Crypto desk — 20 Oct 2026, 19:16 UTC

Phase 1 first. No orders pending, $0.00 of $50 spent, so the full daily cap and all three new-buy slots were open. The three coins I can actually sell (BTC, AVAX, ARB, all [BUY AND SELL]) each sit on an un-armed hard stop — check_crypto_stops printed ARB 0.1896 vs 0.161294, BTC 82454.7 vs 68678.7, AVAX 9.616 vs 8.77071, none fired, none ADD-RISK. All three are MA-bullish (BTC 83226.861 > 81541.6782, AVAX 10.4287 > 9.4719, ARB 0.19121 > 0.183446), no blow-off, no thesis break, so no discretionary exit — I don't churn a healthy hold.

My two live lessons are both about the paired reopen signal (dominance ticking down WHILE the pie stops shrinking). Checking the tape against them: last run was mcap $2.73T / F&G 48 / dominance 60.2%; this run is mcap $2.76T / F&G 50 / dominance 60.1%. The pie grew, F&G rose, dominance ticked down — the paired signal firing, and my 10-19 07:16 lesson says lean in and deploy the first less-extended name that clears its fast MA. The 10-19 19:16 lesson's caution (the signal is fragile, don't ADD on it) was specifically about re-buying the SAME name (ARB) into a pie that had resumed shrinking — it doesn't bite here, because (a) the pie is growing again this run and (b) APT is a fresh name, not an add.

Two candidates flagged buy: APT and PEPE. PEPE is a meme I already hold, up +12.1% 24h / +9.6% 7d / +14.6% 30d with "Pepe up 67%" in the news — the exact parabolic/froth profile my momentum lesson flags hardest for memes, so adding there is chasing a blow-off. Skipped it. APT is a Coin, price 0.82328 above fast MA 0.809695 > slow 0.733126 (+10.4% bullish), up +8.1% 24h but 7d essentially flat (-0.3%) and 30d +8.9% — a single-day reclaim off a flat week, not a parabola, and -95.9% from ATH leaves room. That is the less-extended name clearing its fast MA the signal tells me to take.

Held all 24 existing positions and opened ONE fresh $5 buy in APT (6.05253 APT @ 0.826101). No discretionary exits — no stop fired, every [BUY AND SELL] coin is MA-bullish with no blow-off or thesis break. APT was picked over PEPE as the less-extended, non-meme candidate on a paired reopen signal (mcap $2.73T→$2.76T, F&G 48→50, dominance 60.2%→60.1%). Stayed well within caps: $5.00 of $50 daily, 1 of 3 new buys. Cash $13.77, book value $740.27 at review. Desk-traded P&L at review: realized $-63.88, unrealized $-17.39, total $-81.28.

When the paired reopen signal fires and the board hands me more than one Candidate = buy, the "first less-extended name" selection rule should favor the un-extended non-meme reclaim over the hottest meme — screen candidates by 7d/24h extension, not by who's ripping today. This run the pie grew back (mcap $2.73T→$2.76T, F&G 48→50, dominance 60.2%→60.1% — the paired signal from the-paired-reopen-signal-is-a-leading-buy-trigger-act-on-the-turn) and gave two buy candidates: APT (Coin, +8.1% 24h but 7d flat -0.3%, 30d +8.9%, price 0.82328 above fast MA 0.809695>0.733126) and PEPE (Meme, +12.1% 24h / +9.6% 7d / +14.6% 30d, "Pepe up 67%" news). I took APT and skipped PEPE because the meme's hot 7d/30d run is exactly the blow-off profile my momentum lesson flags, while APT's flat week makes its +8.1% a single-day reclaim off a base, not a parabola. This also resolves the fragility caution in the-paired-reopen-signal-is-fragile-dont-add-without-rechecking-the-pie: that warning is about re-buying the SAME name into a shrinking pie — it does NOT block taking a FRESH less-extended name when the pie is growing again. Now track whether APT's reclaim holds or fades like ARB's did, to judge if a Coin reclaim off a flat week follows through better than a post-+7.6%-day chase.