This article draws on two Goldman Sachs research reports on tech stocks and AI investing. For accuracy and compatibility, cited content is paraphrased. The original reports can be accessed at the bottom. An English version will also be published on Seeking Alpha.
The most significant risks and strategic responses in technology stocks and AI investments.
(Source: Goldman Sachs)
The first report is titled "AI: Too Much Spend, Too Little Benefit?" Goldman Sachs explores the massive capital expenditure (capex) by tech giants on AI, expected to exceed $1 trillion over the next few years. The report questions whether this huge investment will yield corresponding returns. It includes interviews with industry experts like MIT's Daron Acemoglu, who is skeptical about AI's short-term economic benefits, and Goldman's Jim Covello, who argues that AI is not designed to solve complex problems, making its high costs unjustifiable. Goldman notes that AI technology is becoming a significant investment in the global economy, with capex expected to exceed $1 trillion in the coming years. While there is disagreement on AI's long-term impact—some like Joseph Briggs predict AI will significantly boost productivity and GDP, while Daron Acemoglu is more conservative—the potential for AI to drive economic growth is enormous, especially in productivity. However, the path to realizing these benefits is not smooth. Cost is a major barrier to AI adoption, as AI technology is extremely expensive and needs to solve complex problems to justify its cost. Additionally, market competition, such as Nvidia's dominance in the GPU market, may hinder cost reduction. A Short History Of AI Developments Although AI technology is not yet mature, its long-term potential is emerging as technology advances. No "killer app" has yet emerged to drive widespread AI adoption, but this could change as technology evolves. Beyond that, constraints on AI growth include chip shortages and power supply issues, which may limit AI growth, especially in applications requiring massive computing resources. Despite these challenges, AI infrastructure providers and cloud computing companies may continue to benefit from AI growth. At the macroeconomic level, AI development could affect economic policies and market performance across regions. For example, interest rate decisions by the Fed, Bank of Japan, and ECB may be influenced by AI growth expectations. Currently, returns on AI investment are becoming visible, particularly in semiconductors and cloud computing. The semiconductor industry is poised for growth as demand for AI chips surges. Cloud computing companies are also performing well due to the massive computing power needed to train and run AI models. However, AI development also carries risks and regulatory challenges. AI model performance is limited by training data quality, and misuse of AI, such as deepfakes, raises concerns. In summary, despite disagreements on AI's benefits and costs, the AI sector still has significant growth potential. Whether because AI is starting to deliver on its promises or because any bubble will take time to burst, AI will remain a key growth area for years to come.
The second report is titled "AI: To buy, or not to buy, that is the question." Goldman's report aims to clarify misunderstandings about its stance on AI after the previous report "AI: Too Much Spend, Too Little Benefit?" The new report delves into what they call "tech's rational exuberance," suggesting that while AI investments are significant, they do not represent a bubble akin to the internet era.
Goldman notes that since 2010, the tech sector has generated 32% of global equity returns and 40% of US equity returns. These returns are based on strong fundamentals, not investor frenzy. Tech sector EPS has quintupled from post-financial crisis highs, and the median P/E and EV/sales for the "Mag7" are only half of those for the top 7 companies during the 2000 dot-com bubble (Figures 1, 2 above).
Current tech stocks and AI investments cannot yet be called a bubble. Overall valuation levels are lower than those during the "Nifty Fifty" period, Japan's bubble, and the dot-com bubble. The Mag7's valuations are only half of those during the dot-com bubble (Figures 5, 6 below).
From a capex perspective, although capital spending has increased and expectations for future revenue have accelerated, the cash flow payback period embedded in current valuations is still far below the peak of the 2000 tech bubble. At the height of the tech bubble, TMT stocks spent over 100% of operating cash flow (CFO) on capex and R&D; today, it's about 72%. The 40-year median is 67%.
The biggest risk in tech stocks and AI investing is concentration. While valuations of these companies may not be as extreme as in other narrative-driven bubbles, the scale of market dominance this time is larger. The ten largest stocks account for over a third of the index, the highest ever, and the five largest stocks make up 26% of the S&P 500's total market cap. As the market becomes increasingly dependent on the fate of so few companies, stock-picking errors can have a massive impact on performance.
The chart below shows the current weight of the top 5 and top 10 companies in the S&P 500, far above historical averages. Goldman says this is not necessarily irrational, but advises clients to consider the risk of over-concentration and diversify. Historically, as new entrants emerge, competition forces companies to disappear, merge, or be acquired, and few companies' earnings remain unscathed. From this perspective, a market dominated by a few stocks is increasingly vulnerable to disruption or antitrust regulation. Even experienced companies...
Since 1955, only 51 companies from the Fortune 500 have survived to today.
To address this, Goldman created a portfolio called "Ex-Tech Compounder Stocks" (Figure 30 below). This portfolio screens for companies with market cap > $10B, high margins (EBITDA > 14%, EBIT > 12%, net profit > 10%), high profitability (ROE > 10%), strong balance sheets (net debt/equity < 75%, ND/EBITDA < 2x), low volatility (vol < 50), strong growth prospects (sales > 4%, earnings > 8%), and consistent earnings growth over the past decade.
The ETC portfolio has outperformed the global market portfolio over the past year.
Buffett once said: "Diversification is protection against ignorance." Most investors cannot truly understand a company deeply enough to trust their judgment and hold through cycles. Even Buffett's heavy position in Apple is only a small part of his total portfolio. Most of us need to be aware of diversification and build a good margin of safety in our portfolios.
Risk disclaimer: The views in this article are for reference only and do not represent investment advice. Market risk exists; invest with caution.
Like, reply "research" in the background to get the above content. Interested? Follow for more.
