Backtesting Your Algo

This guide walks you through backtesting your algorithmic trading strategies with Datafye. The approach you take depends on whether you're using your own containers or Datafye's managed Algo Container with integrated backtesting engine.

Choose Your Approach

Backtest algos in your own containerized environment by driving the backtest workflow via REST APIs.

Best for:

  • Developers with existing algo infrastructure

  • Teams with custom backtesting frameworks

  • Use cases requiring specific testing methodologies

  • Maximum control over the backtest process

You'll learn:

  • Fetching historical tick data via REST API

  • Resetting environment state between backtests

  • Starting and controlling tick replay via REST API

  • Implementing your own backtest logic and metrics

  • Managing backtest execution programmatically

Scenario: Foundry: Data Cloud Only

Backtest algos using Datafye's managed Backtesting Engine, with parallelized execution, genetic algorithm optimization, and comprehensive scorecarding.

Best for:

  • Developers using the Datafye Algo Container

  • Teams wanting integrated backtesting tools

  • Use cases requiring parameter optimization

  • Focus on strategy refinement, not infrastructure

You'll learn:

  • Configuring backtest parameters in algo descriptors

  • Running single and parallelized backtests

  • Using genetic algorithm-based parameter optimization

  • Analyzing performance scorecards

  • Interpreting backtest results and metrics

Scenario: Foundry: Full Stack

What You'll Need

Before starting, ensure you have:

  • Foundry environment provisioned - Either Data Cloud Only or Full Stack

  • Historical data configured - Data provider with sufficient history for backtesting

  • Algo ready to test - Your trading strategy implemented and deployed

  • Performance expectations - Metrics you'll use to evaluate success

Key Differences

Aspect
Own Container (Data Cloud Only)
Datafye Container (Full Stack)

Workflow

Manual via REST API

Automated by backtesting engine

Control

Complete programmatic control

Declarative configuration

Execution

Single backtest at a time

Parallelized batch execution

Optimization

Build your own

Genetic algorithm built-in

Metrics

Define your own

Comprehensive scorecarding

Infrastructure

You manage

Datafye manages

Last updated