Backtesting is the process of evaluating a trading method using historical data. By looking for statistical significance of how the method performed in the past, one can make an attempt to predict how a trading method might perform in the future.
Example Equity Curve during backtesting
Some popular trading platforms that include the ability to backtest trading methods include: NinjaTrader
, and MultiCharts
While backtesting can reveal a strength or weakness in a method, the results are not always correlated to future outcomes. You can use a large sample size
such as 5 years of market data, or more, and still fall victim to over fitting or curve fitting
your method to fit the historical data, instead of merely data mining
to a positive outcome.
[top]In-depth 2 hour webinar on backtesting
Big Mike and Manesh Patel discuss Automated Trading and Backtesting in-depth in a 2-hour webinar dated Thursday, May 10th 2012. Topics included: Human vs Automation, Data for Backtesting, Data Mining, Curve Fitting, In-sample vs Out-of-sample, Process Diagram, Discussion of prerequisites and common pitfalls, discussion of important performance metrics to consider during a backtesting, and much more.
The webinar is available here (for Elite Members):
This article is a stub
. Please edit the article to improve it and add additional details.
There are not currently any references cited in the article.