Delphi EU Investment Update: Technology & AI Insights – Q3 2025 Update

This Delphi EU investment update highlights that since our last report, no material changes have been made to our investment portfolio. Our outlook on the profitability and growth potential of the companies we hold remains steady, and we continue to invest in the same companies, offering strong potential returns while maintaining minimal business risk.

Unlike most companies in our investment portfolio, about which we have no unusual news to write, we do want to elaborate slightly on the positive developments at Google, in which we are invested, and also address an anomaly we are identifying in the general market.

We have often written in the updates we send to you that we like to invest in companies where the potential-to-risk ratio is asymmetrical, investments where the risk is already priced in, meaning the price is low enough that even if the company sustains some damage, it would still be a reasonable investment at the very least. The significance is not only a reduction in risk, but also a high potential for return when the risk does not materialize. For this reason, Google has been, and still is, a significant investment in our portfolio.

During the last quarter, we witnessed two new developments that led to a significant price increase of about 37% in the last quarter.

The first development is a material reduction in the legal-regulatory risk. Following lawsuits, which claimed, among other things, that Google was misusing its (alleged) monopolistic position in internet search, many investors feared severe punishment if Google were found guilty. One of the central fears was that Google would be forced to sell Chrome, the popular internet browser. This could potentially harm Google’s search business.

These fears led to the stock underperforming relative to the company’s actual profits. The pessimistic scenario was already priced into the stock. Alongside the low price, our view was different from the prevailing opinion, even if the court ruled in favor of the plaintiff, it was not at all certain that selling Chrome was such a bad thing for Google. Furthermore, Google has other areas of activity, such as YouTube, which is currently the largest TV platform (viewing time on YouTube is higher than viewing time on Netflix, for example), its cloud computing services, and a growing fleet of autonomous vehicles it operates, a field in which it is a global leader. Google’s stock price was low enough to make the investment attractive even if the search business was harmed. In fact, the search business continues to grow, and therefore, alongside reduced risk, we saw a high potential return.

Following the court ruling, Google lost the battle but won the war. Although the court determined in April that Google acted anti-competitively, the verdict issued in August did not include significant punishment and, among other things, did not force Google to sell Chrome. After the “threat” was removed, the stock price jumped.

Another factor that weighed on Google’s stock price stemmed from increasing competition, in which people seek information using artificial intelligence tools. It was recently published that Gemini, Google’s chatbot, reached the top of the download charts in app stores, alongside other Google AI tools, such as NotebookLM and Google Labs, which are gaining an increasing number of users. It seems that the artificial intelligence threats to Google are also not as terrible as many investors expected.

 

Artificial Dynamics 

Regarding the adoption of artificial intelligence tools, it is important for us to express a certain reservation, mainly concerning language model-based chatbots (ChatGPT, Gemini, Claude, and others) and their impact on the general stock market.

It is clear that these tools are widely used by private users and within business organizations. What is less clear is how much the use contributes to the improvement of productivity in the organizations that use these tools. Studies often indicate the difficulty in measuring the improvement, and other studies indicate a decrease in productivity, due to receiving products that may look and be phrased very nicely, but lack real content, substance, and creativity.

Since the benefit of using these tools is not sufficiently evident in many cases, Open AI and similar companies are finding it difficult to charge high enough prices for licenses to use their chatbots. The continued development of these products is very expensive, and the required computing infrastructure is also very expensive. Insufficient revenue alongside high expenses creates a financial loss (Google has certain advantages because it is also a supplier of computing infrastructure and because it designs unique chips for itself, which lowers costs).

The widespread use of artificial intelligence tools has created high demand for chips dedicated to AI computations. The company Nvidia has benefited greatly from this soaring demand and has earned huge sums in recent years because of it. But if their customers do not make money, how will they continue to purchase these chips?

We are probably not the only ones asking this question; Nvidia is also aware of the situation. It is likely the reason why various “strange” deals have begun to be signed between Nvidia and its customers. For example, Nvidia will invest in Open AI so that it continues to purchase chips. On the other hand, Open AI wants to take some of the money invested in it and purchase shares in AMD, an Nvidia competitor.

Another transaction of a dubious nature was signed between Nvidia and CoreWeave, a provider of cloud computing infrastructure dedicated to AI applications. As part of the deal, the company will purchase chips from Nvidia, which in turn agreed to buy back the cost of use for any chip not subsequently sold to an end customer. Essentially, in a way, Nvidia is paying its customers to buy from it.

It is important to emphasize that not all of Nvidia’s sales are like this. A large part of them are “real”. But given the existence of transactions of the type we described and the fact that Open AI and other companies in the field are still losing money, it is very important to pay attention to these market dynamics, which are generously pricing companies that are involved, one way or another, in the field of artificial intelligence.

 

Summary

In summary, this Delphi EU investment update shows that no material changes were made to our portfolio during the last quarter, and we did not observe significant developments in the companies we hold, except for the positive progress regarding Google.

The portfolio continues to consist of leading companies in growing sectors that rely on internal resources rather than external debt.
As highlighted in this Delphi EU investment update, we remain focused on businesses with minimal risk while ensuring their valuations offer attractive medium, and long-term returns.

more articles

Delphi 2025 Annual Update

Delphi’s 2025 annual update summarizing portfolio performance, investment positioning, and our outlook on market trends, tariffs, and artificial intelligence going into 2026.

Throughout 2024 – 2024 Annual Update

Delphi’s performance was primarily driven by a significant improvement in the profits of the companies we invest in. According to third-quarter 2024 data (as fourth-quarter reports have not yet been released), the average operating profit of the companies in our portfolio...

Delphi delivered – 2023 Annual Update

In 2023, Delphi delivered a significantly nice returns. The main contribution to this performance came from the fact that the stock prices of the companies we invested in at the beginning of the year were absurdly low...