Every oil shock generates the same consensus response: inflation up, rates up, growth down. That framing is correct as far as it goes. It does not go far enough. The current shock is not primarily a price event. It is a balance sheet elongation event.

The Strait of Hormuz normally carries approximately 20% of global oil and LNG flows.^[1] Disruption there does not simply raise prices — it lengthens routes. Each reroute increases transit duration, ties up more cargo in motion, raises insurance premia, and forces more working capital into the physical movement of energy. The same barrel requires more balance sheet to deliver. Tanker rates from the U.S. Gulf Coast surged above $300,000 per day.^[2] Available vessels dropped approximately 41%.^[2] Crude held near $110–112.^[3]

This is not an inflation transmission. It is a liquidity drain that precedes any CPI print. Foreign official holdings fell by more than $90 billion over five weeks in NY Fed custody data,^[4] even as TIC data showed strong headline foreign buying. TIC captures the purchase. Custody captures what stays. When those diverge, the buyer is more conditional than it looks.

The market has one mental model for AI and private credit: AI is disrupting software businesses whose cash flows are the collateral inside private credit vehicles. The model is correct. It is incomplete. There are two private credit stress vectors from the same AI shock — running in opposite directions through the capital stack.

Vector one — demand side: Blue Owl capped withdrawals after Q1 redemption requests hit 40.7% in its tech-focused fund and 21.9% in its larger credit vehicle.^[6] KKR restricted redemptions.^[6] BDCs trading at approximately 78 cents on reported NAV.^[7] Dimon named it April 6.^[8]

Vector two — supply side: AI infrastructure itself is now a private credit borrower with a structural profile the market hasn't recognized. Datacenters carry large fixed costs, long amortization curves, high sensitivity to energy price variance, and financing via structured private credit. They are not software businesses. They are LNG terminals in disguise.

The full sequence was mapped on March 25 — ten days before Liberation Day fired it. Phase one has cleared. Phase two is active. Phase three has not arrived. The trigger for Phase 3 is not a Fed announcement. It is a data regime shift — consumer delinquency rate inflection, retail sales adjusted for energy inflation, and 2Y–10Y curve behavior aligning simultaneously.

Leg one — Japan: USD/JPY near 160.^[9] JGB 10Y at multi-decade highs near 2.39%.^[9] Domestic triple sell-off forcing reallocation. They don't sell because they want to. They sell because domestic conditions force the hand.

Leg two — Leveraged capital: Hedge funds at worst monthly drawdown in 4+ years.^[10] Treasury bid-ask spreads widened. Executable depth at auction shrank. UK custody proxy for leveraged basis capital is conditional, not stable.

Leg three — the one nobody is naming: Sovereign reserve managers financing longer commodity settlement cycles due to route elongation are temporarily reducing available capital for Treasury absorption. The same Hormuz disruption from Section 01 is thinning the duration bid through a mechanism that does not appear in rate models or geopolitical analysis. Three forcing functions. One simultaneous result.