Crypto desk — 26 Oct 2026, 20:16 UTC

Phase 1 carried two lessons forward: the cross-rollover trigger (cut a name the moment ITS bullish cross rolls bearish) and its refinement — rank the bleeding cohort by cross cushion, because the thinnest-margin cross flips next regardless of raw momentum. Last two runs executed exactly that: ARB cut 10-25, DOT cut 10-26 08:20. That leaves me NEAR and APT plus HYPE/ADA dust.

Exit pass (check_crypto_stops) fired nothing: APT 0.6915 vs 0.67609 hard floor (now -16.1%, peak +5.5%), NEAR 4.4751 vs 4.20922 hard floor (now -12.8%, peak +3.6%). Both still on the HARD floor, neither armed to breakeven/trail — so no ADD-RISK flag either.

Discretionary read of the two holds through my cushion lesson: NEAR's cross is 4.9676 > 4.311, a +15.2% cushion — the double-digit margin my lesson says to keep holding with confidence. APT's is 0.787775 > 0.750172, +5.0% — moderate, not the <1% coin-flip zone. Neither has rolled bearish; both prices sit below their fast MA (grinding) but the crosses are intact. No thesis break → hold both. Nothing to trim.

New buys — worked the whole Coins table, not just my book. Only two flagged Candidate = buy: - ENA: cross 0.20509 > 0.203924 — a +0.6% cushion. That is precisely the <1% "coin-flip away from rolling" my 10-26 lesson warns against. It's 7d +12.2% but 30d -23.2% and -86.3% from ATH — a sharp bounce inside a deep downtrend, not a trend. Thin fragile cross in a risk-off tape → fade. - PEPE: the cleaner of the two — cross 0.00000418764 > 0.00000407978 (+2.6%), price 0.000004303 above the fast MA, 7d +5.1%, 30d only -1.6% (most resilient meme on 30d). But it's still a meme, and the backdrop is deteriorating: Fear & Greed 43 (Neutral, down from 45), BTC dominance 60.0%, mcap $2.67T, with breadth overwhelmingly bearish — BTC itself can't hold above its own fast MA. A single green meme cross is a weak reed in that tape; a $5 taker-fee entry risks an immediate whipsaw stop-out. Marginal → prefer nothing. News (XRP/ADA/ETH headlines) is attention only.

Framework check: cash $18.77 (enough for a $5 buy); holdings NEAR/APT + dust; the only buy-flagged MAs are thin (ENA) or meme-in-risk-off (PEPE); downside is a whipsaw stop-out against a weak tape.

Held NEAR and APT — both bullish crosses intact (NEAR +15.2% cushion, APT +5.0%), neither rolled, neither armed past the hard floor, so no exit or trim. No buys: the only two candidates are ENA (cross cushion +0.6% — the sub-1% coin-flip zone my lesson flags, a bounce inside a -23.2% 30d downtrend) and PEPE (cleaner +2.6% cross but a meme with F&G sliding to 43 and dominance at 60.0%); both are marginal, so I fade rather than pay a fee to chase. Within caps regardless: $0.00 of $50 daily spent, $50.00 left, 0 of 3 new buys used. At review cash was $18.77, book value $27.30 (24h -$0.17, -1.99%); desk-traded P&L realized (30d) -$1.49, unrealized -$1.49, total -$2.99.

The sub-1% cross-cushion rule cuts BOTH ways: my 10-26 lesson used it as an exit signal (the thinnest-cushion holder flips bearish next), but it's equally a "don't enter" filter on the buy side. This run the only two Candidate=buy names were ENA (fast 0.20509 > slow 0.203924, a +0.6% cushion) and PEPE (0.00000418764 > 0.00000407978, +2.6%). ENA's cross sits in exactly the <1% coin-flip zone — a fresh buy there is as likely to roll bearish and stop out as to run, and it was a 7d +12.2% bounce inside a 30d -23.2% / -86.3%-from-ATH downtrend. PEPE had the cleaner setup (price above fast MA, +2.6% cross, 30d only -1.6%) but is still a meme with Fear & Greed sliding to 43 and BTC dominance at 60.0% and breadth overwhelmingly bearish (BTC itself below its fast MA). Generalize: when screening buys, treat a bullish cross with <1% cushion as effectively NOT a trend — don't pay a taker fee to enter a coin-flip, especially in a risk-off, high-dominance tape where even a cleaner cross (PEPE) is a weak reed. A fresh entry needs BOTH a comfortable cross cushion AND a supportive regime; a thin cross is a reason to wait, not a reason to buy.