i had the following read to send last week, but i never got around to it. bummer! i certainly couldn’t have predicted what transpired this week, but i think i nailed the “waiting for a sudden correction” and “buckle up for the comings weeks” part:

markets need to be told what to do, and right now they are sitting in the frog-boiling-pot treading water. they do not know how much hotter the kettle will get, or whether they are already being cooked. or maybe they’re not being cooked at all and are in fact enjoying a nice spa, funded by record earnings??

investors are still buying, passively and/or strategically, but they’re (apparently) frowning while doing so. buying is happening, but investors are not at all euphoric.

markets are currently mad that they have to pay for years-in-advance payoffs while visibility into the future currently maxes out at about a month (or less), and the companies taking part in the AI buildout have plans that need to near-perfectly execute well into the 2030s.

which means earnings beats get sold due to uncertainty about future uncertainty, among other things, rather than celebrated at face value. and forward multiples have compressed over time via earnings growth even as indices make highs. it’s a quiet, silent selloff, of sorts…

credit is also forcing a repricing of equities, with private marks moving first: OpenAI's latest raise cleared only after the target was cut to $1.25 trillion, and Anthropic's valuation talk has compressed from $2 trillion toward $1.5 trillion, with other issuers holding their own pricing until Anthropic prints.

nearly $500 billion of AI debt has been issued this year alone according to Goldman, and hyperscaler credit protection keeps printing records. the capex story transmogrified over time from a positive funding story to a negative credit story.

horizon compression hits multiples until something gives, & the candidates to give it are plentiful: the Anthropic IPO, the Fed, Iran, Ukraine, oil & diesel, the midterms, capex & the debt funding it, France, {insert one of many AI risks here}. one or more catalysts will have to send a clear signal to the market which direction it should head next.

in our previous W39 issue (September 28), we laid out our base case as up then down, or down then up. in other words: not flat:

"we'll keep [cash] on the sidelines until we get either a sudden correction, or an unwarranted rally and post-rally correction. in other words, flatlining from now until midterms/EOY is not our base case"

since then, the tape ran to highs into mid-October and this week delivered a solid drop. the psych/uncertainty/reevaluation dip is real and happening (in our opinion).

buckle up for the next ~2-3 weeks!

now, here’s what we have to say for this most recent week. and hopefully we’ll have a regular cadence moving forward.

a friend recently asked for our thoughts on markets. we said this:

❝

good earnings are getting sold (they’re priced for extreme expectations & long future but nobody has visibility into next month, let alone 2030+), capex is relatively high + increasingly debt-funded + circular (eg. $ORCL ( ▲ 5.85% ) / $CRWV ( ▲ 7.16% ) borrowing to build for their own investors — their credit yield is blowing up bc confidence is shaken), efficiency gains hurt some parts of tech stack while helping others, inflation persistent so hikes coming but who knows when (always next meeting, never this one), Warsh watching capex's second derivative while giving no forward guidance, AI not paying off (enough?), AI paying off too much (hackers/etc), software showing AI selloff was stupid (rekt from fear then un-rekt on growth and outperforming again recently), bonds up to record levels => stocks vs a 5%-paying safe asset, gold down in a war year, "horizon compression" (valuation multiples compression bc less than usual visibility when capex/AI wants visibility into next decade), everyone long so no cash + no washout, 10 stocks = the entire index so money/ants just moving within index from one stock/sugar cube to next, no analog => price discovery is vibes, AI sycophancy + outsourced analysis certainly factoring into things but idk how, ...

basically, nobody knows what to think because plenty of data exists that supports multiple conflicting viewpoints, visibility is cloudy, & there's an unprecedented tech revolution running through everything (while the talking heads on both sides of the pro-/anti-AI argument spew the dumbest nonsense, leaving the sensible middle-ground neglected.)

a day or so later, we read the following about worry, which carries a similar message but is communicated differently/better…

QUICK THOUGHTS

OUR WALL OF WORRY | RUSSYARROW

you should read the above-hyperlinked Our Wall of Worry for yourself. it’s good.

but basically, in our own words:

imagine everyone spent the entire railroad & dotcom booms braced for the bubble.

that’s sort of what’s happening now. we’re living during a post-Internet, peak mental health crisis era, where every marginal bad thing in any corner of the globe is available (and all but forced on people via their addicting phones) 24/7.

so, we’re wondering, that perhaps this time around the widespread ambient anxiety keeps a lid on the euphoria and prevents an outright pop?

though, of course, that doesn't rule out long, orderly drawdowns.

midterms may help mitigate some anxiety. but it also could just take until the 2028 presidential election until we get a societal dose of anxiety medication. or at some point prices could bottom enough that they're seen as a deal, even in an uncertainty-riddled market.

and we're not saying the next president should be from one party or the other, just that whoever it will be will undoubtedly be less of a loose cannon with no predictable trajectory.

Sherwood ( $HOOD ( ▲ 2.2% )’s media arm) said there was the biggest retail dip buying in GOOGL on Friday, by a lot. interesting.

as oil blows away tech on YTD/1-year timeframe in year of our lord 2026, the year of anti-technology & pro-fossil fuels.

someone shared a chart showing just how much of the future can be baked into prices. we don't think this is necessarily true of AI, but just something to keep in mind.

stock valuations are a chaotic, inconsistently-defined mixture of past, present, future, and the future part is getting heavily discounted as earnings fail to exceed extreme expectations + visibility into the future sucks.

tech swapped places with oil, slogging along with some sort of albatross burden around its neck the industry feels it doesn't deserve.

as long as tech continues to spend lots of its undeserved cash on public-benefiting infrastructure (whether at the expense of share prices or not), and as long as AI continues to play close to 0% role in creating some of the best art/music/games we've seen in years (despite claiming it "solved" art multiple times already), the future of the humans not holed up in a death cult in SF will be better off in a few years. (🤞🤞)

PORTFOLIO NEWS

  • 10/15: ASP Isotopes' Quantum Leap Energy signed a binding heads of agreement with Necsa to form a joint venture for HALEU production at Pelindaba. QLE holds about 80% & Necsa about 20%, with TerraPower offtake agreements supporting the venture.

  • 10/15: Redwire won about $13.8M in additional purchase orders to deliver Stalker UAS to the U.S. National Guard, where the aircraft will train soldiers for the Army's new 15X Tactical UAS Specialist specialty.

  • 10/21: Redwire signed an MOU with EnduroSat to combine EnduroSat's high-volume spacecraft manufacturing with Redwire's power systems & prime-contractor experience for European government, defence & commercial missions. the named targets are VLEO defence operations, EU-based orbital data centers & next-generation PNT constellations. no value disclosed.

  • 10/14: Rocket Lab moved Neutron's first launch into 2027 after inspections found voids in the carbon-composite forward propellant dome on the first stage. a replacement dome is built, & full first-stage assembly & testing in Q1 2027 gates a new launch date.

THE FAYE WEBSITE

we’re working on a homepage with some stuff that we find useful, and perhaps others may find it useful as well.

it currently has three pages (four if you count this newsletter), all works in progress.

  1. the dashboard tracks the portfolio, watchlist, indices, & some relevant sector ETFs. it also includes the portfolio & watchlist news (updated daily)

  2. the daily news scan is categorized news, generated daily from a hand-picked set of feeds

  3. consensus is an experiment that we may or may not continue to explore. it aims to track one thesis against a changing landscape of data, tracking a “narrative” as it ebbs, flows, or changes course. we’re testing this out with the Strait of Hormuz closure. for the Hormuz case, for example, it aggregates sources and ship data and attempts to retroactively reconstruct the implicit (deals, routes, etc) then explicit (public statements) shift from “we’ll wait for Hormuz to open again soon” to “we are actively re-routing as best as we can around Hormuz.” in the image below, we track shipping company $FRO ( ▲ 0.71% ) against that “consensus.”

CLOSING

for this issue’s song, a new single from Cloud Nothings’ forthcoming album:

see you next time.

AURUM NOSTRUM NON EST AURUM VULGI 🪱

THE REST OF THE NEWS

🌐 software

🤖 hardware

🧬 biotech

🏗️ infrastructure

💰 finance

🏛️ society

⏳ perspectives

🎨 art

Reply

Avatar

or to participate