Bitcoin’s March 10 rebound faded by the UTC close

Bitcoin’s attempted recovery on March 10, 2018 had reversed by the end of the UTC trading day. The historical BTC-USD candle now distributed through Coinbase Exchange’s API records a $9,255 opening price, $9,518.77 high, $8,720.08 low and $8,795.44 close for the 24-hour interval beginning at 00:00 UTC.

Calculated from those venue-specific figures, bitcoin declined 4.97% from open to close. It finished 7.60% below the session high and only 0.86% above the low. Coinbase reports 17,254.85640436 BTC of volume within that candle.

The result mattered less as another threshold crossing—Coinburn’s March 9 archive already documents bitcoin’s move below $9,000—than as an example of how quickly a continuously traded cryptocurrency market could change within one calendar date. A contemporaneous report published on March 10 described bitcoin back above $9,000 and 99 of CoinMarketCap’s top 100 assets showing positive performance at its observation time. That was an attributable intraday snapshot, not the day’s final condition.

The end-of-day picture looked different

CoinMarketCap’s historical ranking for March 10 places bitcoin at $8,866, down 4.97% over its rolling 24-hour comparison, with reported 24-hour volume of approximately $5.39 billion. Ether was listed at $686.89, down 5.31%, while XRP was $0.7966, down 6.11%.

Every asset in CoinMarketCap’s displayed top ten carried a negative 24-hour change at the end-of-day snapshot. Among those ten, Monero had the largest listed decline at 9.66%. That breadth supports the bounded conclusion that the apparent recovery recorded during March 10 did not survive across the leading assets through the end of the UTC day.

The Coinbase and CoinMarketCap bitcoin marks differed by $70.56, or about 0.80% of the Coinbase close. That comparison is a Coinburn calculation, not a quoted spread available for execution. Coinbase’s figure represents one BTC-USD venue and one daily bucket. CoinMarketCap’s figure is an aggregated market reference drawn from multiple trading markets. The two observations therefore measure related but non-identical instruments and should not be expected to match exactly.

What the data can and cannot establish

Crypto markets do not conduct a universal closing auction. Coinbase’s documentation defines a daily candle as a grouped bucket and warns that historical rate data may be incomplete where no trades occurred. Its API can also return buckets preceding a requested start time. For this reconstruction, the relevant record is specifically the bucket timestamped 1520640000, corresponding to 2018-03-10 00:00:00 UTC.

CoinMarketCap describes historical listings as end-of-UTC-day ranking snapshots. Its percentage fields use rolling comparison windows, so they are not necessarily calculated from the same opening trade used in Coinbase’s candle. The contemporaneous March 10 report used an earlier observation window that was not preserved with an exact timestamp in the article. Its claim that most assets were positive is therefore evidence of an intraday condition, not evidence that CoinMarketCap’s later snapshot is erroneous.

The records establish a failed intraday rebound and broad weakness by the selected UTC boundary. They do not establish why traders sold, identify a single authoritative global bitcoin price or prove that regulatory developments documented on March 7 and March 8 caused the March 10 reversal. Those developments belong to separate dated archive files, and causal attribution would require evidence beyond coincident timing.