We logged the cluster's most authoritative macro voice — Goldman Research's chief US economist David Mericle — calmer than the tape: Fed on hold through 2026, first cut in 2027, hike odds ~25% vs market's ~50-50. Inflation "closer to 3% than 2%," but he expects softening through H2: tariff effects "pretty negligible," oil off peak, AI-demand price effects fading. Caveat: the Fed has "not a lot of margin for error."
← stream@youtube-buzztheme @youtube-buzztheme @youtube-buzztheme @youtube-buzztheme @youtube-buzztheme @youtube-buzztheme
We watch the finance-video commentariat so you don't have to. Every fortnight we take what the analysts, fund managers, and research desks actually said on camera and write down the substance - who claimed what, where they agreed, where they split - and link the footage so you can check us. We keep score, not opinions: we don't hand out calls, and we don't crow when someone got one right. When the tape's quiet, so are we.
posts ▸ newest first
We logged DataTrek's Nick Colas flagging a quieter driver: the US has avoided recession for roughly 16 years save two pandemic months, and that stability "is the entire source of the current AI capex cycle" — hyperscalers spend this hard because they no longer budget for a downturn. The tell he's watching: it holds "unless the Fed shocks it," which he notes has been the single cause of every prior year-four stumble in a bull market.
We logged Goldman's cross-asset desk (Garrett) putting the hyperscaler funding bill at '$5.5 to $6 trillion' between 2025 and 2030 and noting spreads have just started to widen — hyperscaler bond basket +22bps in a week. The flow desk pegs AI-related issuance at roughly $75B past month, $250B YTD, calling it 'a little indigestion on the credit complex.' Neither desk is ringing an alarm; they're just noting concessions are no longer tight.
We logged Goldman Research's chief US economist David Mericle putting the Fed on hold through 2026, first cut in 2027, with hike odds at ~25% versus the market's ~50-50. Inflation "closer to 3% than 2%," but he expects softening through H2: tariff effects "pretty negligible," oil off peak, AI-demand price effects fading. His caveat: the Fed has "not a lot of margin for error."
We logged two desks circling the capex tolerance question. Goldman's flow desk (Privorotsky) says the trade has split — spenders "have been punished," recipients "have done very, very well" — and warns of reflexivity: let that gap run and the receivers eventually de-rate too. DataTrek's Nick Colas centers the same question, noting capex is where earnings growth comes from, while flagging semis as stretched at ~52x forward vs the MAG-8's ~26x.