Optimal Bitcoin Trading Strategy Development Using Quantitative Models for the Current Market Regime

This thesis investigates the development of an optimal Bitcoin trading strategy tailored to the prevailing market regime observed in early 2025. Characterized by heightened uncertainty, potential shifts in trend dynamics, and persistent volatility, the current regime presents challenges for traditional quantitative models. A systematic review of recent academic literature on cryptocurrency market regimes, factor investing, arbitrage techniques, and machine learning applications was conducted. Based on this review and the identified regime characteristics, two distinct quantitative strategies-a Volatility-Adjusted Momentum strategy and a Mean Reversion strategy based on Bollinger Bands-were developed and rigorously backtested using historical daily BTC-USD data from 2018 to 2025, incorporating transaction costs. Performance was evaluated against a passive Buy-and-Hold benchmark using standard metrics including annualized return, volatility, Sharpe ratio, and maximum drawdown. The results indicate that while the Volatility-Adjusted Momentum strategy demonstrated potential for drawdown reduction compared to Buy-and-Hold, its risk-adjusted returns (Sharpe ratio) were suboptimal due to highly restrictive volatility filtering leading to infrequent trading. The implemented Mean Reversion strategy failed to generate trades, highlighting parameterization challenges. Neither strategy, in its current form, offered a demonstrably superior risk-adjusted alternative to Buy-and-Hold over the tested period. The findings underscore the difficulty of systematically outperforming Bitcoin passively with simple quantitative models and suggest that future research should focus on parameter optimization, dynamic regime adaptation, and the implementation of more sophisticated machine learning models, which were identified as promising but deferred in this study.