Vyzora Lens
Methodology
How our editorial framework — wave model, conviction grades, and signal classifications — is derived, applied, and refined. Read this before drawing conclusions from anything in the dashboard.
Last refreshed: April 19, 2026 · Version 1.0
Important: this is editorial opinion, not investment advice. Every framework described below produces editorial research only. Vyzora is not a registered investment advisor. The outputs are not recommendations to buy or sell any security. Past performance shown anywhere in the Service is historical and may reflect selection bias. Consult a licensed financial professional before acting on any content.
1. Editorial focus
Vyzora Lens covers the publicly listed companies powering the artificial-intelligence value chain — from semiconductors, memory, and optical networking through power infrastructure, robotics, monetizers, and quantum. We focus on this universe because it is the largest capital-formation cycle of the 2020s and because it is poorly understood by mainstream sell-side research.
We do not cover: private companies, crypto assets, FX pairs, sports outcomes, or any non-equity asset class.
2. Universe of coverage
Our active universe is approximately 110 publicly traded names across 12 AI sectors. Names enter the universe when they (a) generate or are projected to generate at least 25% of revenue from AI-related demand within 24 months, and (b) have sufficient public disclosure to evaluate. Names exit when they no longer meet (a) or when corporate actions remove them from public listings.
Sector taxonomy:
- Compute & Accelerators (GPUs, custom ASICs)
- Memory & Storage (HBM, CXL, NAND)
- Networking & Photonics (transceivers, switches, CPO)
- Foundry & Equipment (wafer fabs, lithography, EDA)
- Energy, Nuclear & Gas (power for data centers)
- Power & Cooling Infrastructure (transformers, liquid cooling)
- Sovereign AI / Data Centers (national infrastructure plays)
- Robotics & Physical AI (cobots, autonomy, humanoids)
- AI Monetizers (ad-tech, enterprise SaaS earning AI revenue today)
- Software & Tooling (platforms, MLOps, data pipelines)
- Quantum Computing (call-option-style exposure)
- Adjacent Beneficiaries (industrial enablers)
3. The wave model
We organize the AI capex cycle into eight named investment waves. Each wave reflects a structural bottleneck in the value chain that, in our view, will or has shifted alpha to a different cluster of names. Waves are not predictions about price — they are capital-flow narratives that we use to organize coverage.
Wave 2
Memory
2024–25 · Maturing
Wave 3
Power
2025–26 · Active
Wave 4
Optics
2026–27 · Active
Wave 5
Robotics
2027–28 · Loading
Wave 6
CXL Memory
2027–28 · Loading
Wave 7
Monetizers
2026–27 · Active
Wave 8
Quantum
2028–30 · Speculative
Wave assignment is editorial — it reflects our reading of supply-chain data, capex commitments, and management commentary, not a quantitative model. Reasonable analysts will disagree on timing.
4. Conviction grade scale
Every theme and selected name in our coverage carries a conviction grade reflecting our editorial confidence in the underlying thesis, not a price target. Grades reflect (a) clarity of the bottleneck, (b) competitive moat, (c) management track record, (d) optionality, and (e) downside containment.
A
High conviction
Multiple independent data points support a clear, durable bottleneck. We believe the thesis is well-formed.
B
Above-average
Thesis is sound; one or two key inputs require additional confirmation in the next 1–2 quarters.
C
Tracking
Theme is interesting but evidence is incomplete. We are tracking it, not advocating it.
D
Skeptical
Visible interest but with structural concerns we cannot resolve from current disclosures.
F
Avoid (editorial)
In our editorial view, the risk/reward is unfavorable at present prices.
A conviction grade is not a recommendation, a rating, or a price target. It is an editorial label intended to help readers compare our level of confidence across multiple themes within the dashboard.
5. Signal classification
The Performance leaderboard tags each tracked name with one of four signal labels. These describe our editorial focus on the name within our framework — not a recommendation to transact:
| Signal | Editorial meaning |
|---|
| FOCUS | Name receives the most editorial attention in the current period; thesis is fully developed and actively monitored for catalysts. |
| TRACK | Name is in our active universe with meaningful coverage but is not the primary editorial focus this period. |
| MONITOR | Name is on our watch list pending additional data; coverage is light until inflection points emerge. |
| SPECULATIVE | Name is in a speculative wave (e.g., quantum); should not be considered with the same weighting as core themes. |
6. Performance disclosure
Where we cite year-to-date or cumulative performance figures (for example, on the Performance leaderboard or within wave-card commentary), the figures reflect the price return of the named security over the stated period, sourced from public market data feeds.
Important caveats on performance figures.
- Cited performance is historical. Past performance is not indicative of future results.
- The leaderboard displays our currently tracked names. Names that we covered, then later removed, are not shown — meaning the leaderboard inherently displays survivorship bias.
- We do not maintain or publish a hypothetical model portfolio's track record, and we do not claim "alpha generation."
- We do not adjust returns for taxes, transaction costs, slippage, or position sizing.
- Conviction grades and signal labels are revised over time. Reading a current grade against a historic price implies a backtest we have not performed.
7. Sources and inputs
Our research draws on:
- Public company filings (10-K, 10-Q, 8-K, S-1, proxy statements).
- Earnings call transcripts and management commentary.
- Industry conference materials (CES, Hot Chips, OFC, Computex, GTC, etc.).
- Sell-side research (publicly disclosed only) and trade press.
- Market data via Polygon.io, Finnhub, and E*TRADE APIs.
- Supply-chain and capex data from publicly available trade data and component-level disclosures.
8. Update cadence
Our intent is a regular publication rhythm:
- Weekly: Live Tracker quote refreshes and wave-status updates as warranted by significant news.
- Monthly: Conviction-grade review for each active sector.
- Quarterly: Sector report refresh following each earnings cycle.
- Ad-hoc: Material events (M&A, product launches, major catalysts) trigger published updates within 5 business days.
All published versions are date-stamped within the dashboard.
9. Conflicts of interest
Authors and affiliates of Vyzora may from time to time hold long or short positions in securities discussed in the Service. Where applicable, per-ticker holdings disclosure will accompany the relevant theme page. We do not accept payment from any company, broker, or other third party in exchange for coverage of specific securities, favorable framing, or any change to our editorial process.
10. Limitations and what we do not do
- We do not provide personalized investment advice. We do not consider any individual subscriber's financial situation, objectives, or constraints.
- We do not manage any subscriber's money, hold custody of assets, or execute trades on behalf of subscribers.
- We do not produce price targets, rating changes, or buy/sell calls in the sell-side research sense.
- We do not guarantee accuracy, timeliness, or fitness of our content for any purpose. Markets and underlying data change rapidly.
- We make no claim that following our editorial framework will produce positive investment outcomes.
11. Changes to this methodology
We may revise this methodology as our framework evolves. Material changes will be reflected by an updated version number above and noted on the dashboard. Prior versions are not retained publicly but are available on request to research@vyzora.ai.