MonteCarlo

Monte Carlo simulation is a powerful testing technique that can be employed to predetermine the probable limits of likely future system performance. It is particularly useful for testing non-linear payoff functions like those exhibited in mechanical trading systems. Metronome HistoryMaker is able to generate price data over any date range, past and future. Hence a Monte Carlo simulation of trading system performance may be obtained as unlimited price path scenarios may be generated and tested against. The ability to create new data takes walk-forward system testing to a new level; a system may be "walked" into the future! A Monte Carlo simulation involves using today’s price as a seed, then creating many possible future scenarios. These scenarios are then traded by the system under test and appropriate performance measures are made for each scenario. The estimated future result is then determined by calculating the expected value and variance of each performance measure after performing numerous simulations. The following diagram explains the process involved in using generated scenarios to test a trading system's likely future performance.

Scenario: Generated by HistoryMakerä Result: examples are; Annual return (%) Winning Trades (%) Average Winning trade / Average Losing Trade ($/$) Maximum Drawdown ($) Volatility of returns (%) Consolidated Result: Average and Standard Deviation of results.