DCA means buying the same amount on a schedule instead of all at once. We ran three years of weekly BTC and ETH buys against a lump sum on day one, with real Hyperliquid prices, to see when each one actually wins.
November 16, 2026
13
min to read
The short answer
DCA, dollar-cost averaging, means buying a fixed amount on a schedule instead of all at once. It smooths your entry price and removes the decision of when to buy, but it does not protect against a market that keeps falling. In our three-year backtest of weekly BTC and ETH buys, a lump sum on day one finished ahead on both.
Dollar-cost averaging, DCA, means investing a fixed amount at regular intervals instead of all at once, whatever the price is doing that day. FINRA, the US brokerage regulator, defines it plainly: “With dollar-cost averaging, you invest your money in equal portions, at regular intervals, regardless of current market conditions.” In crypto, that usually means an automatic weekly or monthly buy of BTC, ETH or another token, running for months or years without you deciding each time.
The alternative is a lump sum: putting the whole amount in on one day. A “DCA bot” is just software that places the recurring buy for you, on a schedule, so you cannot forget or hesitate; our guide to trading bots covers the other kinds, and our grid trading backtest tests the most common one. The question this guide tests is simple: over real crypto prices, how often does the recurring version actually beat buying it all on day one?
FINRA gives several reasons. It can “remove some of the emotion from investing,” since there is no day you have to decide the market looks right. And because a fixed dollar amount buys more of an asset when its price is low and less when it is high, FINRA says this “sometimes results in paying a lower average price per share over time.” Fidelity’s explainer adds a behavioral angle: it can help “mitigate the effects of investor psychology, as it relates to trying to time the market,” including panic selling after a bad entry.
On average, no. FINRA is direct about the trade-off: spreading your money out “often produces lower returns than lump sum investing, especially over longer periods,” because “you could also miss out on some gains since part of your money stays in cash” waiting for its turn to be invested. Fidelity makes the same point from the other side: uninvested cash waiting to be deployed typically earns very low returns while it waits.
Vanguard’s research puts a number on it. Testing cost averaging against investing a lump sum immediately, across historical and simulated markets, Vanguard found that the lump sum came out ahead roughly two-thirds of the time: 68% of the time in global stocks from 1976 to 2022, against investing the same money in three monthly parts and comparing a year later. The reason is the one above: over those years, U.S. stocks and bonds beat cash most of the time, so money sitting on the sidelines has a real cost. That is a statement about the average outcome across many markets and periods, not a promise about any one asset over any one stretch, which is why we ran the comparison on real BTC and ETH prices below rather than taking the average on faith.
We ran two rules on the same $10,000, over the same three years, on the same two markets, with the same costs:
One thing we deliberately left out: funding. Ethy’s own DCA and TWAP automations buy spot tokens on Base, not a perpetual contract, so funding never applies to them, and nothing in this backtest is charged or credited for it either. It matters only to someone holding a perp position, which a buyer of spot BTC or ETH is not.
Here is where each one ended up, and how far each fell from its own peak along the way:
| Lump sum, day one | Weekly DCA | Lump sum’s worst drop | DCA’s worst drop | |
|---|---|---|---|---|
| BTC | +83.4% | −10.8% | −53.0% | −44.2% |
| ETH | −7.5% | −29.4% | −67.6% | −56.5% |
On both markets, the lump sum ended up ahead of weekly DCA, and by a wide margin on BTC. BTC’s day-one price of $35,580 turned out to be far below where the market spent most of the next three years, so DCA kept buying at prices that average out to $73,265, +105.9% above that first day. ETH tells a different story: both approaches lost money, because the window’s last close sat well below both ETH’s day-one price and DCA’s average cost of $2,597.57.
Which dates favored each approach has a short answer for both markets: weekly DCA’s curve was worth more than the lump sum’s never once, on any of the 1,097 days in the window on BTC, and on 15 of the 1,097 days (1%), between November 16, 2023 and April 21, 2025 on ETH. The rest of the three years belonged to the lump sum.
BTC spent this window mostly climbing, then gave a lot of it back. It opened the window at $35,580, and no later close ever fell back below that opening price: the window’s lowest close, $35,737, came on November 21, 2023, just days after the start. From there it climbed to a highest close of $124,711 on October 6, 2025, and closed the window at well under that high. A lump sum bought on day one caught, in effect, the cheapest BTC would close for the next three years, and rode the whole climb from there. DCA, by design, kept some of its money in cash during that early stretch, then kept buying through 156 Mondays, up to the peak and back down, which is why its average cost of $73,265 sits +105.9% above the day-one price: most of its buys landed well above where the lump sum got in.
What DCA actually protects against shows up in a different comparison. A lump sum unlucky enough to buy at the open on October 6, 2025, the day BTC peaked, would have finished at −47.2%. DCA, which also bought plenty of its coins near that same peak, still finished far better than that, at −10.8%, because its earlier, cheaper buys cushioned the blow. That is the real protection DCA offers here: not a better result than the lump sum that happened to buy on the actual best day, but a much smaller loss than the lump sum that could have bought on the worst one.
ETH shows what DCA cannot fix: a market that is lower at the end than it was for most of the time you were buying into it. It opened at $1,981.70, fell to a lowest close of $1,472.10 on April 8, 2025, rallied to a highest close of $4,834.80 on August 22, 2025, then fell back hard enough that both the lump sum and DCA closed the window underwater. DCA’s average cost, $2,597.57, sits +31.1% above the day-one price, because most of its 156 weekly buys landed above the day-one price, not at the bottom.
DCA was only briefly worth more than the lump sum: on 15 of the 1,097 days (1%), between November 16, 2023 and April 21, 2025. The first stretch was right at the start, before the lump sum’s single entry had time to either gain or lose much. The second was around ETH’s lowest close in early April 2025, when about half of DCA’s money was still in cash and had missed the fall, a temporary edge that ETH’s summer rally and DCA’s continued buying into it erased for good.
Both markets above say the same thing from two different angles: DCA takes the single worst-case outcome off the table, and it also takes the single best-case outcome off the table. What it actually does:
DCA fits best for money you add over time anyway, such as part of each paycheck, where the real alternative is not one lump sum but sitting in cash until you feel ready, which Fidelity and FINRA both note is its own kind of cost. It also fits someone who has a lump sum but knows they would agonize over the single day to deploy it, and would rather accept a smaller expected return for a smaller range of regret. Our wider look at automated setups covers weekly buys alongside five other rules traders run on a schedule, with the same honest result: easy to follow is not the same as the best performer on any one stretch of the market.
It fits less well as a way to “time the bottom” gradually. Nothing about buying weekly tells you whether an asset is cheap or expensive; it only spreads your bet across more days. If you believe a lump sum is a bad idea because the price looks high today, DCA does not resolve that belief, it just spends several months acting on it a little at a time.
Ethy is an AI trading agent you ask before you trade. Its Strategy Library includes a DCA template and a Smart DCA template, alongside a TWAP template that splits a single purchase into smaller pieces over a shorter window instead of buying it all at once. All three run on spot tokens on Base, not a perpetual contract, which is why funding never enters into them. Ethy’s DCA bot shows the three side by side.
Ethy is software for research and analysis. It does not give investment advice, and a backtest, on Ethy or in this guide, is an estimate on past data, not a result you are promised.
Usually not, on average. In our three-year backtest a lump sum bought on day one finished at +83.4% against weekly DCA’s −10.8%, because most of BTC’s three years traded above its day-one price, so DCA kept buying in above where the lump sum got in. Vanguard’s research puts the lump sum ahead roughly two-thirds of the time in stock and bond markets.
It reduces one specific risk: putting all your money in on a single bad day. It does not reduce your exposure to the asset falling overall, and it can still lose money in a market that keeps falling, as our ETH result shows.
There is no single right cadence; weekly, biweekly and monthly are all common. Vanguard did find that the longer a plan takes to invest the whole amount, the more it tends to give up against investing at once.
No. A DCA bot only automates when it buys, on a fixed schedule; it has no view on whether the price is cheap or expensive that day. Automating the schedule removes hesitation, not market risk.
Yes. Ethy’s Strategy Library has a DCA template, a Smart DCA template and a TWAP template, all running on spot tokens on Base on a schedule you set, with a paper mode that trades on live prices without moving money.
DCA repeats a buy on a recurring schedule, such as every week, for as long as you let it run. TWAP splits one purchase into smaller pieces over a shorter, fixed window, to get one position in more gently rather than build a position over months.
Trading involves risk, including the loss of your money. Ethy AI is a research and trading tool and does not give investment advice or recommendations. Backtests are simulations on historical data and do not predict future results.
Deploy your AI trading agent in one click.
