Show HN: Open-source Go repo comparing DCA vs. Grid with reproducible backtests
I built a small open-source repo in Go to compare two classic trading strategies (Dollar-Cost Averaging vs Grid). This is educational/research only — no real keys, no promises of profit.
Why I built it:
I wanted a clean architecture to experiment with strategies.
Most examples online are either black-box or not reproducible.
What’s inside:
Strategy interface for DCA/Grid.
Unified exchange client (Binance/CCXT).
Minimal backtester with seeded RNG for reproducibility.
Metrics: Profit Factor, Max Drawdown, Win Rate, Sharpe.
Sample metrics (BTC 2020–2024):
DCA → PF 1.42, MDD -18.7%, WR 62%
Grid → PF 1.67, MDD -25.3%, WR 58%
Repo: https://github.com/Zmey56/crypto-trading-strategies
I’d love feedback on:
Architecture (Go interfaces, package layout).
How you approach reproducibility in backtests.
Ideas for extending this (filters, regime classification, DeFi/AMMs).
No comments yet