To learn more about EpiX Analytics' work, please visit our modeling applications, white papers, and training schedule.

Page tree



In most discounted cashflow models of capital investment projects we will have a number of time series that we wish to project over the life of the project. As risk analysts, we want to include any uncertainty about those forecasts, of course. We would also like to include any interactions between these forecast variables: for example, that if the exchange rate with the currency a client purchases in goes up, the client can afford more of your product.

In this guide, we have developed a number of time series forecasts to give you some ideas of how to produce a risked forecast model of sales volumes. The models cover a range of situations you might find yourself in:



Selling into a finite demand for a product


There is a maximum possible number of sales that you could make over the entire life of the project. Model Sales Projection for a Finite Market gives a way to model what proportion of those sales you might eventually make, and combines it with an estimate of how likely you are to convert a remaining potential buyer into a sale in each year. This type of model produces an eventual decline of sales as the market becomes exhausted.

The links to the Sales Projection for a Finite Market software specific models are provided here:


Selling a new product that may take off spectacularly, or fail, or something in-between

Offering a new product on the market carries the unpredictability of consumer reactions. Model New Product Sales offers four different approaches to model a sales growth curve whose rate of acceleration is given by a probability distribution, as an elegant way of reflecting consumer reaction.

The links to the New Product Sales software specific models are provided here:


Selling a new product where a competitor may emerge, taking some of the market share


If your new product does very well, chances are that one or more competitors will produce a similar, or even slightly better, version of the same product. The trigger for the introduction of a competitor will be whether they can develop a competing product (maybe you have a crucial patent that will have to expire, maybe they just need to tool up a factory), whether they see it making them a profit or perhaps see a strategic advantage to keeping up with you. Model NPV of a Capital Investment takes its trigger from the total sales that are made. If the market gets to a certain size, a competitor emerges and begins to eat into your market. Then, if the market gets bigger still, another competitor enters. The model has a neat trick for allowing the market to be shared out.

The links to the NPV of a Capital Investment software specific models are provided here:


Selling a product whose demand is a function of economic and other factors


There are often interactions in the real world between economic factors like exchange and interest rates, and sales volumes, plus perhaps politically-driven variables like sales tax rates, or market variables like raw material prices which in turn affect sales price which then affect sales volume. These factors may influence more than one variable in the model which means that we need to explicitly describe their inter-relations to capture the correlation effects they produce between our model variables. For example, if exchange rate to the US$ might affect our sales in the US, as well as being a component of the cost of some raw materials we buy. Market Growth Model offers some techniques for modeling these types of inter-relationships.

The links to the Market Growth Model software specific models are provided here:


Launching a new product whose launch timing is still uncertain


Valuing an investment into a the development and launch of a new product often comes with an uncertain launch timing of the new product. This example model describes how including risks and key correlations in a financial model provides important insights for decision-makers.



  • No labels