Would an EMA crossover strategy have actually beaten just holding?
Pick two EMA periods, a coin, and a window. This runs the classic walk-forward crossover strategy against real daily prices and shows exactly where it would have bought and sold — no cherry-picked backtest window, no fabricated equity curve.
What this is testing
A crossover strategy goes long when the short EMA crosses above the long EMA ('golden cross') and exits to cash when it crosses back below ('death cross'). This is walk-forward — every decision only uses price data up to that point, never future data — and it's compared directly against simply buying on day one and holding.
Different coins, periods, and time windows will show the strategy winning sometimes and losing sometimes — that's real, not a bug. A strategy that always wins in a backtest is usually a sign the backtest is flawed, not that the strategy is good. Past performance is not predictive of future results.