Models and Measurement: (S,s) Inventories and Investment Under Adjustment Costs

该提案包括两个项目，探讨调整成本与投资之间的关系。他们都使用新的模型和方法来研究对理解经济如何随着时间的推移而变化至关重要的问题。第一个项目是与费城联邦储备银行的Aubhik Khan联合开展的，研究的问题是：“库存是否会加剧商业周期？”在一个具有资本积累的动态随机一般均衡模型中，考察了内生广义(S，S)库存投资的总影响，重点研究了中间产品库存。首先，评估了模型再现库存投资、最终销售和生产之间观察到的周期关系的能力。接下来，我们研究了这类库存对整个商业周期的影响。第二个项目考虑的是，“关于公司贴现率的不同假设如何影响资本调整成本的结构性估计？”它探索了研究人员在不同假设下估计调整成本和解释投资波动的波动性和持续性的能力，这些假设涉及管理对公司利润的冲击的过程与管理其跨期利润估值的过程之间的关系。这两个项目都涉及到植物之间相当大的异质性。这使得均衡价格的确定变得困难，因为决策规则通过价格进行标准反馈。均衡价格是通过允许代理使用聚合状态向量的子集预测未来状态来求解的。企业的决策规则是高度非线性的，用多元样条插值法求解。智力优势：库存问题长期以来一直是商业周期讨论的重点，但它在很大程度上被现代商业周期理论所忽视。大多数宏观经济库存研究都集中在线性二次型代表性企业模型上，这些模型在设计上是局部均衡的，而且其经验预测通常令人失望。相比之下，领先的库存微观经济模型--最初由斯卡夫(1960年)求解的(S、S)模型，几乎没有受到宏观经济的关注。该项目纠正了这一重要遗漏。更广泛的影响：这项研究可能会极大地改变政策制定者对库存在总体波动中的作用的理解。例如，初步结果表明，库存管理最近在减少平均库存规模方面取得的进展，可能是商业周期受抑的原因。当这些库存来自非凸订购或交付成本时，这种摩擦的减少导致平均库存销售比降低，库存投资波动受到抑制，实际上可能会提高GDP的整体波动性，因为库存积累的减少伴随着最终销售的更大波动。

This proposal includes two projects exploring the relation between adjustment costs and investment. Each uses new models and methods to examine questions central to understanding how the economy changes over time. The first project, joint with Aubhik Khan of the Federal Reserve Bank of Philadelphia, considers the question: "Do inventories exacerbate the business cycle?" It examines the aggregate implications of endogenous generalized (S,s) inventory investment, focusing on inventories of intermediate goods in a dynamic stochastic general equilibrium model with capital accumulation. First, the model's ability to reproduce observed cyclical relations between inventory investment, final sales and production is assessed. Next, the effects of such inventories for the overall business cycle are studied. The second project considers, "How do differing assumptions about firms' discount rates influence structural estimates of capital adjustment costs?" It explores a researcher's ability to estimate adjustment costs and explain the volatility and persistence of investment fluctuations under varying assumptions about the relation between the process governing shocks to firms' profits and that governing their intertemporal profit valuations. Both projects involve considerable heterogeneity across plants. This makes the determination of equilibrium prices difficult, due to the standard feedback of decision rules through prices. Equilibrium prices are solved by allowing agents' to forecast the future state using a subset of the aggregate state vector. Firms' decision rules, which are highly nonlinear, are solved using multivariate spline interpolation.Intellectual Merits: The topic of inventories has long been prominent in discussions of the business cycle, yet it has been largely ignored by modern business cycle theory. Most macroeconomic inventory research has focused on linear quadratic representative firm models that are partial equilibrium by design and generally disappointing in their empirical predictions. By contrast, the leading microeconomic model of inventories, the (S,s) model originally solved by Scarf (1960), has received almost no macroeconomic attention. The project corrects this important omission.Broader Impacts: The research may substantially alter policymakers' understanding of the role of inventories in aggregate fluctuations. For example, preliminary results suggest that recent advances in inventory management, in reducing the size of average stocks, may be responsible for dampened business cycles. When these stocks arise from nonconvex ordering or delivery costs, a reduction in such frictions that leads to reduced average inventory-sales ratios and dampened fluctuations in inventory investment may actually raise the overall volatility in GDP, because reduced inventory accumulation is accompanied by larger fluctuations in final sales.