Résumé en anglais

The history of accounting for private railway companies in Germany shows that these companies played a major role in the diffusion of historical cost accounting principles and gave birth, together with big other joint stock companies, to the dynamic" or second stage of capitalist accounting, at least in continental Europe. If the representatives of such railway companies did not develop new concepts of accounting, notably as concerned depreciation, they had, by 1875-1879, elaborated a new theory of accounting (historical cost or dynamic theory).This theory had a profound impact at least on the German theorists of the late 19th century and early 20th centuries such as Simon, Rieger and Schmalenbach. This new theory was needed to justify a new law favoring shareholders in a hurry for returns on their investments rather than company creditors. It also defeated the ideology of public finance and patrimonial (or static) theories of accounting. This new theory preceded the law which promulgated the new approach and clearly defended the private interests of shareholders as opposed to those of the public in the strict sense. It appears to contradict Watts and Zimmermanns basic hypothesis of the «theory of market excuses». Agency theory seemingly does not to apply either, for the new theory was proposed by managers allied to shareholders, specifically those «hurried share-holders", against creditors. This is why a kind of «theory of cJliance» appears to be more consistent with these developments. The main reasons for developing the new accounting theory were linked to the issue of dividends. It was necessary to find an accounting approach which would allow the distribution of dividends at the very beginning of an investment cycle. It was also intended to find an accounting approach which would ensure that profits were distributed as evenly as possible throughout the entire investment cycle and among the different shareholders who had financed the investment. Hence, the second stage of the capitalist accounting development was not connected to measure of performance or information problems (monitoring and bonding) but seems to have been caused by the need to regulate profits and dividends in the interests of managers and shareholders. However, as this change took place within the framework of prudence, it was impossible, at that stage of capitalist accounting, to achieve a perfect smoothing of the rate of accounting profit. The solution to this problem was only to be found at the end of the 20th century with the onset of the third or actuarial stage and the "discovery" of fair value.