Kevin Paffrath

Window is pre-mega-cap earnings (Google Jul 22, Microsoft Jul 29, Meta ~Jul 29, Amazon Jul 30 as listed on tape): 'the nervousness for cash the dash for cash maybe or even the desire to take profits maybe so extreme that we could knock on the door of 675' @13:19. Hedged in real time: 'I'm not saying that's gonna be you know today who knows we could end up going green today.'

Hyperscaler capex already priced to go from ~$450B to ~$725B ('maybe that'll break a trillion next year that's already priced in we need to beat those expectations'); until Meta/Google/Microsoft guide higher, 'the market is going to be skittish because SpaceX kind of burned them' @11:23. IBM's spending prioritization and Meta's compute-sale reports read as possible early capex warnings.

Iran's 'pickaxe mountain' enrichment site (Trump's comments moved up his 2-3-year worry) drove the downgrade, yet: 'longer term though still buying stocks' @21:03 and 'I expect to buy stocks and buy the different stocks between now and next week when earnings come out' @21:09. Reminds viewers he called the SpaceX-IPO '30% rise and then a bleed' which he says played out.

On Zuckerberg's reported offer to sell excess compute: bulls read it as opportunistic monetization, bears as overbuild regret; Kevin thinks 'a mix of a little bit of both,' credits Meta's cancellation clauses on off-balance-sheet leases as 'great' for Meta but likely to 'leave a lot of bag holders' among leased-compute providers.

Continuation of his 'SpaceX suckening' thesis: 'We said that SpaceX has a risk of being a top for the market, mostly because it's a liquidity suck, right?' @2:53: a ~$75-85B raise draining market cash; notes SpaceX bonds pay ~150bp over Treasuries vs ~80bp for other hyperscalers, stock 'tanked off of its highs' with Apple retaking most-valuable-company. On memory: 'we've priced in the selling out of memory hardware for the next year or two.'

Said right after Netflix's Q2 miss (revenue ~-0.6%, EPS ~-3.8% per his figures) and engagement worries; he frames ad revenue as a new growth vertical still 'insignificant' vs total revenue, taking time to show up, with a PEG caveat if growth keeps slowing.

'advertising, frankly, recession or not, is going to be one of those sectors that just absolutely dominates with the power of artificial intelligence' @7:31: AI extracting 'any kind of penny out of the consumer that they have left.' Google also praised as 'another fantastic advertiser' now at a 1.3 PEG.

His 'Engels Pause' thesis: 1801-1841 industrial-revolution productivity accrued to companies, not workers, for 40-60 years ('it's not the workers who made money... companies extracted the profit' @12:46); 'I think it's too early to say that we're net creating jobs.' @13:40. Downstream: falling tax revenue could push bond yields higher and force social policies.

A brief aside inside his buy-the-dip mantra ('buy the dip on quality names that you are comfortable holding for the next decade'); he explicitly identifies with the rates-coming-down camp while framing the alternative for viewers building their own thesis. Note Home Depot 'battered because of high interest rates' earlier in tape.

On the day's tech-to-consumer rotation, after his Target-at-$86 call 'almost doubled': consumer names are 'really a cool medium term like hide from AI' @5:54, but 'if AI goes to crap, the consumer is going to go to random crap' @6:02; cites 13% of card balances 90-day delinquent (Q1 2026, highest in 15 years) and chargebacks +29% since 2021 as consumer stress.