Every few weeks someone tells me the same thing, usually with the confidence of a man who has never opened a spreadsheet: you cannot time the market, so stop trying — just dollar-cost average into Bitcoin. Put in the same amount on the same schedule, rain or shine, and let the math do the work.
I believed it. For years I said it out loud. Dollar-cost averaging crypto is the sober, adult, non-degenerate way to own Bitcoin. It is the advice you give your brother-in-law so he stops buying the top.
Then one Sunday I got tired of repeating a claim I had never actually checked. So I checked it. And the advice, as it usually gets stated, broke.
Let me be blunt before I go further: this is not a hit piece on DCA. It is a hit piece on one specific sentence — “DCA beats a lump sum.” I want to show you my own numbers, including the part that embarrasses my own prior. Then I want to explain what DCA is actually for, because it is not what that sentence claims.
The advice everyone repeats (and that I believed)
The pitch is always the same three-line poem:
- You cannot time the market.
- So don’t try. Buy on a schedule instead.
- A schedule removes emotion, smooths your entry price, and beats lump-sum buying.
That third line is the load-bearing one. It is not just “DCA is safer.” It is a claim about returns: that dollar cost averaging bitcoin will, over time, outperform dropping the same money in at once. “DCA vs lump sum” gets treated as a settled contest, the way people treat “index funds beat stock picking.” Nobody argues it. Everybody repeats it.
And look — two of those three lines are just true. You really can’t time the market. Buying on a schedule really does remove a lot of emotion. I will defend both of those to the end. It’s the third line I have a problem with, because returns are a measurable thing, and a lot of people were asserting them without ever measuring them.
So I stopped arguing and measured.
So I stopped arguing and ran the numbers myself
Here is exactly what I did, so you can call me out if I cheated.
I pulled Bitcoin-USD daily closing prices from Coinbase Exchange public market data — every single day from 2017-01-01 through 2026-10-08. That’s 3,568 daily closes. The first close was $992.95 on 2017-01-01. The last close was $82,171.43 on 2026-10-08.
I ran three tests.
Test 1 — one-year windows. I slid a one-year window across the whole history, starting a new one every 7 days. That gave me 458 windows, from 2017-01-01 to 2026-10-05. In each window I compared two people putting in the same total money:
- Option A: a $5,200 lump sum, all in on day 1.
- Option B: $100 every week, for 52 weeks.
Both valuations taken 365 days later.
Test 2 — four-year windows. Same idea, longer horizon: 71 rolling four-year windows. Slow money. Patient money. The kind of timeline people actually claim DCA is built for.
Test 3 — the real, boring plan. What an actual human does: $200 a month, from 2017-01-01 to 2026-10-08. That’s 119 buys, $23,800 invested in total. Compared against dropping the same $23,800 in on day 1.
The caveats matter and I won’t hide them. This is Bitcoin only — one asset, not a portfolio. No trading fees, no taxes, no slippage, no bid-ask spread modelled. “Winning” here means raw arithmetic on a closing price, nothing more. And the same dataset can be sliced to flatter either side, which I’ll get to, because that’s the trap of every backtest ever run. This is not investment advice. I am not your financial advisor, and you should not do anything because a stranger on the internet once wrote a Markdown file.
Okay. Here’s what came out.
What the numbers said (the ugly part)
Test 1, across all 458 one-year windows:
| Metric | Lump sum ($5,200 on day 1) | Weekly DCA ($100 × 52 weeks) |
|---|---|---|
| Average return | +110.2% | +44.2% |
| Median return | +48.1% | +25.5% |
| Best 1-year window | +1,530.8% | +633.8% |
| Worst 1-year window | −82.0% | −55.2% |
DCA beat the lump sum in 163 of those 458 windows. That is 35.6%.
Read that again. Dollar-cost averaging lost to a single lump-sum purchase roughly two-thirds of the time. On average the lump sum returned +110.2% while DCA returned +44.2%. On the median, +48.1% versus +25.5%. The lump sum didn’t squeak past — it ran away with it. The best window is almost comical: +1,530.8% for the lump sum against +633.8% for the DCA buyer.
And it gets worse over longer horizons, not better. Test 2, the 71 four-year windows:
| Metric | Lump sum | Weekly DCA |
|---|---|---|
| Average return | +621% | +206% |
| Median return | +300% | +130% |
DCA won 18 of 71 windows. That’s 25.4%. Longer timeline, worse result for the schedule.
Then Test 3, the boring monthly plan — the one that actually resembles a real life:
- $200 a month from 2017-01-01 to 2026-10-08 = 119 buys, $23,800 invested → $213,887. That is 9.0x.
- The same $23,800 dropped in on day 1 → $1,969,565. That is 82.8x.
Let me sit with that for a second, because it stung. I invested $23,800 slowly and ended up with $213,887 — a nine-bagger, objectively a great outcome — and I still lost to myself by a factor of roughly nine. The slow plan turned a life-changing number into a merely very good one. That is a strange sentence to type, but it is what the arithmetic says.
Why does the lump sum win so often? It’s almost mechanical. Bitcoin spent most of this window going up. When the thing you’re buying rises over the whole period, the person who bought earliest owns the most of it for the longest. Every week DCA sits in cash instead of holding an asset that’s climbing, it pays a quiet opportunity cost. DCA is a standing bet that the price will be lower later. Over the last eight years, almost every week, that bet lost.
The one thing DCA actually wins (the honest part)
Now the part where I defend the thing I just spent 800 words dismantling, because my numbers do contain a real win. It’s just not the win anyone advertises.
Look at the tail. In the worst one-year window — the one starting 2017-12-17, for anyone keeping score — the lump sum lost 82.0%. The DCA buyer in that same window lost 55.2%. Both are ugly. But one is “I want to throw up,” and the other is “I can breathe.” That gap is DCA doing its actual job: not making you money, but not breaking you.
Now zoom out to the drawdowns that genuinely broke people:
- From $67,555 (2021-11-08) to $18,949 (2022-06-18), Bitcoin fell 72%.
- From its all-time-high close of $124,720 (2025-10-06) to $68,112 (2026-04-01), it fell 45%.
- The lowest close since 2024 was $39,524.
Answer honestly: who survives those? The person who put in $5,200 once and watched it turn into a crater is far more likely to panic-sell at the bottom than the person who is still buying every week, whose cost basis keeps crawling down toward the price. The lump-sum investor is watching one decision go wrong. The DCA investor is executing a habit. Those are not the same psychology, and psychology — not price — is what decides whether you’re still holding when the recovery shows up.
That’s the honest win. Not returns. Regret management.
Same one-year window, same $5,200 total — the window that started 2017-12-17
Both lost. One was survivable. That gap is the actual product DCA sells.
Why the advice survives anyway
Here’s where I stop being coy and say what I actually think.
DCA is not a return strategy. It is a regret-management tool wearing a return strategy’s clothes.
That’s the whole thesis, and you can prove it straight from my own tables. If DCA were really about returns, Test 1 alone would have killed it years ago. It survives because it solves a completely different problem: it keeps people from doing the one catastrophic thing.
Think about what lump-sum investing actually asks of you:
- You must have a lump sum in the first place — most people don’t, they have income, not capital.
- You must commit it all on a day you picked, at a price you cannot possibly know.
- You must then not touch it through a 72% drawdown.
Step three is where the theory dies. Everyone is a fearless lump-sum investor inside a backtest. Almost nobody is one at 3 a.m., with the chart bleeding red and the timeline screaming at them. DCA is what you build when you admit you can’t be trusted with a decision, so you automate it out of your own hands. That is a completely rational reason to do it. It’s just not the reason people give. The reason people give is “it beats the lump sum,” and on Bitcoin’s last eight years, that sentence is simply false.
One more honest note, because this warning cuts both ways. I could take the exact same data and make DCA look brilliant. Pick only the windows where a crash lands in the first few months — say a window starting near the 2021 top — and DCA wins by miles, because the lump-sum buyer got crushed early while the DCA buyer kept buying into the wreck. Flip it, keep only the smooth uptrends, and the lump sum wins every single time. The dataset will tell you whichever story you were already paying for. That’s the danger of backtests, and it’s exactly why I’m showing you all 458 windows and not just the flattering ones.
What I actually do now
So what changed for me? Less in the mechanics, more in the framing.
I still buy on a schedule. I’m not going to pretend I now sit on a pile of cash waiting for the perfect entry, because that’s precisely the behavior my own data warns against — and precisely the behavior I’m not disciplined enough to pull off. What changed is that I stopped telling myself I do it for a better return. I do it because I know what I’d do to myself otherwise.
If I had to compress it into rules:
- If you have a lump sum you genuinely won’t need, and you can stomach a 70% drawdown without selling, then investing it early has historically won. The last eight years of Bitcoin say so, loudly.
- If you have income instead of capital, DCA isn’t a compromise — it’s the only mechanism that matches how you actually get paid.
- If DCA keeps you from panic-selling, it’s doing its real job, even if it costs you returns along the way. Pay that tax. Stay in the game.
The last thing I keep coming back to is the reason the DCA advice survives at all. It isn’t that it’s a better strategy. It’s that it’s a strategy a normal, scared, imperfect person can actually stick to. The best plan is the one you don’t abandon. DCA doesn’t win because it beats the market. DCA wins because it beats you.
I ran the numbers expecting to write a clean takedown, and I mostly did. But I came out the other side respecting DCA more than when I started — just for different reasons than the ones I used to parrot. That feels like the honest place to land.
Related reading: Uptober Is a Base Rate, Not a Law · The Blockchain Won’t Know You Died
(The End)
