DCA vs lump sum Bitcoin: every start date tested
Buying Bitcoin all at once beat 52 weekly buys on 68.14% of 3,581 start days since 2016. See the start-year split, Ether, 4 years and the limits.
On Coinbase BTC-USD, putting all your money in on day one ended the year with more value than 52 weekly buys on 68.14% of 3,581 start days from January 2016 to October 2025, with a median gap of +0.36 per $1 invested. But the answer swung hard by start year: the lump sum was ahead on 13.15% of 2018 start days and on 100.00% of 2016, 2020 and 2023 start days. All those overlapping days add up to only about 10 separate years.
These are past cases, not a forecast, and not advice.
What are DCA and a lump sum?
A lump sum means you put all your money in on one day. DCA (dollar-cost averaging) means you split the same money into equal buys over time. Here that means 1/52 of the money every week for 52 weeks.
The question is simple: for any day you could have started, which plan was worth more one year later?
What did Bitcoin show over one year?
We looked at every start day from January 1, 2016 to October 20, 2025 (3,581 days). Each plan was valued 364 days after its start. Each buy lost a 0.05% fee. Money waiting for a later buy earned nothing.
"Per $1" means for each $1 of money invested. A gap of +0.36 means the lump sum ended $0.36 higher than DCA.
| Coinbase BTC-USD, 1-year plans | Result |
|---|---|
| Start days tested | 3,581 |
| Lump sum ended higher | 68.14% of start days |
| Median lump-sum value per $1 | $1.64 |
| Median DCA value per $1 | $1.31 |
| Median gap (lump minus DCA) | +0.36 |
| 10th and 90th percentile gap | −0.32 and +2.78 |
| Lump sum ended below $1 | 30.05% of start days |
| DCA ended below $1 | 28.43% of start days |
| Both ended below $1 | 23.51% of start days |
The percentiles show the spread. About a tenth of start days had a gap below −0.32. About a tenth had a gap above +2.78. DCA did not avoid losses by much: it ended below $1 on 28.43% of start days, against 30.05% for the lump sum.
The extremes are large both ways. The best start day for a lump sum was December 17, 2016: $24.91 per $1 against $10.89 for DCA. The worst was July 24, 2018: $1.17 for the lump sum against $1.86 for DCA.
Did the answer depend on the start year?
Yes, a lot. Here is the share of start days where the lump sum ended ahead, by the year the plan started.
| Start year | Days tested | Lump sum ahead | Median gap per $1 |
|---|---|---|---|
| 2016 | 366 | 100.00% | +1.30 |
| 2017 | 365 | 88.22% | +1.53 |
| 2018 | 365 | 13.15% | −0.29 |
| 2019 | 365 | 66.57% | +0.12 |
| 2020 | 366 | 100.00% | +1.94 |
| 2021 | 365 | 29.86% | −0.14 |
| 2022 | 365 | 46.30% | −0.05 |
| 2023 | 365 | 100.00% | +0.68 |
| 2024 | 366 | 97.81% | +0.36 |
| 2025 | 293 | 32.08% | −0.10 |
Four start years had a lump sum ahead on every day or almost every day: 2016, 2020, 2023 and 2024. DCA was ahead on most days in 2018, 2021 and 2025, and on just over half in 2022. In 2021 the median lump sum ended at $0.60 per $1, and DCA at $0.66. In a year when the price falls, buying a bit at a time lost less.
How many independent cases is that?
Fewer than it looks. Start days January 1 and January 2 share almost the same path. The 3,581 days are really about 10 separate one-year stretches.
To see the effect, we also picked start days a full 364 days apart, which do not overlap. We did this for every possible starting offset. Each offset had 9 or 10 start days. The share where the lump sum won ranged from 50.00% to 88.89%, with a median of 70.00%.
So the result moves by almost 39 points depending only on which day of the year your yearly plan began. Treat 68.14% as a rough size, not a precise rate.
What about Ethereum?
On Coinbase ETH-USD, from January 1, 2017 to October 20, 2025, we tested 3,215 start days. The lump sum ended ahead on 61.62%. The median gap was +0.17 per $1, and the median lump-sum value was $1.22 against $1.13 for DCA.
The lump sum ended below $1 on 43.52% of ETH start days, and DCA on 35.93%. Both were below $1 on 28.21%. The worst start for a lump sum was June 27, 2018: $0.76 per $1 against $1.90 for DCA.
For ETH, the non-overlapping stretches ranged from 33.33% to 77.78%, with a median of 66.67%. That rests on about 9 separate years.
The two coins did not always agree. For 2024 starts, the lump sum was ahead on 97.81% of Bitcoin start days but on only 35.79% of Ether start days.
What changes over four years?
We repeated the test with 208 weekly buys (four years of buying), valued 1,456 days after the start. Starts had to be at least four years before October 2026, so the last one is October 24, 2022.
| Coin | Start days | Lump sum ahead | Median gap per $1 | 10th and 90th percentile |
|---|---|---|---|---|
| Coinbase BTC-USD | 2,489 | 81.44% | +2.36 | −0.58 and +16.38 |
| Coinbase ETH-USD | 2,123 | 64.30% | +0.89 | −2.57 and +13.37 |
Over four years the lump sum was ahead more often for Bitcoin, but the sample is much smaller. Only one or two non-overlapping four-year windows fit in this history. By start year for BTC, the lump sum was ahead on 100.00% of 2016, 2019 and 2020 starts, and on 18.08% of 2021 starts.
Why did a lump sum win so often?
Bitcoin and Ether rose a great deal over most of this period. When the price drifts up, money that waits for later buys gets in at higher prices. A lump sum gets exposed to the rise sooner.
DCA did better when the price fell soon after the start, as in 2018 and 2021. In 2021 the median DCA plan lost less than the median lump sum. This is the usual trade-off between the two plans: a lump sum carries all the risk on day one, while DCA spreads it out but leaves some money idle.
Our guide to why past crypto returns are not forecasts explains why a market that rose in the past does not promise the next stretch. For how deep the falls can go, see crypto drawdowns and recovery times. If you split buys by week, our day-of-week test found no weekday that stood out.
What does this data not show?
- It does not include taxes, transfers, storage or the gap between buy and sell prices. We used one flat 0.05% fee per buy.
- Waiting money earned no interest. With interest, DCA would end a little higher.
- The last DCA buy happens 357 days after the start, so DCA is only fully invested for the last week.
- It has no 2026 start days, because a full year was not yet complete.
- Only two coins, both of which rose a lot. A coin that fell for the whole test would give a very different answer.
How we measured this
We used daily closing prices (UTC) from Coinbase BTC-USD from January 2016 and Coinbase ETH-USD from January 2017, up to October 19, 2026. For each start day, the lump sum buys at that day's close. The weekly plan buys 1/52 of the same money at that day's close and every 7 days after, for 52 buys. Both are valued at the close 364 days after the start. Every buy loses 0.05% as a fee. The gap is lump-sum value minus DCA value per $1.
Try this in Comparand
Ask your connected AI agent: "Get Coinbase BTC-USD daily bars from January 1, 2024 to January 1, 2025. Compare one buy at the first close with 52 equal weekly buys, both valued at the last close. Include a 0.05% fee per buy."
- venue_bars returns the open, high, low and close price plus volume for each completed bar. Use the 1d interval for daily bars. It is free on every key.
- Your agent works out the two plans from those prices. The tool returns prices, not a DCA report.
- You can change the dates, the number of buys or the coin (for example ETH-USD).
You can try 3 coins free, see the tool docs, or check pricing. These are past cases, not a forecast.
The short version
- On Coinbase BTC-USD, a lump sum ended a year ahead of 52 weekly buys on 68.14% of 3,581 start days, with a median gap of +0.36 per $1.
- By start year, that share ran from 13.15% (2018) to 100.00% (2016, 2020 and 2023).
- Ether was 61.62% of 3,215 start days. Over four years, Bitcoin was 81.44% and Ether 64.30%.
- The start days overlap, so the real sample is about 10 one-year stretches, and fewer for four years.
- Past cases, not a forecast, and not advice.
Questions people ask
Is DCA or lump sum better for Bitcoin?
In our test on Coinbase BTC-USD, a lump sum ended a year ahead of 52 weekly buys on 68.14% of 3,581 start days from 2016 to 2025. But those days overlap and make about 10 separate years, and in some start years DCA won most days.
How often did dollar-cost averaging beat a lump sum?
For one-year Bitcoin starts, DCA ended ahead on the rest of the start days, about a third. It won on most start days in 2018, 2021 and 2025, and on just over half in 2022.
Did lump sum also win for Ethereum?
On Coinbase ETH-USD, a lump sum ended a year ahead on 61.62% of 3,215 start days from 2017 to 2025. That is a smaller share than Bitcoin, and it rests on about 9 separate years.
Does this mean I should put all my money in at once?
No. This is a test of past cases, not a forecast or advice. Bitcoin rose a great deal over most of this period, which helps a lump sum. In a flat or falling market the answer can flip.