Firms have exited the market since the start of the COVID-19 pandemic. To evaluate the number of firms exiting the market and their exit rate, we construct a simple model, in which firms optimally choose stopping time for their exit. We estimate the model using firm-level data on firm exits before the pandemic. Subsequently, using recent survey data on firm sales growth, we simulate potential firm exits during the pandemic under the condition that the institutional background, represented by activities such as bankruptcy procedures and government rescue plans, did not change the exit option value. Our main findings are as follows. First, we find sizable heterogeneity with respect to the number and rate of firm exits across industries and regions. Second, in aggregate, the pandemic potentially increased firm exits by around $20\%$ compared to the previous year under the assumption that the recent reduction in firm sales is temporary and, thus, partially incorporated into firms’ expectations for future trend sales growth. In two extreme cases in which the recent sales reduction has a full or no impact on firms’ expectations for future sales, firm exits increased by $110\%$ and $10\%,$ respectively. Third, these increases are mainly due to the decrease in the expected sales growth rate, rather than the increase in uncertainty. Finally, we quantify the hypothetical amount of government subsidies needed to prevent excess increases in potential firm exits, which is around ${10}^{-3}$ of Japan’s GDP.

自 COVID-19 大流行开始以来，企业已退出市场。为了评估退出市场的企业数量及其退出率，我们构建了一个简单的模型，其中企业最佳地选择退出的停止时间。我们使用大流行前公司退出的公司层面数据来估计该模型。随后，我们利用近期企业销售增长的调查数据，模拟了在以破产程序和政府救助计划等活动为代表的制度背景没有改变退出期权价值的情况下，疫情期间企业潜在的退出情况。我们的主要发现如下。首先，我们发现不同行业和地区的企业退出数量和比率存在相当大的异质性。其次，总体而言，疫情可能使企业退出增加约$\mathrm{<span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">20</span>}<span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">\%</span>$与上一年相比，假设最近公司销售额的减少是暂时的，因此部分纳入了公司对未来趋势销售增长的预期。在最近销售减少对企业未来销售预期完全或没有影响的两种极端情况下，企业退出增加了$\mathrm{<span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">110</span>}<span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">\%</span>$和$\mathrm{<span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">10</span>}<span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">\%</span><span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">,</span>$分别。第三，这些增长主要是由于预期销售增长率的下降，而不是不确定性的增加。 最后，我们量化了防止潜在企业退出过度增加所需的假设政府补贴金额，约为${\mathrm{<span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">10</span>}}^{<span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">-</span>\mathrm{<span\; data-immersive-translate-walked="96e594a3-3425-40fc-8334-975ea1fdb532">3</span>}}$占日本GDP的比重。