Crypto DCA Backtester

Backtest dollar-cost averaging (DCA) against lump-sum investing on any coin's real historical price data, using the free CoinGecko market_chart API (no API key required, up to 365 days of daily history). Built for crypto traders and passive-income investors deciding between DCA and lump-sum entry strategies before committing capital. Simulates weekly contributions and compares final portfolio value, invested capital, and percent return against a lump-sum baseline of the same total capital deployed on day one. Complements trading, whale-tracking, and DeFi yield skills by covering the entry-strategy decision, which is distinct from picking what to buy or where to farm yield.

Simulates weekly dollar-cost averaging vs a lump-sum investment of the same total capital, using real historical daily prices from CoinGecko's free market_chart API. No API key needed.

When to use

• Deciding whether to DCA into a position or deploy capital all at once
• Backtesting how DCA would have performed on a specific coin over the last N days (up to 365)
• Comparing entry strategies across multiple coins before allocating
• Building intuition for how DCA smooths volatility vs a single entry point

Commands

bash

python3 scripts/dca_backtest.py --coin bitcoin --days 365 --weekly-contribution 100
python3 scripts/dca_backtest.py --coin ethereum --days 90 --weekly-contribution 250
python3 scripts/dca_backtest.py --coin solana --days 365 --weekly-contribution 100 --json

--coin takes a CoinGecko coin id (e.g. bitcoin, ethereum, solana, matic-network) — use CoinGecko's /coins/list if you need to look one up.

Output

Total invested, end portfolio value, and return % for both DCA and lump-sum, using the same total capital for a fair comparison, plus a declared winner for that specific historical window.

Notes

• Data source: CoinGecko free market_chart endpoint, daily granularity, max 365-day lookback on the free tier.
• This is a historical return comparison only — it does not model volatility-adjusted or risk-adjusted returns, and past performance over one specific window is not predictive of future results. Not financial advice.
• Requests use a curl-style User-Agent; some public API providers block the default Python User-Agent.