Skip to content

Trading context

Backtest jobs

A backtest job evaluates a selected strategy revision across a historical time range. It uses historical market data and simulated execution through the Venue Simulator.

Backtest jobs help you compare revisions and settings before live operation. They are evidence under recorded assumptions, not performance guarantees.

A backtest job is a finite evaluation that:

The runtime path is:

Historical market data server -> Strategy engine -> Target Position Executor -> Venue Simulator

In the current target-position execution path, the selected revision and Target Position Executor are both part of what the job evaluates.

Backtest job path Backtest jobs are finite evaluations that bind job settings, replay historical market data, and record results under simulator assumptions.
Job settings
Selected revision Revision-owned venue and instrument
Strategy Variable values Run-provided exposed node parameters
Historical time range Finite start and end
Starting balances Initial simulated account state
Fees, slippage, latency Recorded simulator settings
Data source
Historical market data server Historical data for the requested range
Data completeness Complete segments and gaps
Evaluation and execution
Strategy engine Evaluates the selected compiled revision
Target Position Executor Turns target positions into order activity
Venue Simulator Applies simulator assumptions
Result
Simulated fills and account state Recorded under assumptions
Result metrics PnL, drawdown, Sharpe, positions, events, completeness
Finite lifetime A backtest job terminates when the requested historical evaluation completes or fails.
Interpretation Backtest job results are evidence under recorded settings and assumptions, not performance guarantees.
Legend
  • Market data Live or historical market data sources.
  • Strategy engine Runtime evaluation of the selected compiled revision.
  • Target Position Executor Target-position intent converted into order activity.
  • Venue Simulator Simulated execution destination and assumptions.
  • Account state Live or simulated positions, fills, balances, and order state.
  • Results Recorded metrics, telemetry, and review context.
  • Backtest path Finite backtest job boundary.
  • Revision-owned Immutable strategy revision logic and settings.
  • Run settings Run-provided values, balances, ranges, or assumptions.

A strategy is the long-lived container for a trading idea. A revision is one saved version of that strategy.

Backtest jobs attach to revisions:

  • One strategy can have many revisions.
  • One revision can have many backtest jobs.
  • Each job points back to the selected revision and its fixed logic.
  • Each job records the Strategy Variable values and job settings used for that evaluation.

This makes results easier to interpret. A result is not just a chart; it is a recorded relationship between the selected revision, run-provided values, historical range, starting balances, simulator settings, and data coverage.

Backtest job settings include:

  • Historical time range.
  • Strategy Variable values.
  • Starting balances.
  • Fees.
  • Slippage.
  • Latency.
  • Simulator assumptions.

Venue and instrument belong to the revision. They stay visible with the job because they affect data selection, simulator assumptions, and result interpretation.

Backtest jobs move through finite lifecycle states:

StateMeaning
draftingThe user is filling out the job parameters.
provisioningStructure is preparing the selected revision and runtime path.
runningHistorical data is driving the selected revision.
completedThe job finished and results are available.
failedThe job stopped before completion and exposes an error reason where available.

The UI mirrors this lifecycle so you can monitor progress, inspect failures, and review results when the job completes.

A completed backtest job records the result and the context that produced it.

Common result fields include:

Metrics describe what happened under the recorded settings. Review assumptions and data completeness before comparing jobs or promoting a revision.

Read result fields together:

  • PnL describes profit and loss under the recorded settings.
  • Max drawdown and high-water mark describe the shape of the evaluated result path.
  • Continuous-time Sharpe ratio is a risk-adjusted comparison aid.
  • Rolling position shows how simulated exposure changed over time.
  • Trade count and event count help explain activity level.
  • Data completeness shows whether the requested historical range had enough available data for interpretation.

For detailed result interpretation, see Backtest result metrics.

Backtest job results are affected by:

For simulator behavior, see Venue Simulator methodology. For assumption and coverage details, see Simulator assumptions and data completeness.

Backtest jobs support the research and promotion workflow:

  • Evaluate one revision over a historical time range.
  • Compare revisions with job settings attached.
  • Review result metrics, assumptions, and data completeness.
  • Create a new revision when the result points to a strategy change.
  • Start a live deployment only after selecting the revision, Strategy Variable values, connected account, venue context, permissions, and deployment settings.

A backtest job does not become a live deployment by itself. Promotion is an explicit user action.

Use backtest jobs as one input to your decision-making process.

Backtest job results do not predict future results, guarantee returns, guarantee risk reduction, or guarantee live execution quality. Market conditions, venue behavior, liquidity, fees, latency, and fills can differ from the recorded simulation.