CoinGecko constate que le Bitcoin connaît ses plus fortes hausses de valeur lors des jours fériés aux États-Unis

[라온]

Bitcoin
BTCUSD
delivers its strongest single-day returns on US federal holidays, according to a CoinGecko study covering May 2013 to May 2026, with New Year’s Day producing an average next-day return of +2.01% and an 84.6% win rate.

The research analyzed 4,753 daily price observations and found that US holidays produced an average next-day return of +0.77 %, roughly four times the +0.19% baseline for non-holidays.

Holiday Effect Skews Heavily Positive

On the win-rate side, Columbus Day also hit 84.6%, with a +1.70% next-day average. Christmas Day produced a smaller +1.46% gain on a 53.8% win rate, while Labor Day registered +1.22% across a 69.2% win rate, according to CoinGecko.

Two holidays buck the trend. Martin Luther King Jr. Day averages -0.84%, dragged down by a -18.65% Bitcoin drop on January 15, 2018. Independence Day averages -0.26%, with both holidays posting win rates below 50%.

CoinGecko researchers attribute the New Year’s Day signal to fresh January capital allocations and December tax-loss selling reversals.

The effect held even as BTC prices ranged from $313 in 2015 to $93,507 in 2025, despite the split 2026 price outlook between bulls and bears.

Day-of-Week Effect Fades Over Longer Horizons

Within the trading week, Monday and Wednesday tied at +0.38% average next-day returns. Thursday is the only day to post a negative average at -0.09%.

The weekday-weekend gap was just 0.01%, far narrower than the documented Uptober seasonality effect.

On a 365-day horizon, every weekday produced returns between 142.15% and 144.56%. CoinGecko called the spread negligible relative to Bitcoin’s volatility.

The data suggests holiday timing may add marginal value at short horizons. Whether the Santa rally pattern extends into next year’s January setup remains an open question.
source: https://www.tradingview.com/news/beincrypto:07201c6d3094b:0-bitcoin-pumps-hardest-on-us-holidays-coingecko-finds/