The Bespoke Nature of Financial Models
I have built and worked with countless financial models over the years. Financial modelling requires bespoke models, predominantly built to evaluate or assess a specific type of project. This doesn’t imply that they are static, in fact that are often quite dynamic and enable the user to run a multitude of scenarios and sensitivity tests. The bespoke nature does however, mean no two models look the same. Whether building a model to value a business, assess an investment opportunity or run scenario analysis, there are a few key elements that should be kept in mind.
1. One Owner
Although circumstances don’t always permit, it is better to have one person take ownership of the model from start to finish. Having one person responsible for the full build helps to reduce any gaps in the model or error arising from misunderstandings.
2. Clear Structure
It sounds logical but you would be amazed by the amount of poorly structured models I have come across. This often arises because a model has grown overtime and has moved away from its initial humble beginnings but poorly structured models are difficult for others to use and understand. When constructing your model remember to lay it out clearly. Most models can be broken into the following sections:
3. Separate Inputs
It’s vital to clearly identify and separate inputs from calculations. Ideally all inputs are set up on one tab and then the model refers back to these as required. Avoid entering inputs or hard values in formulas buried in the model as this makes changing them in the future very difficult and limits the flexibility of your model.
Formatting should not be overlooked. Having well formatted models that use consistent formatting e.g. shading all input cells in the same colour or laying out reports in the same format will make your model easier for others to use and understand. Consistent formatting applied across an organisation also makes it easier for other team members to get to grips with your models quickly. They intuitively know where to look.
5. Keep it Simple
Above all else in keep it simple! Yes, at times we need to use nested ifs and more complex formulas but where possible break them down into simple steps. This limits the potential for error, helps you to troubleshoot more quickly and makes auditing and understanding the model more straightforward. We don’t need to dazzle with complexity, we need to produce robust and accurate models.