Google Q2 & Tesla Q2 – Yin and Yang

Google Q2 25 – Holding off OpenAI.

• Google reported good results but had to increase capex to keep up in the AI arms race in what was otherwise an excellent quarter where Google was able to handsomely confound its critics.
• Q2 2025 revenues-ex TAC / EPS were $81.7bn / $2.31 ahead of estimates of $79.6bn / $2.18 with Search, YouTube and Google Cloud all beating expectations.
• This is an important set of results as there are constant fears that generative AI is better at search than the usual Google Search and that Google’s revenues will start to suffer as a result.
• For the moment, there are no signs of this as Google Search is both better and much cheaper for the usual types of searches that the vast majority of users carry out on a daily basis.
• This means that if OpenAI was to steal all of Google’s traffic, it would probably bankrupt the company, implying that for now, Google is pretty safe.
• Google is also not sitting still, and its AI Overviews and AI Mode are gaining traction and are helping to stop traffic leaving Google and going to the generative AI start-ups.
• Gemini is certainly good enough to prevent its users from going to other providers for more complicated requests, and with its installed base, it has the ability to push these services directly to them and set them as default.
• This will require good execution on Google’s part, and despite a slow start, execution and storytelling have both significantly improved.
• However, this requires investment and demand from customers has forced Google to increase capex for 2025 by $10bn to $85bn.
• With each quarter that goes by, Google is demonstrating that it can keep the generative AI upstarts at bay, making its discount to the peer group look more like an opportunity as opposed to a threat.

Tesla – Just another car company?

• Tesla reported reasonable results, but the ending of EV subsidies is going to cause some pain which, combined with overoptimistic expectations for both robotaxis and humanoid robots leads me to think that the valuation of the shares remains outlandish.
• Q2 25 revenues / EPS were $22.5bn / $0.40 in line with expectations of $22.3bn / $0.40 but the bad news was in the commentary.
• Here, Mr Musk said that the ending of EV subsidies was likely to cause “a rough few quarters”, meaning that sales and profits could easily decline.
• This is not unexpected, and given that the shares only fell by 4% in after-hours trading, this had already largely been priced in.
• However, the hopes of the bulls and the valuation of the shares are supported by the notion that Tesla has an edge in AI that will allow it to come to market with robotaxis and humanoid robots long before anyone else.
• This is where the market and I diverge, as all of the available evidence suggests that Tesla’s AI is no better than anyone else’s and the edge that the market is paying for simply does not exist.
• Hence, I continue to think that Tesla will be an also-ran in both robotaxis and robots, the latter of which is likely to take many years more than Mr Musk says to hit a million units per year in sales.
• This is why I think that the valuations being ascribed to these businesses remain way too high, meaning that even here, the shares are overpriced.
• I continue to avoid Tesla.