Bullpen.News From Edwin Dorsey
Get Daily Recap Feedback

Sunday, November 8, 2026

Back to the front page

Important Portfolio Update

China’s housing downturn has cut some flat values 40% since 2021, curbing discretionary spending and sharpening a portfolio focus on essential purchases.

BRK.A $1.2TAAPL $5T
Cannibal Stocks 1 click

From X

@StableBread

Fajasy on X

Meanwhile, here's what Anthropic has signed up to spend on compute: → $GOOGL: $114.5B from April 2026 to July 2033. → $AMZN: $110.0B from May 2026 to April 2036. → $MSFT: $44.9B from November 2026 to May 2033. → Five-year AI chip lease: $125.2B of rent, up to $42B of which Anthropic can fund by selling convertible notes to $AVGO or a financial partner $AVGO designates. → $SPCX's xAI: Up to $120.5B into 2029 for $NVDA GPU compute, cancellable on 90 days' notice after early 2027. As of June 30, 2026, Anthropic had $265.6B of non-cancelable compute purchase commitments and $321.4B of signed leases that hadn't started yet, up from $54.6B and $50.2B six months earlier. Anthropic has signed ~$266B more since June 30, which by my rough math brings the total to ~$850B (the S-1's own total is still blank). Add an estimated $125-150B for TPU (Google's AI chip) compute in 2028-2031 that isn't in those figures.
@StableBread

Fajasy on X

Anthropic's preferred stock price more than quadrupled in ~9 months, from $140.9676/share in its Series F-1 round (August-September 2025) to $259.1364 in Series G-1 (January-February 2026) and $589.0095 in Series H-1 (May-June 2026). After the 10-for-1 stock split on September 17, 2026, the Series H-1 price works out to $58.90/share, which implies a ~$1T valuation for Anthropic's ~16.9B fully diluted shares as of June 30, 2026. $AMZN holds all $6.2B of Anthropic's remaining convertible notes, which convert at the IPO into 1.84B Class N (non-voting) shares at $2.0928 and $5.0478/share, subject to a 12% ownership cap that Amazon can waive. That's why the balance sheet carried the notes at $108.6B as of June 30, 2026, 17.5x their principal.
@StableBread

Fajasy on X

Non-cash fair-value charges of $68.0B more than account for Anthropic's $66.3B net loss in Q2 2026. $59.6B came from revaluing convertible notes as Anthropic's valuation rose, and another $8.4B came from revaluing Anthropic's agreements to sell more notes and stock to $GOOGL, $AMZN, and $AVGO. Excluding those charges and stock comp, Q2 adjusted net income was $3.0B. Anthropic also generated $3.9B of operating cash flow in H1 2026, vs. $1.5B used in H1 2025. So the net loss reflects Anthropic's rising valuation, not its operations.
@calvinblissett

Calvin Blissett on X

I'm sorry it's a pretty bold claim "a better $NVR" I should elaborate. Cury looks like NVR transplanted into a structurally tighter housing market, with even stronger current unit economics. On an apples-to-apples S&P Global methodology, Cury generated 84.2% ROIC in 2025 versus 47.0% for NVR, while Cury's inventory turned 3.77x versus NVR's 3.26x. (https://t.co/pMkkNhSOWD). ROIC is preferable here to ROE or EPS because NVR's enormous buybacks reduce shares and equity and can make per-share growth and ROE look better. The opportunity set is also more extreme: Brazil's official housing deficit is 5.77 million homes, equal to 7.4% of occupied households, while Freddie Mac estimates the U.S. shortage at 3.7 million units against roughly 150.8 million housing units, or ~2.5% of housing stock. FJP Cury then concentrates affordable housing in São Paulo and Rio, precisely where affordability and housing scarcity are most acute because these are the richest cities in Brazil, while using land swaps, customer presales and subsidized MCMV financing to keep remarkably little of its own capital trapped in each project. The result is basically the feature that makes NVR exceptional pushed further: Cury currently turns each real of invested capital substantially harder, while operating against a relative housing shortage several times larger than America's.
@StableBread 1 click

Fajasy on X

My bear, base, and bull cases value Boost Run’s $BRUN stock on year-end 2026 ARR (on the same basis as the multiples of CoreWeave $CRWV, Nebius $NBIS, and $IREN), which $BRUN will report with its Q4 results in March 2027. None of the three cases adds the $160.7M/year from the July order and Cohere, since both start in 2027. The peers’ contracts that start after 2026 aren’t in their year-end ARR either, so they’re priced into the multiples instead. Since the peers’ multiples leave out the debt they’ll take on to reach year-end ARR, I hold $BRUN’s net debt at today’s $244.3M, rather than adding the lease debt from its H2 2026 build. Each case’s value is year-end ARR times the multiple, minus the $244.3M of net debt, divided by the 88.8M fully diluted shares plus any new shares from a stock sale. I start with the $231.4M of ARR from contracts already running, then add (1) the June order’s $69.2M/year if it starts by December 31 and (2) however much of the $99.4M from contracts without disclosed terms is running by then. Bear case ($4.76, -46%) The bear case assumes $BRUN has no new financing to show by its Q3 report, lessors fund 70% of the H2 2026 build instead of 79.5% (below the 72.0% lessors funded in 2025), and the June order’s November 30 start slips into 2027, so 2026 GPU spending stays at the $1.0B low end of management’s range. That implies year-end ARR of $280M, 30% short of the target, a $62.2M stock sale at $8.89/share (7.0M new shares) to cover the funding gap and three months of 2027 payments, and today’s 2.5x, for a value of $4.76/share. I give the bear case about a 20% chance. Base case ($16.17, +82%) The base case assumes $BRUN lands new lease or asset-backed financing by year-end, starts the June order on November 30, and reaches its $400M target without selling stock. That implies $400M of year-end ARR at CoreWeave’s 4.2x, the median of the three peers, for a value of $16.17/share. I use today’s median, since the base case only solves $BRUN’s own financing problem. It assumes no relief on rates or borrowing costs, which would lift the peers’ multiples too. Bull case ($19.78, +122%) The bull case assumes everything in the base case, plus relief on rates and borrowing costs that lifts the peers’ multiples. That implies $400M of year-end ARR at 5.0x, still below Nebius’s 5.9x, for a value of $19.78/share. So the bull case is really a bet on the market, not on anything more from $BRUN. From $8.89/share, you’re risking a $4.13 loss (-46%) in the unlikely bear case for a $10.89 gain (+122%) in the more likely bull case. P.S. These cases use the October 16 close and predate $BRUN’s October 26 release, which says it has closed more than $600M of non-dilutive project financing in 2026.

Investor Letters

All letters

Axon 3Q26 Business Update

Axon cut full-year guidance after delayed federal Dedrone contracts slowed 3Q revenue growth to 28%, while software and services grew 43%.

AXON $31B
Speedwell Memos 3 clicks

AMD Q3 FY26, Annotated

AMD's Q3 revenue rose 45% to $13.4 billion as data center reached $8.5 billion, 64% of sales, driven by EPYC and Instinct ramps.

AMD $660B
Chipstrat 2 clicks

ARM Q2 FY3/27 Earnings Review

Arm reported Q2 revenue of $1.443 billion, up 27%, as licensing rose 50% while royalties grew 8%; Q3 guidance calls for $1.54 billion.

ARM $181B
Cestrian Capital Research

Publications we follow

Every publication on this site, hand-picked.

129 publications

Newsletters 109

YouTube 4

Blogs 15

Seeking Alpha 75 of 279

Nominate a publication

X accounts we follow

Every account on this site, hand-picked.

91 accounts
Nominate an account

Investor letters we follow

359 firms

Sectors

Ordered by how often our editor reaches for them.