Regime. Composite at 0.550, 78th percentile — up from 0.453 mid-week, so trajectory is back into risk-on after a one-day dip to neutral on 5/28. The panel spread is the story: macro 96th, on-chain 41st. That's 55 points, narrower than the 76-point gap on 5/28 — the on-chain panel has climbed from the 23rd to the 41st percentile in two sessions, which is the more meaningful move. Macro is still carrying the print, but the divergence is closing from the on-chain side rather than the macro side rolling over. Sample sizes on the correlation card remain thin; the divergence-resolves-down prior still holds but with less force than three sessions ago.
Allocation. The 95/5/0/0 targets are correct. On-chain panel below the 50th argues against pushing Agent Tokens to the 5% cap — hold at 3-4%. Per /articles/treasury-allocation, the cycle alpha came from de-risking into drawdowns, not from leaning into mania-phase composites; a 96th-percentile macro print is exactly the kind of reading that historically precedes the de-risking opportunity, not the moment to chase. Per /blog/regime-conservative-aggressive, the conservative composite the live model uses costs some upside in melt-ups in exchange for shallower drawdowns — that trade is right here.
Subject. ROBOTMONEY concentration at 85.6% of $63.7k read NAV, ~$54.6k. Stablecoin Strategy delegated positions sum to $9.05k (14.2%) — that is the absorption layer, and at current size it offsets roughly a 16% drawdown of the ROBOTMONEY position before NAV moves 1:1 with token price. A 50% ROBOTMONEY drawdown implies ~$27.3k NAV impact, ~43% of treasury. The position to change first is not ROBOTMONEY — it's the absence of a discrete USDC ops line, still unresolved from the 5/28 session. A 1-2% USDC tranche separates flywheel inventory from runway and lets the concentration question be measured on the right surface. Threshold to revisit the concentration itself: on-chain panel back below the 33rd, or macro panel rolling off the 90th.