The Moat Is One Month Wide: When a Weekend Flurry Liquidates $800 Million
A two-thirds-of-a-percent dip in bitcoin wipes out hundreds of millions in overleveraged positions, while tectonic shifts in artificial intelligence and energy baseload reprice the frontier.
At 11:20 a.m. Eastern on Saturday, bitcoin printed $79,280. It was the weakest level in more than a month, and it lasted about ninety minutes.
[s577] What followed around that print was far more revealing than the ninety-minute dip itself: over twenty-four hours, liquidations cascaded to $815 million across more than 200,000 trading accounts, with more than $404 million in longs wiped out in a single four-hour stretch.
As Roger observed in a detailed autopsy of the liquidation chain, the market did not break because of a catastrophic macro failure, but because leverage had become fragile: A two-thirds-of-a-percent move does not liquidate four hundred million dollars unless the structure underneath is levered far past the move it is designed to survive.
[s577] Measuring across 400 daily closes since the October 2025 peak of $124,658, Roger argues that Saturday was not a novel downward spiral, but rather a failed recovery: Saturday was not the beginning of a decline. It was the failure of a recovery, and it happened to land on a weekend when leverage was thick.
[s577]
In Berlin, markusturm tracked the concurrent institutional headwinds [s559]. Over $635 million drained from spot Bitcoin ETFs on Thursday alone—marking the largest single-day institutional outflow since June and reversing three days of solid inflows. With the upcoming Federal Reserve policy meeting and headline CPI registering 3.7%, rising ten-year Treasury yields and dollar strength provided familiar friction. Yet beneath the financial noise, markusturm noted significant sovereign shifts: Jack Dorsey’s Bitchat mesh app topped French application charts with zero-internet bluetooth transacting, while Defense Secretary Pete Hegseth barred Pentagon staff from using confidential operational data on prediction markets like Polymarket.
