Goldman Sachs examines AI trading: Investors are increasingly concerned about "over-investment," and a downward revision of second-quarter revenue will severely impact valuation

Author: Zhao Ying

Source: Hard AI

Over the past year, AI has become one of the hottest topics in the market. However, this wave of enthusiasm is now under more scrutiny. Despite the strong performance of AI infrastructure-related stocks, investors still have concerns about the return on investment in AI.

Goldman Sachs pointed out in its latest report that looking at the four stages of AI transactions, NVIDIA is a clear beneficiary. Stocks of infrastructure companies in the 2nd stage have performed well, while stocks of software and IT service companies in the 3rd stage have shown volatility, with doubts about stocks of AI applications in the 4th stage.

Goldman Sachs mentioned in the report:

Over the past four quarters, the AI-related capital expenditures of large-scale cloud computing service providers (such as Amazon, Meta, Microsoft, Google) have reached $357 billion, accounting for 23% of the total expenditures of the S&P 500.

Capital expenditures and R&D investments in the TMT industry relative to cash flow are still lower than during the tech bubble period, but analysts' revenue expectations for these large-scale enterprises have not increased correspondingly with the increase in investment spending.

Goldman Sachs believes that the second quarter earnings season will be a crucial test, and investors should pay attention to revenue forecast revisions, which will be key to evaluating the sustainability of AI investment trends. Goldman Sachs expects to achieve a return on investment similar to recent history. Large-scale enterprises need to generate approximately $335 billion in revenue by 2025, otherwise they may face the risk of valuation downgrades.

Strong performance in the 2nd stage, volatility in the 3rd stage, doubts about the 4th stage

Goldman Sachs previously proposed a four-stage framework for AI transactions:

Stage 1: NVIDIA is the most obvious beneficiary;

Stage 2: Focus on AI infrastructure, including semiconductor companies other than NVIDIA, cloud service providers, data center REITs, hardware and equipment companies, security software stocks, and utility companies;

Stage 3: Companies expected to generate incremental revenue through AI, mainly software and IT service companies;

Stage 4: Companies with significant profit growth potential due to widespread adoption of AI to improve productivity;

Specifically, Goldman Sachs stated:

Driven by NVIDIA, the 2nd stage has seen a round of increases, rising 26% year-to-date; while with investors' doubts about the monetization of AI, stocks in the 3rd stage have recently experienced a decline, falling 19% between February and May; the 4th stage has seen almost no change in earnings, valuation, and performance.

The second quarter earnings season will be a crucial test

Goldman Sachs stated that the second quarter earnings season will be a crucial test of investors' optimism:

Although analysts expect NVIDIA's revenue growth to slow from 265% in the fourth quarter of 2023 to 25% in 2025, the stock's valuation remains above its 10-year average level, a trend that is common among large tech stocks Our sales expectations are higher than the general expectations. Most AI-related companies will release their financial reports by the end of July, with NVIDIA expected to report by the end of August, while many software companies in the third phase will release their financial reports by the end of August.

Investors are increasingly concerned about the issue of "overinvestment" in the AI field, especially by large-scale enterprises. Goldman Sachs pointed out that compared to the technology bubble, AI capital expenditures are still relatively modest.

Compared to the technology bubble era, current capital expenditures and research and development investments are still lower relative to companies' cash flows, and the profitability of TMT stocks is stronger.

In the past four quarters, the four super large-scale companies (Amazon, META, Microsoft, and Google) spent $357 billion on capital expenditures and research and development, accounting for 23% of the total spending of the S&P 500, with a significant portion of incremental investment due to AI.

During the most severe period of the technology bubble, due to the "overinvestment" of many telecom stocks, TMT stocks used more than 100% of operating cash flow (CFO) for capital expenditures and research and development.

In contrast, today's leading TMT stocks are very profitable. Although the proportion of capital expenditures and research and development to sales has increased, it is lower as a percentage of cash flow at 72%, and with the recovery of profits, this ratio continues to slightly decrease, compared to a median of 67% over 40 years.

On the other hand, investors also hold a skeptical attitude towards the "return on investment" potential of AI companies, questioning whether massive investments can bring enough sales growth and revenue. Goldman Sachs stated:

Consensus forecasts for capital expenditures and research and development by large-scale enterprises have increased by 7% in 2024 and 9% from 2025 to date. In dollar terms, analysts expect these large-scale enterprises to spend an additional $27 billion and $38 billion on capital expenditures and research and development in 2024 and 2025, respectively, compared to the beginning of the year. However, analysts have only raised sales forecasts for 2025 and 2026 by $17 billion and $19 billion, respectively.

Over the past 5 years, large-scale enterprises have on average converted 31% of their capital expenditures and research and development expenditures from the past 3 years into revenue. This total means that these companies need to generate $335 billion in revenue in 2025 to achieve a similar return on investment as in recent history, with revenue levels needing to grow by 16% compared to 2024.

Large-scale enterprises will ultimately be required to demonstrate that their investments will generate income and profits, failure to meet these expectations will result in significant declines in valuation and stock prices