Inventories and Backlogs: An Analysis of The Business Cycle

该项目涉及一个重要而有趣的主题，即生产库存和积压的决定因素。多年来，经济研究中一直认识到，库存在商业周期的变化中发挥着核心作用。尽管季度库存投资水平通常不到国民生产总值的1%，但在周期性转折点，库存投资的变化往往是重要的。事实上，经验证据表明，在美国二战后的八次衰退中，库存投资的变化占到了商业活动总变化的近一半。二战前的数据有限，但有迹象表明，同样的关系也成立。因此，分析库存变化可能是预测商业活动大范围转变的好方法。尽管了解企业积累库存的原因显然很重要，但几乎没有明确的证据表明，企业在增加库存时，是否在避免生产成本或最终产品销售的意外波动。此外，公司还可以通过积累积压的订单来实现同样的目标。因此，最常用于研究库存积累的模型，即生产平滑模型，存在固有的缺陷。本项目从企业避免积压订单成本的角度来研究生产库存。一般来说，这些成本可能比单纯持有库存的成本更高，因为未完成的积压订单会导致糟糕的客户关系。从技术上讲，韦斯特教授估计了结构性模型，以检验积压成本在数字和统计上是否具有重大意义。假设生产滞后，如果订单在每个生产周期开始时超过生产，就会发生周期内的积压。他还对这个非线性模型进行了随机数值模拟，以解释观察到的有关库存的事实，特别是该模型是否预测到，负面需求冲击，即经济衰退，会导致二战后时期观察到的那种库存清算。

This project concerns an important and interesting topic, namely the determinants of production inventories and backlogs. For many years there has been in economic research the realization that inventories play a central role in changes in the business cycle. Even though the level of quarterly inventory investment is, typically, less than one percent of GNP, changes in inventory investment often are important at cyclical turning points. In fact, empirical evidence shows that changes in inventory investment account for nearly one half the total change in business activity for most of the eight post- World War II recessions in the U.S. Data is limited for the pre-World War II era, but indications are that the same relationship holds. Thus analyzing changes in inventories might be a good method to predict wide shifts in business activity. In spite of the obvious importance of knowing why firms accumulate inventories, there is little unambiguous evidence as to whether firms are avoiding unforeseen fluctuations in production costs or in sales of final products when they increase inventories. In addition, firms can accomplish the same goals by accumulating backlogs of orders. Thus the models most often used to study inventory accumulation, namely the production smoothing models, are inherently flawed. This project studies production inventories from the standpoint of firms' avoiding the costs of backlog orders. These costs can in general be greater than those of simply holding inventories, since having unfilled backlog orders incurs bad customer relations. Technically, Professor West estimates structural models to test whether backlog costs are numerically and statistically significant. A production lag is assumed, and a within period backlog occurs if orders exceed production at the beginning of each production period. He also does stochastic numerical simulations of the nonlinear model, to explain the observed facts about inventories, in particular whether the model predicts that negative demand shocks, that is recessions, induce inventory liquidations of the sort observed in the post-World War II period.