Modeling of Banca d'Italia Back Office System

Modeling of Banca d'Italia Back Office System

Banca d'Italia processes a certain amount of manual credit transfers every year. These transfers cannot be processed automatically and require two divisions of employees in the back office of the bank. The bank wanted to determine if merging these two divisions would be beneficial.


The back office operations included all of the checking and settlement activities (see the figure) related to the trading operations performed by the corresponding front office. Two divisions, or units, were involved in the execution of domestic credit transfers. Employees in the units worked in shifts from 7:30 a.m. to 7:00 p.m. There were three categories of employees in each division (assistants, coadjutants, and officers) and each category performed different tasks.

Officers in unit A authorized, by signature, the transmission of the payments registered within the IT system to unit B so that they could begin working on them. An assistant in unit A could not fulfill the tasks of an assistant in unit B. Unit B started its operations only when unit A had completed its tasks for each single process. Therefore, the tasks of the two operating units were different and sequential.

The domestic credit transfer process was not the only task performed by these units, but it was the most important one, as every payment had to be carried out within a time limit (5:30 p.m.). If this time was exceeded, clients would apply to the bank for the payment of penalties, which is why this process was a top priority for both units.

Bank back office processes simulation

The goals of the simulation modeling project, carried out by Fair Dynamics Consulting, were:


The consultants employed AnyLogic’s unique capability to use different modeling methods and created two models of the system, one in Discrete Event and the other in Agent Based, to ensure the output. The numerical results were the same. Traditionally, such systems are often simulated with the Discrete Event method. In this case, the Agent Based model was easier to use and quicker to build.

Several experiments with the system were carried out:

The efficiency of the “as is” (current situation) and “to be” (possible merger of the two organizational units) scenarios is compared on the graphs (see the graphs). The graphs illustrate what would happen in a normal working day for both scenarios.

The advantage of merging the two divisions was clearly shown in the simulation for an absence of employees. In the current scenario, the units were not able to meet the deadlines for processing all the payments if an employee was absent. The merger would solve this problem.

Business Process Optimization Bank of Italy


The simulation showed that the merger of the two current operating divisions into one would be very beneficial. The merger would produce the following advantages:

Watch Luigi Geppert from Fair Dynamicsof Italy presenting this project at the AnyLogic Conference 2012 or download his presentation or a paper based on this case.

Similar case studies

More case studies

Get a brochure with industry case studies