How do I model a time delay which has a probability distribution?

Simplified example: loan repayments. Making 100 loans a month, I know that in year 1 the repayment time in months is normal(12, 1) and in year 2 it is normal(8, 1).

How do I model the estimated repayments received in each month? In excel I would have a load of manually enumerated probability tables but I wonder if there is a more elegant way in Causal?

1 Like

Hi Adam, you can create a variable New Loans which has a 100 value, Year 1 Repayment and populate it with normal(12, 1) and then create a Loan Repayment variable which references Year 1 Repayment as a time modifier. See snip below for what your formula should look like:

image

A way to make this more powerful is to bring in the New Loans by cohort. You can do this by making a variable New Loans by cohort where the formula is if t = cohort then New Loans else 0. Then if you follow the same logic as above you can track the repayment schedule of each cohort of loans. Hope this helps!

1 Like

I’ve create a cloneable model template you can use here!

1 Like

Thanks so much @Sanjeev_Sandhar! I discover something more awesome about Causal every day!

1 Like

Glad to help and glad to hear you’re enjoying the product!

1 Like