AI’s Speed Crash: The Structural Bull Market Isn’t Over
AI’s Speed Crash: The Structural Bull Market Isn’t Over
YouTube44 min 59 sec
Watch on YouTube
Note: AI-generated summary based on third-party content. Not financial advice. Read more.
Quick Insights

Investors should view the current tech volatility as a "speed crash" rather than a bear market, using the 40% correction in momentum to scale into high-conviction semiconductor names. Micron (MU) is a top-tier buy as rising DRAM prices and structural memory shortages drive significant upward earnings revisions. Avoid legacy software providers like Salesforce (CRM) and Adobe (ADBE), as their downward-sloping 200-day moving averages signal they are being disrupted by the shift to AI agents. Ethereum (ETH) represents the purest play on the "agentic" revolution, as AI agents will likely use its smart-contract infrastructure and tokens as native currency. Expect a major breakout for Bitcoin (BTC) and the broader crypto sector in late Q3 or Q4 as global regulatory shifts and institutional adoption provide long-term tailwinds.

Detailed Analysis

AI Infrastructure & Semiconductors

The current market volatility is described as a "speed crash" rather than a structural end to the bull market. The rapid decline in AI stocks is attributed to an unwind of extreme momentum and leverage, driven by non-emotional AI trading agents and "volatility control" funds.

  • Micron (MU)
    • Sentiment: Bullish (Long-term).
    • Context: Despite a "head and shoulders" technical pattern that scares many traders, the analyst is buying back in.
    • Key Drivers: DRAM prices are rising across all generations (3, 4, and 5) due to a structural shortage. Recent earnings estimates were revised upward dramatically, and new long-term contracts are considered "game changers."
  • NVIDIA (NVDA)
    • Sentiment: Neutral/Cautious (Short-term).
    • Context: The stock has seen a massive 12-fold increase, but recent earnings reactions have been negative even on "beats." It may face a period of "digestion" similar to its 2023-2024 consolidation.
  • Marvell (MRVL)
    • Sentiment: Bullish.
    • Context: It remains the analyst's largest semiconductor position despite a sharp recent decline.
  • Legacy Software (Adobe, Salesforce)
    • Sentiment: Bearish.
    • Context: These stocks are described as "toast" because their 200-day moving averages are pointing down. They face significant disruption from AI agents and must pivot their entire business models to survive.

Takeaways

  • Buy the Digestion: The "firework show" of easy gains is over, but the "agentic AI build-out" has just begun. Use the 40% correction in tech momentum to scale into high-conviction names like Micron.
  • Focus on Memory: "AI equals memory." The shortage in compute hardware is fundamentally a shortage in memory capacity.
  • Watch the 200-Day Moving Average: Avoid companies where the 200-day moving average is sloping downward (like CRM or ADBE), as these are likely being disrupted by new AI startups.

Cryptocurrency & Tokenization

The analyst views crypto as the "purest AI trade" because it cannot be disrupted by AI; instead, AI agents will use crypto as their native currency and collateral.

  • Ethereum (ETH)
    • Sentiment: Highly Bullish.
    • Context: Outperforming Bitcoin recently. It is viewed as the primary beneficiary of the "agent revolution" because agents require smart contracts to function.
  • Bitcoin (BTC)
    • Sentiment: Bullish (Long-term).
    • Context: Described as the "collateral of the future." While currently in a choppy range, it is expected to be a major "beta" play over the next 12 months.
  • PayPal (PYPL)
    • Sentiment: Bullish (Crypto Context).
    • Context: Included in the analyst's "Crypto 40" index. The interest from Stripe in the stablecoin space highlights PayPal's potential value in the emerging stablecoin infrastructure.

Takeaways

  • Prepare for the "Agentic" Shift: As billions of AI agents are deployed, they will transact in tokens, not fiat. This makes Ethereum a critical infrastructure play.
  • Monitor Global Adoption: Significant regulatory shifts in Japan and South Korea (reclassifying Bitcoin as a financial asset) are massive long-term tailwinds that the market is currently ignoring.
  • Q4 Timeline: Expect the real breakout in crypto to be a late Q3 or Q4 story.

Investment Themes & Market Structure

  • The "Speed Crash" Phenomenon: Markets are moving from "manias and panics" to "bubbles and speed crashes." Because of AI-driven trading, moves higher are more parabolic, and moves lower are faster and more violent.
  • Insatiable Compute Demand: Disregard "bear" arguments about a peak in AI spending. The demand for compute to train and run "agentic" AI is exponential and will not be met within the next four years.
  • The "Vera Rubin" Shift: A new era of AI hardware (referencing the 800-volt DC/Vera Rubin architecture) will create a "step-up function" for a specific group of 25 companies involved in power and cooling infrastructure.
  • Economic Strength: Despite the tech sell-off, the broader economy remains strong. S&P 500 earnings surprises are averaging 16%, and sectors like Banking (JPM, GS) are hitting all-time highs.

Takeaways

  • Ignore Recession Fears: The best recession signal (S&P 500 year-over-year returns turning negative) is nowhere near triggering.
  • Diversify into "Physical" AI: Look beyond just chips. The "Vera Rubin" transition suggests opportunities in power management and industrial AI infrastructure.
  • Don't Fear the Volatility: High "volatility of volatility" is the new normal. Use "VWAP" (Volume Weighted Average Price) on the intraday charts to identify when big institutional selling is exhausted.
Ask about this postAnswers are grounded in this post's content.
Video Description
For More Information Visit: https://ai.22vresearch.com/ Email Mark Whaling: mwhaling@22vresearch.com In this week's video, I walk through a brutal week for AI stocks. Weeks ago, I talked about how AI was changing the human driven market structure in a rising world of AI agents. Markets have shifted from Kindleberger's "manias, panics, and crashes" to bubbles, parabolas, and speed crashes, faster up, faster down, and structurally different, driven by momentum, agents, and leverage stacked on top. Like with gold and silver earlier in the year and Bitcoin back in October, we are witnessing the deleveraging of a trend. The Morgan Stanley tech-momentum factor fell roughly 40% in 18 days, the worst unwind in data going back to 1998. Despite the extreme nature of this move, that's the digestion phase I flagged in my June 4th paper, agentic AI moving from discovery to digestion after the Opus 4.5 run. The most levered funds are being eradicated, the odds on the tote board are finally “fair” for many names. Underneath the noise, the early earnings season is blowing the doors off again, the economy is strong, inflation is falling, and I think the next leg of beta lives in crypto. Timestamps • (00:00–02:42) Setup & market structure: A rough week for AI, but the structural bull and insatiable compute demand are intact. Markets have shifted to bubbles, parabolas, and speed crashes, not a business-cycle problem. • (03:00–06:11) The megaphone unwind: MS tech-momentum factor down ~40% in 18 days, worst since 1998. The "firework show is over", agentic AI moves from discovery to digestion; ~40% of the Opus 4.5 move retraced. • (06:11–07:50) Doing little buys: Adding back Micron on new news; even blowout growth (Micron, Samsung, ASML) can now disappoint. The 100-name thematic index held better than single names. • (08:14–11:11) Earnings blow the doors off: 43 S&P names in, sales surprise +4%, earnings +16%. JP Morgan +34% (stock +10% at highs), Goldman +45%. Dimon: "close to as good as it gets," strength across the board. • (11:11–13:36) Ignore the debt scare: Hyperscaler debt-to-equity near zero, RPOs enormous, compute insatiable, leaders protected by government in the China race. Feels like '07 but no widening spreads, no recession. • (13:36–18:35) Agents & the new vol regime: More agent-driven trading (Robinhood open to agents), record factor vol, a Bank of England warning. Gold, silver, and Bitcoin all saw violent momentum unwinds. Tech-momentum vol ~7x S&P vol makes re-entry hard for leveraged/vol-controlled funds. • (19:47–25:46) AI = memory, 200-day still up: 85% of portfolio names have a rising 200-day, unlike toast software (Adobe, Salesforce). Memory shortage, rising DRAM, Micron contracts a game-changer; Nvidia/Micron head-and-shoulders analogy, Micron ~6x cheaper. • (27:38–30:24) Recap = rotation, not systemic: S&P −1.5%, Nasdaq −4%, small caps flat. Industrials and tech momentum simply mean-reverting. Thematic portfolio ~40% correction (normal); Marvell −62%, still top semi after Nvidia; proceeds rotated into Bitcoin, Ethereum, silver. • (30:50–34:06) Strong economy, falling inflation, reformist Warsh: Philly Fed/Empire ~62 PMI; core CPI/PCE and sticky-core to new lows. Warsh is "reform, not hawkishness" — weight forward-looking data over one backward-looking print. • (34:06–37:47) Hassabis & the singularity: AGI "probably only a few short years away," a shift on par with fire or electricity. Open source eventually dominates enterprises; AI-native companies win, incumbents lose. • (37:47–44:34) Vera Rubin step-up & the next beta: The 800V DC / Vera Rubin paper (~25 names) targets earnings step-up functions. Bitcoin still in a bear market; Ethereum the one to watch (best month, +17% MTD). Stripe's PayPal bid, JPM/BlackRock/Goldman tokenizing, Japan and South Korea reclassifying crypto, crypto as the purest AI trade, with a Q4 breakout the signal.
About Jordi Visser
Jordi Visser

Jordi Visser

By @jordivisserlabs

Empowering seasoned professionals to navigate the future of finance, technology, and AI. What We Offer: - Cutting-edge ...