Introduction to Financial Modelling
Principles of Excel as a Modelling Tool
Common Modeling Structures (I)
Further Operations and Shortcuts
Excel Functions (I): Information and Numerical Aggregation
Excel Functions (II): Conditionality, Advanced Aggregations and Arrays
Common Modelling Structures (II)
Excel Functions (II): Lookups and Referencing
Model Planning and Best Practices
1 of 2

1.1 Introduction to Modelling

This Chapter discusses the meaning and purpose of modelling, and illustrates some core aspects with a few simple examples. The first sections discuss modelling from a general perspective, whilst the subsequent materials focus exclusively on financial modelling and data analysis.

“Definition” of a Model

The term “model” is used in various contexts. Whilst there is no universally accepted definition of what is meant by it, in general a model can be considered to be: “A representation (or portrayal) of a real-life situation or system, which identifies the key elements within the system, and describes the relationships between these elements.”

At this general level, models can exist in many forms, including:

  • Physical. Art, sculpture and objects made with Lego® may be considered as models: They express key elements and their relationships between the items in a system.
  • Graphical or Illustrative. Examples include influence (or flow) diagrams, which can be used to depict and summarise the logic, dependency relationships, choices, and possible outcomes within a situation. Decision trees are of a similar nature. (Each of these may also contain numerical calculations, but in general, their use is for visual and conceptual purposes).
  • Descriptive. A text-based document that describes the situation can also be considered as a model, so long as it captures the elements and their relationships in a reasonably precise way.
  • Mathematical. The elements and their relationships in a situation could be expressed using formulas (which inherently represent relationships between the key elements or variables).
  • Numerical. The relationships and behaviour of a system could be captured directly within a set of calculations (very often using Excel). The variables and their relationships are specified implicitly by these calculations (although some external documentation may also be created, containing descriptions, illustrations or mathematical formulas).

Models as Approximations or Simplifications

In general, models do not perfectly capture every aspect of a situation: Indeed, a “model” which did so would essentially no longer be a model! There are almost always features or behaviours of the real-life situation that cannot be captured exactly (even if these are unusual behaviours that could arise only rarely or in highly specific circumstances).

Part of the “art” of modelling is to capture the key elements in a way that is sufficient to fulfil the particular purpose. Some important reflections to bear in mind, in almost all situations (the the simplest to the most complex) include:

  • “Every model is wrong, some are useful”.
  • “Every model should be as simple as possible, but no simpler”.

Objectives and Areas of Application

The objectives of modelling depend on the context, but generally include one or more of:
  • To facilitate a reaction or be thought-provoking (e.g. art, sculpture,…)
  • To generate insights and understanding, by documenting, testing and developing hypotheses about the behaviour of the system.
  • To support a decision process. This could include the economic evaluation of a possible course of action and the development of alternatives.
  • To calculate a numerical quantity whose value is needed for some purpose, but is not known directly from measurement or other data sets.

There are many areas of application, including:

  • In physics and engineering, models are used in the design of aircraft, cars and wind-turbines (e.g. modelling of airflows over the structures).
  • In epidemiology, models are used to forecasting of the spread of infectious diseases.
  • In finance, economics and business, models are used in business forecasting, project evaluation, investment decisions, valuation, credit assessment, risk quantification, and portfolio optimisation, and many other areas. In addition, data analysis is required not only as an input to models, but also (and increasingly) as an activity in its own right, including to make data-driven prediction, based on relationships that are inferred from the data.

The Modelling Process

In general, a modelling process consists of three stages:

  • A concept or specification stage. This focusses on defining the objectives, the overall concept, architecture and structure, as well as identifying the key variables or elements, and defining the relationships between them.
  • An implementation stage. This is the expression of the model in its “physical” form. (Within this, we include Excel workbooks and programming languages).
  • A results or conclusions stage, where the models’ results are used in some way.
In some cases, not all stages are explicitly used or are given much emphasis. For example, quite often models are built in Excel using a “hands-on” approach i.e. the models are directly implemented in Excel, whilst  the underlying conceptual or contextual basis for the model is de-emphasized, or not documented, or is assumed to be implicitly clear. (Needless to say, this is not an ideal situation in many cases!)
Scroll to Top
error: Alert: Content is protected !!