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

Tuesday, November 10, 2026

Back to the front page

From X

@StableBread 1 click

Fajasy on X

Ambarella $AMBA trades below my bear case, and below where it was before the Financial Times reported takeover talks. In my September 9 update on the edge AI chipmaker, my bear case was $72/share, 16.8% above its $61.67 close on November 2, and my base case $99, 60.5% above. Even the $57 case, where the multiple keeps compressing, is just 7.6% below. $AMBA is down 10.4% since that update, less than the 13.4% drop in the PHLX Semiconductor Index (SOX). Neither $AMBA nor NXP Semiconductors $NXPI, the Dutch auto chipmaker, has confirmed the takeover talks the Financial Times reported on July 31. But on October 30, traders citing a Betaville alert said $AMBA’s board had picked $NXPI for exclusive talks after weighing multiple offers. At $61.67, $AMBA trades below its $74.09 close before the FT report, so the stock isn’t pricing in much of a deal, and a bid would be upside none of my cases counts. The company carries no debt and has $272.3M of cash and securities as of July 31, against ~$18M/quarter of FCF burn in the first half of FY2027. That’s more than three years of runway, so it’s unlikely to need to sell stock. The dilution comes from stock comp instead, which ran at 21.9% of revenue over the last 12 months (too high for my liking) and lifted the share count 3.4% in a year. Memory shortages at its camera customers are the risk to Q4, and the Q3 report, which $AMBA hasn’t scheduled yet, will show whether customers are cutting orders.
@SouthernValue95

SouthernValue on X

Lets not resort to mud slinging when we can have a nice conversation. MSFT trades at a premium to other capex spenders partially bc it has a cleaner financial profile, can piggyback on OpenAI IP, but there are lots of other reasons too: - Perception of health of core business (which caused it to be cheaper than others 6mo ago, now w/ office accel its seen as better than GOOG maybe AMZN). These can change obv. - Predictability of incremental ROIC: MSFT is allocating incremental capex as 1) IaaS at predictable mid-teens-20s IRRs, 2) as tokens in Foundry at >30% IRRs, 3) in GTHB/Copilot inference at potentially even higher returns and with more long term franchise value, or 4) in research with longer term paybacks, but causing near-term drag on margins. Relative to META the R&D drag on EBIT is significantly lower and the other uses of compute have more predictable outcomes. When $META was primarily allocating capital to its core business, driving acceleration and extending its advantage vs other advertisers, it was rewarded w/ a premium mult. Now its multiple is depressed bc people dont know how to think about the ROIC or timeline for Muse, fresh off -$100B burned on FRL, so META has some Zuck yolo premium, whether you think that is warranted or not, it exists. AMZN capex is perhaps the most predictable but it doesnt have the really lucrative owned-token economic streams the others could have, or an AI business as moated as Muse could potentially be if META pulls it off. GOOG is crushing it in cloud, but more dependent on labs as % of incremental sales than MSFT AMZN. And its core biz is (rightly or wrongly) viewed as more susceptible to disruption at the moment. NVDA / AVGO / TSM are cheap but have various degrees of cycle and substitution risk. I'll readily acknowledge investors are short-term oriented and MSFT cleaner visible near-term profile has given it a safety premium vs other capex spenders. And there are many scenarios where those other capex spenders can achieve better IRRs from here if things go well for them. But that doesnt make MSFT overvalued. All stocks have tradeoffs. I think saying MSFT is over-earning bc they dont have to spend as much on R&D and therefore dont deserve a premium mult to other capex spenders is a bit reductive. Why reduce the world to 4 companies? Is MSFT cheap vs. 4 other companies with their own specific dynamics, idk, how about vs. the broader mkt? How about measuring its IRR through 2032 w/ EPS forecast and terminal mult and then just comparing the IRRs?
@convequity

Convequity on X

Larger AI chip packages are pushing manufacturers toward glass substrates. We regard LPKF Laser & Electronics ($LPKFF) as one of very few suppliers that can make the tiny, closely spaced holes through the glass reliably and in volume. 1. The silicon chips and the organic (resin-based) substrate beneath them expand by different amounts as the package heats and cools. A larger package makes that mismatch harder to manage, raising the risk of warping and damaged connections. Glass expands more like silicon. 2. $NVDA is requiring its Taiwanese and Korean suppliers to deliver glass substrates by 2028. 3. China's BOE is reportedly meeting with Samsung on a glass substrate partnership. 4. Producing these holes in volume is difficult. The holes, called through-glass vias, must be made without cracking the glass and then filled with copper. Our view is that LPKF's process, a laser pulse followed by a chemical etch, is built for that hole-making step. We are watching for its customers to move from development into production.
@SouthernValue95

SouthernValue on X

$Msft can leverage OAIs IP through 2032, >6yrs, so what would be the point of duplicating all of that R&D along the way, probably not as well? This is a structural efficiency advantage to $MSFT that is unique among leading AI companies. They can sell premium tokens with owner economics without the elevated training cost. That is worthy of a premium multiple. And why they are growing EBIT >20% with positive FCF and no dilution while META is growing EBIT single digits with negative FCF and significant dilution. I like $META and think it can produce a better IRR than $MSFT from here. But the risk premium is logically higher bc of a messier financial profile and the unproven size and business case for consumer AI vs MSFT in the enterprise where unit economics and demand are clearly compelling today. Research intensity will be higher for MSFT far in the future, but by then the rate of model change may decline, and it will have a larger AI business to absorb the cost.

Investor Letters

All letters

Quick Value #337 - Dana Inc (DAN)

Dana is combining with Eaton’s Mobility unit via a 1Q27 Reverse Morris Trust after selling a division, delevering and accelerating buybacks.

ETN $141B
Value Don't Lie 2 clicks

Roblox: Capture This Moment

Roblox’s Q3 FY26 DAUs fell year over year for the first time, while bookings declined 12% to $1.7 billion.

RBLX $35B
TSOH Investment Research 1 click

DoorDash and Google

DoorDash and Google.

DASH $79B
Stratechery by Ben Thompson – On the business, strategy, and impact of technology. 4 clicks

NEW POSITION: Google LEAPS

Alphabet bought long-dated GOOGL calls after a 12.6% earnings drop, betting AI monetization and distribution overcome capex pressure and…

GOOGL $3.5T
Hated Moats 1 click

Publications we follow

Every publication on this site, hand-picked.

129 publications

Newsletters 109

YouTube 4

Blogs 15

Seeking Alpha 78 of 281

Nominate a publication

X accounts we follow

Every account on this site, hand-picked.

91 accounts
Nominate an account

Investor letters we follow

360 firms

Sectors

Ordered by how often our editor reaches for them.