Thomas PurdelEmpirischer Ansatz zur Umsatzprognose von Einzelprojekten am Beispiel US-amerikanischer Kinofilmproduktionen

About this book deutschenglish

When deciding whether to invest in a project, generally the investment comes with the risk that the forecasted operational and/or financial success could be less than expected or even absent. Focusing on unique projects this risk gets a value added because predicting the financial success of these kinds of special projects seems to be more difficult in general. There are only few processes that can be repeated, when producing an unique item. Methods of capital budgeting which are in common use – e.g. calculating the net present value of an investment – need some several input factors like the cost of capital, the height of the expected cash flows and additionally the point in time when these cash flows will occur to help the investor to find acceptable projects. In the past several approaches (e.g. CAPM) to find the right cost of capital were developed. Furthermore there have been a lot of approaches that focused the prediction of the cumulative revenues of an investment, ignoring the points in time when the cash flows will appear.

The key topic of this thesis is to create an approach which can be used to predict the economic success of unique projects. The focus is thereby set to the prediction of the point in time and the height of single cash flows a certain project comes with. About 1800 movies which have been shown in US-American cinemas during the period of time from 2000 to 2012 represent the empirical database, which is used for the models and methods in this thesis. To answer the questions of this economically topic the author combines methods of the multivariate analysis (cluster analysis) and methods of artificial intelligence (artificial neural networks) whereby the thesis gets a strong interdisciplinary motivation. The created models and methods are focused to projects of the media industry but they might be adapted for projects in other industries.