BTC-USDT Trading Strategies
Walk-forward backtested strategies on BTC hourly data from 2017 to present. Each strategy is optimized independently with out-of-sample holdout validation.
Hidden Markov Model
7-state Gaussian HMM for market regime detection. Trades based on regime transitions between bull, bear, sideways, and capitulation states. Long and short strategies with expanding and rolling walk-forward windows.
MA + Stochastic Direction
Dual moving average crossover with stochastic oscillator confirmation. NSGA-II multi-objective optimization (Sharpe / Drawdown / Profit Factor). Separate long-only and short-only with equal-weight portfolio.
Multi-Asset Scanner
MA+Stoch strategy applied to all crypto spot pairs with sufficient history. Tests optimized parameters across 1H (19 pairs) and 4H (17 pairs) timeframes.
Spike Fade (Mean Reversion)
Fades large intraday BTC spike bars (>2%) expecting mean reversion on 15min/30min timeframes. Grid search over trigger, target, stop and session filter parameters. Based on statistical analysis of BTC microstructure autocorrelation.
Spike Continuation
Long-only momentum after large up spikes (>1-3%) on 15min/30min with daily MA trend filter. Grid search over trigger, trend MA, hold period, stop and target. 68.8% continuation rate for 15min spikes >3% in bull markets.
Candlestick Patterns
Statistical reliability analysis of all 61 TA-Lib candlestick patterns across 227 Binance USDT spot pairs. Forward returns at 1–10 periods, volume analysis, and timing breakdowns across 1H, 2H, 4H, 8H, 12H, and 1D timeframes.
Candlestick + Indicator Filters
Portfolio backtest of 61 candlestick patterns with 10 technical indicator confirmation filters (RSI, Stochastic, ADX, MACD, Bollinger Bands, Volume). Compares signal quality, risk-adjusted returns, and alpha vs BTC buy-and-hold.
Economic Regime Detection
Distance-based non-parametric regime classification across 7 macro variables. Identifies historically similar periods and generates long/short signals for major assets. Based on Mulliner, Harvey et al. (2025).
Candlestick Inverted Signals
Portfolio backtest of inverse candlestick signals where weak bullish edges are flipped into bearish trades (and vice versa). Includes Kelly/equal-weight sizing modes, drawdown controls, and detailed trade logs.
Combined Portfolio
Normal + inverted candlestick signals combined into diversified portfolios. Two studies: split capital ($50k each) and unified signal pool. Return correlation of -0.10 provides natural hedge and Sharpe boost.