BTC / ETH / SOL DCA Backtest Tool
Want to verify: if you had started DCA 2 years ago, what would your account be worth today? This backtest charts the real DCA curve for BTC, ETH, and SOL over the past 2–4 years and compares "lump-sum" vs. "DCA". Source: CoinGecko monthly closing prices (2022-01 to 2026-05), 100% browser-side computation, no user input collected.
Three parameters →
—
Numbers convincing? To actually start DCA — OKX Auto-Invest supports weekly or monthly schedules from $10.
Go to OKX to enable DCAPrice trend across the backtest window
Blue: asset price · Orange: DCA average cost (if the orange line stays below the blue line, DCA is winning)
Three things to read from this backtest
1. DCA is not a "guaranteed-return tool"
If you start DCA on BTC from the November 2021 bull-top (try switching to that start date), your actual return through May 2026 may be less than you expect. DCA only has meaning on long-term-rising assets over long-enough horizons. Anyone selling "DCA = guaranteed return" is not telling the whole story.
2. DCA vs. lump-sum: which wins depends on market shape
- One-sided uptrend (e.g. 2023-10 → 2024-04 ETF rally): lump-sum beats DCA decisively, because every later buy is more expensive.
- Fall-then-rise (e.g. 2022-06 → 2026-05): DCA beats lump-sum decisively, because DCA buys more units at lower prices.
- Sideways chop: DCA slightly ahead or roughly tied.
- One-sided downtrend: both lose; DCA loses slightly less.
3. Read the orange average-cost line
The orange line on the chart below (your DCA average cost) sits in the middle of the blue line (market price) range — that is "averaging" visualised. DCA does not buy you the floor or the peak; it averages out across the whole range.
Applying these conclusions on OKX
If after the backtest you decide to start DCA:
- Use OKX Auto-Invest — set it up once, auto-debited weekly or monthly. See the OKX Auto-Invest hands-on guide.
- BTC + ETH 50/50 — more stable than a single coin.
- Size at 5–10% of monthly disposable income — anything above that becomes gambling, not DCA.
- Set it and don't change it — the hardest and most valuable discipline in DCA.
- Move any portion untouched for 30 days to self-custody — see the three-layer custody approach.
FAQ
What data does this DCA backtest use?
BTC / ETH / SOL monthly closing prices over the past 5 years (2021-01 to 2026-05), sourced from CoinGecko and Glassnode public history. All data is hard-coded into the page — no API calls, no data collection.
Why does DCA sometimes beat lump-sum and sometimes lose?
Simple rule: in a one-sided uptrend, lump-sum wins (every later buy is more expensive); in chop / downtrend / fall-then-rise, DCA wins (more units bought at lower prices). This is the essence of DCA — trading market-timing anxiety for certainty.
Can past performance predict the future?
No. The backtest only answers "if you had done this in the past, what would your account look like today". It does not answer "how much will doing this earn going forward". Crypto's historical volatility is severe; a 5-year backtest cannot span all market regimes. Treat it as a "tool for understanding the DCA mechanism", not an "investment-decision basis".
Why no USDT DCA option?
USDT is a stablecoin pinned near $1, so DCA into it is meaningless (no "average cost" concept). If you want to "park funds in USDT periodically", that is plain periodic savings, not DCA.
Does this tool collect my input?
No. 100% browser-side JavaScript computation, with the price data hard-coded into the HTML. Whatever amount you type stays in your browser. See the privacy policy.