The main purposes of this study were first, to develop a flexible empirical framework that allows to test the implications of different aspects of the more recent investment literature, and second, to present an empirical application that illustrates the advantages of this approach. The framework chosen was that of (discrete) Markov decision processes, for two main reasons. As discussed in Section 3.1, this framework is very flexible. It allows, for example, to incorporate various non-standard constraints into models of firm behavior, related to both “real” and “financial” aspects of investment decisions. Also, there are powerful methods available for structural econometric estimation of discrete decision processes; these were discussed in Section 3.2. These tools were put to work in an empirical model of firms’ investment and exit decisions in Section 3.3, and in Chapter 4, I presented an empirical application of this model (extended to allow for financial constraints).
Unable to display preview. Download preview PDF.