TABLE OF CONTENTS A good model is easily recognisable it has clearly identifiable results based on clearly defined inputs. The relationship between them can be tracked through a logical audit trail. There is little empirical research into the needs and expectations of model users, but our experience suggests that most users want to know the location of the key results. The ability to perform sensitivity and/or scenario analysis is also very important, so the location of the key inputs should be explicit.
In this chapter, we will consider some of the general conventions concerning model structure. It is tempting to refer to them as rules, but in almost every case the suggestion that 'we must always do this ' can be immediately countered by the observation 'except when we don't'. It is important to recognise that when setting out a rule-based methodology we should have techniques for proving conformance with such rules and for locating and identifying exceptions. This forms the basis of Chapter 2.
The demonstration workbooks for this chapter are located in the Chapter 1: Model structure folder on the CD-ROM.
In a book about financial modelling it may seem obvious that we are talking about spreadsheets, but remember that this isn't always the case. Very recently I met with a client who was trying to design a model that would be manipulated in several ways to generate management information relating to the operational costs of a number of business units. The calculations were...
