Overview
Bank deposits are modeled in order to predict deposit cash balances in the future. There are several ways to do this, differing in accuracy and complexity.
One approach to modeling balances is to just model the total balances as a time series. A simple way to do this is to fit the data to a geometric brownian motion by setting a drift and a volatility.
Modeling Accounts
A more complex and accurate measure would be to model individual deposit accounts. There are some complexities here. You need to model how often an account opens, the balances over time, and how often an account closes. In effect, this becomes three separate models.
- Account Openings : an opening is a random event that occurs when a customer decides to open an account.
- Modeling Balances : An account balance is a dollar amount that varies over time. Ro simplify the modeling, it is common to separate modeling the account balances from the account closures.
- Account Closures : Account closures are the flip side of account openings. It represents a one time event that occurs, after which the account balance is zero.