Jenna Harrick @HarrickJen

Is there anyone in San Jose who’d like to have dinner with me tonight? Eating alone is a little boring…

What is the best Photonics ETF? $LITE $COHR $AXTI $AAOI $LWLG $TSEM This is another question I get alot. I'm not an ETF guy, I'm a stock picker. So I do not find any ETFs with optics exposure attractive (I don't even know if there are any) So what I am going to do on my Substack is create a model Photonics ETF With all the companies that I would put it, with weightings/percent allocation It will be a fun project and I think you would all find it useful! Will be posted here when I get to it! cruxcapitalgroup.substack.com

Gaetano @crux_capital_

5 months ago

What Photonics companies should I buy?! $LITE $COHR $AAOI $POET $LWLG I get asked this question often. I can't tell anyone what to buy. So instead, let me lay out all the companies I track. The reality is is that if it has decent optics exposure, I'm probably tracking it

24 68 480 136K 728

Palantir’s @MabreyTed says private equity companies are “culturally not built to do technology.” “To do this technology right: you have to presume that you're wrong at the outset— not presume you're right.” “That's very hard from my experience working with folks in private equity.” Via @tbpn

In 2026, when it comes to $META and its AI efforts, one of two things will happen: either $META delivers a SOTA AI model (can also be SOTA in their focus areas - consumer personal assistant) or $META quits the AI model race and partners with a model provider. The interesting thing is that either way, investors will be happy about it. If $META quits the AI model race, its CapEx needs drop substantially, FCF blows up again + they might get a very lucrative distribution deal with a model provider as EVERYONE wants to be the $META model provider with the distribution that $META has. If $META makes a SOTA model, the market starts pricing in a new AI model provider premium. Honestly hard to say what the long-term better path is at this point. On one hand, having a SOTA model can give you a long-term edge in terms of not being dependent on one provider on the other hand if AI models commoditize and the real value layer is in distribution and having the infrastructure to efficiently serve the models ( $META has that already) for $META it might be better to not spend all that cash on chasing the SOTA model, especially if their targeted use-case is the consumer segment with personal assistants (doesn't need God like AI intelligence TBH).

Every week I send a letter to tens of thousands of investors covering the most important charts and themes in markets. Join Here: bilello.blog/newsletter

Agentic Commerce Update | 1Q2026 Recap - 3 most interesting developments in agentic commerce over the past few months… 1/ @ChatGPTapp scaling back from an end-to-end shopping destination and pivoting to a shopping entry point that routes transactions to specialized commerce apps (source: The Information). 2/ All commerce platforms embedding agents directly in their products ( $BKNG $EXPE $AMZN). The most interesting example to me may be AMZN’s “Buy for Me” - agent can complete purchases from other brands inside the Amazon Shopping app when Amazon itself does not carry the product. 3/ B2B commerce is seeing early agentic adoption on the sourcing side, where the workflow is highly research-intensive. --- Guiding principles - what likely holds regardless of how agentic commerce evolves 1/ AI is all about democratizing information and resources Commerce has historically been closer to a seller’s market. Consumers are largely shown what platforms choose to surface - through ads, rankings, and distribution channels. Each layer adds another “tax” in the form of marketing spend, platform fees, or payments. AI agents should gradually shift power toward the buyer. Instead of navigating layers of marketing and platform incentives, consumers can theoretically access the best product for their needs directly. This dynamic may be even more powerful in B2B commerce, which historically has been opaque, fragmented, and relationship-driven. 2/ Faster, cheaper, better rules Faster & cheaper: Price competition should intensify across retailers and even payment providers; Scale advantages may matter more, not less - AMZN’s logistics and fulfillment advantage still remains structurally difficult to replicate Better: more personalized and context-aware - agents matching products to a buyer’s specific needs, constraints, and preferences. 3/ Control of the top of the funnel matters (still) Today $AMZN generates roughly $80B+ (source: Visible Alpha consensus number) in ads revenue, largely because shopping intent starts on $AMZN. If discovery shifts upstream toward AI interfaces (e.g. ChatGPT or GOOGL’s Gemini), the economics of the top of the funnel could shift meaningfully. --- h/t Bernstein analysis: More analysis on each company in here: open.substack.com/pub/robonomics…

The latest energy infrastructure damage from the weekend - covered in our Saturday Macro Note - and much more...

MacroEdge @MacroEdgeRes

5 months ago

The latest on the evolving situation in the Middle East with oil & gas data, Strait of Hormuz traffic, and more below. Saturday Macro Note: Strait of Hormuz & Oil Supply Update, Fracking Activity, War Impact on Aviation and Hospitality Linked below ⤵️ #MacroEdge

3 8 39 11K 8

sam recently made another important prediction: "a breakthrough beyond transformer architecture may be coming, with current models helping discover it" i think openai likely has internal agents whose job is to read papers and explore new ideas. so i would not find that surprising

In 1972, Warren Buffett purchased See’s Candies for $25M At the time: Revenue: ~$30M Pre-tax earnings: ~$4M Since then, See’s has generated more than $2B in cumulative pre-tax earnings for Berkshire — while requiring only ~$32M of additional capital. Buffett has called See’s the “prototype of a dream business.” Why? 1. Strong pricing power 2. Minimal capital required to grow 3. Generates large amounts of cash 4. Durable brand moat That’s what a great business looks like.

This is crazy. $SATS just became the FIRST pure play space stock added to the S&P 500. $ASTS $RKLB $PL holders pay close attention. Yesterday the committee looked past a $14B GAAP loss to add SpaceX proxy $SATS to the flagship index. Alongside the other adds, critical AI infra growing at hyperspeed $LITE, $COHR, $VRT, $SATS stands out like a sore thumb. It's a former dinosauric telco operating business (Sat TV + Mobile Carrier) that recorded a $14.5B net loss for 2025. But beneath a convoluted financial structure mired in debt and subsidiary blur.... $SATS is really just a treasury company for scarce space economy assets that's trading significantly below NAV. The most scarce of all being a 2.2% stake of SpaceX at a $400B valuation the company got when it sold its spectrum to Starlink. That stake alone is worth $79.70 per share based on a $1.25T SpaceX/xAI valuation. For context, $SATS closed at $106 on Friday. The vast majority of this company's value is just the SpaceX equity it's holding on its balance sheet. To add further proof that $SATS has fully embraced its treasury strategy, they cancelled their Q1 earnings call. They're basically saying - we'll see y'all after SpaceX goes public so you can value us properly. So why did the S&P 500 committee add Echostar? Well they pass the market cap minimum of $22B. But their financials are a mess like I pointed out and the S&P 500 rules explicitly require positive GAAP earnings over the trailing four quarters AND the most recent quarter. $SATS posted a $14.5B net loss for 2025. The only way this works is if the committee stripped out the network impairment and decommissioning charges as extraordinary items. The committee had to work hard to see the real story. - $22.65B in cash coming from the AT&T spectrum sale in H1 2026. - ~$11B in SpaceX equity at $400B closer to $27.8B at today's $1.25T SpaceX/xAI valuation. - $6B+ in remaining spectrum (paired AWS-3, 700 MHz, CBRS). - 7.5M Boost Mobile subscribers generating real wireless revenue. Then strip out the huge subsidiary debt that EchoStar can structurally walk away from, Hughes + DISH. And you have a very financially healthy company. The S&P committee doesn't add companies out of charity. They added $SATS because it's one of the most undervalued and liquid space economy vehicles in public markets. Every index fund tracking the S&P 500 now HAS to buy $SATS. That's ~$8 trillion in passive assets that need to allocate. Institutional ownership will explode. Analyst coverage will follow. Options liquidity deepens. And I think the market gets more efficient on this name. Now zoom out to the bigger picture. The S&P 500 just endorsed, whether intentionally or not, a company whose primary value is a SpaceX stake. That's a signal. It tells the market that the committee thinks space economy exposure belongs in the most important equity benchmark on the planet. $ASTS and $RKLB both trade above the $22.7B market cap minimum. Both are pre-profit or barely profitable which historically would've kept them out. But the $SATS inclusion just set a precedent. The committee showed it's willing to exercise discretion and look through messy financials for companies with massive strategic positioning in space. If they bent the rules for a satellite TV company holding SpaceX equity... $ASTS and $RKLB may be closer to S&P 500 inclusion than anyone thinks. And if that happens, the same passive flow tsunami hits them too. Combine this with the idea that $ASTS and $RKLB are extraordinarily undervalued if SpaceX is worth nearly $2T. The S&P 500 just opened the door for space.

My favorite comment I heard this week was from the team at $NOW: "The stock market seems to think we don't know how to do basic math if they think that our pricing of a subscription seat versus AI tokens isn't going to yield a material uplift as we deliver more value for customers". Having caught up with most of the SaaS teams this week, I feel pretty good that their reality is way better than perception. And this sort of separation often leads to the best returns over time.

U.S. CONSIDERS SPECIAL FORCES RAID TO SEIZE IRAN’S URANIUM The U.S. and Israel have discussed sending special operations forces into Iran to secure its stockpile of highly enriched uranium, according to officials familiar with the talks. The target is roughly 450 kilograms of uranium enriched to 60%, which could be upgraded to weapons-grade within weeks. Preventing Iran from obtaining a nuclear weapon is a key objective of President Donald Trump. Any mission would likely involve U.S., Israeli, or joint special forces, possibly alongside nuclear experts and inspectors from the International Atomic Energy Agency. Options include removing the uranium from Iran or diluting it on site. Officials say the operation would only happen after Iran’s military is sufficiently weakened, due to the risks of entering heavily fortified underground facilities. U.S. officials believe most of the uranium is buried at nuclear sites in Isfahan, with smaller amounts at Fordow and Natanz. If fully enriched to weapons grade, the stockpile could theoretically produce up to 11 nuclear bombs. The plan under discussion is not a large invasion, officials say, but small targeted special-operations raids designed to secure or neutralize the material.

Iranian oil depot

Michael Mauboussin on the 2 ways you can make money using Competitive Advantage Period (CAP) 1. Identify higher-than-expected excess returns: earnings surprise & better balance sheet management 2. Expanding competitive advantage period: changes in management, industry structure, or investor psychology

$ORCL will report Q3 FY2026 earnings on March 10 after the market close. Oracle’s reports are becoming one of the best ways to gauge the health of the AI infrastructure ecosystem. Massive debt. Massive capex. Massive exposure to OpenAI. In many ways, $ORCL earnings are becoming what $NVDA earnings used to be 👀

9 green bars in a row for IBM $IBM after the "AI obsolescence" scare. Think it got a little overdone?

The news about $ORCL and OpenAI ending plans to expand the Texas Data Center site is great news for $MSFT, $AMZN and $GOOGL as OAI will have to lean more towards those compute partners.

$PLTR +14.5% this week. Nice bounce off Fib 38.2% but monthly BLUE candle hasn't printed yet and those don't come often. Get 270+ charts daily. Try it free: patreon.com/posts/master-m…

Easy. Lots of GPUs

Fiscal.ai @fiscal_ai

5 months ago

Here's the quarterly Average Revenue per User for social media companies. Pinterest: $2.16 Snap: $3.62 Reddit: $5.98 Meta: $16.56 What makes Meta so much better at monetization? $PINS $SNAP $RDDT $META

25 8 133 52K 28

Gemini really stunk things up fact-checking Claude output: