Faculty
Social Science
Supervisor Name
Charles Saunders
Keywords
Bankruptcies, Heterogeneity, COVID-19, Pandemic
Description
This paper studies Canadian monthly bankruptcy data from January 2014 to February 2022 with an aim towards identifying the existence of underlying heterogeneity in the decision-making of firms across different industry sectors during periods of economic adversity. The data used include provincial two-digit NAICS bankruptcy level data, provincial pandemic-related data concerning the evolution of cases and stringency of adopted policies, and external factors pertaining to the domestic and foreign economies such as industry GDP, the overnight rate target, exchange rates, imports and exports, prices, and bond liquidity premium. The method is two-fold. First, we identify changes in bankruptcy trends caused by the pandemic both provincially and by industry sector, and then contrast the findings with the results from a zero-inflated Poisson GLM model that assesses the probability of bankruptcies occurring at different points in time given the set of inputs described above. Second, the model is used to predict bankruptcy numbers for three different intervals during the reference period: 2019, 2020-2022, and 2021-2022. It is noted that despite the model’s ability to accurately predict bankruptcy values for 2019, its accuracy decreases for the pandemic period, even when pandemic-related variables are controlled for. The conclusion is that the drop in Canadian bankruptcies during the pandemic seen in the data must be accounted for some other factor that upheld firms’ continued operations. The presence of underlying heterogeneity is inferred from the fact that if heterogeneity was not present, then controlling for industry sector would suffice to capture the effect of such a program, thus increasing the accuracy of predictions for the pandemic period.
Acknowledgements
I thank my supervisor, Professor Charles Saunders, for his amazing support and guidance during this project.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License
Document Type
Paper
Included in
Econometrics Commons, Finance Commons, Macroeconomics Commons
A Study of Canadian Bankruptcies, 2014-2022
This paper studies Canadian monthly bankruptcy data from January 2014 to February 2022 with an aim towards identifying the existence of underlying heterogeneity in the decision-making of firms across different industry sectors during periods of economic adversity. The data used include provincial two-digit NAICS bankruptcy level data, provincial pandemic-related data concerning the evolution of cases and stringency of adopted policies, and external factors pertaining to the domestic and foreign economies such as industry GDP, the overnight rate target, exchange rates, imports and exports, prices, and bond liquidity premium. The method is two-fold. First, we identify changes in bankruptcy trends caused by the pandemic both provincially and by industry sector, and then contrast the findings with the results from a zero-inflated Poisson GLM model that assesses the probability of bankruptcies occurring at different points in time given the set of inputs described above. Second, the model is used to predict bankruptcy numbers for three different intervals during the reference period: 2019, 2020-2022, and 2021-2022. It is noted that despite the model’s ability to accurately predict bankruptcy values for 2019, its accuracy decreases for the pandemic period, even when pandemic-related variables are controlled for. The conclusion is that the drop in Canadian bankruptcies during the pandemic seen in the data must be accounted for some other factor that upheld firms’ continued operations. The presence of underlying heterogeneity is inferred from the fact that if heterogeneity was not present, then controlling for industry sector would suffice to capture the effect of such a program, thus increasing the accuracy of predictions for the pandemic period.