Eagle-eyed agents may see that sometimes the Months Supply of Inventory (MSI) for a total region, as defined by the time series, is greater than (or smaller than) the MSI for any of the different price buckets in the Price Distribution chart. It's subtle, but it can happen.
The Months Supply of Inventory is defined as the number of active listings at the end of the month divided by the number of closed listings for that month.
For the time series calculations, and thus for the main metric box in Market Insights, this includes all listings in all price bands for the active inventory, and divides that by all closed listings in all price bands.
For the Price Distribution chart, it does the same math---except by price band. It divides all listings that were active at the end of the month in a defined price band by all listings that closed in that month in that defined price band.
The difference occurs when there are no closed sales in a particular/multiple price bucket but there were Active listings. The overall MSI will use those active listings in its calculation, when obviously the individual price buckets ignore all other price buckets.
Mother Nature gets very upset if we try to divide by zero. It's basically undefined. And that's okay. In this example, the Price Distribution chart isn't showing an MSI when there's no closed sales; that column is by definition undefined. So the MSI of the price buckets won't add up to the overall MSI.
You can validate this keeping the same selection filters, and changing the metric to Active Inventory (or End of Month Inventory). You'll see those active units appear and show you where the undefined infinite-weight MSI calculations were hiding!