EMA 9/21 Crossover Backtest: It Beat Buy and Hold, Then Broke
We tested the classic 9/21 EMA crossover, long only, on Bitcoin, Ethereum and Solana using daily and four hour candles, with fees and slippage included. On daily candles it beat buy and hold on every coin, with Ethereum returning 4480.3 percent against 456.0 percent. On four hour candles it mostly lost to buy and hold, and on the newest data it lost money on all three coins. It looks like a decent trend filter on daily charts, not a money machine.
Method: signal at candle close, trade at next candle open, long only, no leverage. Costs: fee 0.10% + slippage 0.05% per side. Data: Bitget spot candles.
This is a backtest on historical data, not financial advice.