Bitcoin Broke Below $50,000 as a Global Selloff Hit Crypto

Bitcoin fell through $50,000 on Coinbase on August 5, 2024, as a global flight from risk hit digital assets and forced leveraged positions out of the market. The venue’s BTC-USD daily candle, measured from 00:00 to 23:59 UTC, opened at $58,131.31, traded as low as $49,050.01 and closed at $54,029.12.

Those figures translate to a 15.62% drop from the UTC open to the intraday low and a 7.06% open-to-close decline. They are Coinburn calculations from Coinbase Exchange data, not a universal crypto-market close. Bitcoin trades continuously, and highs, lows and daily percentage changes vary by exchange, pair and time-zone boundary.

Ether absorbed the sharper shock

The same Coinbase window shows ETH-USD opening at $2,688.38, reaching $2,116.02 and closing at $2,420.37. That was a 21.29% open-to-low fall and a 9.97% open-to-close decline. On this venue and measurement window, ether’s drawdown was materially deeper than bitcoin’s.

The difference mattered because the selloff was not merely a change in the quoted value of one asset. It tested leverage across centralized derivatives and decentralized lending while U.S. spot bitcoin exchange-traded products had already connected crypto more directly to conventional portfolios. Coinbase Institutional wrote on August 5 that roughly $280 billion had been removed from total cryptocurrency market capitalization over the weekend, leaving about $1.9 trillion. It also reported, using an unspecified liquidation-data source in that note, $459 million of bitcoin long liquidations and $381 million of ether long liquidations over 24 hours.

Those liquidation and market-cap figures are contemporaneous Coinbase estimates, not exchange-candle measurements and not independently reproducible from the cited note. They establish the scale perceived by an institutional market participant, but should not be combined with the Coinbase spot candles as though all figures share one methodology.

A macro shock, not a proven single cause

The chronology points outward from crypto. On July 31, 2024, the Bank of Japan raised its target for the uncollateralized overnight call rate from about 0%–0.1% to around 0.25%. On August 2, weaker U.S. employment data intensified concern about economic growth. By August 5, Asian and European equities were falling sharply and volatility had spread across asset classes.

Coinbase’s August 5 research said an unwind of yen-funded carry trades “may” have affected crypto decisions. Reuters likewise described recession fears and broad selling of risky assets. These accounts support a macro-driven interpretation, but they do not prove that the Japanese rate decision, U.S. data, Middle East tensions, bankruptcy distributions or any rumored institutional sale caused a specified portion of bitcoin’s decline. The event-day record supported multiple interacting pressures, not a clean attribution model.

That distinction is important. Bitcoin’s ability to trade throughout the weekend made it an immediate venue for repricing before many traditional cash markets opened. It did not make bitcoin immune from the same liquidity demands, positioning reductions and volatility affecting conventional risk assets. Ether’s larger venue-specific fall also showed how collateral use and leverage could amplify stress differently across assets.

What the August 5 record established

The strongest conclusion is narrow and verifiable: Coinbase’s BTC-USD market printed below $50,000 during the August 5 UTC session, while ETH-USD suffered a larger percentage drawdown on the same venue and window. Contemporaneous reporting placed both moves inside a global risk-off episode.

The record did not establish a single cause, an industry-wide closing price or the final amount of forced liquidation. Follow-up research should compare synchronized candles across additional spot venues, obtain timestamped liquidation data with exchange coverage disclosed, and separate spot selling from derivatives and on-chain collateral liquidations. Those limitations prevent the day’s dramatic numbers from being treated as more precise than their underlying datasets allow.