Alphabet Q2 24 – The disconnect

Big disconnect between words, numbers, investments and revenues.

• Alphabet reported solid Q2 2024 results, but anyone would think from the commentary that all of the revenues were coming from its new AI products rather than advertising (75%), cloud (12%) and subscriptions (11%).
• I suspect the generative AI services make up a tiny percentage of the revenue despite massive investment and even Mr Zuckerberg over at Meta has voiced the possibility that everyone is currently overinvesting.
• This is a much more serious problem than when everyone overinvested in internet capacity in 1999 and 2000 because the dark fibre simply sat there until it was needed.
• The problem with overbuilding AI capacity is that the technology is rapidly evolving and so when it comes around to finally needing the overbuilt capacity, it could already be obsolete.
• The downside to not getting on board with the mad rush now is to end up being behind when the technology finally comes into its own, but I suspect that the correction will give the laggards ample time to catch up.
• This is especially the case as there is mounting evidence that progress is already slowing as the differences between the different foundation models are getting smaller implying the law of diminishing returns is starting to have an effect.
• Against this backdrop, Alphabet is no longer a clear leader in AI which is why it is using the Android ecosystem and its dominance in internet search to distribute its AI offerings to users.
• This is evident in its results where Q2 24 revenues ex-TAC / EPS were $71.4bn / $1.89 just ahead of estimates at $70.7bn / $1.84.
• The real bright spot of these numbers was Google Cloud which grew revenues by 27% YoY and profits surged by 197% YoY to $1.2bn.
• This implies that Google Cloud has finally hit critical mass and in this game of scale, this puts it in a good position to finally compete effectively with Microsoft and Amazon.
• This result also sends a positive signal for both Microsoft and Amazon where I expect that revenue growth will remain strong with robust profitability.
• These are the 2^nd order beneficiaries of the AI boom while the 1^st order and main beneficiary remains, without question, Nvidia.
• However, there was less good news when it came to investments as capex was even higher than expected at $13.2bn in Q2 2024 bringing the annual run rate of capex to over $50bn.
• 2024 is supposed to be the year when we begin to see revenues from generative AI and here we are halfway through and there is precious little sign of any returns starting to materialise.
• Instead, investments are increasing beyond what was expected, and valuations continue to climb.
• RFM research concludes that the going rate for generative AI is now somewhere around 142x revenue to sales which is absurd by any reasonable measure.
• Fortunately for Alphabet, its core advertising business remains strong giving it the ability to ensure that it is Google AI services that users turn to first although there is strong competition for that position.
• The other good news is that Alphabet is still not crazily valued at 24.2x 2024 PER meaning that if the AI bubble pops it will have a lot of downside protection.
• However, there is very little to excite me or the market in these results and so I remain fairly indifferent to Alphabet as an AI investment.
• I continue to remain happy with my AI adjacencies of inference at the edge where I own Qualcomm and nuclear power where I have a basket of physical Uranium funds, ETFs and Uranium miners.