Study · re-run every night · updated
DCA or lump sum? Bitcoin all at once or $100 a week, every first day since 2014, tested
$5,200 into Bitcoin at once, or $100 every week for a year? Over 4,305 first days from Jan 1, 2014, all at once held more Bitcoin once the year was in on 65% of them - 20% more in the middle case. The deepest each sat under the money put in during its first year: −21% for the weekly plan against −21% all at once in the middle case, −64% against −83% at worst. Each Jan 1 as one case, 12 years: the weekly plan ended with 9.0% fewer coins than the lump sum on average - a luck test of −0.7, not past the bar of 3. Begun a year ago, on Oct 14, 2025: $5,200 at once is worth $3,758 on Oct 14, 2026, $100 a week $5,530. Past patterns, not financial advice.
All at once ahead
65%
of 4,305 first days
Middle lead
+20%
coins, all at once over weekly
Deepest dip, weekly
−21%
the first year, middle case
Deepest dip, at once
−21%
the first year, middle case
§ 01 · Today
A year ago today
as of · Bitcoin's daily closes, Jan 1, 2014 to Oct 14, 2026
Two ways to put $5,200 into Bitcoin: all at once, at one day's close, or $100 every 7 days for a year - 52 parts, each at that day's close. The second is dollar-cost averaging (DCA): no guessing the day, more coins when the price is low and fewer when it is high.
Begun a year ago, on Oct 14, 2025, with Bitcoin at $113,028: all at once is worth $3,758 at the close of Oct 14, 2026 ($81,682 a coin), the weekly plan $5,530 - both from $5,200 put in.
Once the last week is in, both plans only hold coins, so which ends ahead is settled by that first year: the plan with more coins stays ahead by the same share at every later date.
§ 02 · Every first day
Every first day
as of · 4,305 first days, Jan 1, 2014 to Oct 14, 2025
Each point is a first day (one a week drawn): how many more coins all at once held than the weekly plan once its year was in (above the line), or fewer (under it). All at once held more on 65% of the 4,305 first days - 20% more in the middle case, 28% more on average. Its best first day was Mar 24, 2017 (249% more coins), its worst Dec 16, 2017 (61% fewer).
How deep each sat under the money put in so far, at its worst close of the first year: −21% for the weekly plan against −21% all at once in the middle case, −64% against −83% at worst.
§ 03 · Year by year
Year by year
as of · first days in each calendar year
| First days in | Days | All at once ahead | Middle lead | Deepest dip, weekly | Deepest dip, at once |
|---|---|---|---|---|---|
| 2025 part | 287 | 33% | −13% | −37% | −40% |
| 2024 | 366 | 98% | +26% | −15% | −13% |
| 2023 | 365 | 100% | +40% | −10% | −7% |
| 2022 | 365 | 46% | −4% | −23% | −32% |
| 2021 | 365 | 30% | −20% | −50% | −49% |
| 2020 | 366 | 100% | +98% | −8% | −4% |
| 2019 | 365 | 67% | +10% | −42% | −38% |
| 2018 | 365 | 13% | −23% | −46% | −53% |
| 2017 | 365 | 88% | +125% | −22% | −12% |
| 2016 | 366 | 100% | +59% | −4% | −2% |
| 2015 | 365 | 95% | +32% | −16% | −11% |
| 2014 | 365 | 0% | −36% | −57% | −66% |
Middle lead: all at once's extra coins over the weekly plan for that year's middle first day (under zero: the weekly plan held more). Deepest dip: the plan's worth under the money put in so far at its worst close of the first year, for the middle first day. All at once wins the years the price climbs; the weekly plan wins the years it falls or swings, because each $100 gets more coins at the lows.
§ 04 · Later
One, three and five years on
as of · what each plan was worth, over the money put in
| First days | Worth, at once | Worth, weekly | Under water, at once | Under water, weekly | Worst, at once | Worst, weekly | |
|---|---|---|---|---|---|---|---|
| A year on | 4,305 | $8,376 | $6,960 | 33% | 30% | $878.89 | $1,862 |
| Three years on | 3,575 | $21,428 | $19,939 | 0% | 4% | $4,696 | $4,411 |
| Five years on | 2,845 | $59,588 | $40,448 | 0% | 0% | $4,586 | $7,110 |
Worth: what $5,200 became, for the middle first day. Under water: the share of first days on which the plan was worth less than the money put in, that many days after its first day. Worst: the lowest it was worth then.
§ 05 · How it is tested
How it is tested
as of · methodology
A luck test (t) says how far a result sits from zero, in units of its own noise: about 2 is rarely luck for one test, but many tests at once need more, which is what the first cut is for.
The numbers were fixed before any was read: $5,200 at one close against 52 parts of $100, one every 7 days, each at that day's close, on every day of Bitcoin's record with a full first year after it (Bitstamp's daily closes before our own, from Jan 1, 2014, then ours); worth one, three and five years on. Money waiting for its week earns nothing.
Neighbouring first days share almost all of their year, so the 4,305 of them are not 4,305 separate cases. The luck test takes each Jan 1 as one case - no two share a day - and asks whether the weekly plan's coins over the lump sum's (as a log) average away from zero: a plain test over 12 years, read on Student's t since the cases are few. A result needs 3 or more and the older and the newer half of the years on its side. Here the weekly plan ended with 9.0% fewer coins than the lump sum on average over 12 years (1.8% fewer in the older half, 15.7% fewer in the newer): a luck test of −0.7, not past the bar. Prices only, before trading costs.
§ 06 · Revisions1 dated
Revisions
as of · every change to its headline numbers, dated
This study is re-run every night on the whole history: Bitcoin's daily closes from Jan 1, 2014 to Aug 16, 2017 are Bitstamp's, Europe's oldest exchange, and ours from the next day on (on the 3,346 days both have, the two closes differ by 0.06% in the middle case). It was first written on Oct 15, 2026; its numbers last moved on Oct 15, 2026. Each night its headline numbers moved is listed here, newest first.
- All at once ahead on 65% of first days, 20% more coins in the middle case
§ 07 · More
More research
as of