GTC Berlin & Meta – All About Confidence

GTC Berlin Day 2 – A Very Civilised Affair

  • GTC in Berlin is much larger than it was in Paris, but the civilised nature of the conference has continued with free water, coffee, lunch and even pretzels being made available for all delegates.
  • At the same time, it is also well attended, as all the hyperscalers, neo clouds, infrastructure companies and a large number of start-ups are present, and there is an optimistic tone even if more than 70% of the infrastructure is being built in the USA.
  • GTC Paris in 2025 felt very much like a toe in the water, but 2026 has the same vibe as the main show (San Jose) but at a somewhat smaller scale.
  • There were also plenty of robots on display, but it is clear from the demos and the stands that progress towards a commercially viable humanoid robotic market remains pretty slow.
  • I continue to think that this represents a great long-term opportunity for the Western car OEMs as there is plenty of overlap between making cars and making robots, and these companies badly need
  • This confirms that there is no shortage of interest in rolling out AI in Europe, just as long as the regulators can get out of the way.
  • Here, the news is not good, as when anyone who deals with these sorts of things is pushed for an update, very little is forthcoming.
  • Consequently, the AI Act is still on track to start enforcement in 2027, and the new CADA Act threatens to make the European AI sector so expensive that it is unable to compete with cheaper imports.
  • While this will cause consternation to anyone who wants to develop an AI business in Europe, it is music to the ears of the Middle East, who will be readying their tents for the European trade shows in 2027 with the promise of low regulation, lots of electricity, no tax and plenty of sunshine.

Meta Platforms Q3 26 – Not what was hoped

  • Meta Platforms reported good results, but its plans to raise vast amounts of money to invest in AI over the coming years undid all of the good work Jensen did in Berlin when it came to reassuring the markets that there is good money to be made in AI.
  • Q3 26 revenues / Adj EPS were $64.6bn (up 26% YoY / $7.40 nicely ahead of expectations of $63.3bn / $6.54.
  • This was largely due to the cost reductions as Meta has finally been able to get its AI to work properly and now it no longer needs a lot of the humans that it had hired over the last 7 years when its AI was languishing.
  • This is the main reason why headcount was down 15% YoY.
  • Revenue guidance was slightly ahead of expectations, and so I think that the $8.37 consensus estimate for Q4 26 can be beaten, but this is not where the hand-wringing was to be found.
  • The problems all centre around investments, where 2026 capex will come in at the top end of the $130bn – $145bn with more to come in 2027 and beyond.
  • This is where the share split comes in as only one of the two new shares will carry a vote (C class).
  • This enables Mr Zuckerberg to stay in California should the wealth tax pass the popular vote, as well as give the company a currency to raise money without Mr Zuckerberg losing control of the company.
  • This is what triggered the sell-off, and the market is now concerned that the consequence-free currency will trigger more profligate spending and poor returns.
  • However, the reason for the excellent returns that the company has reported in Q3 26 is due to these sorts of investments being made, which, combined with the superb returns that are currently on offer in the data centre business, lead me to give the company the benefit of the doubt.
  • I am hopeful that Meta will soon announce deals similar to those announced by xAI which is what allowed SpaceX to handily beat expectations when it reported earlier this week.
  • Muse continues to top the download charts, implying that it is more than just a flash in the pan, increasing my confidence that Mr Zuckerberg has recovered from his Llama 4 implosion.
  • Hence, I am inclined to raise my medium-term expectations for Meta rather than cut them, especially as the chance of the company getting into real financial distress remains tiny.
  • I have a position in Meta, and if I did not already have a full position, I would be inclined to add some more here.

RICHARD WINDSOR

Richard is founder, owner of research company, Radio Free Mobile. He has 16 years of experience working in sell side equity research. During his 11 year tenure at Nomura Securities, he focused on the equity coverage of the Global Technology sector.

Leave a Comment