New Ask Claude or ChatGPT about any stock, using Bullpen. Bullpen in Claude or ChatGPT. 7 days free
Bullpen.News From Edwin Dorsey
Get Daily Recap Feedback

Saturday, November 14, 2026

Back to the front page

From X

@calvinblissett

Calvin Blissett on X

I listened to a Bill Gurley @bgurley interview on @TKPPodcast, great interview by the way Bill. He compared the US AI labs to farmers who go to the market and just trade goods, whereas in China the farmers go to the market, trade goods, and are forced to share best practices and help each other implement them. In his words: “The competitive dynamic in China is more intense. Because it is more intense, everyone has chosen to go open source. That creates a system that is capable of innovating far faster than the competitive system we have here.” China’s way results in exponential learning and growth, which is pretty clear. Charlie Munger was a notorious China bull late in life: “The reason that I invested in China is I get so much better companies at so much lower prices.” Sentiment has also got to account for something. China’s CSI 300 just confirmed a bear market, down 20.2% from its June peak, and 19 years later it has still never closed above its 2007 high. On October 23 the Hang Seng Tech index fell 5% to a two-year low and the CSI AI index fell 3.3% in a single day. I am expressing this macro view through a bottoms-up investment I found: a shareholder-friendly, founder-led company paying an 11%+ dividend and authorized to buy back about 3% of its market cap. Revenue grew 33% in the first half of 2026, net income grew 34%, and paying users went from 3.8 million in 2021 to 18.2 million in 2025, a 48% CAGR. It has been preparing for AI since the beginning. The founder: “From the very first day of the company’s founding, we had our own AI algorithm engineers and big data team… we began applying these cutting-edge technologies as early as 2019.” Today about 70% of staff are R&D, AI-generated code is approaching half of all R&D code, and net margins run near 30%. All for the grand price of about 2.5x earnings, with a ~$494M market cap sitting below roughly $630M of unencumbered cash. Did I mention there are forced sellers selling for completely non-economic reasons because they are VC investors? VCs including Hike, Northern Light, Qiming, SIG and Sequoia China owned about 55% of the company after the IPO. Qiming alone has sold 43% of its stake in nine months, from 19.4M shares to 11.1M, per its 13G filings. Here is my writeup on Yuanbao $YB:
@valuedontlie

Colin King on X

Ugly chart at $APTV Maybe one of my favorite setups right now despite "tough industry" dynamics... Lapping one-time VGNT spin costs so FCF should jump to $1B+ next year (on $8B market cap = 12.5% FCF yield) and committed to buyback stock with 100% of FCF through 2027 Assuming no operational/industry improvement, it's not hard to get $7-8 EPS by 2028 (vs. $5.70 in '26) As a bonus, this management team has done 2 spins in 10 years and wouldn't be surprised if they spin the (better) engineered components business
@convequity

Convequity on X

I think the Oracle credit story could look very different in 2028 and 2029 from what the market is pricing today. $ORCL has priced $10 billion of stock and $10 billion of bonds to fund its AI buildout. Management expects free cash flow to turn positive in the second half of fiscal 2028. The interesting part is how Oracle gets there. AI customers need the capacity, and those customers also have relationships with the suppliers building the infrastructure. Customer prepayments and supplier financing can move some of the funding further upstream rather than leaving Oracle to finance everything itself. I'd also watch whether support can extend further up the supply chain. The final cost of a data center is not the same as the underlying cost of the components, so there may be room for suppliers or their suppliers to provide credit backing with less exposure than the headline project value. The actual contracts and who bears losses matter here. Oracle's five-year CDS has been trading close to 300 basis points. I'm curious about the other side of that trade: who is underwriting the protection, and what happens if Oracle reaches positive free cash flow and the fear around AI infrastructure financing eases? Selling protection is still a risky bet. A premium can look attractive right up until the underlying credit deteriorates, and spreads can widen before the trade works. But if Oracle delivers the capacity and its customers pay for it, I think the picture in 2028–29 could be very different from today's gloomy consensus.
@InvestSpecial

Dalius - Special Sits on X

While everything is down, $RSVR keeps delivering. Another solid Q2, another guidance raise. The review of the founder-led bid keeps dragging on. Given the steady improvements in the business, I wouldn't be surprised if the special committee is holding out for a higher price. A bump certainly looks possible. But with the spread down to just 1.5%, the risk/reward isn't nearly as compelling anymore. Up 17% since SSI write-up. Trimming heavily 👇

Investor Letters

All letters

Is Nebius Fucked?

Nebius’ Q3 beat relied on $325 million of one-off partnership fees as it raised 2027 CapEx to $40–45 billion and may need $6–11 billion of…

NBIS $32B
Daniel Romero 6 clicks

Qualcomm Custom ASICs?

Qualcomm’s Alphawave SerDes, CPU cores and HBC could support custom AI ASICs as hyperscalers build heterogeneous fleets beyond GPUs.

QCOM $185B
Chipstrat 4 clicks

Publications we follow

Every publication on this site, hand-picked.

130 publications

Newsletters 110

YouTube 4

Blogs 15

Seeking Alpha 88 of 293

Nominate a publication

X accounts we follow

Every account on this site, hand-picked.

91 accounts
Nominate an account

Investor letters we follow

376 firms

Sectors

Ordered by how often our editor reaches for them.