Market Recovery & AI Investment Strategy Post-Crash

FinanceThe AI Crash Is Over. Here's What Comes Next

TL;DR

After a historic AI crash in July, markets have recovered with a 4-day rally 6% larger than the 3-month consolidation range, signaling the bottom is in and the beginning of sustained AI-driven growth across tech, industrials, and emerging AI-native companies.

Key Takeaways

1The Nasdaq rallied 5% in a week while the thematic AI portfolio jumped 6.5%, with the S&P up 3.6%, marking the second-biggest week of the year after breaking above a 3-month pennant formation formed between May-July.
2Factor volatility has collapsed following the crash, indicating investors are now hedged; the 50-day moving average slope on 80%+ of AI thematic names is pointed upward, the hallmark of a true bull market versus a bear market rally.
3Rate-of-change signals peak at 50% over 50 trading days on the thematic portfolio; when combined with breadth breakdown inside the 100-name index, this indicates crowding and requires hedging—a tool the speaker has used since 2007 to identify inflection points.
4Palantir's commercial revenue grew 149% year-over-year (from $306M to $764M in Q2), closing 220+ enterprise deals of $1M+; this adoption rate shows AI agents are compressing economic time and disrupting Fortune 500 business models.
5Memory, not compute, is now the binding constraint: demand grows 200% annually while supply grows only 20%, creating sustained pricing power; SpaceX and Eaton confirmed this bottleneck mirrors the paper the speaker published in June.
6The Federal Reserve and Treasury intervened in the Japanese yen market for the first time since 1998, signaling the fiat system is under stress; this coordination suggests rate hikes are unlikely, making debasement trades (gold, silver, Bitcoin) attractive.
7Aggregate weekly payrolls show the weakest 6-month rate of change since 2012 (excluding COVID), labor force participation is falling, and wage growth has collapsed back to 2019 levels—indicating AI job losses are beginning despite low unemployment claims.

Notable Quotes

“The speed crash has occurred. It's over. The bottom is in as far as I'm concerned on a probability basis.”
— Speaker

“We are in acute compute shortage. And at the same time, 500,000 people, maybe 250,000 are using agentic AI. What happens when we go from 500,000 to 1%, to 100 million to 500 million? There will never be enough compute ever.”
— Gavin Baker (referenced)

“Nvidia, as we record this, is at the lowest forward PE of the last 10 years... not a single instance of deceleration. Nothing. In fact, every metric is accelerating.”
— Gavin Baker (referenced)

“The memory output is increasing by around 20% a year. The demand is increasing by 200% a year, maybe higher. If you've got demand increasing faster than supply, economics 101 would suggest the price increases.”
— Elon Musk (referenced)

Chapters

1. Market Recovery & Technical Setup

After a historic July crash with factor volatility at 45-year highs, the market rallied 6% over 4 days—larger than the entire 3-month May-July consolidation range. The S&P gained 3.6% (Q's +5%), with all sectors positive year-to-date; the Cosby is up 61% and tech +116%. The 50-day moving average slope is up 80%+ on AI thematic names, confirming a real bull market, not a bear trap.

2. Hedging Framework & Rate-of-Change Warning System

The speaker built turbulence models and breadth-based hedging signals because factor volatility (which includes long-short leverage components) no longer correlates with index volatility. A 50% rate-of-change over 50 trading days on the 100-name thematic portfolio signals crowding; when breadth deteriorates, hedging is required. This system generated recession warnings in September 2007 and identified Micron's July peak.

3. Palantir's Adoption Explosion & AI Agent Impact

Palantir's commercial revenue surged 149% YoY to $764M (from $306M), closing 220 deals of $1M+, 98 of $5M+, and 78 of $10M+. This validates AI-agent adoption is accelerating and compressing economic time; new business formations have gone parabolic since Opus 4.5, with million-dollar companies operating as one-person teams using AI agents, disrupting Fortune 500 employment models.

4. Memory Bottleneck & GPU Economics

Memory demand grows 200% annually versus 20% supply growth, confirmed by SpaceX's Elon Musk and the speaker's June paper. GPU rental for B200 clusters doubled from $2/hour to $4/hour in 7 months; Nvidia's forward PE is at 10-year lows despite accelerating demand across all metrics (GPU availability, rental pricing, DRAM spot prices).

5. Hyperscaler Capex & Structural Risks

Hyperscalers (Microsoft, Meta, Amazon) are spending heavily on AI compute but free cash flow won't turn positive until far into the future. While cloud revenues exist, they won't sustain capex; CDS spreads widening is justified because eventual compute surplus (from algorithmic breakthroughs or hardware shifts) could compress multiples. Google's loss of Jeff Dean and Demis Hassabis signals talent frustration with bureaucracy, driving a shift toward private AI-native companies by 2030.

6. Labor Market Deterioration & AI-Driven Displacement

Aggregate weekly payrolls show the weakest 6-month rate of change since 2012 (excl. COVID); labor force participation has collapsed, wage growth is back to 2019 levels (3.2% average hourly earnings, Atlanta Fed wage at 2019 baseline). Only healthcare jobs were created in the last 18 months. AI agents haven't yet caused mass firings but are suppressing new hires; as million-dollar solo-founder companies grow, larger firms will shrink, triggering worker recycling.

7. Fiat System Stress & Fed/Treasury Coordination

US intervened in the Japanese yen market (first coordinated move since 1998) to prevent BOJ from selling Treasuries to fund yen purchases. This signals both the fiat system and Treasury market are under strain. Kevin Worsh gave no rate-hike guidance; Fed odds are 45% for September despite payroll weakness, suggesting the fiscal train cannot stop and debasement is the path forward, benefiting gold, silver, and Bitcoin.

8. Inflation Metrics & Rate-Hike Skepticism

Trueflation (which weights recent prices more heavily) has collapsed violently over the last 4 weeks while headline CPI core is flat; the speaker argues economists are ignoring superior data. With labor force shrinking and wages flat, traditional inflation metrics are misleading. 45% Fed rate-hike odds in September seem disconnected from weak payroll and labor data, favoring the debasement thesis.

9. Gold/Silver Breakout & Bitcoin Setup

Gold broke above the 50-day moving average for the first time since March (third longest stay below); it rallied 7.2% in one of the largest printing-press weeks ever. Silver and platinum also broke above their 50-day moving averages. Bitcoin remains in a bear market setup below key resistance, but the ecosystem (measured by a 40-name crypto index) is holding well despite Clarity Act setbacks and Cold Card exploits.

10. AI Agents & The Knowledge Economy

The speaker's college-age son (not a CS major) built a personal AI chief of staff that spins out sub-agents for his internship; he prefers ChatGPT, finds Gemini Coding 'insane.' At 250,000 agentic AI users globally (of 8 billion), scaling to 100M–500M users will create exponential compute demand. Agents are tireless workers; once deployed, demand for compute is unstoppable and will never plateau.

Key People & Entities

Gavin Baker: Investor with deep AI industry knowledge; argues the AI selloff is disconnected from quantitative data, sees no deceleration in GPU demand or hyperscaler capex
Kevin Worsh: Federal Reserve Chair; gave no rate-hike guidance, sparking market confusion; faces pressure to support the Japanese yen and Treasury market rather than tighten policy
Scott Bessent: US Treasury Secretary; coordinated with BOJ to intervene in yen market, first coordinated move since 1998; attempting to prevent bond-market stress
Elon Musk: SpaceX founder; confirmed on earnings call that memory (not compute) is the limiting factor for AI adoption, with demand growing 200% annually versus 20% supply growth
Jeff Dean: Google AI researcher; recently left Google, signaling talent frustration with bureaucracy at large public companies restricting AI innovation
Demis Hassabis: DeepMind co-founder; shifted roles within Google, reflecting organizational tensions between free innovation (Alphafold) and commercial AI revenue requirements
Nvidia: GPU manufacturer; forward PE at 10-year lows despite accelerating demand for AI chips; B200 rental prices doubled in 7 months
Palantir: AI/data analytics software company; commercial revenue grew 149% YoY to $764M; closed 220+ enterprise deals of $1M+, validating rapid AI adoption

Glossary

Factor Volatility: Volatility measure that includes long-short leverage components; it diverges from index volatility (S&P VIX) during deleveraging events and signals whether funds are hedged or holding gross leverage.
Breadth: The percentage of stocks in an index trading above their moving averages; breadth breakdown (deterioration) signals weakening participation despite price gains and is a hedging signal.
Rate of Change (ROC): A technical indicator measuring the percentage change in price over a fixed period (e.g., 50-day ROC); extreme ROC levels (>40-50%) indicate crowding and overbought conditions requiring hedging.
Pennant Formation: A chart pattern where price consolidates in a narrowing range after a sharp move; a breakout from a pennant is usually followed by continuation in the original direction with accelerated momentum.
Gross Leverage: The total notional exposure in a portfolio (long + short combined), which multiplies both gains and losses; high gross leverage increases vulnerability to volatility spikes and requires deleveraging after crashes.
Agentic AI: AI systems (like ChatGPT with plugins or sub-agents) that autonomously perform tasks, make decisions, and iterate without human intervention for each step; they require continuous compute and memory resources.
Thematic Portfolio: A diversified 100-name index covering AI adoption across sectors (tech, industrials, utilities, chemicals, power, construction); it spans global markets and excludes concentrated indices like the Nasdaq.
Trueflation: An inflation metric that weights more recent prices more heavily than historical prices, providing a real-time inflation signal that differs from traditional CPI calculations.
CDX: Credit Default Swap index measuring the cost of insuring corporate debt; widening CDS spreads indicate rising default risk; the speaker notes hyperscaler CDS are widening due to massive capex commitments.

Explore