Is AI’s bubble about to burst?
October 16, 2025
-
4 min
The technological transition, driven by AI and automation, is reshaping the global economy. But will AI be a repeat of the dot-com bubble? Jens Peers, CIO of Mirova US and portfolio manager of the Mirova Global Sustainable Equity sees similarities but discusses below the key differences in today's tech giants. He also highlights the ongoing opportunities in cloud computing and data storage, as well as the next big trend which excites him.
You describe your global equity investing strategy as multi-thematic. Can you briefly tell us about these four trends?
Jens Peers: We invest in companies with meaningful exposure to economic tailwinds from long-term transitions that are affecting the global economy: demographics, environmental, technological and governance. The technology transition includes artificial intelligence, digitalisation of the economy, and increasing connectivity and automation which are expected to transform industries, improve efficiency, and create new opportunities.
And roughly what percentage of your portfolio is invested in the technological transition? Do you have set limits or targets?
Jens Peers: We don't really have set targets. We create the portfolio from the bottom up, based on stock opportunities and valuation opportunities from these stocks as well. While we were a lot more focused on environmental factors, let's say about ten years ago, we're now a lot more biased towards technology. Technology as a transition and the companies that offer solutions there roughly represent about 45% of the portfolio today. It has to be said that some companies obviously contribute to and benefit from multiple trends.
Technology, particularly AI, has been the hottest investment theme of recent times. There are concerns about overinflated valuations, though, and comparisons to booms and busts of the past, like the tech wreck of the early 2000s. How exposed are you to the tech sector, and how do you manage the risks of investing in tech, robotics, and AI alongside all these exciting opportunities?
Jens Peers: It is true that there are many similarities today, specifically given the outperformance of mainly the AI names, with the dot-com crisis. Ultimately, that turned out to be a bubble that burst. But don't forget that the Internet really drove big changes in our economy, and even throughout that bubble back in those days, the most important companies today and the best performances over the last 20 years were actually found in companies with exposure to that theme.
In terms of the boom and bust after that, we see many similarities. The market was back then and certainly is today very concentrated in terms of the highest weights of individual companies, themes and sectors, like tech. Right now, valuations of many of those AI names, specifically a lot of startups, are very high, and expectations are huge as well. The timing, even if AI is really changing our world significantly, is also dependent on many other things that need to be done. Such as investing in better energy infrastructure and getting access to water.
We saw the same happen about 25 years ago, where there was a boom first, followed by a bit of a bust.
However, there are also differences. The main companies active in this space now, compared to 25 years ago during the dot-com crisis, are very well capitalised. They have strong balance sheets, generate a lot of cash flow, and continue to invest significantly in this area, mainly through the creation of that free cash flow. That should create a bit of a buffer on that valuation.
But it's true that these AI companies have outperformed the other part of the technology industry. However, we believe software companies today are very attractive after a strong underperformance versus those AI companies and the semiconductor companies that also benefit from AI. The performance gap and the valuation gap between software companies and the market – which thinks these companies might actually suffer from the fierce competition with AI - are at a very high level. So, in our portfolio, we have some exposure to some of those semiconductor names, like NVIDIA and Taiwan Semiconductor, but we also have software companies in the portfolio, like Salesforce, where we see a lot more valuation opportunities, and there's certainly no question about bubble valuation.
AI is very closely tied to cloud computing, and we've seen huge profits and growth in cloud computing in recent months. How far into that move to cloud computing do you think we are?
Jens Peers: AI depends a lot on data, and the way that the AI models work depends heavily on access to the data in the cloud as well. Roughly around 50% to 60% of all the data in the world today is stored in the cloud already, so we've seen a big move to cloud computing. But we're only about halfway there, so there's still a lot more opportunities. We've seen that this requires a lot more investments in data centers— not just for learning but also for storing the data as well. There's still a lot of opportunities in this space.
Looking a bit further into the future now, are there any emerging themes tied into these big megatrends that you're watching closely or that you think will grow into exciting investment opportunities?
Jens Peers: Well, there are plenty. AI will change the way we deal with everything, really. In the short to mid-term, AI will help develop solutions in the pharmaceutical space, specifically in areas like oncology, and probably also Alzheimer’s and Parkinson’s.
One area that excites me a bit more long-term is the concept of a self-driving car. This will only happen very slowly, given the replacement cycle of cars. But we are at the beginning of huge changes in this industry, and the adoption of electric vehicles, having more data around electric vehicles, and greater data connectivity through satellites combined with AI will lead to a huge breakthrough with autonomous vehicles. That's one of my most exciting longer-term new themes where we currently have no exposure, but it's an area we are certainly looking at a lot and expect to bring into the portfolio in the future.
