In this FSD episode, titled The We're So Back Moment, Ram opens on the drive home with a story about running team performance reviews with Claude Code, giving it his Slack channels, his DMs, and his leadership framework so it could map each person against buckets like vision, standards, and pace. Then it's straight into the tape. Markets are overbought and staying overbought, Microsoft just landed on Goldman's conviction buy list, and Goldman, Morgan Stanley, and Bank of America all came out the same day with the same message: cloud is back and the return on AI spend is real. Software, semis, and cloud rallying together is a melt-up condition, and with earnings growth this strong and valuations still reasonable, Ram thinks S&P 8,000 is a breeze.
He rewinds to what the smart money was saying a year ago. Howard Marks published Bubble Watch, Jeremy Grantham was doing what he does best as the most articulate bear in the world, GMO turned cautious, and Warren Buffett was raising cash and already out of Apple. The gray-haired money moment came and went, and now everyone who stepped aside has to get back in. A study of this exact setup found one historical match, March 2007, a correction followed by a stair-step rally. There is still value out there: he bought The Trade Desk back at 9.7 times forward earnings ahead of a political ad cycle he expects to reach 20 billion dollars, and Nvidia and TSM are still growing into reasonable multiples. The caveats are junk taking off, which usually precedes a cooling, and the September fade in midterm years.
The back half is the AI thesis and everything around it. Sentiment turned on a dime once Microsoft's CFO noted growth is coming from beyond the frontier labs, and Ram reframes the worry about OpenAI and Anthropic: they are not counterparty risk for the clouds, they are aggregators of end demand that Microsoft and Google would serve directly anyway. Financial services leads AI adoption yet the big banks are still in POC mode, which is why he sees an AI front-to-back bank as a bigger opportunity than Erebor's five billion dollar crypto bank raise, and why Lumida is building on the Lumida Invest app. Along the way: a peptides update (BPC-157 energy, DSIP for sleep), Trump's Jedi mind trick on Iran, Citadel's couple billion dollar week as a preview of return-on-time businesses, and the closing challenge for founders to move from AI productivity world to AI delegation world.
[00:00] Performance reviews with Claude Code
[00:51] Overbought and staying overbought
[01:21] Cloud is back: the melt-up
[02:20] AMD, Apple, and stocks behaving as they should
[03:13] The S&P at 8,000
[03:28] Last year's bears: Marks, Grantham, Buffett
[05:19] The March 2007 study
[07:18] The Trade Desk and political ad spend
[08:53] Nvidia, TSM, and why 8,000 is a breeze
[10:33] Bikes, trails, and a peptides update
[12:10] South Korea's mini 1987 reset
[12:44] Iran, Trump, and the Jedi mind trick
[14:06] Sentiment on a dime: the aggregator thesis
[15:30] Banks are still in POC mode
[16:45] Erebor and the AI bank opportunity
[18:06] Cash sweeps and the Lumida Invest app
[19:57] Citadel's return on time
[21:40] Productivity world vs AI delegation world
About the show: Non-Consensus Investing is Ram Ahluwalia's running commentary on markets, where he shares how he's actually positioning capital and talks through the ideas most investors are missing. Real-time analysis, specific names, and a bias toward what's overlooked rather than what's crowded.