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

Strategy Backtesting

Evaluate a strategy on historical data — reproducibly and fully isolated from live trading — and get the trades, equity curve and performance metrics.

Where you will find this

Backtest setup (instrument, period, params), Backtest results (equity curve, metrics, trades).

Purpose

A backtest replays a strategy version over historical bars and reports what would have happened: each simulated trade, the equity curve, and a full metric set. Runs are immutable and reproducible — the same inputs always produce the same result.

No live orders, ever

Backtesting never touches a broker. Only non-draft strategy versions can be backtested, so you are always testing something concrete.

The Strategy Backtesting module enables users to evaluate trading strategies using historical market data.

It provides a controlled environment for simulating historical trading activity and measuring strategy performance without participating in live markets.

The module serves as the primary research and validation platform for trading strategies.

Objectives

The module will:

  • Execute trading strategies against historical market data.
  • Simulate historical trading activity.
  • Measure strategy performance.
  • Produce reproducible backtesting results.
  • Support comparison of multiple strategy executions.
  • Preserve historical backtesting records.
  • Provide performance information to authorised business modules.

Functional Scope

The module includes:

  1. Backtest Definition
  2. Historical Data Selection
  3. Strategy Selection
  4. Backtest Configuration
  5. Backtest Execution
  6. Trade Simulation
  7. Performance Measurement
  8. Result Management
  9. Backtest History
  10. Backtest Audit

Backtest Model

A backtest represents the execution of a trading strategy against a defined historical market dataset using a specified configuration.

Each backtest is an independent and reproducible business activity.

Backtest execution does not interact with live trading.

Strategy Selection

You can select one or more trading strategies for backtesting.

Strategy selection will use strategies maintained by the Strategy Management module.

Only approved strategy versions is eligible for execution.

Historical Data Selection

Every backtest will reference historical market data.

Selection criteria may include:

  • Instrument
  • Exchange
  • Date Range
  • Timeframe

Historical data is supplied by the Historical Market Data module.

Backtest Configuration

A backtest will maintain its execution configuration.

Typical configuration includes:

  • Strategy Version
  • Historical Dataset
  • Initial Capital
  • Trading Period
  • Transaction Cost Model
  • Slippage Model

The design remains extensible for additional configuration parameters.

Trade Simulation

The platform will simulate trading activity based on strategy-generated trading decisions.

Simulation may include:

  • Entry Transactions
  • Exit Transactions
  • Position Lifecycle
  • Order Lifecycle

Simulation remains independent of live broker connectivity.

Performance Measurement

The platform will calculate performance metrics for every completed backtest.

Typical measures include:

  • Total Return
  • Annualised Return
  • Net Profit
  • Gross Profit
  • Gross Loss
  • Win Rate
  • Profit Factor
  • Maximum Drawdown
  • Sharpe Ratio
  • Calmar Ratio
  • Number of Trades

The platform remains extensible for additional performance metrics.

Backtest Results

Every completed backtest will produce a complete result set.

Typical outputs include:

  • Trade History
  • Position History
  • Equity Curve
  • Performance Metrics
  • Execution Summary

Backtest results remains available for future analysis.

Backtest Comparison

You can compare multiple completed backtests.

Comparison may include:

  • Performance Metrics
  • Trade Statistics
  • Risk Statistics
  • Return Characteristics

Comparison does not modify historical backtest results.

Backtest History

Historical backtests is preserved.

History includes:

  • Configuration
  • Strategy Version
  • Historical Dataset
  • Execution Timestamp
  • Result Summary

Historical records remains immutable.

Business Rules

  • Every backtest will reference one strategy version.
  • Every backtest will reference one historical dataset.
  • Backtesting will never interact with live trading.
  • Historical market data remains unchanged during execution.
  • Backtest results is reproducible using the same inputs.
  • Completed backtests remains available for future analysis.
  • Historical backtest records will never be deleted.

Audit Requirements

The platform will maintain an audit trail for backtesting activities.

Audit events includes:

  • Backtest Creation
  • Backtest Execution
  • Configuration Changes
  • Result Generation
  • Result Comparison
  • Historical Archive

Audit records remains immutable.

Out of Scope

The following are outside the scope of this module:

  • Live Trading
  • Broker Connectivity
  • Order Execution
  • Live Risk Management
  • Portfolio Management
  • Strategy Optimisation

These responsibilities belong to their respective business modules.

A backtest is gross unless you say otherwise

Transaction costs are off by default, so a run reports gross figures until costs are switched on. This matters most for strategies that trade often, where costs change not only the total but the shape of the equity curve.

  • The cumulative P&L curve leads from zero, so the first trade’s contribution is visible rather than absorbed into the starting point.
  • A result over one period describes that period. It is not evidence that the rule will make money, and nothing in a backtest models the queue position or the slippage it was not given.
  • Past intraday option bars are largely unavailable from vendors, so an options backtest over intraday data may have gaps rather than results. A gap is often the vendor’s, not a fetch that failed.

Run a backtest

  1. Pick the strategy (and version) and the instrument.
  2. Choose the date range, timeframe and starting capital.
  3. Run it — the engine replays the data and produces trades, an equity curve and metrics.
  4. Review the results, then optimise, validate or compare against other runs.