How to create a generic View for different Models?

For those that create financial models and program, why not create a model programmatically instead of with Excel?

• I used to create financial valuation models myself with Microsoft Excel, so I am familiar with it as a modeling tool. However, after learning some functional programming, I came to the realization that Excel is largely a table-based tool for basic functional programming through a visual interface. The unfortunate thing about Excel is that it very quickly results in legacy financial models that are hard to use by anyone other than their creators because referring to constants like beta, risk free rate, WACC, etc. by cell names (e.g. A9, B52, JJ11, etc.) is not very intuitive, especially once a financial model has become large. For those that have programmed in a functional language (or even in an imperative language), why not program you financial models instead of using Excel?

• Answer:

  This is a rather simplistic answer, but I can sum it up with the following: Standardization - Excel has become the de-facto standard for viewing financial data.  Everyone is comfortable with reading charts, looking at tabular data, but eventually it gets copy/pasted into Excel - or worse, as I have seen it, transcribed manually back into a spreadsheet.  In addition, Excel contains the entire runtime and editing environment that allows people to pry and look at the underlying code. Data portability - your code, despite how beautiful or insightful it is, is hard to transfer from one colleague to another.  You can certainly keep generating CSV outputs or export it in an excel-macro friendly manner, but now you're back at square one which begs the question maybe that's where it should have been done anyways. Excel is easy to forward in email, people can open and view it quickly - and unless in the most secure environments, there's no difficult passwords or usernames or login accounts or hoops to jump through to get to the data. Accessibility - this is very close to data portability, but in general, people want to poke at the assumptions, make changes to formulas, or key in different inputs or extend the data into other views.  A functional programming model makes this difficult as they have to understand the logic, functions, and libraries that make simple changes hard even though it may make hard things possible. Legacy Code Base - by now, there are thousands of plugins, libraries, formuals and functions expressed in Excel or VBscript.  There are a lot of codependence and userbase around these extensions and is hard to replicate.  The ecosystem around Excel is just too powerful to ignore. Having said all that, it's good to note that a lot of the large datasets, complex modeling, machine learning and projection systems are usually written in heavy duty platforms (functional or not) - Excel just happens to be the "last mile" for less programming intense renderings.