After — days in Capitulation, Bitcoin has broken above the range that Note No. 2 identified as a typical Phase A structure and is approaching a possible transition into Repair. Only two completed Repair transitions exist in the post-2018 record, and neither moved supply back into profit this quickly. The rally has lifted a large block of supply above water. 68.5% of circulating bitcoin now sits above its cost basis, up 16.9 percentage points in one week.
The breakout had been building beneath the surface. Months of accumulation placed a dense band of supply below the market. Washington also provided a more constructive policy backdrop, with Treasury advancing stablecoin rulemaking and the SEC proposing tailored exemptions for crypto markets. A record short squeeze in perpetuals supplied the final acceleration. Much of that buying was forced, not evidence of broad demand. On the peak liquidation day, Hyperliquid generated short-liquidation volume at roughly seven times the share implied by its open interest. The price move was broad. Its fuel was concentrated.
The breakout is constructive, not conclusive. This issue first separates the forced buying from the underlying move, then turns to the on-chain signals that would show voluntary demand taking over. Those signals will decide whether the market completes the transition into Repair or the run-in fails with Capitulation still intact.
Last week's compression broke upward. Price has moved. The regime has not.
Note No. 3 left Bitcoin trapped in a volatility squeeze, with the balance of risk leaning toward a break into Phase B of Capitulation as macro vulnerability met a holder base already under pressure from time. The squeeze resolved higher instead. Over the seven days to today, spot gained — and cleared both cost-basis levels the value-regime framework uses to separate Capitulation from Repair. Price first reclaimed the Short-Term Holder cost basis at —, then pushed through the True Market Mean at —.
The framework still reads Capitulation. As Note No. 1 established, spot does not trigger the regime change. The 1d–1w cost basis at — does. It measures the average price paid by coins that moved within the past week. Repair confirms only when this recent-buyer cost basis rises above the Short-Term Holder basis and clears the model's deadband. It still needs to climb roughly — from today's level to reach a threshold near —.
Earlier Repair labels confirmed only after spot spent most of a week above the Short-Term Holder basis. The delay is mechanical because the confirming cohort covers one week. Price must hold while higher transaction prices enter the average and newly profitable buyers decide whether to sell into strength. If the recent cost basis settles above the Short-Term Holder basis, higher prices are being accepted rather than immediately rejected. Recent buyers also move further from break-even, reducing the pressure to exit there.
Until that happens, the regime remains Capitulation, now on day —. With spot already trading beyond the threshold while the confirming cost basis still catches up, we describe the current state as Capitulation leading Repair.
Price has cleared the level associated with Repair. The confirming cost basis has not. Time now carries as much weight as price.
After — days in Capitulation, spot broke from a volatility compression often seen in the first half of these episodes. It reclaimed the Short-Term Holder cost basis, then pushed through the deadband above it near —.
Price is now leading the regime signal. Spot (—) sits roughly — above the threshold the 1d–1w cost basis (—) still needs to reach. Holding this zone allows higher transaction prices to enter the average and gives the ledger time to confirm the move. Losing it would leave another failed escape from Capitulation behind.
Three questions carry the rest of the analysis. What drove a move of this size? What changed on-chain when earlier episodes reached Repair? Can the market hold long enough for the transition to complete?
Washington provided a more constructive policy backdrop on 19 August. At a 2:30 p.m. ET White House meeting with technology and crypto leaders, President Trump urged Congress to advance the CLARITY Act. At the same time, an already fragile derivatives market began to break.
Perpetual venues force-closed $1.3B of shorts against just $48M of longs, the largest daily short-liquidation event on record. The next two days added waves of roughly $430M and $720M, helping carry spot through $79.5K.
A short liquidation is a market buy without a valuation decision behind it. The position must close, so it takes the available offer. That purchase lifts price and moves the next short closer to its liquidation threshold, allowing one forced close to trigger another. Futures can then run ahead of the underlying market. Index pricing, arbitrage, and dealer hedging transmit that pressure into spot, dislocating price until the liquidation chain is exhausted or voluntary sellers absorb it.
A liquidation cascade begins with positions accumulated before the move. During a long, compressed range, open interest can rise while price barely moves. The surface looks calm, but a growing leveraged book leaves more positions exposed to the same break.
Once price turns against the crowded side, liquidations unwind the structure. Open interest contracts, compulsory market orders accelerate the move, and each close can pull the next threshold into range. A larger and more leveraged imbalance gives the cascade more fuel.
The imbalance was most visible on Hyperliquid. From 16 August, the venue accounted for 40.6% of all short liquidations and peaked at 96.2% of one four-hour bucket, despite representing only 6.2% of perpetual open interest. Hyperliquid was not carrying the market's largest book. It was carrying the most fragile one.
Open interest reveals the build before liquidations reveal the side. Hyperliquid's book grew 31.2% into 17 August against 7.2% at Binance and 4.5% across the whole market, then gave back 20.7% to finish 4.0% above where it started. No other venue moved like that. Open interest does not name a side. The liquidation split does, and the positions that came off were disproportionately short.
Open interest measures the size of the book. Funding shows which side is paying to hold it. Over the past two weeks, funding across most major venues traded below the 10.95% annualised neutral level, consistent with persistent taker selling and a growing short bias.
At the peak of the open-interest build, Hyperliquid funding fell to −10.0% annualised, while Binance held at +3.7% and the broader market sat near zero. Negative funding means shorts pay longs. Hyperliquid traders were leaning much harder against price than the rest of the market before the breakout.
Once liquidations swept through the market, that pressure reversed. Funding across most venues has since returned toward the 10.95% baseline, while Hyperliquid has overshot it six times in the past two days, peaking at +25.2%. Binance, by comparison, has remained close to neutral.
The leverage risk has changed sides. The short crowd that helped fuel the breakout has largely been cleared. Aggressive long positioning is now rebuilding beneath price.
Hyperliquid's liquidation heatmap shows where the rebuilt leverage is vulnerable. It maps price zones containing clusters of positions near forced liquidation. Reaching one of those zones can turn latent leverage into market orders.
During the range, a dense short-side liquidity band formed between $66.1K and $67.7K and remained in place for roughly nine days. It was cleared on 19 August, while fresh short clusters were repeatedly squeezed as price pushed through $79.5K.
The structure has since flipped. New long-side liquidation pools have formed below spot, with the nearest concentrated zone around $74.0K to $75.6K. A slide toward the Capitulation-to-Repair threshold would meet these shelves first. Forced selling there could accelerate the decline, mirroring the mechanism that drove the short squeeze.
Derivatives explain how the market escaped the range. They cannot show whether it will stay out. With much of the forced bid now spent, the burden shifts to voluntary demand and the ledger.
The squeeze explains the launch. The ledger will decide whether the move can mature into Repair.
This Capitulation episode has not entered its typical Phase B. All three prior episodes eventually broke into that deeper stage, where the most severe damage followed, usually alongside a negative macro shock. The current episode briefly reached −20.4% on 30 June, but the break did not hold. By 21 August, spot had recovered to close 7.0% above the Capitulation entry level.
The path has been different so far. Repair still requires the 1d–1w cost basis to clear the Short-Term Holder basis and stabilise above it. At current levels, that translates into spot holding above roughly —.
The chart places the current move beside the final two weeks of the earlier Capitulation episodes and the Repair regimes that followed. The comparison contains only two completed Repair transitions, so it frames the path rather than forecasting the outcome. The current rally has so far followed a similar route into the threshold. Voluntary demand must now absorb spending from newly profitable holders and support price long enough for the cost-basis signal to confirm.
The complete post-2018 record fits in two rows. Both episodes rallied into Repair confirmation, then spent time inside broad ranges. In 2019, price held within a roughly 16-point range over the following month. In 2023, it moved through a wider 28-point range over eight weeks. The current run has gained +22.7% in seven days, nearly matching the 2019 run-in after fourteen days but remaining below 2023. The comparison now depends on whether spot can hold the ground already covered.
| Episode | Day 0 (T−14) | Repair begins | Repair ends | Repair days | Run-in | In Repair | Total |
|---|---|---|---|---|---|---|---|
| 2019 | 2019-03-20 | 2019-04-03 | 2019-05-13 | 41 | +23.0% | +56.8% | +92.9% |
| 2023 | 2023-01-03 | 2023-01-17 | 2023-08-22 | 218 | +27.7% | +22.5% | +56.5% |
| 2026 · open | 2026-08-16 | — | — | — | +22.7% | — | — |
The rally has pulled a large block of supply back above water. From 1 June to 15 August, an average 52.5% of circulating bitcoin was in profit, leaving almost half below its cost basis. The recent +22.2% rally lifted that share to 68.5%.
Even so, the recovery remains thin for a market trading above the Short-Term Holder cost basis. Historically, the median share of supply in profit under the same condition is 86.9%, placing the current reading in the bottom tenth of observations.
The gap leaves fresh profit in the hands of recent buyers. Many have only just moved above water after a long Capitulation interval, and some will spend into strength. Their selling is the next test. If voluntary demand absorbs it without surrendering the breakout, the transition has room to continue. If not, the market can slide back into the range while the label remains Capitulation.
Profitability recovered unusually quickly. The short squeeze pushed price through a dense band of supply in only a few days. If this week becomes the final leg before Repair, it will be the fastest transition in the post-2018 record.
Price and supply in profit naturally move together, but not by a fixed amount. The response depends on how much supply sits inside the range price crosses.
Over the chart's most recent week, spot rose from $63.1K to $77.1K, a +22.2% gain, while supply in profit jumped 16.9 percentage points. Before Repair in 2019, price fell 1.9% and profit share declined 5.3 points. In 2023, price gained +3.2% and profit share rose 6.1 points. This week's move was far faster than either approach to Repair, and leverage supplied part of that speed. Voluntary demand now has more profitable supply to absorb.
The supply that just moved back into profit also has the clearest reason to spend. Coins last moved during this Capitulation fall largely within the Short-Term Holder cohort. That cohort has accounted for 80% of all realised profit this week.
Recent buyers are now the pressure point. The same cohort dominated profit-taking through both completed Repair regimes, accounting for 88.4% of realised profit in 2019 and 65.4% in 2023, compared with 46.5% across the full historical record. This week's share is already inside that range.
Repair requires the market to absorb those gains without surrendering the levels just reclaimed. Profit-taking that remains contained while price holds would show demand meeting the sellers. Spending that rises steadily with price would show more of that demand being used to fund exits rather than build support.
Short-Term Holder SOPR shows whether recent coins are being spent above or below their cost basis. A reading of 1.0 is break-even. Readings above it show the cohort realising profit, while readings below it show loss-taking.
The 7-day average stands at —, its highest level since —. A reading above 1.0 does not mean recent buyers have stopped selling. It means the coins they spend are profitable. If SOPR remains above break-even while price holds, demand is absorbing that spending. A return below 1.0 would put the newest cohort back into realised loss.
The evidence leaves the market close to Repair but not inside it. Price has moved, profit share has jumped, and Short-Term Holder spending has returned above break-even. Confirmation now rests on time, the cost-basis crossover, and the market's ability to absorb profitable spending.
Direction alone will not settle the regime. The path through cost basis, holder spending, and realised profitability will decide whether the squeeze becomes Repair or fades into another failed break.
Two levels define the path. Under QFI Labs' regime framework, Repair confirms when the 1d–1w cost basis at — rises above the Short-Term Holder basis and clears the model's deadband. In practical terms, that requires spot to hold above roughly — for about a week.
This is a minimum condition, not a price target. With spot near —, price has already done its part. The remaining test is time.
The opposing path opens below the — Short-Term Holder cost basis. Losing that level would push this week's buyers back underwater and drag the 1d–1w cost basis lower, undoing the condition needed for Repair to confirm.
Realised profitability has a second gate, and it remains closed. The 30D average of the realised profit-to-loss ratio stands at —, below both parity and the 365D average at —. The market is still realising slightly more loss than profit, and the short-term measure has remained below its annual baseline since October 2025.
The direction has improved sharply. The 30D ratio has more than doubled from 0.42 in late July as the rally pulled more supply above cost basis. The 365D measure now sits only slightly above 1.0, bringing parity and the annual baseline close together.
A sustained move in the 30D average above the 365D average would mark two changes at once. The market would be realising more profit than loss, and short-term profitability would be running above its annual baseline. That is the profitability profile Repair needs.
| Indicator | Scenario | Interpretation |
|---|---|---|
| Realised Profit / Loss Ratio Figure 9 |
The 30D profit-to-loss ratio clears parity, then overtakes the 365D measure, currently — against — | The market is now realising more profit than loss, while the short-term trend has overtaken its annual baseline. Together, these shifts suggest the emerging Repair signals are being supported by a genuine revival in demand rather than a random fluctuation |
| The 30D measure stalls below parity while the 365D measure keeps falling toward it | Capitulation's loss-dominant profile remains intact, and the rally has not changed it | |
| Regime Figure 8 |
Spot holds above the Short-Term Holder cost basis plus its deadband, currently —, for about a week | The market has met the minimum cost-basis condition for Repair, carrying the 1d–1w basis over the line. The next hurdle is the True Market Mean at — |
| Spot loses the Short-Term Holder cost basis at — | This week's buyers move underwater and the 1d–1w basis follows lower. The run-in fails rather than completes | |
| Short-Term Holder SOPR Figure 7 |
The 7-day average of Short-Term Holder SOPR holds above 1.0, currently — | Spot demand has caught up with the rally, absorbing the sell pressure from holders who have only just moved back above water |
| It breaks back below 1.0 | The market has failed to meet that sell pressure, and the newest cohort returns to selling at a loss | |
| Hyperliquid Liquidity Heatmap Field note · 5 |
Price holds above the long-side shelves at $74.0K to $75.6K | Long-side leverage remains untested, and the liquidation book stays behind the move rather than feeding a reversal |
| Price retests them | Liquidations run the other way. A long-liquidation cascade becomes the mirror of the short squeeze, concentrated on the same venue |
One of the largest weekly moves on record lifted a dense band of summer supply back above water. Coins accumulated between June and mid-August sat just below spot. The rally has pushed 68.5% of circulating bitcoin into profit.
The perpetual market supplied the extra force. A liquidated short must be bought back at market, turning crowded bearish positioning into a chain of compulsory bids. On the peak day, Hyperliquid generated roughly seven times the forced buying its open-interest share implies. That fuel is finite. Once the shorts are cleared, the bid disappears.
The next test comes from newly profitable holders. Short-Term Holders now account for 80% of realised profit, echoing the cohort dominance of both completed Repair regimes. Their 7-day SOPR crossed back above 1.0 on 19 August and sits at —. Remaining above break-even while price holds the reclaimed levels would suggest spot demand is absorbing profitable spending rather than being overwhelmed by it.
The simplest line remains the Short-Term Holder cost basis. Holding above it keeps the run-in alive and gives the 1d–1w cost basis time to catch up. Losing it would return this week's buyers to loss and weaken the case for Repair.