Backtesting

What does your historical test leave out?

Replay strategy rules while keeping data quality, costs and execution assumptions in view.

Specialist inspecting historical trade markers and test assumptions

A test record, not a forecast

Backtesting applies your configured rules to historical market data and records the signals and simulated orders they would have produced under stated assumptions. Setup usually takes several working days once the strategy and data source are defined. We check instrument coverage, time zones, transaction-cost inputs and rule versions before generating a report. You receive the test configuration, an event-level record and a summary of assumptions. Historical tests cannot reproduce every live fill, outage or market condition, and past performance does not predict future results.

Analyst comparing historical events and time-zone settings

A test that exposed a timing mismatch

An independent trader wanted to examine a rule built around the close of a daily price bar before connecting it to a broker. During a one-week configuration review, we compared the rule's stated timing with the data source and found that its bar close used a different time zone. We documented the difference, adjusted the test setting with the trader's approval and reran the event log. The deliverable was a traceable test configuration, not a recommendation to trade.

Scope
One daily-rule configuration
Duration
One working week
Reviewed
Time zone and bar timing
Delivered
Configuration and event log

When a test is worth running

A test is useful when a rule can be written precisely and the available data covers the instruments and period you want to examine. It is less informative when the rule depends on unrecorded judgement, thin historical data or assumptions about fills that cannot be checked. We show those limitations alongside the output rather than treating a chart as a verdict. You should compare multiple periods, inspect individual events and decide whether the remaining uncertainty is acceptable before considering live execution.

Read beyond the summary chart

Need a test with stated assumptions?

Tell us which rules, instruments and historical period you want to examine.

Ask about backtesting