Vol. I — No. 1The First IssueSummer MMXXVI
Hysteresis Research迟滞研究
--:--:-- UTC中文

Decoupled, on one side

Notes · BTC · 2026-05-05

Through March, BTC traded as a high-beta NQ proxy. Four weeks later, half of that correlation is gone — only on the upside.

Fig. — Correlation collapse

Rolling 7-day correlation, 2026-03-25 → 2026-05-05.

BTC vs equity, four weeks of correlation collapseBTC vs equity, four weeks of correlation collapse+0.0+0.2+0.4+0.6+0.803-2504-0804-1504-2204-2905-05+0.71 → +0.34BTC – NQBTC – NVDARolling 7-day correlationHYSTERESIS RESEARCH

By the count — on 2026-04-08, the rolling seven-day BTC-NQ correlation was +0.71. On 2026-05-05, it was +0.34. The BTC-NVDA pair collapsed further1 — from +0.68 to +0.21. By the conventional reading, BTC has decoupled.

The reading is incomplete. On the hours when NQ fell, BTC’s beta to it was still +0.92. The decoupling is in the other half — on hours when NQ was flat or up, BTC contributed +9.28% of independent gain over thirty days. More than half of its total return.

The mechanism is unilateral. Funding has been negative on 66% of all eight-hour periods since 2026-04-01, with a trough of −13.4% annualized. That is one-directional fuel — squeeze potential, no symmetric counterpart in equities. As long as it persists, BTC has an axis they do not.

Decoupling is not symmetric until it is.

  1. 1The BTC-NVDA pair ran from +0.68 on 2026-04-08 to +0.21 on 2026-05-05 — a deeper collapse than the BTC-NQ pair over the same window. ↩