Premium

The equity risk premium (earnings yield minus the 10-year Treasury yield) indicates the marginal return investors receive for investing in stocks over Treasury bonds. On a forward basis, this premium is currently around 4%, which is not as low as during the dot-com bubble but is the lowest it has been in some time. This suggests investors are taking on equity risk for a lower marginal return, partly due to the recent spike in 10-year Treasury yields. Higher treasury yields can draw capital out of the stock market, particularly impacting AI companies reliant on investor capital for growth.

Conclusion

While analysts anticipate significant earnings growth from AI, which could justify current valuations and lead to solid stock performance, the current market performance is historically abnormal. A lot of upside is already priced into stock valuations, presenting considerable risk to investors. It's challenging to definitively declare whether the market is in a bubble or not, as various metrics offer different perspectives. A nuanced understanding of these factors is crucial for investors navigating the current market landscape.