Skip to content

Operating paths

Paper deployment methodology

A paper deployment is a long-running evaluation path for a selected strategy revision. It uses live market data and sends order activity to the Venue Simulator instead of a live venue. In the current target-position execution path, a paper deployment evaluates both the selected revision and the Target Position Executor behavior under simulator assumptions.

The runtime path is:

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

Paper deployments let users observe real-time strategy behavior under simulated execution assumptions before deciding whether a revision is ready for live operation.

Paper deployment path Paper deployments use live market data and the Venue Simulator as the execution destination.
Data source
Live market data server Current venue market data
Evaluation
Strategy engine Selected compiled revision plus Strategy Variable values
Target positions Desired exposure emitted by the revision
Execution model
Target Position Executor Turns target positions into order activity
Simulator assumptions Fees, slippage, latency, fill model, liquidity assumptions
Simulated destination
Venue Simulator Shared simulated execution destination
Simulated fills and account state Evidence under assumptions
Lifetime A paper deployment is long-running and continues until the user stops it or a failure path ends it.
Interpretation Paper deployment results are evidence under explicit simulator assumptions, not guarantees of live execution behavior.
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.
  • Paper path Long-running paper deployment boundary.
  • Run settings Run-provided values, balances, ranges, or assumptions.
  • Target positions Desired exposure emitted by strategy evaluation.

A paper deployment evaluates how a selected revision behaves as live market data arrives.

It keeps the strategy path close to live operation:

The execution destination is the key difference from a live deployment. Paper deployments do not send order messages to the live exchange gateway.

The paper deployment path stays close to live operation because live market data still drives the selected revision. The difference is that account state, fills, and order state are simulated by the Venue Simulator rather than produced by a live venue account.

For the shared executor boundary, see Target-position execution model.

Paper deployments use revision context and deployment context.

ContextExamples
Revision contextStrategy, revision, fixed node parameters, validation state, compile output, venue, and instrument.
Run contextStrategy Variable values, selected simulated balances or account state, deployment owner, start time.
Simulator contextFees, slippage, latency, liquidity assumptions, fill model, and simulator limitations.
Live data contextCurrent market data availability and any venue data issues visible during the deployment.

Strategy Variables are concrete values supplied when the deployment starts. They let users evaluate a revision with run-specific values without changing fixed node parameters in the revision.

Paper deployments continue running until the user stops them or a failure path ends the deployment.

Common lifecycle states:

  • drafting: the deployment parameters are being specified by the user.
  • provisioning: Structure is preparing the selected revision and runtime path.
  • running: the deployment is active.
  • stopping: Structure is shutting the deployment down.
  • stopped: the deployment is no longer running.
  • failed: the deployment could not start or continue.

Because paper deployments are long-running, review them through health, logs, simulated account state, positions, target exposure, event volume, latency, and status changes over time.

Paper deployments show real-time strategy behavior under simulated execution assumptions.

They are useful for answering:

  • Does the revision behave as expected as live data arrives?
  • Do Strategy Variable values produce the intended target positions?
  • Does the Target Position Executor produce order activity consistent with the strategy’s target positions?
  • How do simulated positions, PnL, event volume, latency, and logs evolve during the deployment?
  • Which simulator assumptions affect the observed result?

Paper deployment results are not live execution evidence. They use the Venue Simulator, and the current Venue Simulator does not model the full exchange order book. Treat paper deployment behavior as evidence under explicit assumptions, not a guarantee that a live deployment will produce the same fills, account state, or order behavior.

For assumption details, see Simulator assumptions and data completeness.

Paper deployments and backtest jobs both use the Venue Simulator.

PathData sourceExecution destinationLifetime
Paper deploymentLive market data serverVenue SimulatorLong-running deployment
Backtest jobHistorical market data serverVenue SimulatorFinite job

Use a paper deployment when you want to observe real-time behavior without sending orders to a live venue. Use a backtest job when you want to evaluate a revision over a historical time range.

Paper deployments and live deployments both use live market data.

PathData sourceExecution destinationLive venue orders
Paper deploymentLive market data serverVenue SimulatorNo
Live deploymentLive market data serverLive exchange gatewayYes

Use a live deployment only after selecting the revision, Strategy Variable values, connected account, venue context, permissions, and deployment settings intended for live operation.

A paper deployment does not automatically promote itself to a live deployment.

Promotion is an explicit user decision. The user reviews paper deployment behavior, simulator assumptions, logs, simulated positions, and run context before selecting a revision and connected account for live operation.