Since 2011, Verizon has a regulator-mandated MVNO with Charter and Comcast that lasts into perpetuity Moffett estimates Charter and Comcast are using this to resell Verizon service at a 70% gross margin… aka higher than the margins of Verizon’s own wireless business Crazy

For all the complaining that people do about Ticketmaster’s usurious fees, why does their parent company $LYV always look like such a crappy business? 5% EBIT margins for a monopoly, really?

For those bidding up $AMD in the aftermath of $NVDA last night, I present to you: superhacker George Hotz’s review of AMD’s current ML software stack

Apparently the best leading indicator of Chinese stock market returns is… how much dumb foreign money is fleeing the market Forward 3 months returns have spiked nearly every time northbound flows (ie foreign inflows) went negative in the last 5 years

Just another reminder that the auto industry sucks - in aggregate, European auto OEMs burned about $15B more cash than their businesses actually generated from 2010 to 2019

China has historically accounted for 50-60% of German OEM profits, far higher than reported numbers due to JV and royalty accounting 😬

Seems bad - while German and Japanese OEMs have historically dominated the Chinese car market, they’re basically absent from the EV leaderboards, where they’ve been replaced by BYD and local startups. EVs are now 25% of the Chinese market

I feel like it’s under discussed how quickly Netflix’s business became… boring? I remember bears and bulls knifing each other online over $NFLX content amortization schedules in 2017, but today even bears like MoffettNathanson are projecting 20% FCF margins in 3 years

This has always felt like the left tail bear case for NVIDIA - CUDA doesn’t actually touch ML customers, it’s always intermediated by PyTorch. If Facebook wanted to build a CUDA alternative that supports AMD or their own ASICs, this would be their avenue to disintermediate NVIDIA

I’ve found it! The one place left in global equity markets that’s still using “Price to 2030 earnings discounted to present”: Chinese enterprise software stocks

Ideal use case for Google Lens

Oh wow what a total surprise

More on this from Bernstein - high demand naturally creates new vacation rental supply, as opposed to supply constraining demand

(Whether you view this chart as bearish for India or as, I guess, bullish for Mexico and the Philippines probably says a lot about your outlook on life)

This is sourced from the State of Indus Valley report by @BlumeVentures, who are probably the most sober minded India VCs willing to speak in public

This is what always bugs me about India tech. Only 220M people in India currently transact online. Their wealth roughly adds up to the GDP of Mexico + the Philippines. Yet this group is propping up startup equity equal to 10x (20x?) their Mexican and Filipino counterparts 🤔

Craig Moffett interviews Verizon’s head of FWA - isn’t this basically the most bearish interview possible for cable companies? Verizon claims FWA has “almost unlimited capacity” even in big cities. 50-60% of net adds come from cable. Fuckkkk $CHTR $CMCSA

IMO online marketplaces come in a 2x2 matrix: - It’s harder to price services than goods (because services need a different price for every hour) - It’s harder to price low volume purchases than high volume ones $ANGI chose the hardest possible intersection

I think about this random 2012 comment on @John_Hempton’s blog a lot: a 8 paragraph argument on how the levered discount grocer SuperValu is a better investment than Google at 16x earnings Six months later, SuperValu dropped -70% You don’t get points for complexity in this game

I can’t believe I never thought to look at this - the BEA’s PCE dataset tracks monthly US spending on taxis and ridesharing (from MN’s Uber initiation). The story this chart tells is pretty extraordinary. Uber didn’t just cannibalize taxis, it expanded the market by 3x

US productivity growth if we all stopped watching TV

Constantly in shock every time I relearn that the average US household paying for cable still spends 250 hours per month watching TV That’s like… 8 hours a day. And that’s down 20% from peak historical levels. Literally how??

Constantly still in shock every time I relearn that the average US household paying for cable spends 250 hours per month watching TV? That’s like… 8 hours a day. Literally how??

Michael Nathanson absolutely roasting $SNAP here “Next week, we will get 4th quarter updates from the world’s leading digital advertising companies… and SNAP”

Never forget your base rates - 50% of growth stocks stop being growth stocks after just 4 years (from Bernstein’s recent growth purgatory note)