- The Coinbase Bitcoin Premium Index has stayed negative for 50 consecutive days, the longest streak since the metric launched.
- US spot Bitcoin ETFs lost over $2.6 billion in nine trading days before flows turned positive in early July.
- The Mt. Gox estate moved 10,422 BTC ahead of its October creditor repayment deadline, keeping a supply overhang on the market.
- The previous record streak in January and February ended with a rally from $72,000 to six figures.
The Coinbase Bitcoin Premium Index has now spent 50 consecutive days below zero, the longest negative stretch since Coinglass began tracking the metric. The streak started on May 19 and the latest reading sits at -0.0742%, meaning Bitcoin continues to trade cheaper on Coinbase than on Binance. The previous record, a 40-day run between January 16 and February 24 this year, already exceeded the roughly 30-day discount seen during the October 2025 crash. A persistent discount of this length points to one thing: the selling, or at least the absence of buying, is coming from American capital specifically.
Why a Negative Premium Points Straight at US Money
The index measures the percentage gap between the Bitcoin price on Coinbase Pro and the price on Binance. The split matters because the two exchanges serve very different clients. Coinbase handles the bulk of regulated US institutional flow and custodies the assets behind most American spot Bitcoin ETFs, including BlackRock’s IBIT and Fidelity’s FBTC. Binance, which controls somewhere between 33.8% and 42% of global spot volume depending on the source, is dominated by international retail traders. Coinbase holds around 5.5% of global spot share, a fraction of Binance’s footprint, but it concentrates nearly all US compliance-bound capital in one venue.
When Coinbase trades at a discount, US-based sellers are hitting bids harder than global buyers can absorb them on that specific order book. International traders on Binance are effectively holding the floor while American institutions either distribute or stay out entirely.
ETF Redemptions Did Most of the Damage
Fund flow data lines up almost perfectly with the depressed premium. Between June 18 and July 1, US spot Bitcoin ETFs posted net outflows on every single trading day, shedding a combined $2,627.6 million, according to Farside Investors. Two sessions did most of the work: June 24 and June 25 accounted for over 44% of the entire streak’s redemptions. BlackRock’s IBIT led the exodus with nine consecutive days of withdrawals.
| Date | Total ETF Net Flow | Detail |
|---|---|---|
| June 24 | -$469.0M | First of two capitulation days |
| June 25 | -$691.7M | Heaviest single-day outflow of the streak |
| June 26 | -$444.5M (IBIT alone) | BlackRock’s largest standalone exit |
| July 6 | +$265.7M | Flows flip positive, IBIT takes in $209.4M |
The turn came in early July. Across July 2 and July 6, institutional buyers put $489.2 million back into the products, cutting the net loss for the full period to $2,138.4 million. One week of inflows does not erase a nine-day bleed, but the direction changed for the first time since mid-June.
Mt. Gox Keeps a Known Supply Overhang on the Table
ETF redemptions are not the only source of pressure. On June 2 the defunct Mt. Gox estate moved 10,422 BTC, worth $739 million at the time, to new addresses ahead of the October 31, 2026 creditor repayment deadline. Nobody knows how much of that supply creditors will actually sell, but the market prices the risk regardless, and it prices it primarily on the venues where large distributions historically land. Stack that overhang on top of nine straight days of ETF outflows and the record discount stops looking mysterious.
The January Streak Ended With a Rally to Six Figures
The 40-day negative run earlier this year offers the most useful precedent. When that streak began on January 16, Bitcoin traded near a local peak around $96,000.
Persistent US selling dragged the price to a bottom near $72,000 by February 5, pushing the daily RSI into deep oversold territory below 30. The price then coiled between $72,000 and $76,000 for close to three weeks, capped by the 50-day and 100-day moving averages. Around February 24, the exact day the negative streak ended, Bitcoin broke through both averages and ran toward $100,000 in the following weeks.
| Negative Streak | Duration | Price Action | Aftermath |
|---|---|---|---|
| October 2025 crash | ~30 days | Sharp liquidation-driven selloff | Gradual premium recovery |
| Jan 16 – Feb 24, 2026 | 40 days | Drop from $96,000 to $72,000, then three-week base | Breakout above 50 and 100 SMA, rally past $100,000 |
| May 19 – ongoing | 50+ days | Discount deepened alongside $2.6B in ETF outflows | Unresolved |
For readers less familiar with the technical side: a moving average simply smooths the price over a set number of days, and traders watch the 50-day and 100-day lines as levels where rallies tend to stall or accelerate. An RSI below 30 signals that selling has been so one-sided it often precedes a pause or reversal. Neither guarantees anything. In the January case, both markers fired near the bottom while institutions were still dumping.
Seoul Is Flashing the Same Exhaustion Signal
South Korea provides an unusual parallel. The Kimchi Premium traditionally shows Bitcoin trading above global prices in Korea, a product of aggressive local retail demand and strict capital controls that block arbitrage. Since early June, that premium has inverted, sinking to a discount between -1.3% and -3.5% according to data from CryptoQuant and community trackers. Korean retail, historically the most FOMO-prone cohort in crypto, is paying less than the global price.
Two structurally different buyer bases, US institutions and Asian retail, rarely capitulate at the same time. Right now both premiums sit in negative territory simultaneously, which suggests broad exhaustion rather than a regionally isolated problem.
What Has to Happen Before the Discount Closes
Three markers deserve attention over the coming weeks. First, whether the July 6 ETF inflow of $265.7 million marks a genuine reversal or a single relief print; a second and third consecutive green day would carry far more signal. Second, the premium index itself, since the February precedent shows the flip back to positive coincided almost exactly with the price breakout. Third, the Mt. Gox calendar, because the October 31 repayment deadline keeps a known supply overhang on the table through the autumn regardless of what ETF buyers do.
The honest reading is that a record discount describes the present, not the future. It confirms US sellers have dominated for 50 days. Whether that pressure is finally exhausting itself, as it did in February, or merely pausing before another leg down, will show up first in exactly these three data points.