Nov 19, 2026
•
Insights

Does Crypto Go Quiet on Thanksgiving? Nine Years of Bitcoin Volume

Does Crypto Go Quiet on Thanksgiving? Nine Years of Bitcoin Volume

Lately, yes: in CoinQuant's check of Binance daily candles, Bitcoin volume on Thanksgiving Thursday and Friday was 60.32% to 87.43% of the prior 30-day average in 2022 to 2025, but above average in 2017 to 2021. Thin days are where a backtest fill and a real fill can differ most.

Bitcoin Thanksgiving history is usually told as a price story. This article checks one narrower claim, that crypto trading is slow on the holiday, with a method you can repeat. Then it shows what a thin day means when you read a backtest.

Bitcoin trades every day, holidays included. US stock markets close on Thanksgiving, which falls on 26 November in 2026. The idea is that crypto traders step away too, so the market goes thin.

The Answer First: Is Thanksgiving Thin for Bitcoin?

In the last four years, yes. Volume on Thanksgiving Thursday and Friday ran at 60.32% of the prior 30-day average in 2022, 87.43% in 2023, 61.48% in 2024 and 65.54% in 2025.

Before that, no. From 2017 to 2021, the same two days traded above the prior 30-day average every year, at 117.04% to 190.30% of it. So "crypto goes quiet on Thanksgiving" describes 2022 to 2025, not Bitcoin Thanksgiving history as a whole.

Price gave no clear signal either way, which the section on price shows. The useful takeaway is practical: a thin day is where a backtest fill and a real fill can drift apart.

The Claim and How We Checked It

The claim is simple: is crypto trading slow on holidays like Thanksgiving? Here is the check, step by step, so you can repeat it:

  1. Take the closed Binance BTCUSDT daily candles (UTC) for Thanksgiving Thursday and the Friday after, for every complete holiday from 2017 to 2025.
  2. Average the base volume of those two days. Base volume is counted in BTC, so a change in price alone does not move it.
  3. Divide by the average daily base volume of the 30 days before the Thursday. A 30-day baseline keeps one unusually busy or quiet week from setting the bar.
  4. Read anything under 100% as a thinner holiday than usual.

Quote volume, counted in USDT, tells the same story for 2022 to 2025: 53.41% in 2022, 92.44% in 2023, 69.67% in 2024 and 61.42% in 2025.

One caveat on the clock. A UTC day starts at 19:00 New York time the evening before, so the "Thursday" candle covers Wednesday evening and most of Thursday in the US. That shifts the hours a little, but not which week we measure.

Bitcoin Thanksgiving History, 2017 to 2025: What the Numbers Show

YearThanksgivingThursday and Friday base volume vs the prior 30-day average
201723 Nov173.19%
201822 Nov190.30%
201928 Nov117.04%
202026 Nov154.04%
202125 Nov124.06%
202224 Nov60.32%
202323 Nov87.43%
202428 Nov61.48%
202527 Nov65.54%

The split is clean: five holidays above the average, then four below it. None of the last four came close to the average; the nearest was 2023 at 87.43%. The volume data shows when the pattern changed, not why, so we do not offer a cause.

Nine holidays is also a small sample. Two regimes in nine years means the 26 November 2026 holiday could look like either one. Treat the 2022 to 2025 pattern as a description, not a rule.

Did Price Move More or Less on Those Days?

If you came here for the Bitcoin Thanksgiving price, the answer is less tidy than the volume. Thursday's daily close was below Wednesday's in seven of the nine years. Three of those seven, though, were within 0.30% of flat, in 2022, 2023 and 2024.

A close that barely moved is not a falling market. With three closes that were almost flat in a sample of nine, the data supports no up or down claim for the holiday, so this article makes none.

What Thin Days Mean for a Backtest Fill

On CoinQuant, an AI trading platform, a daily strategy fills its orders at the close of the day. The trade log stamps each row with the start time of its candle, so a row stamped 24 Nov 2022, 00:00 UTC was filled at the close of 24 Nov UTC.

Two trade logs show holiday fills. BTC HMA 55 Close Cross 1D 2021-2026 bought on Thanksgiving Day 2022 at 16,598.95 and sold on 16 Dec 2022. That was a thin holiday: volume ran at 60.32% of the prior 30-day average.

A trend rule bought on Thanksgiving Day 2022: in the BTC HMA 55 Close Cross 1D 2021-2026 trade log, the row stamped 24 Nov 2022, 00:00 UTC was filled at the close of 24 Nov UTC, and the next row is the sell on 16 Dec 2022.

A trend rule bought on Thanksgiving Day 2022: in the BTC HMA 55 Close Cross 1D 2021-2026 trade log, the row stamped 24 Nov 2022, 00:00 UTC was filled at the close of 24 Nov UTC, and the next row is the sell on 16 Dec 2022.

BTC RMA 50 Close Cross 1D 2021-2026 sold at the close of 26 Nov 2021, the day after Thanksgiving, at 53,726.53, and bought again at the 28 Nov close. In 2021 the holiday traded at 124.06% of the prior 30-day average, so that fill came on a busier day than usual.

Sold at the close of 26 Nov 2021, the day after Thanksgiving, and bought again at the close of 28 Nov 2021: two rows of the BTC RMA 50 Close Cross 1D 2021-2026 trade log, stamped 00:00 UTC.

Sold at the close of 26 Nov 2021, the day after Thanksgiving, and bought again at the close of 28 Nov 2021: two rows of the BTC RMA 50 Close Cross 1D 2021-2026 trade log, stamped 00:00 UTC.

A backtest assumes the whole order filled at that close, with fees and nothing else. On a thin day, a real market order can move the price further than on a normal day, so the gap between the backtest price and your price can be wider. For how order types change that gap, see market vs limit orders in backtesting.

To check your own rule, describe it in plain English and run it. For example:

Buy BTCUSDT at the daily close when the close crosses above the 55-day Hull Moving Average, and sell at the daily close when it crosses back below. Use a 0.1% fee per trade.

Then open the Trades tab and look for rows dated on or just after Thanksgiving. The settings behind every test on this blog are listed in how we test crypto strategies.

Common mistakes when reading holiday fills

  • Reading 00:00 as a midnight fill: on a daily rule it is the close of that UTC day.
  • Treating a fill on a thin day as a price you would have got: the backtest models fees (0.1% taker fee on entry and exit) but no slippage.
  • Reading one run as a holiday effect: both rules are long only, one position at a time, with no leverage. Past results do not predict future results.
  • Mixing Total Return with yearly return: BTC HMA 55 Close Cross 1D 2021-2026 made a Total Return of 9.84% over five years, not per year; see CAGR vs Total Return.

The Practical Lesson

  • Holiday volume is a reason to check fills, not a signal: nine years show two regimes and no reliable price direction.
  • Know what the time stamp means: a daily row stamped 00:00 UTC was filled at the close of that UTC day.
  • Expect more drift on thin days: if a rule trades on a holiday, real fills can differ more from the backtest.

If your strategy trades around late November, open its trade log and find the holiday rows before you trust its results. In CoinQuant every row carries a date, a time and a price, so this takes minutes. If a large share of a rule's profit comes from a handful of holiday fills, treat that profit with extra care.

Build trading strategies on real data. No coding required.

Frequently Asked Questions

Does crypto trade on Thanksgiving?

Yes. Crypto markets trade every day of the year. US stock markets close on Thanksgiving (26 November in 2026) and close early the next day.

Is Bitcoin volume lower on holidays?

On Thanksgiving, lately yes: Thursday and Friday volume was 60.32% to 87.43% of the prior 30-day average in 2022 to 2025. From 2017 to 2021 it was above average, at 117.04% to 190.30%. We did not check other holidays.

Does Bitcoin go up or down on Thanksgiving?

There is no reliable answer. Thursday's close was lower in seven of nine years, but three of those were within 0.30% of flat.

What does the time in a CoinQuant trade log mean?

It is the start time of the daily candle. A row stamped 24 Nov 2022, 00:00 UTC was filled at the close of 24 Nov UTC.

Check how your own rule handled the holiday weeks, row by row.

Start your first backtest on CoinQuant

Try it yourself

Have an idea worth testing?

Describe it in plain English. CoinQuant writes the rules and shows how it would have done on real prices.

Plans from $12.99 a week. Cancel anytime.

Disclaimer:

This content is for educational and informational purposes only and does not constitute financial, investment, or trading advice. All strategies and examples are for illustrative purposes and do not guarantee results. Always conduct your own research before making financial decisions.

Key Takeaway