Will a stock market bubble burst due to AI soon?
Facing the hype around Artificial Intelligence (AI), some experts are warning of a bubble. Fund manager David Wehner of Do Investment, the asset manager, sees at least an overvaluation, as he explained in an interview with ntv.de. What investors should pay attention to.
ntv.de: Critics see an AI bubble on the stock market, the latest being the prediction of an MIT professor causing a stir: Goldman Sachs had Daron Acemoglu calculate, why he expects much lower productivity growth (0.5 percent in the USA) and GDP (0.9 percent) from AI in the next ten years than optimistic observers, who expect a leap of up to 10 percent. What's your assessment: Is it a bubble?
David Wehner: I would speak of a certain overvaluation. The sector is dominated by very few players, namely Nvidia, AMD, and ASML in the semiconductor industry, as well as the large tech companies like Microsoft, Google, or Apple. They have experienced enormous growth in market capitalization in the past one and a half years, mainly because many investors want to invest in this area, but there are only a few players. The big question now is whether the narrative in their earnings or at least forecasts will materialize.
Are you counting on it?
Since the beginning of stock market history, new developments have been overvalued at first and undervalued in the long run. Every innovation generates a clear advantage, but not in the initially given time. In general, it takes not two, three, but ten to twenty or even thirty years for the potential to manifest itself. At the moment, AI stocks are relatively expensive, which means there is potential for a correction: If expectations do not materialize in the next two, three years and there are perhaps setbacks in research or computing power. I would not compare it to the Dotcom bubble, back then the affected companies were not as economically strong as the AI players today, and there was also a hype about stock market manipulation. But: The AI sector could be subjected to a reality check.
MIT Professor Acemoglu's core of his prediction - based on the analysis of studies - is that he does not believe in such a high tempo of transformation through AI. AI will have effects on less than five percent of all tasks in the next ten years, only a quarter of the tasks affected by AI will become cost-effective to automate within ten years. Do you share this assessment?
Yes, the infrastructure required to be built is very capital-intensive. Currently, only one company - Nvidia - can produce enough chips and computing capacity. There is hope that others like AMD or Intel can become competitors, which would lower prices and increase productivity. However, it is not yet clear that this bottleneck will resolve itself in the short term. To accelerate development, chips would need to be more widely available.
Which AI stocks are particularly overvalued?
I'm thinking of the top players in the semiconductor industry and of the tech companies with particularly high market capitalization. In addition, there are some niche players with low market capitalization that have also profited from the hype and are now facing potential correction with price drops.
Should investors therefore sell their AI stocks now?
Investors who were in early, i.e., those who had the right nose for it, do not have to sell everything right away. But they could consider realizing profits and diversifying their capital.
Is it worth jumping on the AI bandwagon now and buying the already expensive papers?
We, as a company, are currently in a waiting position. We are waiting for the quarterly results of the next few weeks and observing if the euphoria is starting to show up in the results or at least in the forecasts. Our portfolio is currently very diversified, spread across almost all essential sectors, with no clear over or underweights. The majority lies between five to ten percent in weight.
You manage a fortune worth over a billion Euros for wealthy private individuals, middle class entrepreneur families, foundations and state-owned companies - how much of that have you invested in AI stocks fueled by hype?
We follow a very pragmatic approach and adjust our portfolio to the market environment. During the stock market boom in the last three years, we had up to about 30 percent of our investments in tech stocks, as we focus mainly on large capitalized companies with a corresponding high stock market value. Now we have reduced that share to approximately 15 percent.
Which AI papers have you invested in?
Since we focus on market heavyweights, nearly all major technology companies could be in our portfolio.
What is the overall stock percentage in your depots?
This varies depending on the investment goal - a large portion of our clients have 50 to 80 percent of their assets in stocks.
Do you have an underdog tip that has been overlooked in the context of the AI hype?
Beneath the top performers, which mostly come from the technology sector, there are now many companies and sectors that are undervalued and have not been able to profit from the hype of the last months. From their assessment and history, they have enormous catch-up potential, for example in the areas of consumer goods, pharmaceuticals or industry, as well as the automotive industry. It's worth taking a look at them.
How did it come about that there is an overvaluation in the AI sector, why is generative AI overhyped in your opinion?
In the case of new innovations, there is initially a gold rush mentality, large investments flow into a sector. In this phase, assumptions are made about growth, productivity increases, efficiency, and of course profits. Just like now with AI, it's often only a few companies that can represent the innovation and profit in this phase, attracting a lot of capital. When the expectations for an innovation are only partially met, it becomes widespread, a competitive situation arises. Prices fall, capital is withdrawn and invested in other companies. In the case of AI, the focus is still very much on the innovators of the first hour, who still bind a lot of capital. Their shares have risen strongly, but the profits have also followed suit, making the price-earnings ratio fairer. The overvaluation has somewhat normalized. However, these companies are still relatively highly valued, and now the question is whether they can fill their high valuations with enough profit - or there will be disappointments.
What role does AI really play in your own work?
At our company, the human being is in the focus, and therefore AI currently plays a very subordinate role. We do not use investment tools that generate buy signals, evaluate companies, or detect valuation anomalies, but our own understanding. We base our investment decisions on fundamental valuation, both conjuncturally, geopolitically, and for individual titles. We make our investment decisions independently and represent them to our clients and investors.
Interview with David Wehner conducted by Christina Lohner
Given the context provided, here are two sentences that incorporate the given words and follow from the text:
Despite the warnings of an AI bubble, some investors continue to invest heavily in AI stocks, with companies like Nvidia, known for its work in AI and stock trading, seeing significant growth in market capitalization.
In light of the predicted lower productivity growth from AI by MIT Professor Daron Acemoglu, it remains to be seen if the high stock prices of AI-focused companies like Nvidia, Microsoft, and Google will be sustainable over the long term.
