// research
Bitcoin after the halving: 2020 and 2024 backtested
Buying bitcoin the day after the halving and holding for a year returned +554.04% after the 2020 halving and +30.45% after the 2024 one. Against every start day from 2020 to 2025, the year after the 2020 halving beat 97.5% of them and the year after the 2024 one only 45.2% — two halvings, two very different outcomes, and no pattern to lean on.
5 min read
The halving — the block every four years when the bitcoin mining reward is cut in half — is one of the most quoted reasons for a bitcoin rally. This platform's data covers two of them, 2020 and 2024. Buying right after each and holding for six, twelve or eighteen months gave one exceptional result and one ordinary one.
The halving dates
- Block 630,000, 2020-05-11 19:23 UTC — reward from 12.5 to 6.25 BTC.1
- Block 840,000, 2024-04-20 00:09 UTC — reward from 6.25 to 3.125 BTC.2
- The halving before them, block 420,000 on 2016-07-09,3 cannot be tested here: Binance spot candles on this platform start on 2017-08-17, a year later. The one before that, in 2012, predates Binance altogether.
Data and method
- Backtests: Binance spot BTCUSDT, daily candles, the strategy "Lump sum: buy once, hold" (#35, also used in DCA or lump sum). It buys with 9,880 USDT of a 10,000 USDT account at the first daily open after the halving day — 2020-05-12 and 2024-04-21 (the 2024 block came nine minutes after the April 20 open, so that open was still before the halving) — and sells at the daily close six, twelve and eighteen months later. Fee 0.1% and slippage 0.05% on each side, no leverage. Six runs, #698–#703.
- Baseline: the same trade from every start day 2020-01-01 → 2025-12-31 whose window ends by 2026-10-19: buy at the next day's open, sell at the close 6, 12 or 18 months later, computed from the daily candles without fees. That is 2,192 six-month windows, 2,119 twelve-month and 1,936 eighteen-month. The halving windows are placed in this spread by their own price move, also without fees.
Results
| Halving | Hold | Backtest | Max drawdown | BTC price move | Beat, all start days | Beat, start days in the same year |
|---|---|---|---|---|---|---|
| 2020 | 6 months | +81.77% (#698) | 17.41% | +83.18% | 86.2% | 40.4% |
| 2020 | 12 months | +554.04% (#699) | 25.12% | +561.88% | 97.5% | 85.8% |
| 2020 | 18 months | +647.42% (#700) | 53.07% | +656.53% | 98.6% | 92.3% |
| 2024 | 6 months | +5.92% (#701) | 24.24% | +6.30% | 38.1% | 19.7% |
| 2024 | 12 months | +30.45% (#702) | 27.91% | +31.16% | 45.2% | 24.0% |
| 2024 | 18 months | +68.96% (#703) | 27.91% | +70.20% | 48.9% | 73.2% |
"Beat" is the share of start days whose window of the same length did worse. A run shows slightly less than the price move: fees and 120 USDT left in cash.
2020: an exceptional year to own bitcoin
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2020-05-11 → 2020-11-11
- Return
- +81.77%
- Max drawdown
- 17.41%
- Sharpe
- 2.63
- Win rate
- 100.0%
- Trades
- 1
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2020-05-11 → 2021-05-11
- Return
- +554.04%
- Max drawdown
- 25.12%
- Sharpe
- 3.12
- Win rate
- 100.0%
- Trades
- 1
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2020-05-11 → 2021-11-11
- Return
- +647.42%
- Max drawdown
- 53.07%
- Sharpe
- 2.21
- Win rate
- 100.0%
- Trades
- 1
2024: an ordinary one
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2024-04-20 → 2024-10-20
- Return
- +5.92%
- Max drawdown
- 24.24%
- Sharpe
- 0.48
- Win rate
- 100.0%
- Trades
- 1
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2024-04-20 → 2025-04-20
- Return
- +30.45%
- Max drawdown
- 27.91%
- Sharpe
- 0.78
- Win rate
- 100.0%
- Trades
- 1
Backtest
Lump sum: buy once, holdBTCUSDT · 1d · Spot · 2024-04-20 → 2025-10-20
- Return
- +68.96%
- Max drawdown
- 27.91%
- Sharpe
- 1.00
- Win rate
- 100.0%
- Trades
- 1
Against random start dates
For context, here is what holding bitcoin from an arbitrary day of 2020–2025 did:
| Hold | Windows | Median | 10th percentile | 90th percentile | Ended higher |
|---|---|---|---|---|---|
| 6 months | 2,192 | +21.35% | -41.13% | +110.32% | 1,438 of 2,192 |
| 12 months | 2,119 | +49.55% | -45.08% | +301.97% | 1,295 of 2,119 |
| 18 months | 1,936 | +74.98% | -51.43% | +324.59% | 1,282 of 1,936 |
The 2020 halving windows sit at the top of this spread. But look at the last column of the results table: against the other start days of 2020 alone, the six months after the halving beat only 40.4% of them. The median six-month window starting anywhere in 2020 made +101.2%, the median twelve-month window +322.8% — March 2020's crash bottom was in that year, and so was the start of the 2020–2021 bull run. 2020 was a great year to buy bitcoin on almost any day; the halving day was one of those days.
The 2024 halving windows landed below the median of all start days on every horizon, and on six and twelve months they trailed most other 2024 start days too — 80.3% and 76.0% of them did better. Only the eighteen-month window beat most of its own year: it ended on 2025-10-20, two weeks after the highest daily close in this data (2025-10-06) and about 11% below it.
Conclusion
- Buying the day after the 2020 halving and holding for 12 months returned +554.04%; after the 2024 halving, +30.45%.
- 2020's windows were in the top few percent of 2020–2025, but mostly because 2020 itself was an exceptional year to buy — the six months after the halving were below the median of that year's start days.
- 2024's windows were ordinary: below the median of all start days on every horizon.
- Two halvings cannot establish a pattern, and the third one in history is outside this data.
How these runs are built: how to backtest a crypto strategy. How to read the cards: backtest metrics explained. Another calendar story checked against chance: Uptober.
Notes
-
Block 630,000 on mempool.space (timestamp 2020-05-11 19:23:43 UTC, checked 2026-10-20): https://mempool.space/block/000000000000000000024bead8df69990852c202db0e0097c1a12ea637d7e96d ↩
-
Block 840,000 on mempool.space (timestamp 2024-04-20 00:09:27 UTC, checked 2026-10-20): https://mempool.space/block/0000000000000000000320283a032748cef8227873ff4872689bf23f1cda83a5 ↩
-
Block 420,000 on mempool.space (timestamp 2016-07-09 16:46:13 UTC, checked 2026-10-20): https://mempool.space/block/000000000000000002cce816c0ab2c5c269cb081896b7dcb34b8422d6b74ffa1 ↩
Test your own idea
Create a free account, start from a template or from your own rules, and read your first report.
More articles
- Does high funding predict crypto crashes? 8 coins, 2020–2026BIS economists found that high crypto carry predicts crashes. On 8 Binance perpetuals, 2020–2026, crashes came more often — but after bigger gains.
- Crypto momentum after the paper: a weekly test on 8 coinsLiu and Tsyvinski found weekly momentum in bitcoin up to 2018. On 8 coins in 2020–2026 the one-week signal lost to holding; four weeks cut drawdowns.
- Does price move the hour before funding? 8 coins, 2020–2026Do traders sell before paying funding? The hour before each settlement on 8 Binance perpetuals, 2020–2026: no reliable move, and fees took what was left.
- Weekends and time of day: when does bitcoin move?Do bitcoin gains come on weekends or outside US trading hours? Hourly BTC and ETH data since 2020, and session backtests: with fees, -99.87%.
- What leverage does to one strategy: 1x, 2x, 3x, 5xThe same golden cross on BTC and ETH perpetuals, 2021–2026: BTC +9.78% at 1x, -7.76% at 2x, -99.77% at 5x. Funding times leverage moved the liquidation price.