The $87k move cleared the door, not the room

As of 22 September 2026, 3:06 p.m. Central / 20:06 UTC, Coinbase BTC-USD last traded at $86,276. The current UTC daily candle is still open. The last completed daily, 21 September, closed at $86,595 after reaching $87,397.

The move in one minute

• Bitcoin first cleared $81k intraday on 18 September; the first completed close above $81.5k came on 21 September, not on the first push.
• ETF demand helped: Farside shows +$1.59B across the reported sessions from 17–21 September, including +$999M on 21 September. But the tape also carried a short squeeze and rising leverage.
• The desk grades this partial confirmation. $86k was crossed and then traded below intraday on 22 September; one more completed daily close and better-quality demand evidence would turn the level into support.

What the original thesis said

The earlier map was conditional. The $77k–$80k stall was not a confirmed floor. The failed $81k–$81.5k push remained resistance until price could close above it and hold. Cash remained valid until a decisive daily close made the direction observable. A clean break above the May-high area, backed by spot demand rather than excessive leverage, opened the road through the $81k–$86k supply zone. This was a scenario map, not a dated prediction.

What actually happened

Coinbase’s UTC candles show the sequence. On 18 September, Bitcoin traded to $81,388 but closed at $80,875. It closed at $81,234 on 19 September and $81,160 on 20 September: above $81k, not above $81.5k. The decisive step arrived on 21 September, with a high of $87,397 and a close of $86,595. From the 17 September close of $76,349, that was about 13.4%. Coinbase candle data

That answers the daily-close question with an important limitation: as of this letter’s retrieval, there is one completed daily close above $81.5k. The 22 September candle is open. The original two-close rule is therefore not fully satisfied. Bitcoin has demonstrated acceptance above the old band once; it has not yet demonstrated that the acceptance survives a second completed UTC session.

The move was a combination, not a single-cause victory lap. Farside’s US spot ETF table records outflows of −$450.4M on 15 September and −$295.9M on 16 September, then inflows of +$159.5M on 17 September, +$433.0M on 18 September, and +$999.0M on 21 September. That is supportive vehicle-level demand, but “persistent” is too strong after one outsized session and a weekend gap. Farside Investors’ Bitcoin ETF flow table

Derivatives added fuel. OKX’s BTC-USDT perpetual open interest rose from $2.14B on 15 September to $2.53B on 18 September and $2.66B on 20 September, then eased to $2.61B on 21 September. Average realized funding stayed positive, moving from 0.0048% per eight hours on 15 September to 0.0100% on 19 September and 0.0071% on 21 September. Leverage participated, but these rates are not, by themselves, evidence of an extreme long crowd. OKX open-interest history · OKX funding history

Short liquidations were likely the spark. CoinDesk, citing CoinGlass, reported $746.6M of crypto liquidations in a 24-hour window on 21 September, including $647.9M of shorts and $277.5M of Bitcoin. The Block reported a similar broader total but $360.7M in Bitcoin positions. The windows differ; the shared signal is more useful than false precision: a large short book was forced to buy into the break. CoinDesk · The Block

Graph 1 interpretation. $86k acted as a level price moved through cleanly on 21 September, then as a live retest on 22 September—not yet as proven support. The current candle has traded below $86k and recovered. A final close below it would make the move look more like a fast range escape than settled acceptance.

Graph 2 interpretation. The original spot-demand condition was not wrong; it was simply not sufficient as a one-variable explanation. ETF flows improved before the largest price extension and were positive on the breakout day. The desk can grade spot demand as supportive, not yet persistent enough to carry the whole thesis without help from forced buying and derivatives positioning.

What the desk got right

The resistance rule held until it did not. The $81k–$81.5k band was tested, rejected, retested, and then cleared. Cash was valid while closes were missing. The path through supply did open once acceptance arrived. Separating a conditional map from a prediction survived contact with the tape: the level was useful even though the timing was not forecast.

What the desk missed

The framework under-modeled the short book. It treated “spot demand versus leverage” too much like a pass/fail gate, when ETF demand can improve, shorts can be forced out, and new leverage can chase momentum in sequence. It also treated $86k as the end of the supply map when price crossed it in one session. The missing variable was the speed at which positioning changed once $81.5k stopped attracting sellers.

The map from here

Bull case. Two completed UTC closes hold above $86k, ETF inflows persist beyond the $999M session, and OI/funding stop accelerating. That upgrades the move from squeeze-assisted to broader acceptance. Base case. Price holds the $81.5k–$86k box while flows and leverage normalize; the breakout remains unproven, and cash remains valid while evidence matures. Failed-breakout case. A close below $86k followed by loss of $81.5k/$80k, renewed ETF outflows, and weak spot participation would put the prior $77k–$80k range back in charge.

Scenario	Trigger	Confirmation evidence	Invalidation	Implication
Bull acceptance	Two UTC closes >$86k	ETF inflows persist; OI stabilizes; funding cools or stays orderly	Daily close <$86k, then loss of $81.5k	The $87k move becomes a supported breakout rather than only a squeeze
Base / consolidation	Hold $81.5k–$86k	Mixed flows, lower liquidation intensity, repeated closes inside the box	Two closes <$81.5k or renewed heavy ETF outflows	Acceptance is pending; observe and preserve optionality
Failed breakout	Close <$81.5k, then <$80k	Spot volume/CVD weakens while price loses the band	Reclaim $86k with flow repair	Previous range returns; the $77k area becomes a test, not a forecast

Final conclusion

The move to approximately $87,000 partially confirms the original hypothesis. Bitcoin cleared $81k, then printed its first completed close above $81.5k and moved through $86k. ETF flows supported the move, but short liquidations and derivatives positioning supplied torque. The next fact is whether $86k survives a completed daily close with ETF demand present and leverage no longer doing all the work.

Sources

• Coinbase Exchange BTC-USD ticker and UTC daily candles — live price and price pathway; retrieved 22 September 2026 at 20:06 UTC.
• Farside Investors — US spot Bitcoin ETF flows — daily net flows in US$ millions; reported sessions through 21 September.
• OKX contract open-interest history and funding-rate history — venue-specific positioning; retrieved 22 September 2026 at 20:06 UTC.
• CoinDesk / CoinGlass liquidation snapshot and The Block liquidation snapshot — secondary reports with different windows; used directionally and flagged as non-identical.

Educational-not-financial-advice disclaimer. This letter is educational scenario research, not personalized financial advice. It does not recommend buying, selling, leverage, or any position size. Levels are observation points, not instructions; cash is a valid position when confirmation is absent.