Continued from previous post http://www.forexfactory.com/showthre...48#post4243148 .....
The result of the optimization (image below) shows that the ideal simulation it found was to set the stop amount at 138 pips and the limit amount at 178 pips which would result in a net profit of about $2046. Not bad when you compare this to the previous backtest with the default settings.
Next, add the new stop and limit into the strategy and run the backtest. Below is an image of the results which we can compare to the previous backtest with the default inputs.
As you can see, the net profit has jumped to $2,046, an increase of nearly 7X previous backtest, while the number of trades placed is a little less than double the previous backtest. I hope this example of changing the inputs and optimizing the numbers has shown you how the inputs you select can greatly impact how an indicator performs.
A word of caution!!!!
This is not the key to finding the holy grail of trading!!! This type of backtesting optimization will find the optimal results over the past period selected on your chart. This by no means indicates that future performance will be the same as past performance. It simply indicates the optimal settings over the previous period selected. Additionally as you increase or decrease the amount of data you backtest over, it will also change the backtest results. To give a quick example, I doubled the historical data on the chart, and the results of the backtest for the previously optimized data turned negative!
Nonetheless, I hope this gives you some insight into how changing indicator variables will impact the performance of the indicator. There are tools out there to help you such as optimization in Strategy Trader. In the end, backtested results will not mean the same results can be expected in future performance. What I would do is look at what type of market condition my strategy performed well in during the first backtest, then look at what happened in the second back-test. Identify what caused the losing trades and see if you can create some type of filter that negates trading during the periods that tend to cause drawdowns in your strategy.
That was a long post. Thanks for reading!
The result of the optimization (image below) shows that the ideal simulation it found was to set the stop amount at 138 pips and the limit amount at 178 pips which would result in a net profit of about $2046. Not bad when you compare this to the previous backtest with the default settings.
http://img30.imageshack.us/img30/549...4201012290.png
Next, add the new stop and limit into the strategy and run the backtest. Below is an image of the results which we can compare to the previous backtest with the default inputs.
http://img510.imageshack.us/img510/5...1420101234.png
As you can see, the net profit has jumped to $2,046, an increase of nearly 7X previous backtest, while the number of trades placed is a little less than double the previous backtest. I hope this example of changing the inputs and optimizing the numbers has shown you how the inputs you select can greatly impact how an indicator performs.
A word of caution!!!!
This is not the key to finding the holy grail of trading!!! This type of backtesting optimization will find the optimal results over the past period selected on your chart. This by no means indicates that future performance will be the same as past performance. It simply indicates the optimal settings over the previous period selected. Additionally as you increase or decrease the amount of data you backtest over, it will also change the backtest results. To give a quick example, I doubled the historical data on the chart, and the results of the backtest for the previously optimized data turned negative!
http://img153.imageshack.us/img153/1...rgerperiod.png
Nonetheless, I hope this gives you some insight into how changing indicator variables will impact the performance of the indicator. There are tools out there to help you such as optimization in Strategy Trader. In the end, backtested results will not mean the same results can be expected in future performance. What I would do is look at what type of market condition my strategy performed well in during the first backtest, then look at what happened in the second back-test. Identify what caused the losing trades and see if you can create some type of filter that negates trading during the periods that tend to cause drawdowns in your strategy.
That was a long post. Thanks for reading!