Running Backtest Strategies in Simulated or Broker-Linked Trading

Summary

The document describes a platform feature for running strategy research code as a live or simulated trading service. A strategy is supplied with its source, name, frequency, starting capital, and optional benchmark. Without a trading API it runs as a simulation; with a broker-facing API it can route orders to an account. A signal mode determines whether strategy state and order queries come from the simulator, with simulated fills forwarded to the account, or directly from the brokerage interface. A recovery setting controls whether execution begins from the current time, a prior stop point, or a specified session point.

The article also lists order and trade-log utilities, including canceling all orders and retrieving open orders. It cautions that startup timing affects which session callbacks run, and that an explicit market or limit order policy matters because computed prices may not satisfy broker price precision. The example demonstrates configuration and queries, but provides no performance evidence or operational safeguards such as monitoring, failure recovery, or reconciliation procedures. Readers should treat it as platform guidance, not evidence that backtested behavior will match live execution.

Key ideas

• Research code can be run in simulation or connected to a brokerage trading API.
• Signal mode controls whether strategy state comes from simulated trading or the brokerage account.
• Recovery settings determine how execution resumes after a strategy starts.
• Order policy, session startup timing, and account data synchronization affect live operation.
• The document explains platform mechanics but offers no evidence that backtest results transfer to live trading.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.