> Optimizing Your Expert Advisor
The Universal BOSS has multiple input parameters that must be optimized for each currency pair. No two currency pairs will have the same identical setup. We will optimize each, for maximum profit and minimum drawdowns to our account.
Different platforms or brokers may also react differently to a certain set of input parameters. One set of successful input parameters on one broker, may not reproduce the exact same results on another. Cause - The Strategy Tester Back-Testing Data is not identical.
The process of optimization is a vital part of owning and maintaining a good Expert Advisor. Keeping your EA tuned and working well is very important. You should never use the default settings of your EA for automated trading without running the full spectrum of optimization backtests!
Let's Keep It Simple
The staff at www.forex-expert-advisor-free.com realizes that most individuals are new to Forex Trading and may not know how to optimize their Expert Advisor. Therefore, we have made every attempt to make this process as simple as possible.
Step 1 - Open up your Strategy Tester
At the top of your trading platform menu - click on the VIEW tab. Then Click on the STRATEGY Tester tab or simply hit CTRL+R. This will open the STRATEGY TESTER at the bottom of the page. Feel free to expand the size. Should look like this:

Step 2 - Setting it up to OPTIMIZE

Important to note that the following setup will determine the quality and time of the optimization backtest. Average Full Optimization should take approximately 4.5 hrs per currency pair.
1. Expert Advisor -- Lists the current expert advisors you have stored on your platform (in your experts folder)..
2. Symbol -- Currency Pair to be optimized.
3. Model -- There are 3 different backtesting (optimization) model's we can use. The difference between them is speed (time) vs. accuracy. We will begin using the CONTROL POINTS model which gives us the median of speed vs. accuracy. Once we have identified the top two or three input parameters....write them down and begin testing one by one with the EVERY TICK model. This will give us the most accurate backtest available. Remember to unclick the Optimization button (6) when running the final (accurate) phase. To ensure a high quality backtest, be sure your modeling quality is 90%. Anything else will have too many data error's and inaccurate calculations. Please visit the FAQ (in the main menu) for a walkthrough on how to achieve 90% modeling quality .
4. Date -- Enter beginning and ending dates for backtest optimization. We have set up a 1 year backtest above. Shortening / lengthening this parameter will slow / speed up your optimization greatly.
5. Period -- Set to H4.
6. Optimization Tab -- Check Box for full optimization of multiple parameters. Uses Optimization Results & Optimization Graphs. Uncheck Box for single optimization backtest of one set of parameters. Uses the Graph & Report functions.


Step 3 - Expert Properties Tab & Input Parameters to Optimize
Next, Click on the Expert Properties Tab. Look at the TESTING & INPUTS tabs. They should look like this:


Step 4 - Variable, Value, Start, Step & Stop - How they work
Simply Check or Click on the Variable Boxes that you want to optimize (arrows on the left of diagram - Variable Column). The Value Tab (Green Highlight) of Input parameters will run a single Graph and Report. Whereas, the Start, Step and Stop (Yellow Highlights) will run multiple backtests simultaneously to find the best settings for that currency pair (Optimized Settings).

In the Start area, input your beginning range to be tested. The Step area describes the increase in value to that of the Start variable. The Stop area is the ending value to be tested. For example, lets look at the MASetting - Start 2, Step 1, Stop 7. The results received will be 2,3,4,5,6 and 7 of the MASetting Variable. When we check additional boxes, the calculations will increase exponentially as the Strategy Tester computes the results of multiple 'VARIABLES' and the START, STEP's and STOP ranges associated there with.
Step 5 - Sample Test - With the Optimization Box Checked
This process will run multiple backtests to determine the best possible settings for Expert Advisor and Currency Pair.

Input the variables as provided in the Recommended Backtest Ranges. Click on the Optimization Graph, Optimization Results & Settings to view progress. The Green Optimization bar will become full upon completion of test. Next, the results of the backtest: - Diagram Below:

The first arrow (top left) represents how many passes: are " in progress vs total passes to be completed"
The second arrow represents the estimated time to completion: current time vs time needed to finish job
Optimization Results will list the optimization findings. Best Profit, Trades, Profit Factor, Drawdown, etc. You may cycle through each column by clicking on the corresponding tab at the top to view the best result for each column.
Optimization Graph will give you a quick view graph
Step 6 - Retrieving Best Data
Now that our strategy tester has finished the full backtest, let's look at our results.

1. Best Profit -- Click on the PROFIT tab (once or twice - it lists A-Z and Z-A).
2. Lowest Drawdown % -- Click on the DRAWDOWN tab.
3. The COMPLETE SET of INPUT Variables -- All input variables (SET) associated with desired result. If you cannot see them all, simply move the cursor up and down on that set and it will highlight.
FIND THE BEST COMBINATION OF THE TWO
Step 7 - Let's Plug It In
Now that we have our best set, let's test it individually. Click Settings and Uncheck Optimization Box on your Strategy Tester.
Click Expert Properties and Input the SET into the VALUE column. Click OK when complete.
On your main Strategy Tester screen, click START.
Results will be shown in the GRAPH & REPORT tab.
Step 8 - Final Step
It is crucial that we now take our best backtest set's and test again using the "EVERY TICK MODEL" to derive a high quality backtest. Modeling quality should be maxed out at 90%. If the modeling quality is less, please refer to the FAQ - " How can I achieve better backtest modeling quality?"
The CONTROL POINTS method back test & EVERY TICK method will not yield exactly similar results. Imagine, if you will, that the best optimized input's for a particular currency pair is a picture of a landscape. The CONTROL POINTS method will allow us to see the picture, but it's blurry and not clear. The EVERY TICK method will show us the true color and high definition in that picture. We use the CONTROL POINTS method because it will narrow down the parameter's much faster than the EVERY TICK method.
IMPORTANT TO NOTE: Take the BEST PROFIT &/or BEST MIN DRAWDOWN SETTINGS. Other unchecked boxes in your back tester will remain constant. Right-Click on Optimization Results and use the "Copy ALL" function. Paste to an excel sheet for later review. You may save the "BEST PROFIT" settings & "BEST MIN DRAWDOWN" settings seperately by clicking on respective column tabs (top) and using "Copy ALL" function. This may save you time when sifting through 100's of back tests.
|
|
!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!-->!--> !-->!-->!--> >>> |

