Ex-Equity PM & Quant.
I built an AI engine to apply 10y+ institutional rigor to the open web. ๐ข
Scanning for hidden gems. ๐
I invest my capital using this.
๐Introducing the Tailwinds Database - a living, continuously-updating knowledge base of recurring investment themes.
When you read 100s of open-web investment articles daily for long enough, something becomes obvious:
The best individual ideas are rarely isolated. They sit on top of deeper structural forces:
๐น rearmament cycles
๐นenergy sovereignty
๐นAI infrastructure buildouts, etc.
that keep generating new opportunities.
Nobody was mapping these forces systematically. So we built it.
Macro themes โ sub-themes โ catalysts, risks, beneficiaries. Every claim grounded in cited source material from the pitches we evaluate.
It grows as we read. New evidence reinforces or extends what's already there. A compounding knowledge base.
I haven't seen anything like this shared publicly. Now it's live at our platform (link in commet and bio)
๐๐ข
๐GEM STOCK PITCH IDENTIFIED
From an "ungovernable digital commune" in 2023 to the most essential dataset in the agentic economy.
REDDIT ( $RDDT) by @Hiddenmarketgem
๐ THE PIVOT:
The company is productizing the "search reddit" habit. Its AI search product, Reddit Answers, grew from 1M weekly users to 15M in just three quarters. It is evolving from a forum into a high-intent commercial engine where 40% of conversations are now commercial in nature.
๐ก WHY THE MARKET IS LAGGING:
- Stuck in the "social media" drawer next to Snap and Pinterest.
- Licensing revenue ($130M/year) is viewed as a one-off rather than a compounding annuity.
โก CATALYST STACK:
- Translation tools unlocking India, Brazil, and France ARPU.
- Dynamic pricing for AI licensing as models require "fresh" human data.
- Transitioning search into shoppable modules for high-intent queries.
โ๏ธ THE ASYMMETRY:
A year-to-date drawdown of 30%+ implies the market lost its nerve on the multiple, yet the fundamentals (40% EBITDA margins) have never been cleaner. The downside is anchored by a $1B buyback funded entirely by cash flow.
MY TAKE
The "free option" on search is what makes the risk/reward skew so heavily toward the bull case. Even if the search product only monetizes at a fraction of Meta's rate, the current entry provides a significant margin of safety. I'm watching the Anthropic litigation closely, as a legal win there would effectively wall off the garden for good.
๐MY FAVORITE STOCK PITCH FROM THE LAST 24H
CHURCHILL CHINA ( $CHH ) by @dirtcheapeurope
๐ THE SETUP
The market treats this as a cyclical bet on restaurant openings. In reality, ~80% of revenue is replacement -plates break regardless of the macro cycle. As long as restaurants stay open, they have to reorder the same pattern to match their existing sets.
๐ THE NUMBERS
โ EV/EBITDA: 3.0x (vs 8.2x historical median)
โ P/B: 0.61x (vs 2.2x historical median)
โ Dividend: 6.5% yield with ยฃ11M cash to support it.
โก CATALYST STACK
The implementation of a 79% duty on Chinese ceramic imports in 2026 creates a massive price umbrella. Combined with the bankruptcy of its two largest domestic rivals, Churchill is walking into a near-monopoly for high-end UK hospitality supply.
๐ฏ THE RISK
Stock is illiquid and management ownership is low (CEO at 0.2%). There is a risk the board remains passive with the cash instead of buying back shares at these depressed levels.
MY TAKE
The thesis rests on the "gravity and wet floors" annuity. Even if new restaurant installations stay at zero, the replacement base generates enough cash to cover the dividend and wait for the cycle. The risk/reward is heavily skewed when you're buying a survivor at 60% of book value. Itโs a bet on the floor holding while the competition clears out.
SYNTHESIS:
5 of 6 pitches reach the same destination from different starting points.
The combined three-leg thesis:
โ Leg 1 (Value): Sub-20x forward P/E once one-time costs are stripped - lowest in a decade
โ Leg 2 (Catalyst): Alani Nu scanner growth triple-digits; if durable, the FY26 revenue guide looks conservative
โ Leg 3 (Floor): Pepsi's $34.08 blended cost basis makes a buyout accretive at current prices - a rational backstop no analyst disputes
MY TAKE
The stack would be a clear ๐๐๐ if one pitch had definitively answered BizModelMastery's Alani demand question using channel check data or repeat cohort analysis. None did. The thesis is strong - but its weakest link is also the loudest unresolved variable, and that is exactly the kind of gap worth flagging to subscribers before they size a position.
Read the originals: ๐
๐ Jun: https://t.co/VyGqsKkiiI
๐ May (1): https://t.co/6cJJuBuRRd
๐ May (2): https://t.co/RwSSowNX2e
๐ Apr (1): https://t.co/zVV9PuNS4N
๐ Apr (2): https://t.co/VdxsnSQSDY
๐ Mar: https://t.co/lweqjOC04m
We surface the signal inside pitch clusters like this daily at https://t.co/aDpDe7FZye ๐๐ข
Six pitches. One beaten-down name. 90 days.
Celsius Holdings ($CELH)
This density on a mid-cap consumer brand is rare enough to pay attention.
๐ Jun - Bullseye Picks: "Monster playbook + hidden 20x forward multiple"
๐ May - Business Model Mastery: "Survives the Kick Out Step โ barely"
๐ May - Philosopher Investor: "CELH +0.30% in a week of macro noise"
๐ Apr - WinterGems (@Govro12): "Alani Nu re-accelerating at triple digits post-acquisition"
๐ Apr - ็พ่ก่ๅธๆฉ: "Adding conviction on Pepsi-backed multi-brand platform"
๐ Mar - Stock Talk Newsletter (@bmb21): "Lowest P/E in a decade - Pepsi should just buy it"
What do six independent reads, ranging from conviction adds to a stress test, actually agree on? ๐งต
๐ฏ CONSENSUS - what all six land on:
โ PepsiCo's 11% stake, board seat, and $34.08 blended cost basis creates a structural floor no standalone valuation fully prices in
โ Alani Nu, not core Celsius, is the growth engine - triple-digit scanner growth post-acquisition is the number that matters
โ The valuation headline (70x GAAP P/E) is distorted by one-time integration costs and misleads casual screens
โก ADDITIVE - one insight no other pitch brought:
โ [Mar] Stock Talk: Kirkland's $0.07/can entry specifically does NOT threaten Alani Nu's female-skewed DTC-origin brand - 52% of Celsius buyers repeat 5x+ and those demographics don't overlap
โ [Apr] WinterGems: Alani's sales growth actually *accelerated* after the Celsius acquisition, disproving the "brand dilution" concern at the most critical measurement window
โ [May] BizModelMastery: This is the *only* pitch that formally stress-tests the durability question - is Alani's demand real consumer pull or shelf expansion loading? Verdict: uncertain, not confirmed
โ [Jun] Bullseye: The Monster Energy historical analogy is the *only* pitch to frame this as a category-expansion story rather than a share-gain story - those are different TAMs
โ๏ธ WHERE THEY SPLIT:
โ MOAT: Bullseye/Stock Talk treat PepsiCo distribution as near-permanent; BizModelMastery explicitly calls the moat "fairly weak overall" โ brand-sensitive, competition-sensitive
โ ALANI DURABILITY: 4 pitches assume it; 1 (BizModelMastery) makes it the thesis-killer question; PhilosopherInvestor doesn't engage at all (it's a macro brief, not a CELH deep dive)
โ MANAGEMENT: BizModelMastery flags SEC 2021 charges, lack of guidance, and comp tied to revenue not ROIC - no other pitch raises this
๐MY FAVORITE STOCK PITCH FROM THE LAST 24H
$RAY.TO is a textbook case of a "misfiled" compounder.
The market sees a shrinking radio business; the numbers show a capital-light advertising machine with double-digit organic growth and a founder owning 21% of the equity.
๐ THE SETUP
โ Owns a global music library (sunk cost) monetized 3x over.
โ Recurring subscriptions from cable/telcos provide a stable floor.
โ Advertising (FAST and Retail Media) is the primary growth engine.
๐ฏ THE NUANCE
The "backfill" ad inventory is owned by partners like Vizio and Samsung. This creates a gatekeeper risk, though Stingrayโs niche in non-on-demand audio keeps royalty costs low enough to share revenue where larger rivals cannot.
๐งฎ THE MATH
Trading at ~6.7x forward FCF. Analyst targets sit near $20, implying 40%+ upside if the market stops extrapolating the radio decline onto the growth engine.
โก WHY NOW
The TuneIn acquisition provides the programmatic tech to fill C$600M of empty ad slots automatically. Management reports the strongest start to a year in the company's 18-year history.
MY TAKE
Founder Eric Boykoโs background as a professional accountant is visible in the FCF conversion, which often clears 100%. While the Newcap radio acquisition was a clear overpayment, the pivot toward high-margin digital inventory looks better calculated. The risk/reward is asymmetric here as long as the programmatic run rate continues to scale toward the $275M annualized target. It is a classic "cheap for a reason" stock where the reason is starting to matter much less than the cash flow.
The market thinks Stingray $RAY.TO is a dying radio company. It isn't one.
Three-quarters of the business grew 43.6% and threw off record cash. The radio write-down spooked everyone into marking down the whole thing.
New deep dive ๐(free)
https://t.co/HsxUjCmTNL
Weekly stock gem summary is out!
The GPU is the trade everyone made. The transceiver manufacturer behind it trades at a fifth of the price of the brands it builds for. The PCB supplier inside Nvidia's networking switch trades at a fraction of the holding company that owns it. The copper mine feeding AI's physical infrastructure is days from first production and priced as if it's still a developer. The precision optics platform already delivering into fusion and space solar hasn't moved.
Same week. Four different asset classes. The same underlying phenomenon: the market is pricing the story while the mechanism sits one layer down, unmoved. This week's gems are the mechanism.
link in comments
Three signs a pitch is human-written and worth your time.
A falsifiable thesis. An admission of what the author doesn't know. And a number that doesn't appear on any data terminal, like a channel check or a primary-source estimate, the kind of calculation only someone who'd lived with the name for a year would bother to do. ๐๐ข
Process before outcomes. A great pitch that didn't work is still a great pitch if the framework was sound. Judge the decision, not the result. ๐๐ข
Three things every quality open-web author does, without exception. They state the thesis in one sentence. They show the evidence. They define the disconfirming case, the scenario where they'd be wrong.
If any of the three is missing, keep scrolling. It's not a quality filter, it's the quality filter. ๐๐ข
The bar for my selections:
โ "Adobe is cheap after earnings" โ Generic
โ "AI infrastructure is still a good trade" โ Lazy
โ "Six authors hit $ADBE before the gap between fundamentals and price becomes impossible to ignore โ here's the specific number driving them" โ Specific
โ "Three independent governance-arbitrage setups, same mechanical mispricing, calendar catalysts now inside 30 days" โ Testable
I track 1,000+ authors and read hundreds of financial pieces weekly to surface the deep-dive pitches actually worth reading.
Above was a brief summary. If you want the detailed view:
๐: https://t.co/h5wme2HbOn
๐: https://t.co/8u4jsNlFol
๐๐ข
๐๐ช YOUR WINDOW INTO OPEN WEB INVESTING CURRENTS THIS WEEK (quick summary)
At least 3 interesting things happened this week:
โฎ๏ธ US and Iran signed a preliminary peace deal. Nasdaq surged 3.07% in one session.
๐ฆ Warsh held rates. 18 of 19 Fed officials now see a hike coming before year-end.
๐ Adobe beat earnings last week. Stock is still down 37% YTD. Six authors spent this week trying to explain why.
174 pitches on the open web spent the week hunting mispricings.
Here are the broad strokes of the hive mind.
๐งต๐
๐งฌ Theme 5: Biotech Binary Events and the GLP-1 M&A Window
(8 pitches, 7 tickers)
These aren't pipeline bets. They're regulatory calendars and M&A windows with specific close dates.
The argument: the market has trained itself to discount binary biotech events. That discount is systematic and occasionally wrong โ especially when the catalyst date is known and the downside is bounded.
๐น $QURE (uniQure) โ already +90% since January on the FDA's Accelerated Approval reversal for AMT-130 in Huntington's. US BLA submission expected Q3 2026. $586M cash, runway through late 2029. Author argues the re-rating isn't done.
๐น $VKTX (Viking Therapeutics) โ dual GIP/GLP-1 oral candidate. Current valuation $3.5-4.5B vs acquisition target range of $5.5-9B+. M&A window open now, before Phase 3 data makes the asset too expensive for acquirers.
๐น $ABVX (Abivax) โ selloff on a "thin" malignancy signal the author argues is non-systemic. Rinvoq's regulatory history cited as the comparison case for misread early safety flags.
The falsification test: for $QURE, FDA rejection on BLA review. For $VKTX, Phase 3 data release before a deal closes. Both reprice the setup immediately.