Article
The last mechanic is the one that saves the most time: every output comes from the same run.
One Riverbend run produces:
- The financial model: a workbook in live formulas, with its Checks tab
- An investment teaser and an investment summary deck
- An information memorandum
- An indicative term sheet
- A credit approval memorandum and an IC memorandum
- A lender pack
- A Model Book: structure, assumptions register, checks
- A snapshot that reloads the deal in the app
Then every document is read back. The figures in each one are extracted and compared with the model, and the build fails on a mismatch, a missing disclaimer or a real company name.
The timings:
| Loading the deal and running the engine | 0.6 seconds |
|---|---|
| Workbook and documents | 25.8 seconds |
| Converting the Word documents to PDF | 23.8 seconds |
| Reading all 15 files back against the model | 3.2 seconds |
| Total | 53.5 seconds |
Change an input, and the whole set is rebuilt and re-checked in under a minute. That is the point of building the mechanics carefully once: the term sheet, the credit memo and the workbook can't disagree, because none of them is typed.
Riverbend, as it stands:
That's the series: one ordinary deal, the mechanics a lender tests, every number from one model.
Riverbend Solar is an illustrative case study: a fictional deal with no real counterparty.