BTC/USDT + ETH/USDT + BNB/USDT + SOL/USDT | 1-2x | 1000-10000 USDT | Risk level: low This model uses an N-rho reinforcement learning model to locate trades. It averages 40-120% annual return with a low mean drawdown, making it a great choice for people who have expectations of reasonable risk-adjusted returns. It has undergone regularization, k-fold cross validation and has evolving walk-forward optimization embedded such that the best parameters are selected. Use responsibly, cap leverage at maximum 2. Best used for 1x long and short with the position sizes split 25% between BTC, ETH, BNB and SOL.