AI Trading Phases: Infrastructure Matures, Beyond Applications Looms

Over the past six months, the volatility of AI-related stocks has intensified. After a recent rebound, Goldman Sachs pointed out in its latest research report that the AI "second phase" trading will tend to mature, and investors need to approach the "third phase" investment prudently.

Half a year ago, we cited Goldman Sachs' "four-phase theory" on AI investment in an article. Among them, as the most clear beneficiaries of AI, NVIDIA belongs to the "first phase", the "second phase" includes companies focusing on AI infrastructure, the "third phase" focuses on companies with the potential to monetize AI by generating incremental revenue, and the "fourth phase" mainly includes companies with the greatest potential for profit improvement in the widespread adoption of AI and productivity enhancement.

Goldman Sachs believes that the current investment focus is on the second phase of infrastructure. Typical stocks in this category have risen by 27% since the beginning of the year, and their valuations are 0.4 standard deviations higher than the 10-year average, gradually entering a high valuation phase. Goldman Sachs warns that the future growth of stocks in this phase may depend more on actual earnings performance rather than further improvement in valuation:

We expect the second phase of trading to mature, with stock price increases driven by earnings growth rather than valuation expansion.

Advertisement

The initial valuation is higher than the average, NVIDIA's sales exceed expectations, the capital expenditure of large-scale cloud computing service providers exceeds expectations, and the frequency of enterprises mentioning AI has slowed down.

The upcoming third quarter earnings season will be an important test.

For the third phase of AI application stocks with the potential to generate incremental revenue, Goldman Sachs suggests being cautious. However, Goldman Sachs believes that in this phase, "platform" type stocks show strong investment appeal:

Although the valuation is below the average, the timing of AI application landing and monetization is still unclear.

Investors who are confident in the subsequent development stages of AI can choose some "platform" type stocks.Phase Two: Growth Challenges Under High Valuation

Currently, investment hotspots are primarily concentrated on infrastructure stocks in Phase Two. These stocks have seen a 27% increase so far this year, with valuations 0.4 standard deviations higher than the 10-year average.

Among them, large-scale AI investments by tech giants are an important factor supporting these stocks. Tech giants such as Amazon, Google, Meta, Microsoft, and Oracle have announced large-scale AI-related investments, with capital expenditures expected to reach $215 billion and $250 billion in 2024 and 2025, respectively, a significant portion of which is AI-related.

However, high valuations also bring growth challenges. Goldman Sachs analysts pointed out that the growth of these stocks in the future may rely more on actual earnings performance rather than further increases in valuation:

We believe that there is still room for Phase Two stocks to rise, but future gains will be more driven by earnings growth rather than valuation increases.

The current stock prices of these companies have exceeded expected earnings per share, reflecting investors' optimistic expectations for AI potential. Some Phase Two stocks have recently risen due to AI-related positive news (such as increased shipments by Advanced Micro Devices and the nuclear energy agreement between Constellation Energy and Microsoft).

However, due to the already high starting valuations, future returns may be limited, even though valuations have limited predictive power for the short-term performance of large-cap stocks.

It is worth mentioning that the company also noted in the report that the degree of AI spending exceeding expectations is narrowing, and the frequency with which companies mention AI is also leveling off, which may suggest that future gains for Phase Two stocks may be more moderate:

The degree of AI spending exceeding expectations is narrowing, which suggests that future gains for Phase Two stocks may be more moderate. Although current valuations are slightly high, if the AI investments of tech giants exceed expectations, they could still drive stock prices higher. At the beginning of 2023, demand for Nvidia chips greatly exceeded expectations, and expectations for capital expenditures by cloud giants in the first half of 2024 were continuously revised upwards. However, these exceed expectations have been shrinking recently. The upcoming third-quarter earnings report will be an important test.Similarly, the frequency with which companies mention AI and the willingness of management to spend on IT, while still above historical levels, have become stable. The proportion of companies mentioning AI in the second quarter of 2024 reports is slightly lower than in the first quarter. Our IT spending survey in September also shows a slight decrease in spending intentions compared to January.

Third Stage: Prudent Selection of Application Stocks

At present, the third stage of AI application stocks is in a horizontal period, with valuations below the mean. However, Goldman Sachs believes that the timing for AI application development and monetization is still unclear, and it is difficult to fully deploy in the short term:

Based on our conversations with investors and recent stock performance, the third-stage stocks with potential AI-enabled revenue have mostly failed to meet expectations in application development and monetization. We expect this uncertainty to continue in the short term.

According to the Goldman Sachs IT spending survey, management expects only 3% of the IT budget to be allocated to generative AI next year. By 2025, the willingness to spend on AI products may improve, as management will have a clearer understanding of the direction of Federal Reserve policy and election results.

However, the company believes that "platform" stocks show strong investment appeal in this stage:

"Platform" stocks may represent an attractive intermediate subset of the third-stage stocks.

Our analysts believe that "platform" stocks, including databases and development tools, will become the main beneficiaries of the next wave of generative AI investments.These platforms are not only capable of making the best use of AI infrastructure but also provide foundational modules for building the next generation of applications.

Phase Four: Still a Ways to Go

For phase four stocks, which are companies expected to gain the most significant productivity improvements from the widespread application of AI, Goldman Sachs analysts believe that their performance is still a ways to go. Data shows that currently, only 6% of businesses are using generative AI in production, a proportion far below market expectations. Analysts at Goldman Sachs point out that investors are likely to gain confidence in these types of stocks only after substantial progress is made in the adoption of AI applications in phase three:

We believe that the launch of applications in phase three stocks is a necessary condition for investors to gain confidence in phase four stocks (these stocks have the greatest potential for profit growth due to AI-related productivity improvements).