> OK, financial modelling using Excel only gets messy what are the alternatives?
Good question, and here ius an answer based on where we are in 2022:
- If you like the methodology of modelling with functions, you could write your own:
- You could use Visual Basic VBA to build a few standard projections functions, and there is some material on the download page that should help you. Be aware that VBA is maybe 20 times slower, so you need to minimize the data transfer between Excel and VBA (eg building array functions) and you need to limit some functionality - try to keep the amount of looping low.
- Develop your own functions in a fast language using excellent toolkits like xlw and xll+ that we know of, and try other toolkits that we havent tried like Excel DNA and Excel Add-In Express. You may even be able to use C# instead of C++. You will need someone on your staff with programming experience ready to hand to fix bugs at fairly short notice. If you have the resources, there is a lot to be said for building functions that uniquely suit your business, and bank trading operations do this all the time, you can see what they do on the forum at Wilmott.com.
- You could use Visual Basic VBA to build a few standard projections functions, and there is some material on the download page that should help you. Be aware that VBA is maybe 20 times slower, so you need to minimize the data transfer between Excel and VBA (eg building array functions) and you need to limit some functionality - try to keep the amount of looping low.
- You could develop formulae that work in a standard way across the budgetary timebase. You could look at our seminar material on the download page where we tried to doing things in a number of ways ie with BF, VBA, and native Excel formulae. Native formulae are very fast, but hard to debug and understand. But when they work, they are very efficient.
- You could look outside of Excel to modelling products like Quantrix. But this of course introduces a whole new world of skills and cost. Have a read on their website! It's very interesting, but for me I think its too much of a leap into basically a new modelling paradigm.