Brad Lyons

Including the full August 2025 tweet that I never posted below. TLDR: I argued that the AI boom was structurally unsustainable: too much capital, too many undifferentiated “wrappers,” and foundation models burning billions to subsidize usage. Since then, the bubble has inflated

@sv_techie Well said. Public optics are the only thing temporarily delaying the rug pull at most companies. Although Block's decision to cut 50% may be the first domino to fall. We shall see... what I do know is that most people are largely unaware of what is about to hit them.

@blyons151 Capex spend is now funded by opex savings 🤷‍♂️

No surprise. Infrastructure isn't cheap and available cash is finite. Once AI streamlines operations, labor becomes the easiest lever to pull. Public companies will have no choice but to pursue mass layoffs to fund the infrastructure required to stay competitive. This is just the

@SaaSletter Thanks Matt... hope all is well!

👀 I've found @blyons151's work to be particular thoughtful + rigorous -> worth reading vs most AI takes:

$AAOI animal

In August, I wrote this but never sent it. Publishing it felt like bad business. The very funds I was warning might lose were, and still are, key clients for Crossover Research. So I stayed quiet. But staying quiet no longer feels right. With software multiples down more than 30% across the board, and analysts calling this the SaaSpocalypse, the reckoning I expected has arrived. This is not a macro correction. It is not rates, inflation, or demand softening. It is structural. AI is not just competing with enterprise software; it is replacing it. The per‑seat model that powered twenty years of SaaS growth is collapsing as agents bypass the interface entirely and operate directly on the data. Salesforce, ServiceNow, Adobe, and Workday are all down 40% or more from recent highs. Thomson Reuters fell 16% in a single session after Anthropic released its legal agent. The room I once hesitated to rattle has already been rattled. The math has not changed since August. Only my willingness to say it out loud has. The Thesis AI acceleration is collapsing the cost of creation and narrowing the gap between “build” and “buy.” The winners will be those that: > Own mission‑critical workflows: controlling the system of record where business logic and risk live. > Capture proprietary, permissioned data feedback loops: continuously refreshed, high‑signal data that compounds advantage over time. > Convert trust and embeddedness into pricing power: turning reliability, compliance, and integration depth into premium retention. Everything else will be repriced toward zero. Four structural realities: 1. Commoditization crushes undifferentiated software. Vendors competing on price or easily cloned features face accelerating margin compression as AI drives time‑to‑parity toward zero. Only those with differentiated ROI, deep workflow embed, or regulatory trust sustain pricing power. 2. Enterprise exposure is a time moat, not a permanent one. Integration and compliance slow churn but do not stop it. As agentic AI removes implementation friction, retention will flow toward vendors that own the workflow, not those that simply serve large customers. 3. Build‑cost compression redefines survival. Stand‑alone tools and UX‑first point solutions are first to fall. Platforms that control data, compliance, and execution layers, the true systems of record, will outlast the rest. 4. Proprietary data feedback loops are the modern moat. Durable software compounds advantage through exclusive, self‑reinforcing data capture that directly improves outcomes and compliance intelligence. Raw data volume is no longer defensible; uniqueness, context, and feedback velocity define resilience. What this means for diligence This is exactly the question Crossover Research was built to answer for PE and growth investors: not whether a vendor looks sticky on paper, but whether customers prove the moat through workflow embeddedness, data defensibility, pricing leverage, and displacement risk. We have built a Voice of Customer diligence engine to make that visible [crossoverresearch.com]. If you want to read the full piece I wrote in August ("The AI Investment Supercycle Hypothesis - Mon, Aug 11, 2025") DM or email me: brad@crossoverresearch.com

INTRODUCING CROSSOVER CATALYST - THE FIRST DUAL-SIDED BANKED PROCESS INTELLIGENCE PRODUCT Crossover Catalyst delivers customer intelligence from banked transactions before formal processes begin. Investment banks commission us to conduct customer fieldwork to win mandates. We

@GavinSBaker @FT Please do. Doubling down after he gets backed into a corner. Gotta love journalistic malfeasance. Hope all is well!

My most extensive X argument in years. Has been kind of fun. The @ft really needs to either retract the article or correct it along the lines I have repeatedly described. Should I make this a daily crusade until it happens? And I think the poor A100 has a defamation case!

14 months later and $PI has gone from $50 to $220… was a high conviction play, longs got paid. Congrats. Some others that I’m focused on: $PCOR - long $ONON - short $OS - neutral NT; but bullish LT DMs open for any additional comments or if you have other ideas

$AAOI up 17% today, and up a whole lot more than that since my initial write up.

Nike $NKE Restoring the “old guard” is driving renewed confidence in the turnaround story this AM. Nike getting a much needed catalyst overnight with the return of 30-year Nike veteran, Elliott Hill. While there is a lot that needs to be worked through to get back into competitive form (i.e.addressing: (1)North America underperformance (2) the stalled innovation engine and (3) China weakness), one could only hope that the shake-up of executive leadership brings a renewed focus to what historically made Nike great… product innovation, wholesale partnerships, and brand positioning/storytelling. One of the first things that came to mind after seeing the headline, was the derivative impact on Foot Locker $FL and whether Elliott’s return will drive an even greater shift in focus to restoring wholesale channel relationships. While that dynamic remains to be seen, one can only presume that it's a logical move he makes given the challenges Nike has seen leaning into its DTC channel.

@tomas_maj 1) Clearly ya didn’t read the post 2) it was never intended to be a short term play (in the actual piece we implied that 1H25 is when this should flow through in a more meaningful manner)

@BLyons151 Ok! You were in the red since your post - and finally positive after an upgrade.. Seems like a stretch to be congratulating yourself… I would cash out:)

$AAOI Nice move on an upgrade from RJ. Stock up >20% since my initial post in May. Overlapping cycles have presented some compelling tailwinds. Some execution risk (...given AAOI's history) but the upside opportunity seems fairly compelling from here, in my view. Raymond James upgraded Applied Optoelectronics to Outperform from Market Perform with a $17 price target. The firm expects the shares remain volatile as investors debate both the size of Applied's opportunity and its ability to execute in both datacenter and cable television. While execution risk remains, the company is among the most likely companies in the sector to experience upward estimate revisions. Access my bi-weekly write up here: t.co/mfw3YRLzN6

RT @heartbreakout: Interesting snippet for $AMZN AWS in this morning's TheInformation article on $META's llama: Meta is also a longstandin…

@cole_barcia @LogicalThesis Yes, they've led with CTV angle which is where dollars have flowed over past 3 years... so $MGNI has been the outsized beneficiary Retail / commerce media is another good growth vector so agree with you there re: $PUBM

@BLyons151 @LogicalThesis Upper hand in CTV? Or do you see other competitive advantages beyond partnerships?

I'm definitely not trying to sell $PUBM here but when I think about $PUBM vs $MGNI $MGNI has better competitive positioning (CTV), $NFLX partnership, and chart is way more intact Perhaps its better to just consolidate into $MGNI Gotta imagine $PUBM is buying back heavier here

$TKO did not disappoint

$ABNB - the concerns I highlighted in the quoted tweet are now playing out. My view in 2h23 was that valuation would be tough to support given confluence of headwinds impacting Airbnb’s room night growth. I also thought margins would be pressured as company would need to increase s&m to maintain competitiveness.