Step 1: Get Data
The first step is to use the Data window to get data about the security you wish to trade and other data you think may drive (influence) your traded security.  Profit enables you to access data in Metastock format or text files where each variable is in a column and each row is a day.  You can access multiple text files and Metastock formatted data simultaneously.  Profit synchronizes the data to date align it and fills gaps where data is missing, particularly if you use data from different exchanges in various countries, thus having different holidays.  The data window tells you information about each security and the securities together and enables you to set the dates for the period you wish to import.  By default, Profit will import up to 4 years of data, which offers a balance of recent quantity.  You can export data here to use in other software, whether it be trading packages or Excel.  For our international friends, Profit is international settings sensitive so you can import text data in your local format.
 

Step 2: Create Indicators
The next step is to create indicators.  You create a predicted variable, something similar to the change in price of the traded security 2 days in advance.  Why 2 days?  This is so that you can download data today, take a position tomorrow and then take gains the day after that.  Profit's Indicator Building window offers you a palette of functions you can apply to your imported data series so that you can create a wide variety of indicators, with permutations in the billions.  To make it easier, Profit has built in equity based indicator optimization to find combinations of mathematical functions to apply to your data to maximize the equity power of your indicators.  You can also chart the indicators and see their equity performance (pro/elite versions only) as though you were trading the indicator itself directly.

In Profit Pro and Elite, you have more capabilities and indicators to choose from.
 

Step 3: Key Indicators (Optional)
In Profit Pro and Elite you can perform highly sophisticated "Causal Analytics" (CA) and "Principle Variables Analysis" (PVA) to see to what extent your indicators drive the desired timing signal.  This advanced analytics walks your variables through time measuring cause and effect between your indicators and your timing signal.  This tells you the truth about the value of your indicators so you are not "spoofed" into believing you have excellence only to learn otherwise when reviewing your account balance.  No other vendor gives you this important capability.
 

This capability offers you a "Causal Map" that shows the strength of the cause and effect relationship vs. time so you can see that variable X influences your timing signal Y bars in the future.  The indicators in the Yellow zone are statistically likely to have a random effect on your market movement forecasts and thus have questionable utility.  Important to know!
 

Step 4: Build Models
Now that you have solid indicators, it's time to build predictive models.  These models translate today's indicators into tomorrow's positions in a manner that is profitable, rather than accurate.  Doing this can be as simple as pushing the "Go" button.  Profit fits the models using an earlier portion of your data and will hold-back, by default, 25% of the most recent data for "out-of-sample" testing.  This is so you can see the unmodeled hypothetical trading performance of your system. 

While modeling, Profit will search through combinations of types and structures of our advanced "Mesh" predictive modeling algorithms and through combinations of indicators to find the most profitable models possible.  While running you can see the equity curve and statistics of the models being built.  You can choose the objective of the modeling, but Profit is generally seeking upward trending smooth equity curves.

Profit brings in a set of models that are combined into a trading "committee" who vote to create a "System Model".  This gives your systems more robustness.

In Profit Pro and Elite, you have the ability to control the types of models built, how many inputs to use, model diversity, date range to use, and can use more than one CPU to do the modeling.
 

Step 5: Evaluate Results
Before using Profit's suggestions in your trading, you need to view your system's hypothetical trading performance on data unseen in the processes above.  Many trading tools call this "back testing", but we test forward by creating systems on older data then seeing how they trade as they "play forward".  The screen to the left is a VIX system that was created four years ago.  The green chart is a graph of the hypothetical equity performance of this system and how it continues to make money to this day (presuming one could directly trade the VIX).  You can also see your performance in more traditional Open-High-Low-Close charts such as below.  The blue lines are the OHLC with trade indicators (red for short, green for long), then the timing signal issued by Profit's System model is in yellow and the equity curve is in red.