Brain Machine Learning Middle-Frequency Signals on Cryptocurrencies
This approach combines rigorous quantitative modeling with flexible parameterization, enabling consistent performance across multiple cryptocurrencies. The strategies exploit market inefficiencies through optimized parametric models that capture time-based dynamics, combining both momentum and mean-reversion effects.
The systematic strategies, driven by hourly signals for the most liquid cryptocurrencies such as ETHUSDT and BTCUSDT, have shown robust performance, delivering a Sharpe ratio well above one.