Uddrag fra Bank of America, bearbejdet til dansk:
“How far do hyperscalers need to fall for market to start trading capex cuts?”
That’s the key “zeitgeist” quote from the latest Flow Show by BofA’s Michael Harnett, and it captures not only a recurring question that Hartnett has been posing for much of the past month, but also reflect the dramatic divergence between hyperscalers (check/capex payers) and chip stocks (check/capex receivers) we have been highlighting for the past month.
This is also what Goldman’s Delta One head Rich Privorotsky discussed in his latest, Friday note (available to pro subs here) where he wrote the following:
The hardware beneficiaries continued to fly but, as feared, it was the spenders…the hyperscalers and consumers of memory…that materially underperformed. Apple raised prices across parts of its MacBook lineup while Microsoft did the same for Xbox hardware (all memory related), reinforcing the idea that higher AI infrastructure costs are increasingly being passed through the value chain. Aggressive profit-seeking in memory alongside genuine HBM supply shortages has created a very acute value transfer from consumers to producers.
This is poised to become one of the defining debates of the next few months. Markets have rarely rewarded companies allocating exceptionally large proportions of free cash flow toward capex during the build phase. This does not necessarily mean management is making poor long-term decisions; it simply means equity investors generally prefer immediate, tangible returns to distant, uncertain cash flows. Multiple expansion typically occurs during the harvesting phase, not the construction phase. The theory of reflexivity remains highly relevant in this environment. If hyperscalers continue to underperform while suppliers rally, boardrooms may increasingly question whether incremental AI investment is maximizing shareholder value. At some point, capital expenditure may slow and if one peer blinks, the market will immediately question whether others should follow.
Indeed, this is precisely the tension point we touched on last week when we rhetorically asked hyperscaler shareholders whether it was “time to have a conversation with your CEO about the 435% increase in memory prices in the Vera Rubin rack which is crushing your returns.”
Dear Hyperscaler shareholders: it’s time to have a conversation with your CEO about the 435% increase in memory prices in the Vera Rubin rack which is crushing your returns pic.twitter.com/ezGqRFyiFb
— zerohedge (@zerohedge) June 23, 2026
While this was tongue-in-cheek, one can certainly make the case that while Apple’s attempt to procure Chinese memory chips – as we reported earlier – may or may not be successful, in the end it will be up to shareholders to put a halt to the unprecedented amounts of capex spending (on memory and other AI components) if they view that the stock is not responding in a favorable fashion to the neverending spending.
The question, of course, is what is the breaking point beyond which hyperscalers will be forced to limit capex spending (which until this point was running at a berserk rate that according to Goldman could go as high as $1.4 trillion in 2027)?
Here, Hartnett offers an answer.
The BofA strategist notes that there are three key catalysts for a “proper risk-off summer”, namely the MAGS (Mag7 ETF) trading <$60…
… UDJPY trading <110…
… and the yield curve inverting.
In the meantime sticky 16% margins…
… reaffirms the market’s love for stocks, and the liquidity pulled from mega-cap AI arms racers simply racing into semis & illiquid cyclicals (small/mid-cap, housing, REITs) to front-run Trump pivot to affordability (which may or may not come at this rate).
AI bubble aside, Hartnett looks at some of the other “new normal ” indicators, and notes that gold <$4k, silver <$60, XBT <$60k as Iran and era of sanctions end and dollar pop. But despite some of the crazy moves in the market so far this year, the strategist believes that the 2020s to remain era of fractured geopolitics and populist politics prioritizing booms over inflation; as such Hartnett says US dollar is a “rent” not an “own”, and any price <$4k is a very good entry point for gold…
… while secular trade remains long EM.
Taking another step back, this time to the biggest picture and one including the bond market, Hartnett writes that since Warsh’s term started on May 22nd, US Treasuries up 3.2%, stocks -1.6%;
The BofA strategist concludes that while it is still early days and the nouveau-hawkish Fed has yet to convince any investor to abandon core “Anything But Bonds” allocation, thus far Warsh is actually mirroring “lower yield” Fed chairs such as Eccles, Volcker, Greenspan, and Bernanke…
… and to Hartnett, long the long-end remains most contrarian secular trade in markets.
* * *
While we would normally cut it off there as the remainder of the note tends to be somewhat boring with little notable signal, in this week’s Flow Show, Hartnett made another observation that appears important from a flow perspective: namely that stocks may have finally hit a local adverse inflection point with US stocks experiencing an $8.5 billion outflow, the first since March 2026 and follows a record $119.2 billion inflow
2 md. adgang for
2 x 49 kr.
Få straks adgang til denne artikel og derefter 2 måneder til alle artikler på ugebrev.dk
- Alle artikler på ugebrev.dk
- Om investering, finans, ledelse, samfundsansvar, life science og Bestyrelsesguiden.dk
- Daglige nyhedsmails med nyheder og analyser
Tilbuddet gælder til 31. juni 2026. Abonnement fortsætter til normalpris på 249 kr. efter bindingsperiode på to måneder. Opsig når du vil - til udgang af den anden måned. Tilbud gælder kun, hvis du ikke har haft abonnement på ØU udgivelser de seneste tre måneder
Allerede abonnent? Log ind her
