Welcome to my blog!

I built a simple amortizing bond model application using python and Streamlit. This is a follow-up to my series of posts on amortizing bond models. The app lets the user quickly examine a bunch of simple models by varying the parameters.

This is version two of the app. Version two added arbitrary revenue input, automatically calculated capitalized interest, arbitrary ongoing fee input, a debt service coverage ratio feature, and date-based inputs for the dated/closing date and the final maturity date. It handles the initial stub period interest payment correctly. Based on the user’s inputs, it outputs a simple graph of the debt service along with a sources and uses table and debt service table. There are download buttons at the bottom to download csv’s for those two tables.

For version three, I’m planning to add even more features:

  • Arbitrary pricing scale inputs, including serial/term bond structures;
  • TIC and arb yield calculations;
  • Annual/semiannual amortization options; and
  • Warning messages pointing out common structuring issues.

I may shift some of those into a version 2.5, depending on how tricky the implementation is. In particular, coming up with helpful and comprehensive warning messages will be time-consuming.

At work, I build these sorts of models in Excel all the time. It’s been very satisfying to relearn Python by working on a project where I’m so familiar with the goal and implementation. It’s been especially helpful in transitioning from thinking in an Excel way (i.e., data-first) into a more traditional coding way (i.e., manipulating objects with functions and then observing the result).