MyAlgoGuru Help
Education Glossary Open the app
Help/Reference/Research & Analytics

Strategy Parameter Optimisation

Sweep a strategy’s parameters over a grid of values and rank every combination by your chosen objective.

Where you will find this

Parameter optimisation.

Purpose

Optimisation runs a backtest for each parameter combination in a grid, then ranks them by an objective — maximum return, best Sharpe, minimum drawdown, and so on — and surfaces the best parameters. It never changes the strategy itself.

The Strategy Parameter Optimisation module enables users to identify parameter combinations that produce the most favourable historical performance for a trading strategy.

The module executes multiple backtests using different parameter configurations and compares the resulting performance.

The module supports systematic strategy tuning while preserving the underlying strategy logic.

Objectives

The module will:

  • Execute multiple backtests using varying parameter combinations.
  • Evaluate strategy performance across parameter sets.
  • Identify parameter combinations meeting defined objectives.
  • Preserve optimisation history.
  • Support comparison of optimisation results.
  • Provide optimisation results for further research.

Functional Scope

The module includes:

  1. Optimisation Definition
  2. Parameter Definition
  3. Parameter Range Management
  4. Optimisation Configuration
  5. Optimisation Execution
  6. Result Ranking
  7. Optimisation Comparison
  8. Optimisation History
  9. Optimisation Audit

Optimisation Model

An optimisation represents the repeated execution of a single strategy using different parameter combinations over the same historical dataset.

The optimisation process does not modify the underlying strategy definition.

Each optimisation will represent an independent research activity.

Strategy Selection

Each optimisation will reference one strategy.

Only approved strategy versions is eligible.

The optimisation process does not alter the selected strategy.

Parameter Definition

You define the parameters to be optimised.

Typical parameter characteristics include:

  • Parameter Name
  • Start Value
  • End Value
  • Increment
  • Allowed Values

The design remains extensible for additional parameter types.

Historical Dataset

Every optimisation will reference one historical dataset.

Dataset selection may include:

  • Instrument
  • Exchange
  • Timeframe
  • Date Range

Historical data remains unchanged during optimisation.

Optimisation Configuration

Each optimisation will maintain its configuration.

Typical configuration includes:

  • Strategy Version
  • Historical Dataset
  • Parameter Definitions
  • Initial Capital
  • Transaction Cost Model
  • Slippage Model
  • Optimisation Objective

Optimisation Execution

The platform will execute multiple backtests using all eligible parameter combinations.

Each execution is independent.

Every execution will generate its own performance results.

Optimisation Objectives

The platform will support different optimisation objectives.

Typical objectives include:

  • Maximum Return
  • Minimum Drawdown
  • Maximum Sharpe Ratio
  • Maximum Calmar Ratio
  • Maximum Profit Factor
  • Highest Win Rate

The design remains extensible for additional objectives.

Result Ranking

Completed optimisation results is ranked according to the selected optimisation objective.

Ranking may include:

  • Overall Rank
  • Performance Score
  • Risk Score
  • Stability Score

Ranking criteria remains configurable.

Optimisation Results

Each optimisation will produce:

  • Parameter Combination
  • Performance Metrics
  • Trade Statistics
  • Risk Statistics
  • Execution Summary

Results remains available for future analysis.

Optimisation History

Historical optimisation activities is preserved.

History includes:

  • Strategy Version
  • Parameter Definitions
  • Historical Dataset
  • Optimisation Configuration
  • Result Summary
  • Execution Timestamp

Historical records remains immutable.

Business Rules

  • Every optimisation will reference one strategy.
  • The underlying strategy remains unchanged.
  • Every optimisation will reference one historical dataset.
  • Every parameter combination will execute independently.
  • Historical market data remains unchanged.
  • Optimisation results is reproducible.
  • Historical optimisation records will never be deleted.

Audit Requirements

The platform will maintain an audit trail for optimisation activities.

Audit events includes:

  • Optimisation Creation
  • Configuration Changes
  • Optimisation Execution
  • Result Generation
  • Result Ranking
  • Historical Archive

Audit records remains immutable.

Out of Scope

The following are outside the scope of this module:

  • Strategy Editing
  • Live Trading
  • Order Execution
  • Broker Connectivity
  • Portfolio Management
  • AI Strategy Generation

These responsibilities belong to their respective business modules.