中文摘要
本文研究金融冲击对厂商保险、厂商动态与宏观经济影响的作用。与文献的一个关键区别在于,厂商在长期关系中向风险规避型工人提供工资保险合同。此类合同内生地对厂商施加了一种刚性债务责任,使得金融净值有限的厂商相比其他厂商更易受冲击影响。在我们的模型中,厂商具有异质性生产率与金融净值,且厂商面临金融摩擦。劳动力市场中的搜寻与匹配内生地决定工资合同的价值,从而决定厂商所承担的负债水平。我们有三个新发现:(1)定量模型的工资动态与实证研究一致,对厂商层面的特质性冲击提供了显著的保险,同时在存在总体冲击时,总体工资与 GDP 之间的联动较弱。(2)金融资源有限的厂商受到金融冲击的打击显著大于其他厂商,对股息、信贷借款、厂商价值与厂商进入的负面影响更大;内生工资保险也变得更加有限,且分离更多。这些对厂商和工人的影响不同于——也无法由——具有代表性厂商或工人的标准模型所解释。(3)针对美国大衰退的定量分析表明,总体生产率冲击与金融冲击发挥不同作用;若没有金融冲击,我们无法解释金融变量的动态,包括信贷借款减少、信贷利差飙升以及股息支付大幅下降。最后,我们为模型含义提供了微观层面的实证支持。
Abstract
This paper studies the impacts of financial shocks on firm insurance, firm dynamics, and macroeconomic implications. A key departure from the literature is that firms provide wage insurance contracts to risk-averse workers in a long-term relationship. Such contracts endogenously impose a form of inflexible debt liability to firms, and make firms with limited financial net worth more vulnerable to shocks than others. In our model, firms have heterogenous productivity and financial net worth, and also firms face financial frictions. Search and matching in the labor market endogenously determines the value of wage contract, and thus the level of liability to the firm. We have three new findings: (1) The wage dynamics of the quantitative model are consistent with those in empirical studies, with significant insurance against firm-level idiosyncratic shocks, and also a weak comovement between aggregate wage and GDP when there are aggregate shocks. (2) Firms with limited financial resources are hit by financial shocks significantly more than others, with larger negative impacts on dividends, credit borrowing, firm value, and firm entry; the endogenous wage insurance also becomes more limited, and there are more separations. These impacts on both firms and workers are different from—and cannot be accounted for by—standard models with a representative firm or worker. (3) Quantitative exercises for the Great Recession in the United States suggest that aggregate productivity shocks and financial shocks play different roles; without financial shocks, we cannot account for the dynamics of financial variables, including the reduced credit borrowing, spikes in credit spreads, and large drops in dividend payments. Lastly, we provide microlevel empirical supports for our model implications.