Bitcoin’s Cost Bases Broke Down, but On-Chain Signals Do Not Confirm a Bottom

In reporting published on November 18, 2026, bitcoin’s decline raised a narrower question than whether sentiment had reached an extreme: had the market reached a durable bottom, or were holders still under pressure to sell? The reported on-chain measures show widespread unrealized losses, but they do not settle that question.

Bitcoin fell from roughly $74,000 to a November 15 low near $61,900 over five days, according to the report. It also cited nearly $2 billion in liquidations over 24 hours, more than 80% of them long positions. The figures describe a sharp deleveraging event, not proof that selling was complete.

Cost bases moved above the price

The report listed four estimated cost bases above bitcoin’s subsequent price: the spot ETF average near $81,500, the market-wide average near $77,500, the short-term-holder cost basis near $75,600 and Strategy’s average purchase price near $75,400. The ETF estimate was attributed to Glassnode data dated October 27; the report did not give equally precise measurement dates for every other estimate. These are different holder groups and calculations, not four independent price supports.

Glassnode data cited in the report put approximately 4 million BTC’s acquisition prices between $85,000 and $118,000. The share of supply in profit reportedly fell from 60.4% to 55.3%, while short-term holders as a group had an estimated 7.4% unrealized loss. Those measures indicate potential selling pressure, but unrealized losses do not establish when holders will sell.

Flows and valuation signals remain mixed

From late October through November 13, spot bitcoin ETFs reportedly recorded more than $2.3 billion in cumulative net outflows, with net inflows on only two of 12 trading days. The report also cited a roughly $14.2 billion contraction in stablecoin supply over 30 days and approximately $55 million in daily net exchange deposits by whales. It attributed to Glassnode the view that those whale deposits were preparation for possible distribution—not confirmation that the coins had been sold.

There was countervailing activity: wallets holding 100 to 10,000 BTC were reportedly accumulating, Strategy was adding to its holdings, and Glassnode characterized cumulative ETF outflows as modest relative to the funds’ total size. The report did not provide a common measurement date or dataset for every flow figure, limiting how precisely they can be compared.

MVRV—the ratio of bitcoin’s market value to its realized value—was reported near 1.2. CryptoQuant analyst Darkfost said his model had issued its third low-range signal of 2026, after signals in February and June–July, while cautioning that three signals do not mean three bottoms. Galaxy research head Alex Thorn’s framework, cited from a June report, looked for MVRV below 1 and a price below realized price at a typical cycle bottom. The November report said neither condition had occurred, though Thorn also noted that this cycle’s lower peak MVRV could make historical bottom thresholds less applicable.

The evidence therefore supports a limited conclusion: holders were under strain and some capital was leaving, while other buyers remained active. Whether bitcoin revisited its reported June low range of $57,700–$58,500 depended on subsequent flows and selling behavior; the cited metrics could not establish that outcome as of the original November 18 reporting date.

Further reading