How to calculate difference bet 2 months if only 1 month has value

[bui195]

I'm trying to create a workflow to dynamically calculate month over month variance. My current formula only calculates the difference if the GL account and Client ID combinations exist in both current month and prior month. In the scenario that client ID and GL account combinations only exist in current month (or vice versa), it should take current month subtract 0, or 0 subtract prior month and show result in the MoM Variance column. 

Example below: 

Input:

Expected output:

[BS_THE_ANALYST · Accepted solution]

@bui195 should be completely dynamic. It also doesn't matter the order in which the effective dates are entered into the workflow.

[binu_acs]

@bui195 Another way of doing this

[bui195]

Thank you so much! This works really well and simplify my actual workflow. I have a follow up question tho. How can I dynamically re-order column if next month I add March and the report should show January, February, March (the end goal is to show January to December as we go toward year end). With the Tile tool, I can see we assign the 2 most current months number 1 and 2 and the previous months will be assigned 3, 4, 5, etc. With March data coming in next month, I see it will show 2/28/2023, 3/31/2023, 1/31/2023 then MoM Varianc column. Do you have suggestions on how I can dynamically move the previous months the beginning so the months will be in the correct order?