Ex Ante and Ex Post Data in Inventory Investment
作者:
Murray Brown,
期刊:
Journal of the American Statistical Association
(Taylor Available online 1961)
卷期:
Volume 56,
issue 295
页码: 518-534
ISSN:0162-1459
年代: 1961
DOI:10.1080/01621459.1961.10480642
出版商: Taylor & Francis Group
数据来源: Taylor
摘要:
Expectational data on manufacturers' inventory investment and sales are analysed. These data are provided byFortuneMagazine's survey of corporations. It is found that firms tend to underanticipate investment changes and that entrepreneurs tend to be conservative in anticipations when economic conditions are depressed and less conservative when conditions are favorable. A theory of investment behavior is then derived and two realization functions are tested, one assuming perfect competition and cost minimization and the second assuming profit maximization and an oligopolistic market. The second performs best. Certain ex ante and ex post estimating forms are tested; the ex post form yielding excellent statistical results with real inventory stock as dependent variable involves the explanatory variables of real unfilled orders, real sales and a distributed lag variable. A major conclusion is that theFortuneex ante data appear to be only marginally useful in explaining inventory investment; and until observation errors are substantially reduced (inter alia), the marginal work units involved in using this data may not be warranted. A positive conclusion is that the real unfilled order variable appears to be highly useful.
点击下载:
PDF (1602KB)
返 回