# The best day of the week to buy Bitcoin: a DCA backtest

> Is there a best day of the week to buy Bitcoin or Ethereum? 351 weekly buys on each weekday since 2020: the gap between days is 2–3 points.

Research · Janorium · Published 2026-10-15  
Web page: https://janorium.com/articles/best-day-to-buy-bitcoin-dca-backtest

Barely. Buying 25 USDT of BTC every week for 351 weeks, the best weekday (Monday) ended 2.08 percentage points ahead of the worst (Wednesday) on a +170% result; on ETH the gap was 3.42 points. Year by year the winner kept changing.

![Six headline numbers: Monday best and Wednesday worst on BTC, 2.08 points apart, Monday best on ETH, 3.42 points apart, and four and three different weekdays winning the six years](https://janorium.com/articles/best-day-to-buy-bitcoin-dca-backtest/scorecard.en.svg 'Seven weekdays, two coins, 351 weeks — every number is from the runs below')

## Setup

- **Rule:** 25 USDT of BTC or ETH every week, always on the same weekday, at that day's daily close (00:00 UTC of the
  next day), held to the end. One strategy per weekday — seven in all.
- **Period:** 2020-01-06 → 2026-09-27, Monday to Sunday, exactly 351 weeks, so every weekday gets the same 351 buys.
  For each calendar year, a 52-week window from its first Monday.
- **Market and costs:** Binance spot, daily candles, 0.1% fee and 0.05% slippage per buy and sell, 10,000 USDT
  account — 8,775 USDT of it invested over the period, the rest idle.
- **Runs:** 98 — 7 weekdays × 2 coins × (the whole period + 6 years). Nothing was tuned.

## The whole period

![Total return of buying every Monday, Tuesday … Sunday for 351 weeks, BTC and ETH, bars from zero](https://janorium.com/articles/best-day-to-buy-bitcoin-dca-backtest/full.en.svg 'Drawn from zero, the seven bars are almost the same length')

On BTC every weekday finished between +168.22% (Wednesday) and +170.30% (Monday); on ETH between +138.85% (Saturday)
and +142.27% (Monday). Spread over 351 weeks, the few days between two possible buys average out.

![Each weekday's total return minus the average of the seven, in percentage points](https://janorium.com/articles/best-day-to-buy-bitcoin-dca-backtest/diff.en.svg 'Zoomed in to the points that separate them')

Monday came first on both coins — by 2.08 points over the worst day on BTC and 3.42 on ETH.

**Backtest: [DCA every Monday, 25 USDT](https://janorium.com/backtests/538)**  
BTCUSDT · 1d · Spot · 2020-01-06 → 2026-09-27

| Return | Max drawdown | Sharpe | Win rate | Trades |
| --- | --- | --- | --- | --- |
| +170.30% | 50.22% | 0.62 | 82.6% | 351 |

*BTC, every Monday: +170.30%*

**Backtest: [DCA every Wednesday, 25 USDT](https://janorium.com/backtests/566)**  
BTCUSDT · 1d · Spot · 2020-01-06 → 2026-09-27

| Return | Max drawdown | Sharpe | Win rate | Trades |
| --- | --- | --- | --- | --- |
| +168.22% | 50.18% | 0.61 | 82.6% | 351 |

*BTC, every Wednesday: +168.22%*

## Year by year

![For BTC and ETH, each year's 52 weekly buys by weekday, with the winning weekday of each year highlighted](https://janorium.com/articles/best-day-to-buy-bitcoin-dca-backtest/years.en.svg 'Four different weekdays won the six years on BTC')

No weekday won every year. On BTC the winner went Monday, Tuesday, Sunday, Monday, Tuesday, Friday; on ETH Monday won
four of the six. Inside a year the best weekday was at most 0.59 points above the average one.

> **What this test does not show**
>
> - **One history.** Monday leading on both coins over 2020–2026 is one observation of one period, not a law — the
>   yearly grid shows how easily the order changes.
> - **The daily close, in UTC.** "Monday" means the candle that closes at 00:00 UTC on Tuesday. Buying at another hour
>   of the day is a different test.
> - **Returns are on the whole 10,000 USDT account,** most of which waited in cash; the yearly numbers are small for that
>   reason, and only their differences matter here.

## All seven days, the whole period

| Weekday   |                        BTC |                        ETH |
| --------- | -------------------------: | -------------------------: |
| Monday    | [+170.30%](https://janorium.com/backtests/538) | [+142.27%](https://janorium.com/backtests/539) |
| Tuesday   | [+169.47%](https://janorium.com/backtests/552) | [+141.65%](https://janorium.com/backtests/553) |
| Wednesday | [+168.22%](https://janorium.com/backtests/566) | [+140.17%](https://janorium.com/backtests/567) |
| Thursday  | [+169.09%](https://janorium.com/backtests/580) | [+141.22%](https://janorium.com/backtests/581) |
| Friday    | [+168.73%](https://janorium.com/backtests/594) | [+139.62%](https://janorium.com/backtests/595) |
| Saturday  | [+168.80%](https://janorium.com/backtests/608) | [+138.85%](https://janorium.com/backtests/609) |
| Sunday    | [+169.27%](https://janorium.com/backtests/622) | [+139.77%](https://janorium.com/backtests/623) |

Each return links to its run. 351 buys of 25 USDT per run.

## Conclusion

- The day of the week you buy on changed a 351-week DCA result by 2–3 percentage points — far less than the choice of
  coin, period or amount.
- Monday led on both coins over the whole period, but it did not win every year, and the gaps inside a year were
  fractions of a point.
- On this evidence the weekday is a detail of a DCA plan, not a lever.

More research: [DCA or lump sum](https://janorium.com/articles/dca-vs-lump-sum-btc-eth-backtest).