OctoBot Tutorial Details Quantitative Trading Strategy Workflow

A comprehensive tutorial details the construction and validation of a quantitative trading strategy using OctoBot and OctoBot-Script, emphasizing an isolated development environment. The workflow involves configuring a rule-based strategy that integrates Relative Strength Index (RSI) oversold signals, Exponential Moving Average (EMA) trend confirmation, and Average True Range (ATR) for adaptive stop-loss and take-profit levels. This strategy is executed via OctoBot's native market-order and backtesting Application Programming Interfaces (APIs).

The process begins with retrieving historical Open, High, Low, Close, and Volume (OHLCV) data through OctoBot's data layer, which includes automatic exchange fallback mechanisms to ensure data availability. Following data acquisition, a multi-parameter grid search is performed over an in-sample period. This search systematically explores various combinations of parameters to identify the configuration that yields the highest excess return compared to a simple buy-and-hold strategy. The selected optimal parameters are then rigorously validated on a completely separate out-of-sample period. This validation step is crucial for assessing the strategy's generalization capabilities and detecting potential overfitting, where a model performs well on training data but poorly on new, unseen data.

The tutorial specifies key parameters for the trading strategy. The symbol for trading is set to "BTC/USDT", with a daily time frame ("1d"). Supported exchanges include Binance, KuCoin, OKX, Bybit, MEXC, and Kraken. The in-sample period for initial parameter tuning spans from January 1, 2019, to January 1, 2023. The out-of-sample period for validation extends from January 1, 2023, to June 1, 2025. The grid search explores "rsi_period" values of 7, 14, and 21; "rsi_threshold" values of 25, 30, and 35; and "tp_atr_mult" (take-profit multiplier based on ATR) values of 3.0 and 5.0. Fixed parameters include "ema_fast" at 50, "ema_slow" at 200, "atr_period" at 14, "sl_atr_mult" (stop-loss multiplier based on ATR) at 2.0, a "position_size" of 20%, a "min_offset_pct" of 1.0%, and a "max_offset_pct" of 40.0%.

Finally, the tutorial demonstrates how to extract OctoBot's backtest report data. This data is then utilized with the Pandas library for data manipulation and Plotly for interactive visualization. The analysis covers parameter sensitivity, portfolio performance metrics, price action, indicator behavior, and execution results. This interactive analysis is conducted within a Google Colab environment, allowing users to explore the strategy's performance dynamically. The setup involves creating a virtual environment directory "/content/octobot_env" and a working directory "/content/octobot_lab", utilizing OctoBot version "2.1.1" and Python version "3.12".