Understanding and Using Trend Charts

After completing a simulation with multiple related forecasts, you can create a trend chart to view the certainty ranges of all the forecasts on a single chart. A trend chart summarizes and displays information from multiple forecasts, making it easy to discover and analyze trends that might exist between related forecasts.

Figure 42. Upward Trending Sales Figures, By Quarter

This figure displays a Trend Chart, showing an upward trending sales figures, by quarter.

Trend charts display certainty ranges for multiple forecasts in a series of colored bands. Each band represents the certainty ranges into which the actual values of the forecasts fall. For example, the band which represents the 90% certainty range shows the range of values into which a forecast has a 90% chance of falling. By default, the bands are centered around the median of each forecast.

The trend chart in Figure 42, Upward Trending Sales Figures, By Quarter displays certainty ranges on a quarterly basis over a three year period. Because the model contains quarterly forecast formulas dependent on the previous quarter’s results, the bands widen in the future. This occurs because the standard deviation of the forecasts increases or widens for each quarter. Trend charts like this one demonstrate the compounding of uncertainty that occurs as predictions are made farther and farther into the future.

You can customize trend charts to display the probability that given forecasts will fall in a particular part of a value range. For example, if a model contains forecasts related through time, you can use trend charts to view the certainty ranges for each forecast side by side. You can also compare at a glance the certainty ranges for an early time period and a later time period.

Trend charts are only meaningful when you have multiple forecasts that are related to each other.

The following sections describe how to create and work with trend charts: