First a message from our friends at Porter & Co (Sponsor) |
Famed historian Yuval Noah Harari recently issued a warning that should send a shiver down the spine of every American. He predicts the emergence of a massive new "useless class." |
These aren't just people who are temporarily unemployed. |
These are people who have become economically irrelevant. |
As Luke Lango and I just exposed in our recent interview, we have reached the "Singularity." For the first time in 250 years, intelligence has been decoupled from labor. |
During America's first 1776 moment, the steam engine replaced muscle. In this new 1776 moment, AI is replacing the human mind. |
If a machine is faster, cheaper, and smarter than you at your specific job... the economic reason to hire you simply vanishes. |
This is why you see the "Magnificent 7" tech giants adding trillions in value while the "real" economy feels like it's in a death spiral. |
And the divide is widening. On one side: The "Useless Class" who cling to old-world skills. On the other: The "New Aristocracy" who own the assets of the Technological Republic. |
Which side will you be on? |
Luke and I have identified the three specific money moves our research indicates you must make to ensure you stay on the winning side of this divide. |
Click here for the full story. |
Good investing, |
Porter Stansberry |
FEATURED ARTICLE |
AI Is Booming—So Why Is Software Getting Punished? |
Tonight's tape gave us a clean, almost rude, message: |
AI hardware is being treated like a cash machine.
Enterprise software is being treated like a promise that hasn't turned into dollars yet. |
Same AI wave. Two totally different report cards. |
NVIDIA just posted $68.1B in Q4 revenue (up 73% YoY) and $62.3B in Data Center revenue (up 75% YoY)—official numbers, not whispers. Meanwhile, the software giants are living in "guidance gap" land: Salesforce guided FY2027 revenue $45.8B–$46.2B, slightly below expectations. Snowflake beat and guided FY2027 product revenue $5.66B (above estimates), yet the stock still got hit because the market is skittish about AI changing software economics. The Trade Desk guided Q1 revenue $678M, below consensus around $688–$689M, and the stock sold off hard. Even "high quality workflow" names like ServiceNow can beat/guide well and still wobble because investors are suspicious of how fast AI turns into durable ARR.
|
Cheap Investor translation: |
The market is no longer paying for "AI is important."
It's paying for "AI shows up in guidance." |
|
|
Let's break down the divergence—and where the mispricings tend to hide. |
|
Scoreboard: What actually happened |
1) Hardware: proof-of-cashflow quarter |
NVIDIA delivered a headline that ends arguments: $68.1B quarterly revenue, +73% YoY. |
The hardware story has three features the market loves in a macro-cautious regime: |
Immediate monetization (sell chips/systems now) Visible demand signals (hyperscaler capex, backlog language, supply constraints) Pricing power (strong margins reported alongside growth)
|
2) Software: guidance gaps and "prove it" math |
Software isn't being punished for missing the AI trend. |
It's being punished for timing mismatch: |
Costs move now (AI R&D, inference costs, support, compute) Pricing models are in transition (seats → usage/actions/outcomes) Customers are cautious (pilots expand slower than the narrative implies)
|
Salesforce is the clean example: AI narrative is strong, but FY2027 revenue guide came in $45.8B–$46.2B, below the Street's expectation (midpoint below ~$46.06B). |
Snowflake is the other side: it guided FY2027 product revenue of $5.66B (above estimates) and talked up AI usage (including "Snowflake Intelligence" adoption), but shares still dipped—because software is trading like a sector whose business model is being renegotiated in real time. |
The Trade Desk is the market's blunt instrument: solid business, but Q1 guide $678M vs Street ~$689M, and the stock reacted like growth durability just got questioned. (continued below) |
|
The Race to Replace China's Metals Is Heating Up(sponsor) |
|
The U.S. and its allies are investing billions to rebuild domestic supply chains for strategic materials. One Rio Tinto-partnered project could help anchor N. America's answer to China's dominance.
Learn how this company is playing a role in restoring supply >
|
(continued…) |
The real reason: AI monetizes differently in hardware vs software |
Hardware monetization is transactional and front-loaded |
If AI demand rises, hardware vendors ship more and bill more. |
That's why NVIDIA's quarter acts like a macro "all clear" signal for AI infrastructure. |
Even AMD is getting pulled into the supercycle via mega-deals (Meta buying ~$60B of AI chips over 5 years, per Reuters' view piece). |
Software monetization is contractual and back-loaded |
Software has to answer harder questions: |
What is the unit of value in an agent world?
Seats are a clean unit. "Actions" or "outcomes" are harder. Does AI reduce seat counts before it increases spend-per-customer?
If AI boosts productivity, CFOs first ask: "Do we need fewer licenses?" Do customers pay twice?
Many are already paying for cloud + data + security + workflow tools. AI is an upsell battle, not a blank check.
|
This is why software can report fine quarters and still get punished on guidance: the market is trying to price a new equilibrium for the whole model. |
|
Company/Theme deep dive: "Guidance gaps" and what they actually mean |
A "guidance gap" is usually one of three things: |
Gap Type A: macro caution (real economy) |
Customers delay expansions, lengthen deal cycles, scrutinize renewals. |
Salesforce explicitly pointed to pressure on enterprise demand amid global uncertainty and tight budgets—hence the underwhelming full-year outlook despite strong Q4. |
Gap Type B: pricing transition (seat → usage) |
AI changes how vendors charge—and how customers budget. |
The market hates transitions because they create temporary opacity: |
|
Salesforce is publicly wrestling with this shift while defending SaaS from "SaaS-pocalypse" fears. |
Gap Type C: competitive compression (walled gardens, bundling) |
In ad tech, the "open internet" model has to prove it can keep share when budgets get pulled toward platforms with first-party data and closed-loop attribution. |
That's the subtext behind why TTD can beat a quarter and still get slammed on guidance—Wall Street is treating it like a share/market-structure question, not a quarterly execution question. |
|
"Is it cheap?" The only Cheap Investor test that matters tonight |
Here's the trap: software looks "down a lot," so it looks cheap. |
But cheap is not "down." |
Cheap is: |
Down + durable + misunderstood + fixable. |
|
|
So use a two-step filter: |
Step 1: Does the company benefit from AI usage even if seats shrink? |
These are the "AI toll collectors." |
Snowflake's case is straightforward: more AI workloads typically means more data processing, more analytics usage, more consumption—hence why it guided FY2027 product revenue to $5.66B (above estimates). |
ServiceNow is similar in spirit: workflow automation + AI embedded in operations. It guided FY2026 subscription revenue above estimates and talked up AI partnerships (OpenAI/Anthropic) as part of demand strength. |
Step 2: Is the market punishing near-term guidance more than long-term unit economics? |
Salesforce is a classic "near-term guide vs long-term narrative" mismatch. It boosted longer-term ambition (2030 target) and buybacks, but the stock reacted to the FY2027 guide. |
That can create opportunity—but only if you believe the monetization path is measurable, not rhetorical. |
|
The Cheap Investor "under-the-radar" angle: don't buy "software," buy the right business model |
Tonight's divergence is pushing money toward models with: |
Usage-linked revenue (benefits from AI activity) Mission-critical workflows (less discretionary) Clear payback (ROI that survives budget scrutiny)
|
Here are the three cohorts implied by the tape: |
Cohort 1: Hardware kings (easy ROI, clear scarcity) |
|
This cohort is expensive-looking—yet keeps winning because the "AI tax" is being paid in silicon. |
Cohort 2: AI toll collectors (software that gets paid when AI runs) |
|
These are the software names with the cleanest path to "AI shows up in dollars," because the meter runs when the workloads run. |
Cohort 3: Monetization question marks (great products, harder budget math) |
|
Not "bad companies." Just priced on a tougher rubric right now. |
|
Bull / Base / Bear scenarios |
Bull case: software re-rates as AI monetization becomes visible |
What would have to happen: |
AI moves from pilots to scaled deployments (not just "interest") Pricing transitions stabilize (clear unit economics) Guidance starts to reflect AI as additive, not defensive
|
Winners: |
|
Base case: choppy tape, hardware leadership persists, software remains selective |
What happens: |
Hardware keeps printing results (capex still flows) Software earnings are "fine," but guidance stays cautious Stock picking beats sector buying
|
Winners: |
|
Bear case: macro tightens and "seat rationalization" becomes explicit |
Triggers: |
|
Winners: |
|
|
Actionable plan (no hype, just process) |
1) Decide which side of the divergence you're trading |
If you want momentum + proof, you're in hardware. If you want mean reversion + valuation, you hunt software—but only with the right filter.
|
2) Use a 1/3 – 1/3 – 1/3 scale-in framework for software |
First 1/3: after a guidance-driven flush when the tape stabilizes Second 1/3: when the stock holds up on a bad market day (relative strength) Third 1/3: only after the next quarter confirms the monetization path
|
3) Don't buy "AI software." Buy "AI budget survivors." |
Ask one question: |
If budgets freeze for two quarters, does this product still get paid? |
|
|
If yes: it's investable.
If no: it's tradable at best. |
|
Cheap Investor checklist / scorecard (next 1–2 weeks) |
NVDA follow-through: does the market treat the $68.1B quarter as "extend the runway" or "sell the news"? IGV behavior: does software stop making lower lows (stability > rebound)? CRM narrative vs numbers: do investors keep punishing the guide, or start rewarding buybacks + long-term AI traction? SNOW consumption signals: does AI adoption translate into sustained product revenue confidence? NOW enterprise demand: subscription revenue trajectory and AI integration momentum. TTD ad budget reality: does guidance stabilize, or do "walled garden" fears keep biting? One red flag: hardware up while software down and breadth deteriorates—means the market is narrowing risk, not expanding it.
|
|
Bottom line |
Tonight's divergence is rational: |
Hardware is booming because AI spending shows up as immediate revenue (NVDA just proved it again). Software is struggling because AI has to show up as guidance, not demos—and the business model is mid-transition.
|
If you want a "cheap angle" that isn't a trap: |
Don't buy the software sector. Buy the software business model that gets paid when AI runs.
That's why the market is more forgiving to usage-linked "toll collectors" than to seat-based suites and ad-tech models facing macro and structure pressure. |
Disclaimer: This article is for informational purposes only and does not constitute investment advice. Investing involves risk, including the potential loss of principal. Always do your own research before making investment decisions. |
|
|
|
|
0 التعليقات:
إرسال تعليق