REAL-TIME GLOBAL RESEARCH
JPM TECH SKETCH: META AI Bull Case Much Clearer, NFLX Engagement Problems, DIS+ Free Tier?, SPOT AI Data Labeling
Research evidence excerpt
JPM TECH SKETCH: META AI Bull Case Much Clearer, NFLX Engagement Problems, DIS+ Free Tier?, SPOT AI Data Labeling
Specialist Sales J P M O R G A N
10 July 2026
Media Q226 Earnings Preview Webinar (DIS, LYV, TKO, FOXA, etc.) Tuesday July 28th at 10AM EST - with David
Karnovsky & Mark Schilsky REGISTER HERE
NEWS – DESK COLOR – JPM TECH RESEARCH
META is Building a Hyperscaler (re: AMZN, GOOGL, SPCX,
OpenAI, Anthropic, etc.):
Technically, Zuckerberg laid out plans to build a neocloud and a leading edge LLM lab, but ‘hyperscaler’ is just so much easier to
use in a title.
Regardless, META rallied +5% yesterday because Zuckerberg clearly laid out a credible plan to monetize his massive
capex investment. This is nearly a 180-degree turn from where we were two weeks ago when investors were still scrambling to
understand how META was going to monetize the much-rumored consumer and business AI agents. Now, META has laid out a
clear plan to monetize its capex investment via two proven routes: Neocloud & LLM API. Investors know these business models
can generate significant revenues at decent margins. Just take a look at SPCX, CRWV, NBIS, OpenAI & Anthropic. These are
proven business models that META shareholders can underwrite with a reasonable degree of certainty. This is a far cry
from the uncertain future of consumer and business AI agents that we have been actively discussing for the past few months. The
debate on META has shifted considerably over the past week. Pay close attention – META’s Q226 earnings call is going to be a
banger.
Investor Feedback: The debate on META is rapidly shifting from ‘How big will the 2027 Capex guide be?’ to ‘How will
META best monetize that huge Capex build?’ That’s a MUCH healthier debate for the shareholder base to be having, IMO.
The English excerpt is extracted automatically from the cited source page and may contain layout or recognition errors. It is never batch translated.
Open report viewer