Jordi Visser's June 2026 Market Framework

AI Rotation, Physical Bottlenecks, and the Compute-and-Energy Cycle

About Jordi Visser

Jordi Visser is the Head of AI Macro Nexus Research at 22V Research and a veteran macro investor with more than three decades across hedge funds, derivatives, ETFs, global macro strategy, and AI/digital-asset market research. His experience matters here because he brings a practitioner’s view to the AI transition: how capital rotates, how markets discount technological change, where momentum can persist, and how investors can remain calm while evidence changes. Readers who want Jordi’s first-person perspective can find him through the links below.

Executive synthesis

Jordi Visser's June 2026 framework is best understood as a rotation-and-selectivity thesis inside an AI-led market, not as a simple bullish or bearish call.

His broader posture is constructively optimistic, but not casually optimistic. Jordi's market style is pragmatic: remain calm, avoid emotional overreaction, and let empirical evidence, back-tested models, price behavior, and validation discipline guide the thesis. In that sense, his optimism is closer to applied stoicism than promotional enthusiasm.

He tends to scan the horizon for the next opportunity set, the next capital rotation, and the next area where technological change may create durable value. He also appears comfortable operating as a momentum investor across intermediate windows — roughly three to eighteen months — while still holding certain multi-year convictions. Bitcoin and Tesla are among the highest-conviction examples in the June corpus, while silver appears as a strategic commodity watch item because of its role in industrial, AI-infrastructure, semiconductor, space, defense, and electrification narratives. BSM treats those longer-term convictions as important signals, but not as conclusions to accept without ongoing verification.

That optimism is not useful if it becomes blind enthusiasm; it becomes useful when paired with evidence, timing, valuation discipline, and a clear sense of what would prove the thesis wrong.

The near-term market question is whether weakness in crowded AI, semiconductor, memory, and high-beta leadership is a normal digestion phase after a powerful run, or the start of broader regime damage. Jordi's source-attributed view across June is that the first interpretation remains plausible while breadth, credit, labor, earnings revisions, margins, and PMIs remain supportive. BSM's interpretation is more conditional: the rotation thesis is useful only if it is falsifiable.

The deeper structural point is more durable than any single week of price action. Jordi repeatedly frames the AI economy as moving from a traditional labor-and-capital cycle toward a compute-and-energy cycle. The scarce inputs are no longer only workers, wages, housing, rates, and credit. They now include electricity, grid capacity, data-center construction, transformers, switchgear, cooling, HBM/DRAM, optical bandwidth, and capex discipline. The outputs are tokens, intelligence, automation, productivity, and application-layer business value.

That shift changes the dashboard. Traditional macro still matters. But for the AI cycle, BSM also needs to track physical bottlenecks, hyperscaler capex conversion, memory intensity, token costs, model efficiency, and whether enterprise AI moves from narrative to measured ROI.

June corpus integration

The June source packet shows an evolution rather than one static thesis. A more useful reading is to ask what Jordi appeared focused on during each June source window, and why that mattered to BSM's market framework. The June 7 assessment packet is included as the first entry because it set the initial thesis architecture that the later June sources then refined, challenged, or extended.

The monthly corpus supports the public report's main arc: AI is not over, but the market is demanding more discrimination. The strongest version is not “chase every bottleneck.” It is: identify which bottlenecks are real, which are priced for perfection, which are delayed, and which convert into measurable cash-flow or productivity.

Key takeaways from the month of June

These are the key findings from the June corpus. BSM frames them as research filters for tracking macro conditions, market regimes, AI-infrastructure risk, and possible rotation signals — not as trade instructions.

1. Rotation is not the same as regime damage

Jordi's June framework treats AI weakness as potentially rotational rather than automatically bearish. BSM's test is straightforward: a market can absorb leadership rotation if credit, labor, earnings, margins, PMIs, and breadth remain intact. The thesis weakens if those variables deteriorate together.

2. The AI trade is becoming more selective

The easy phase of owning the whole AI infrastructure basket appears to be maturing into a harder layer-selection phase. Compute, memory, optical networking, power, storage, servers, models, and application-layer AI need to be judged separately.

3. Physical bottlenecks are now macro variables

Power, grid interconnection, transformers, cooling, construction delays, memory availability, and optical bandwidth can now affect revenue timing, margin expectations, and investment narratives. They belong beside rates, credit spreads, jobless claims, and PMIs.

4. Application-layer AI has to prove economic return

Jordi points repeatedly to healthcare and financials as potential application-layer beneficiaries. BSM's framing is narrower: proprietary data and workflow integration matter only if they convert into revenue, margin, productivity, or defensible ROI.

5. Verification discipline is the edge

The June corpus contains compelling ideas and several spectacular claims. BSM's public value is not repeating every claim. It is ranking them by evidence quality, identifying falsifiers, and separating what can be stated from what should be watched.

Market regime: rotation versus bear-market breakdown

Jordi's near-term framework treats the June pressure as digestion and rotation unless broader stress confirms otherwise. BSM's dashboard should test that view with market breadth, credit, labor, earnings, margins, and PMIs.

The first-pass data are mixed but generally supportive of a rotation framework:

BSM read: broad market performance, equal-weight strength, small-cap strength, and contained credit argue against a simple “AI weakness equals broad bear market” conclusion. But the sector data also warn against overstating a defensive healthcare/financials rotation on full-H1 evidence. If the final public claim is about a specific week or month, it requires the matching weekly/monthly data window.

Macro stack dashboard

The macro stack is the backbone of Jordi's non-bear thesis. The first-pass public data support a “still constructive, but not one-way” read:

BSM read: credit spreads did not show stress, PMIs stayed expansionary, and FactSet supported upward earnings revisions into quarter-end. Labor claims rose modestly, and PMIs decelerated month-over-month. The right public phrase is not “no macro deterioration.” It is: the macro stack was still broadly supportive through June, with pockets of deceleration that belong on the watchlist.

AI mid-cycle slowdown: capex, tokens, memory, and ROI

The June corpus repeatedly points to an AI cycle entering a more discriminating phase. The first phase rewarded resource acquisition: chips, land, power, memory, capital, data centers, model capacity. The next phase asks harder questions:

• Does capex convert into revenue?
• Does utilization justify the spend?
• Do token costs decline enough to support usage growth?
• Does pricing power persist under model competition and open-source pressure?
• Do gross margins remain stable as depreciation, power, memory, and networking costs rise?
• Do customers get measurable ROI from enterprise AI deployments?

This is the practical meaning of the “AI mid-cycle slowdown.” It does not mean AI demand disappears. It means the market starts separating builders, bottlenecks, beneficiaries, and overcapitalized expectations.

Electrons-to-tokens: the compute-and-energy cycle

A useful BSM formulation is:

Traditional macro remains necessary for regime-risk monitoring, but it is no longer sufficient for understanding the AI-capex cycle.

The AI dashboard should add physical and economic bottlenecks:

Named-security and asset dashboard

These rows are descriptive research context only. They are not recommendations.

Tickers and asset proxies are included as research context only. Inclusion does not imply endorsement, recommendation, portfolio suitability, or a suggested transaction.

The most striking data point is dispersion. Memory and AI-hardware proxies posted very large H1 moves, while Fluence, Bitcoin, and silver were negative. That supports the idea that AI-related narratives were not moving together as one simple basket by the end of June.

Unsubstantiated claims

Jordi Visser is often sharing opinions or reiterating subjective assessments shaped by market intuition gained over more than 30 years in financial markets, trading, investment management, and macro research. Some of those claims may be directionally useful or worth tracking, but they are not necessarily corroborated by empirical data in this report, and they are not necessarily endorsed or adopted by BSM. Claims involving JPMorgan data-center capacity, Eli Lilly's decade-end market-cap potential, Lilly compute-cluster details, Korean retail memory-stock crowding, recursive self-improvement solving energy or memory constraints, Bitcoin as the only high-quality asset, tokenization unlocking dormant assets, SpaceX/xAI hyperscaler timing, orbital data-center silver demand, silver supply deficits through 2030, chemicals as “new oil,” Fable/open-source shutdown narratives, Stripe/agentic commerce, indium phosphide, Johnson Redbook, and inflation swaps belong in BSM's watchlist or future-source queue rather than in the load-bearing thesis of this public report.

Monitoring checklist

For the second half of 2026, BSM suggests using the following as a research-monitoring checklist. These items are meant to support context-building and thesis review, not to direct portfolio action.

• Breadth: RSP, QQQE, IWM, sector leadership, 52-week highs/lows.
• Credit: HY OAS, IG OAS, credit ETFs, financing stress.
• Labor: initial/continuing claims and 4-week averages.
• Earnings: revisions, margins, concentration, top-10 versus ex-top-10 earnings.
• AI hardware: MU, SK hynix, Samsung, SOXX, SMH, HBM/DRAM pricing.
• Optical/networking: MRVL, CPO adoption, hyperscaler networking commentary.
• Power/data centers: grid interconnection, energization, storage, equipment lead times.
• Application AI: healthcare/financials adoption, revenue impact, productivity disclosures.
• Crypto/tokenization: separate lane; do not mix conviction language with AI infrastructure evidence.

Static dashboard: end-June 2026 snapshot

The checklist above is forward-looking; the table below preserves the end-June baseline snapshot behind this report. A future weekly dashboard could refresh the same categories on a slower cadence, giving readers an updated regime view without turning the site into a real-time trading terminal.

Conclusion

Jordi's June framework points to a market that may be less broken than rotating, and an AI cycle that may be less over than maturing. The strongest version is not “AI is dead” or “AI wins everywhere.” It is: the AI trade is entering a more discriminating phase where physical bottlenecks, capex ROI, memory timing, optical bandwidth, power flexibility, and application-layer economics matter more than theme ownership alone.

The useful spirit of the June corpus is constructive rather than complacent. Jordi's mindset is to keep looking for where capital may rotate next, where innovation may be underestimated, and where the market may be mispricing the next leg of opportunity. For serious market observers, opportunity may exist, but it has to be earned through due diligence, patience, humility, and a willingness to separate signal from story.

For BSM, the right posture is disciplined monitoring. Named securities should be treated as examples inside a research framework, not as instructions. The value is in separating source-attributed views, verified facts, BSM interpretation, and unresolved claims — then updating the dashboard as evidence changes.

Brandon St-Martin follows Jordi Visser because Jordi’s blend of disciplined optimism, empirical validation, tactical flexibility, and long-horizon imagination reflects the market posture BSM wants to cultivate: calm, evidence-led, and opportunity-aware without ignoring risk. The purpose is not to copy every trade or endorse every conviction, but to study how a serious investor updates views as evidence changes. BSM uses Jordi’s public work as a starting point for structured summaries, dashboards, caveats, and decision-support frameworks that help readers approach market complexity with more composure and better questions.

Final disclosure. This report is for general informational and educational purposes only. It does not provide individualized financial, investment, tax, legal, or accounting advice. It is not a recommendation or instruction regarding any security, asset, strategy, or transaction. BSM LLC is not acting as an investment adviser, broker-dealer, fund, fiduciary, or portfolio manager. Readers should independently verify market and company information and consult qualified professionals before making financial decisions.