Awesome
AbsBox
a structured finance cashflow engine wrapper for structured credit professionals:
- transparency -> open source for both wrapper and backend engine, great variable inspection support.
- human readable waterfall -> no more coding/scripting, just lists and maps in Python !
- easy interaction with Python numeric libraries as well as databases/Excel to accomodate daily work.
installation
pip install absbox
Documentation
- English -> https://absbox-doc.readthedocs.io
- Chinese(inactive) -> https://absbox.readthedocs.io
Goal
- Structuring
- Easy way to create different pool assets/deal capital structures and waterfalls
- User can tell how key variables(service fee/bond WAL/bond cashflow etc) changes in different structure of transactions.
- Investor
- Given powerful modelling language to build cashflow model , user can price bonds of transaction with pool performance assumption
What it does
- Provide building blocks to create cashflow models for ABS/MBS
- Adapt to multiple asset classes
- Residential Mortgage / AdjustRateMortgage/ IO,PO,Balloon Mortgages / Auto Loans
- Corp Loans
- Consumer Credit
- Lease (For CMBS)
- Fix Asset (Solar Panel/Hotel)
- Receivable
- SRT/Siginificant Risk Transfer
- Master Trust
- Features
- Sensitivity Analysis on different scenarios or deal structures
- sensitivity analysis on pool performance assumptions
- sensitivity analysis on capital structures or any deal components
- Bond Cashflow/Pool Cashflow Forecast, Pricing
- Sensitivity Analysis on different scenarios or deal structures
What it takes to master
- Python syntax, nice to have knowledge of functional programming ,or exposure to package
toolz/lenses - Patience & Persistence, but remember , there is a slack community and responsive support !
Missing Features ?
Raise issues or disucssion with the prospectus or spreadsheet how asset cashflow should be projected.
Data flow
<img src="https://absbox-doc.readthedocs.io/en/latest/_images/Intergration.png" width="600" height="347"/>Data Privacy
User can have option to connect to a private calculation engine in his/her own environment