Bitcoin breakout research compares three daily BTC/USD models using Coinbase data. Each saved model requires the current high to meet or exceed the 20-bar highest-high reference and the close to be above EMA(3). The additional trend filter is EMA(50), EMA(70) or EMA(7), depending on the model. All three saved exits use price crossing below EMA(3). These are separate variants, not a rule requiring the close above all four moving averages simultaneously. Reproduction also requires the platform's definition of the highest-high window, including whether the current bar is included, and its signal-execution convention. The study asks whether a price-breakout condition combined with trend filters changes the historical return path. It does not establish that a breakout is early, that resistance will hold as support, or that the best-looking EMA variant will remain best outside the test sample. Historical backtests are research, not a forecast or investment advice. Review sample dates, drawdowns, trade count and execution assumptions. Fees, slippage and losses can materially change realized results.
Over an empirical backtest sample comprising 168 audited executions, this model achieved a cumulative net return of +9,676.6% (+48.6% compound annual growth rate). Trade execution efficiency registered a 43.5% win rate (relying on asymmetric payoff distribution to overcome lower win frequency), paired with an outstanding profit factor of 3.03. Downside risk analysis records a maximum historical drawdown of -24.4% (contained historical peak-to-trough drawdowns within acceptable systematic tolerances), while current drawdown stands at -5.2% from recent peak equity. Risk metrics reflect an annualized Sharpe ratio of 4.50 (exceptional risk-adjusted alpha) alongside an R² statistical fit score of 0.91 (91% consistency, indicating high equity curve linearity with controlled dispersion).
The quantitative profile demonstrates verifiable historical edge and balanced risk distribution suitable for systematic asset allocation.Strategies: 0
