Recognising investment opportunities at the onset of recoveries

This model focuses on the decision to invest in novel fields of activity. Making such decisions implies that managers recognise the potentialities of emerging technological patterns, which is not a trivial ability. Ultimately, it depends on the mental categories that they developed through their working life, which may or may not be appropriate to the situation that they are facing. In this article, investment decision-making is modelled by means of an unsupervised neural network. Its neurons represent firms as decision-makers and their weights correspond to the coefficients of a disaggregated, flexible accelerator. The ensuing formalisation accounts for the often observed inability of firms that used to be highly successful with a certain technology to recognise and exploit novel, substitute technologies. Within economic theory, this formalisation may be seen as a possible micro-foundation for the beginning of recoveries in Goodwin's model of the business cycle. (C) 2006 University of Venice. Published by Elsevier Ltd. All rights reserved.