Google: The BIG Winner of AI
Summary
Tom presents an extremely bullish thesis for Google, shifting from a long-term skeptic to a 'hyper bull.' He identifies the primary bottleneck of the AI revolution as power consumption and infrastructure costs. Tom explains that while the industry relies on Nvidia's versatile but expensive GPUs (which he calls 'supercars'), Google has developed TPUs ('freight trains') that are specifically designed for AI training and inference. This vertical integration allows Google to operate at a fraction of the cost of competitors like Microsoft and Amazon, who are stuck in what Tom calls a 'golden prison' of dependency on Nvidia.
Tom highlights that Google Cloud is now the fastest-growing cloud provider, expanding at nearly 50% annually. He points to Google's massive data advantages and distribution channels like YouTube and Google Search as key drivers for Gemini and other AI tools. From a financial perspective, he notes that Google is fundamentally strong with $400 billion in revenue, $74 billion in free cash flow, and a massive cash reserve, all while trading at a relatively low forward P/E of 23. He suggests a dollar-cost averaging (DCA) strategy to build a position during market dips.
Mentioned Stocks
Reasoning: While Tom recognizes Nvidia as the engine of the AI revolution, he describes their GPUs as a 'golden prison' for cloud providers due to high costs and energy consumption. He highlights that Google is successfully bypassing Nvidia's supply chain, which could pressure other hyperscalers who remain dependent on Nvidia's pricing.
Reasoning: Tom believes Google is 'hyper bull' because its TPUs provide a 10x cheaper and more energy-efficient alternative to Nvidia GPUs, creating a vertically integrated powerhouse. He notes the stock is mispriced at a 23 forward PE despite 50% cloud growth. He sets a 5-year bull case price target of $1,400 and a bear case of $420, suggesting that any entry point below previous peaks is attractive using a DCA approach.