Using a Lookback in Scenarios: Lookback Basics
In the previous post, Scenarios in SAS Visual Investigator, you learned about the three different kinds of scenarios. Let's look in more depth at the Lookback options we have when using the Aggregation Scenario.
When using an aggregation scenario, we have to identify what dates to include in our aggregation. You know if you have ever bought a paper calendar, there are options for viewing by a day, a week or a month. I’m a weekly girl. I like to see my whole week at a glance. But some people prefer to see the whole month at once. And I even have a crazy friend who uses a daily calendar!
Well, the lookback is liking picking your calendar view. What timeframe are you viewing? That is the Lookback Unit, day, week or month, etc. Next you need to decide how many Lookback Units you are grouping together.
Example 1: Let’s say you want to monitor for a Currency Transaction Report. Banks are required to file if any single person deposits or withdraws over $10,000 cash in a single day. You would create a scenario with Day for the Lookback Unit and 1 for the Lookback, to aggregate a single day’s cash transactions.
Example 2: Let’s say you want to monitor for structuring. Typically, structuring occurs when someone wants to avoid a Currency Transaction Report, so there would be multiple cash transactions just under $10,000 within a few days. You would create a scenario with Day for the Lookback Unit and 7 for the Lookback, to aggregate seven day’s cash transactions. *In this example, you would also want to include a count of transactions in your rule, so you aren’t also catching those that had a single transaction over $10,000 and did have a CTR filed.
Now there is another option to use with the Lookback and it’s called Offset. By default, the scenario aggregation starts at the last day of the data. Let’s say our data is for the first month of the year, 1/1/2024 through 1/31/2024.
Example 1: The scenario is only for a single day, so it would aggregate 1/31/2024.
Example 2: The scenario would start at 1/31/2024 and go back 7 days, to 01/25/2024, and aggregate the transactions for those days.
The offset changes the last date of the data by the Lookback value to determine the date at which to start aggregation. So, if our data ends with 1/31/2024, we can aggregate an earlier time period by adjusting the offset.
Example 2: If we add an offset of 1, the scenario will start at 1/24/2024 and go back 7 days, to 01/18/2024, and aggregate the transactions for those days. If we add an offset of 2, the scenario will start at 1/17/2024 and go back 7 days, to 01/11/2024.
Select any image to see a larger version.
Mobile users: To view the images, select the "Full" version at the bottom of the page.
Using an offset in this way is useful for comparing aggregations. For example, using the different offsets, you could write a rule to compare the weekly amounts and see if there is a steady increase or decrease in activity.
To learn more about writing scenarios in SAS Visual Investigator or using the out of the box scenarios in SAS Anti-Money Laundering, the courses Building an Interface with SAS® Visual Investigator and Administering SAS® Anti-Money Laundering in SAS® Viya® are available as eLearning.
Find more articles from SAS Global Enablement and Learning here.
