The need to introduce heterogeneity in the capital-accumulation process and to differentiate capital goods by their vintage or productivity have been widely recognized since the seminal work of Solow [37]. Together with embodied technological progress, the most crucial aspect of the vintage capital models is that capital goods of later date are more productive, or make products of higher quality (see [28], [38], [2] and [7]). Accordingly, recent studies (see among others, [40], [16], [1], [17]) take into account the fact that firms not only have to decide on the volume of the capital goods but also on the optimal age distribution of them, and have been centered around the questions posed by Chari and Hopenhayn [12]: Why are new technologies often adopted so slowly? Why do people of-ten invest in old technologies even when apparantly superior technologies are available? How are decisions to adopt new technologies affected by the prospect that even better technologies will arrive in the future?

Vintage capital models are able to generate different properties and dynamics from the classical capital accumulation models. The so-called “vintage effect”, i.e. the productivity differential between successive vintages of capital due to embodied technological progress, plays a crucial role on the firm’s optimal plans. Such a productivity differential creates an advantage in investments in younger machines. Younger machines are not only endowed with a superior technology but have also a longer lifetime than the older ones (see [5]). However, they have also the disadvantage that the older machines are cheaper and the costs of depreciation and discounting are less. With the presence of these latter effects, [16] provides an explanation for why new technologies are often adopted on a large scale only after a long period of time. Taking into account the vintage effect, [16] analyzes in what way a perfectly competitive firm adjusts current investments to the predictions of technological progress. As current investments do not affect the profitability of investments in future technologies in a perfectly competitive market, predictions of higher technological progress in the future do not influence the current investments. However, considering market power, [17] shows that a ”negative anticipation effect” occurs. Since current investments increase output which decreases the price, this creates a negative effect on the profitability of future investments, so that the anticipated technology shocks will be preceded by declines of investment. This will be followed by a period of higher growth where new capital goods can be purchased without reducing the output price too much.