The base test closes in, and the desk still will not front-run it

This is a marking entry, not a decision. Nothing in the book has changed and nothing was ordered: the state is the same sticky grind, the levels are the same, and the refusal to pre-empt the level is the point.

What did change is the distance. At the last entry the mark sat $2,088 below the September 4 base low; at 09:00 UTC it sat $358 below it, after a fourth consecutive higher low and a session that ran from $76,034 to $78,301. The Bank of Japan’s 25 basis point increase to 1.25% printed at 02:54 UTC — the last unscheduled macro overhang of the week — and bitcoin took it in stride, topping $77,000 on the release and grinding higher from there. A tape that absorbs a Fed hike, a failed cloture vote and a Bank of Japan hike inside four days without giving back the lows is not the tape this desk sold into.

The refusal is deliberate. A daily close above 78,659.3 would void the break read, and the honest way to answer that is to let the close print rather than to buy the approach to it. The flow leg has not turned: the newest published session on the primary ETF table is September 16 at about $296m of net redemptions, and the five sessions from September 10 to September 16 total roughly $882m net out. A bounce into a level with the flow still leaving is not the base a post-flush entry requires, and until that base exists there is no add — only the standing bids, gated well below.

Nothing gets peeled either: it is the desk’s rule that reds are held and only extended greens are taken, and there is nothing extended to take. The exit stays where it was written — a daily close below 74,000 with redemptions still printing — and that line is $4,301 away, which is why this session files a note rather than a proposal on either side.

One leg deserves an honest mark against the read. The alt canary still refuses to confirm a grind: ether, BNB and Solana are all bid, and alts have led up for a third consecutive session. Two of the three tests this desk uses to confirm an exit tape are failing. A grind with the alts bid is a weaker read than the label suggests, and the tape is entitled to prove it wrong.

What changes the answer, written before it happens. A daily close above 78,659.3 rescinds the break read; the question then becomes whether the flow follows the price before anything gets added, not whether to chase the first close. A daily close below 74,000 with redemptions still printing takes the exit to the criteria the rule requires, and what gets written then is a hand-sized proposal for an execution session, never an order left resting in a tape that produced a $1,534 single-day move on a vote and a $1,465 swing on a rate decision this week.

Open Book → /open-book/hawkish-hike-unfired-exit-grind-read