From backtest to live trading

TRADINGPROFITTEVOLE · FREE GUIDE 04/05

From backtest to live trading

The strategy tester is a laboratory. Live markets add execution delays, costs, errors and conditions that no simulation captures perfectly.

Requested price versus execution price

In a backtest, orders are processed according to the model. In live trading, milliseconds or longer may pass between submission and execution while prices move. The difference is called slippage. It can sometimes be favourable, but often matters most during volatile periods. Requotes, rejected orders and partial fills may also change a robot’s behaviour.

Spread, commissions and swaps

The spread is the difference between ask and bid prices. It may widen during news releases, rollover or thin liquidity. Commission and overnight swap charges can add further costs. Strategies with small average profits per trade are particularly sensitive. Before running one, I compare actual account costs with those assumed in the test.

Broker, platform and infrastructure

An identical EA may behave differently with different brokers because quotes, sessions, minimum volumes, stop levels and hedging/netting rules differ. A VPS can help continuity and latency but cannot eliminate execution problems. I check symbol names, decimal digits, contract sizes and the robot’s handling of open and closed markets.

A gradual, controlled transition

Before risking significant capital, I prefer observing the EA on a demo account and then, if appropriate, at reduced live exposure. I compare actual orders with expected orders, check platform logs and record anomalies. I also plan for disconnections, terminal restarts, positions left open and software updates. Automation reduces manual intervention, not the need for supervision.

A simple example

Suppose a robot aims for an average 8 points per trade in the tester. If real-world spread, commission and slippage add 5 points of cost compared with the model, its statistical edge may shrink dramatically. This simplified example illustrates why execution quality matters as much as signal logic.

Checklist before moving on

  • Have I checked the broker, symbol, contract specifications and account type?
  • Have I compared actual spread and commissions with the test assumptions?
  • Does the EA handle rejected orders, disconnections and restarts?
  • Is there a low-exposure observation period?
  • Do I know where to inspect logs, open positions and anomalies?
  • Do I have a safe procedure for stopping the EA?

Moving into live trading is where assumptions meet reality. A convincing backtest deserves a careful verification process, not an act of faith.

This guide is for educational purposes only. Trading involves the risk of losing capital. Historical data and simulations do not guarantee future results.

← Back to all free guides