In this week’s video, I explain why 2026 is shaping up as a year of deleveraging, not bearish, but structurally different from anything since 2007. My turbulence model is flashing signals we typically see after the S&P has already fallen, but this time they’re firing before the drawdown. That distinction matters. Combined with financials below the 200-day moving average (with the 200-day now pointing down for the first time since 2023), credit stress spreading from private markets into data center financing, and oil above $80 with Strait of Hormuz risk, the conditions of this market have fundamentally changed making short term decision making filled with risks.
On the credit side, Blackstone faced $1.7 billion in net redemptions, BlackRock limited redemptions and had a loan marked at par just three months ago now moved to zero, and the private credit reckoning that Lloyd Blankfein and Mark Rowan warned about is a growing risk as headlines seem to come out every day. This is not contained as it’s spreading into the AI infrastructure funding chain, with Oracle scrapping data center expansion plans with OpenAI and announcing thousands of job cuts amid a cash crunch.
Meanwhile, AI progress is outpacing enterprise adoption. Jensen Huang called OpenClaw the most important software release possibly ever, the fastest open-source download in history. But the disruption cuts both ways: SaaS is transitioning from a growth annuity to a technology risk asset, the computing stack is being completely rewritten, and the labor market shows zero job creation outside healthcare. The payroll print was negative, and three-month diffusion has been deteriorating since 2024. For portfolios, I favor IT services/consulting names, cybersecurity, and Palantir on the AI adoption side, while maintaining that energy and materials market caps should eventually overtake software. This is a trading year, have a plan, stay nimble, and don’t assume the old playbook works on a muddy track. The speed of AI progress if causing issues across stocks, enterprises, geopolitics, credit and soon the midterms.
Timestamps
- (00:00–03:14) This is a deleveraging year, not a bearish call, comparing conditions to LTCM and the 2007 Quantquake. Credit is becoming the dominant issue. The old playbook doesn’t work when conditions change, the “muddy track” analogy.
- (03:14–06:29) AI is disrupting everything simultaneously: software, labor, government, commodities, the capital structure itself. Turbulence model shows deleveraging signals firing before the S&P has fallen, showing a clustering of increased cross asset volatility.
- (06:29–08:27) Leverage at recent highs from prime brokerage data. VIX closed near 30, trending higher with potential to reach 50. Covariance-based strategies need to reduce leverage as correlations break down, volatility increases and financial conditions tighten.
- (08:27–11:00) Financials below 200-day MA with the moving average now turned down. Software relative to S&P after a violent decline. Oil above $80, gas up 33 cents, one of the largest weekly moves in 22 years. CPI prints in April will be very high on the headline number.
- (11:41–15:36) Credit crisis deepening: China halts diesel/gasoline exports. Blackstone hit with $1.7B in net redemptions. BlackRock limits redemptions on a loan marked from par to zero. Private credit extending into insurance and retail, Steve Eisman’s podcast on the insurance connection is a must-listen.
- (15:59–18:25) This is not a buy-the-dip environment. Oil, credit, and AI disruption are three headwinds that cannot all reverse simultaneously. Tight range at the start of the year typically breaks and the break isn’t higher with these conditions.
- (18:25–22:08) Technical picture: almost everything in a downtrend. Oversold indicator not yet triggered, breadth hasn’t panicked. Historical pattern: every major drawdown (2015, 2018, COVID, 2022) required a policy catalyst to reverse. Financial conditions are lagging but deteriorating daily.
- (22:24–25:44) S&P only down 2% for the week but momentum unwind was the worst since liberation week. Bitcoin was up as a reversal trade. Jet fuel prices in an extreme move with huge implications for travel costs.
- (26:09–28:12) Negative payroll print, zero job creation ex-healthcare. Three-month labor diffusion has been near historical recession levels since 2024. AI is absolutely impacting jobs through hiring freezes, not mass layoffs. Paper on investor implications of AI labor disruption.
- (28:50–33:12) Year-to-date: energy, staples, industrials, materials, utilities lead. OpenAI raised $110B (4x the largest IPO). Anthropic ARR growing at hyper speed, catching up to OpenAI. Jensen Huang says OpenClaw is the single most important software release, fastest open-source download in history. Viral meetups globally including 1,500 people in Shenzhen on three days’ notice.
- (33:12–40:13) The computing stack is being completely rewritten. SaaS thrived in the old stack, the new agent-driven, AI-native stack is different. Palantir is built for the new world. IT services and consulting firms (Accenture, McKinsey, BCG) will boom as enterprises need help with the transition. The shift from GPUs to ASICs is a major investment theme.
- (40:13–49:00) Oracle is the fragile link, negative free cash flow, scrapped Texas data center plans with OpenAI, thousands of job cuts. SaaS is becoming dead money (the Ford analogy: same stock price since 1990). Buybacks decelerating as cash returns as a percent of net income fall. CTA sell signals approaching.
- (49:00–56:12) Software market cap at $5.4T vs. energy $2.4T and materials $1.5T, the combination should eventually surpass software. Government involvement in AI accelerating (Pentagon, Anthropic, weapons use in Iran/Mexico). This is the most bullish case for Bitcoin: if growth assets become politically constrained, digital assets benefit. Hunker down, trade this market, and prepare for more volatility.
