four wins and a stack of new fires

Morning — yesterday was the kind of day where the desk closed five trades, four of them green, but the story is really about what just opened.

Four winners closed out yesterday: JTO, NEAR, PENDLE, and WLD all hit their +3% targets in the range-short playbook, which is the bread-and-butter trade here — small, orderly, and consistent when the tape cooperates. That's +3.74%, +3.49%, +3.01%, and +3.11% respectively, closed in 19 to 75 hours. One loss: a SHIB dip buy that ran for 24 hours and gave up −6.34%, hitting the three-day max hold before climbing back. The dip-buy strategy aims for quick +3% to +8% moves with no hard stop — you're catching sharp down moves that reverse — but SHIB just wasn't it this time.

The more important thing: six new positions fired in the last day, and three of them are stacked repeats. JTO re-fired its dip-buy signal just 12 hours after closing the range short (0.5 days open now). SHIB fired twice as a dip buy within 15 hours of each other (0.4 days open), then the autonomous bot fired it again five minutes later — three separate buys on the same coin, each on independent logic, each aimed at catching different dips in real time. LDO also came back into the book as a long (range short variant) just 1.2 days after the short stopped out. The backtest read here is clean: when bitcoin moved only +1.96% between the two LDO signals and 0.0% between the SHIB repeats, those second and third entries historically ran slightly better than fresh ones.

That leaves the book with 13 open positions: one deep bounce short on LDO that's 12.5 days in (one day from timeout), a mix of range shorts on PENDLE, PEPE, INJ, and JTO that are all fresh (under three days), and multiple dip buys on SHIB and JTO hunting for quick reversals. It's not chaos — it's just what happens when independent signals agree about the same coins at nearly the same time.

the market from here

Bitcoin is sitting just barely above its 50-day average at $63,364, up +0.1% from the 50-day line. That's as middling as it gets: the tape isn't decisively below any major timeframe's open (it's only −0.52% from today's open), but it is −12% from the 200-day average and −5.3% from the 30-day high at $66,907. Ethereum is in a similar zone, 6 weeks into a move up that's cooled — it's above its 50-day average by 6.9%, but −12.1% from the 200-day average. Both RSIs are middle-neutral: bitcoin at 46.3 and ethereum at 53.7.

The key levels tell the fight: $63,413 (the 50-day average, week low, and today's open all stacked) is acting as a floor right now, and if price rolls below that it has a clear downside path to $58,532 (monthly open) and $57,886 (the 30-day low). Topside, $65,510 (weekly open and yesterday's high) is the first real resistance, then $66,907 (month's high, week's high), then the 200-day at $71,995 is where the longer trend lives.

Yesterday's order flow was seller-heavy across both majors — bitcoin saw the most aggressive seller action of the last ten days (−9.5% of volume), ethereum the same (−5.2%), and PENDLE one of the book's positions was crushingly seller-heavy at −14.7%, the most in its last ten days. But AAVE (also on the book) was the only real buyer showing up, hitting +6.5% aggression, the most in its recent window. So the tape is mixed: macro sellers pressing, but pockets of buyer defense on certain coins.

Futures positioning leans short on both majors — shorts are paying longs funding on both bitcoin and ethereum, which means the crowded side is the short side. That's a squeeze risk if price rallies hard, and you can feel it in the data: bitcoin's open interest just ticked up +1.2% yesterday and +4.3% over the week, meaning new shorts are stacking in on a down move. That's classic crowding. If you've watched a short squeeze before, you know it starts slow and ends sharp.

The desk is reading this as a tired seller's move that hasn't quite committed to rolling over.

one thing worth keeping

Repeat signals — when the same coin fires twice in a short window — sound like a glitch but they're not. The backtest covered this: independent rules running on independent clocks can agree. And when they do agree without an interim pump, they actually perform slightly better than fresh signals, because the coin has confirmed the setup twice. That's not luck; that's validation. JTO and SHIB both proved it yesterday.

the book right now

closed in the last 24h

open positions

all-time