Crypto desk — 13 Sep 2026, 19:16 UTC

My two most recent lessons split on regime direction: a flat level earns one disciplined tranche (2026-09-12), while a downward drift across several readings means hold the powder (2026-09-13 09:32, written when F&G was sliding 68→67→66, mcap $2.64T→$2.62T, breadth broadly red/Mood Soft). The 09:32 lesson explicitly said to reconsider "when the drift stops or reverses (F&G turning back up)." That is precisely what today's tape shows: F&G firmed 66→68 (Greed), total mcap recovered $2.62T→$2.64T, and book 24h is only -0.98 (-0.24%) with Mood Neutral, versus the Soft/-1.84% reading at 09:32. BTC dominance is flat at 58.8% — not the falling-dominance risk-on tell — so this is a flat-to-firming level, not a full firming. That maps to "deploy small, keep dry powder," not "deploy fuller."

Exits first: check_crypto_stops fired nothing — all six [BUY AND SELL] coins (ETH, BTC, BNB, AAVE, TON, ADA) on hard floors, none armed, none ADD-RISK. Discretionary pass found no thesis breaks: BNB +3.7% (peak +12.1%), ETH +0.2%, BTC -2.1%, AAVE -3.5% (MA still +16.8% bullish), ADA -6.8% (MA still +6.5% bullish). TON stays the lone watch item — bearish MA (-0.4%), -1.8% now, 7d -4.4% — but that is shallow chop on a $5.40 position with its hard stop far below at $1.13626; not a decisive break worth a whipsaw fee, so I held it as in the prior two runs.

For the buy, I worked the Decision Framework across the whole Coins table. Cash $15.00; caps $5/buy, $50/day (spent $0 today), ≤3 new buys/run. Candidates above their fast MA and eligible for a fresh entry included DOT (rank 50, 7d +5.2%), LTC (7d +0.1%), BNB (rank 4, but peak +12.1% and getting nearer to arming) and ETH. ETH is the highest-quality name by structure: rank 2, price 2509.63 above fast MA 2471.8285 (spread +13.1%), 7d +0.8% so not extended, and peak only +6.3% on its stop — squarely NOT ADD-RISK. On a flat-to-firming tape the disciplined move is a single quality tranche, and ETH beats the fresh alternatives on liquidity, MA spread and downside risk. News (ETH ETF inflows, "$3K back in play") is attention only — it did not drive the call; the MA screen and the regime turn did.

Bought $5.00 of ETH (0.00198923 @ 2513.53, taker fee taken in-kind). Held everything else; no stop exits, no discretionary exits or trims.

• One disciplined tranche into the highest-quality name above its fast MA and not extended (ETH, rank 2), sized small because the tape is flat-to-firming (F&G 66→68, mcap recovered $2.64T, dominance flat 58.8%) rather than fully firming — deploy small, keep dry powder.
• TON held as the watch item (bearish MA, shallow chop on $5.40) — not a break worth a fee.

Cash after: ~$10.00. Daily spend today $5.00 of the $50 cap across 1 buy. Desk-traded P&L (real fills) at review: realized $-0.53, unrealized $-4.86, total $-5.38.

The reversal of a softening drift is itself the re-engage signal: when the same readings that were sliding turn back up together, the "hold the powder" gate flips to "deploy one tranche" — you don't need a new firming tell on top, the recovery IS the tell. My 09:32 lesson (a softening regime is weaker than a flat level so hold) parked $15 while F&G slid 68→67→66 and mcap fell $2.64T→$2.62T, and told the next run to reconsider "when the drift stops or reverses." This run that condition was met exactly: F&G firmed 66→68 (still Greed), mcap recovered $2.62T→$2.64T, and book 24h improved to -0.24% (Mood Neutral) from the Soft/-1.84% at 09:32 — with BTC dominance flat at 58.8% (no falling-dominance tell). I did NOT wait for a stronger, independent firming signal; I treated the drift's reversal as sufficient and deployed one $5 tranche into the highest-quality above-fast-MA, not-extended, not-ADD-RISK name (ETH, rank 2, spread +13.1%, 7d +0.8%), keeping ~$10 dry. Generalized principle: pair the softening-hold with its mirror — hold while readings drift down together; re-engage with one tranche as soon as those same readings turn back up together (a flat-to-firming level), and let the size stay small (dry powder kept) until dominance also eases into a full firming.