Locked lesson.
About this lesson
Calculate the upper and lower confidence bound.
Exercise files
Download this lesson’s related exercise files.
09-03-Add Confidence Levels-Start.xlsx26.9 KB 09-03-Add Confidence Levels-Complete.xlsx
33.1 KB 9.03 add-confidence-levels - Exercise.docx
43.1 KB Exercise - Add Confidence Levels.xlsx
46.7 KB 9.03 add-confidence-levels - Exercise solution.docx
142.9 KB Exercise Solution - Add Confidence Levels.xlsx
51.4 KB
Quick reference
Add Confidence Levels
Add an Upper and Lower Confidence Bound to the Forecast.
When to use
We add confidence levels to our forecast whenever we want to see the upper and lower range where all predicted values will fall.
Instructions
Excel predicts future values with a confidence level of 95% by default. That means, that 95% of all predicted values will fall between the lower confidence bound and the upper confidence bound.
It is useful to define the confidence level and the upper and lower bound and add this information to our line chart.
Add a Confidence Level
We first need to add the confidence level.
- Add a new column with the heading 'Confidence Level'.
- Scroll down and type =FORECAST.ETS.CONFINT
- Copy the formula down.
- Apply Number Formatting.
We can now use the confidence level to calculate the upper and lower confidence bound.
- Add two more columns to the dataset with the headings, 'Upper Bound' and 'Lower Bound'.
- Scroll down and type the following calculation.
In this formula, we are adding the confidence level to our seasonal forecast value. The result is our upper confidence bound.
- Copy the formula down.
- In the next column, type the following calculation.
This time we are subtracting the confidence level from the seasonal forecast value. The result is our lower confidence bound.
Using the first set of numbers as an example, we can see that 95% of predicted values will fall between 15,633 and 24,774.
- Select the Dates, the Values, the Seasonal Forecast Data, and the Upper and Lower Bound Data.
- Hold down CTRL to make non-contiguous selections.
- From the Insert tab, in the Charts group, click Line Chart.
- Name the Chart, 'Upper and Lower Confidence Bound'.
We can now get a good comparison between all charts.
Login to download
- 00:04 When we're forecasting data in Excel,
- 00:07 something that you'll commonly see on your line chart is an upper and
- 00:12 a lower confidence bound, and the default confidence level in Excel is 95%.
- 00:18 Now what exactly does that mean?
- 00:20 Well it means that 95% of predicted values will fall between
- 00:25 the Upper Confidence Bound and the Lower Confidence Bound.
- 00:30 And we can show this on our Line Chart.
- 00:32 So let's add in another column, I'm going to call this, Confidence Level.
- 00:39 Once again, I'm just going to use the Format Painter,
- 00:42 to copy across this formatting.
- 00:44 Let's give this another color, we're going to choose, let's go for
- 00:48 a purple background fill.
- 00:50 So let's scroll down.
- 00:52 Now, I want this to overlap again so I'm just going to copy and
- 00:56 paste this one across.
- 00:57 And we're going to type in FORECAST.
- 01:00 And this is the one that we want, FORECAST.ETS.CONFINT.
- 01:05 And if you read the little screen tip, it says, returns a confidence interval for
- 01:09 the forecast value at the specified target date.
- 01:12 So if we select this, you'll notice that the arguments that we have are pretty much
- 01:17 the same as what we've had in the other two functions.
- 01:20 So our target date is the first date that we want to predict,
- 01:24 our X value essentially.
- 01:26 The values, these are our Y values so that's the sales in millions.
- 01:31 So let's grab these.
- 01:32 And then our timeline is our known Xs, so it's all of our dates in here.
- 01:38 So let's grab those.
- 01:40 And once again you'll notice that much like with the seasonality,
- 01:44 we have an option to add the confidence level, that's the next argument.
- 01:47 And if we don't add in a confidence level,
- 01:50 Excel is just going to use that default of 95%.
- 01:53 So we're going to use the default.
- 01:55 Let's close the bracket, hit Enter, and then we can drag this down and
- 01:59 apply our number formatting.
- 02:01 So I'm going to add in a comma, and remove the decimal places.
- 02:05 And in fact, for these, we don't actually need that top value in there.
- 02:09 So now that we have these confidence levels, we can use these to work out
- 02:13 the Upper Confidence Bound and the Lower Confidence Bound.
- 02:17 So let's go to the top and add two more columns.
- 02:19 We're going to say Upper Bound and Lower Bound.
- 02:26 So let's widen out these columns.
- 02:29 Once again, we're going to use the Format Painter to apply some formatting.
- 02:33 And we'll make both of these the same color as they are kind of related.
- 02:38 So let's make these a yellow color.
- 02:43 So now what we can do, is we can use the confidence level just here
- 02:47 to work out our Upper Confidence Bound.
- 02:50 And this is just a formula.
- 02:51 We're going to say equals, I'm going to use my seasonal forecasted
- 02:56 value in cell E55, and I'm going to add the confidence level.
- 03:00 Let's hit Enter and let's drag this formula down.
- 03:05 So that's now given me the Upper Confidence Bound.
- 03:09 Now we're going to do something similar for the Lower Confidence Bound.
- 03:12 Again, we're going to take our seasonal forecasted value, but
- 03:16 this time we're going to minus the confidence level, and hit Enter.
- 03:21 And this is going to give us our Lower Bound.
- 03:24 So now looking at these values, let's say for
- 03:29 this first one, I know that 95% of my predicted values
- 03:34 are going to fall between 15633 and 24774.
- 03:39 So that is how those Lower and Upper Confidence Bounds work.
- 03:42 So now we can combine all of this into a line chart.
- 03:46 Now we don't need all of our data here,
- 03:50 what we want is we want our dates and our sales values.
- 03:55 We don't want the Linear Forecast, we do want our Seasonal Forecast.
- 03:58 Let's hold down Ctrl and select this column with those values.
- 04:03 And we don't want the Confidence Level, we want the Upper and Lower Bound.
- 04:08 So we can select both of these.
- 04:11 So now if we go to Insert, we can insert a line chart and check out what we get.
- 04:18 So now you can see the original data showing in blue,
- 04:22 you can see the forecasted values showing in orange in the middle there, and
- 04:27 then we can see what our Upper Confidence Bound is and our Lower Confidence Bound.
- 04:32 So let's add a Chart Title, and then we can move this into place so
- 04:36 we can get a good comparison between all four of our line charts.
- 04:41 So let's drag this over and resize, and
- 04:43 now you can get a good comparison between the different forecast types that you
- 04:48 can create using some of those basic FORECAST functions.
Lesson notes are only available for subscribers.