AI Infrastructure Inversion — OptionsPlay
Home Live Coaching Pricing Courses
📊 Hyperscalers committed $675B+ in AI capex for 2026 — 8.4x what they spent in 2020
AI Infrastructure Inversion — Exclusive Thematic Research

The Market Is
Mispricing AI Stocks.
We Know Which Ones.

Wall Street is modeling for chatbots. The industry is building autonomous agents that consume 3,000x more compute per task. That gap is why the $2T tech selloff is your entry point — if you know which companies to own.

$675B+
Hyperscaler AI Capex 2026E
3,000x
Agent vs Chatbot Compute Gap
Loading

14 days free · No card required · Cancel anytime

AI Infrastructure Exposure Map OptionsPlay Research
IndustryExposureScore
Semiconductor & ComputeBULLISH+10
Power Generation & GridBULLISH+9
Physical InfrastructureBULLISH+9
🔒 All 8 industries + company picks — sign up for free access
$675B+
Hyperscaler AI Capex 2026E
3,000x
Agent vs Chatbot Compute
8
Industries Mapped
$1.5T
AI Market by 2030
0
Black Boxes
Research By
Tony Zhang, Chief Strategist at OptionsPlay
CNBC
📺
Tony Zhang
Chief Strategist, OptionsPlay · CNBC Contributor

"Wall Street is making a historic miscalculation. They are modeling for chatbots while the industry is building autonomous agents that consume 3,000x more compute per task. The current selloff is creating one of the most asymmetric entry points in a generation."

You've Seen the AI Headlines.
This Is What's Behind Them.

Combined hyperscaler capex surged from $80B in 2020 to $675B in 2026. That's an 8.4x increase in six years. We mapped where every dollar is flowing — and which companies capture it.

8 Industries. 7 Bullish. 1 Bearish.

Seven industries are positioned for substantial upside as capital flows into physical infrastructure. One — legacy SaaS — faces structural headwinds as AI disintermediates the traditional software model. Every industry scored from −10 to +10.

📊

The Math Wall Street Is Getting Wrong

75% of knowledge workers use AI as a chatbot consuming ~500 tokens. When those same workers use AI agents, each task consumes 75,000–1.5M tokens. Same users, up to 3,000x more compute demand — without adding a single new user.

🔗

One Top Pick. 8/8 Dimensions.

Our anchor thesis is one company that sits at the intersection of every critical AI layer and self-funds its entire buildout. It's the only company scoring 8 out of 8 across the AI value chain. Full analyst profile in the research.

📖

Full Methodology. Zero Black Boxes.

Every industry score is documented. Every bull and bear case published. No vague ratings, no hidden assumptions. The research is built so you can verify every number and push back if you disagree.

See What's Inside

A preview of what you'll access after signing up.

GOOGLAlphabet Inc.
Strong Buy · 8/8
OptionsPlay Investment View

The only company competing at the highest level across all 8 dimensions of the AI value chain — from custom silicon (Ironwood TPU) to frontier models (Gemini) to consumer distribution at 750M monthly active users.

Forward P/E
27.53x
vs 21.92x sector avg
Net Margin
32.81%
+23.22pp vs sector
🔒 Full analyst profile + valuation + options strategy — sign up →
AI Value Chain Scorecard
8 dimensions · 6 companies compared
Custom Silicon
Cloud Infra
Frontier Models
Consumer
Enterprise
Research
Self-Funded
Data Moat
8/8
Our Pick
5/8
Next Best
2/8
Closest Rival
🔒 See the full competitive analysis — sign up →
The Numbers Wall Street Is Getting Wrong
All figures from the research — confirmed sources on every stat
Hyperscaler AI capex 2026E (from $80B in 2020)$675B+
AI users paying less than $20/month97%
Knowledge workers using AI weekly75%
Inference share of AI compute by 2026E (was 35% in 2024)67%
🔒 Full thesis + investment framework — sign up for free access →
8 Industries — AI Infrastructure Exposure Map
Seven bullish. One bearish. All scored from −10 to +10.
Semiconductor & Compute
BULLISH+10
Legacy SaaS & Enterprise Software
BEARISH−8
🔒 All 8 industries + company picks — sign up for free access →
📋
The Macro Thesis
Why the current selloff is an entry point. The case that Wall Street is dramatically underestimating AI compute demand — and why.
🔁
The Inference Inflection
Training consumed 65% of AI compute in 2024. By 2026, inference consumes 67% and growing. Why this changes infrastructure economics from 'build once' to 'run forever.'
🗺
8-Industry Exposure Map
Every industry scored from −10 to +10. Bullish and bearish positions. The specific logic behind each rating documented in full.
🏆
The #1 Pick — Full Analyst Profile
Financial analysis, competitive positioning, valuation, and options strategy for the company scoring 8/8 across the AI value chain.
📅
Upcoming Industry Deep Dives
Subscribers receive a deep-dive report on each of the 8 industries as it's published — specific companies, specific risk/reward, building toward a complete portfolio.
Options Strategy Overlays
For each position, specific strategies to enhance returns, manage risk, and generate income using OptionsPlay's proprietary tools.

Read the Methodology.
Make Up Your Own Mind.

By the time the rotation is obvious, the best entry points will be gone. Subscribe now to receive each industry deep dive as it's released — and build your AI Infrastructure Portfolio before the rest of the market catches on.

Loading

14 days free · No card required · Cancel anytime