Will COVID-19 economic fallout look like a large-scale natural disaster? The Fed seems to think so.
Introduction
The growth curves for unemployment rates of COVID-19 look an awful lot like those of large natural disasters. After the first three weeks of surges, the NY Fed equated the impact to hurricanes Katrina and Maria. Looking at the data, we agree it is hard to deny the similarities which have only gotten stronger as more claims data became available. AQN also looked into some of the key outcome metrics for consumer financial products after Hurricane Katrina to try and baseline how lenders should be thinking about impacts to their portfolios.
US unemployment claims have been a recurring headline as the US economy reacts to the COVID-19 pandemic. This is because unemployment rates are an easily trackable and relatable statistic. In fact, the Federal Reserve Bank of Saint Louis first proposed methodologies for using the unemployment rate as an economic indicator in a paper published during the 1973-1975 recession.
As businesses and governments began responding to the safety implications of COVID-19, The New York Times produced a series of infographics showing that unemployment claims have dwarfed even the unemployment rates of the Great Recession. The astronomic spike likely caused people's anxiety to spike as they tried to comprehend the severe impact on our economy, the effect on their retirement accounts, or whether they have enough savings to weather the storm ahead.
The New York Times - More Than 3 Million Americans Lost Their Jobs Last Week.
As seasoned consultants and analysts, we turn to the data in search of answers and guidance. Unfortunately, limited data exists on economic outcomes and unemployment rates during the 1918-1919 Influenza. In the past 20 years, only one event in the United States approaches the magnitude and severity of the current unemployment spike: Hurricane Katrina in 2005 (Hurricane Maria was excluded from analysis due to data issues like 0 weekly claims). While the COVID-19 pandemic differs in many ways from a natural disaster, we believe there are several interesting economic parallels that make reviewing the hurricane’s economic impact on Louisiana a worthwhile case study.
Comparing Hurricane Katrina to COVID-19
Before we infer any results and outcomes from the impacts of Hurricane Katrina, it is important to highlight key similarities and differences that could skew inferences we draw between the two events.
Similarities:
Skewed impact towards the travel and entertainment industries
Short term logistical impacts to essential businesses (grocery, big box, & home improvement stores)
Large scale government assistance
Differences:
Katrina destroyed physical infrastructure while COVID-19 has idled the economy
Katrina had a disproportionately higher level of physical Federal assistance (coast guard, FEMA, corps of engineers) than COVID-19
COVID-19 recovery may not have beneficial boost from construction
Hurricane season has a defined endpoint where COVID may continue to impact the foreseeable future
In absolute terms, we see the growth rate in weekly unemployment claims relative to the start of each of the two crises look remarkably similar. COVID-19 appears to have had a slightly higher growth rate, but the most recent reporting (week ending 4/18/2020) shows the potential start of a plateau. The next few weeks will be vital to understanding whether rates will flatline like Hurricane Katrina or continue to climb. Assuming a plateau in unemployment, we can look at how those impacted by Hurricane Katrina fared in the aftermath of the storm to see if there are any outcomes that can help us quantify the impact of COVID-19. Two important hypotheses to keep track of are whether the Paycheck Protection Program will distort the plateau by delaying layoffs until the end of summer or whether the $600 per week incremental stimulus for unemployment check encouraged more people to enroll.
Data based on US Department of Labor: Unemployment Insurance Weekly Claims Data.
Economic Outcomes
For forward looking guidance, we look at key economic outcome metrics after Hurricane Katrina that might help us envision the true severity to the economy. The 2017 article, "Household Finance after a Natural Disaster: The Case of Hurricane Katrina" by Gallagher and Hartley in the American Economic Journal: Economic Policy, quantifies many important outcome metrics for overall impact to household finances. Their study looked at individuals most and least impacted by Hurricane Katrina flooding and compared them to an economically similar baseline group who was unimpacted by flooding.
Unsecured Debt
In the quarter immediately following Katrina, there was an increase of $500 in unsecured card balances representing a 15% increase over pre-Katrina levels
This aligns with a 2008 study which found that only those in the bottom 20-30% of total assets increase card debt in response to unemployment
Lowest decile seems to be underserved by traditional banking products while highest deciles do not need to borrow
Risk
Immediate risk indicators are obfuscated due to grace periods on loans, especially mortgages, and delayed delinquency reporting
90+ DQ Rates increased 10% 1-year after Katrina when grace periods ended
Risk scores dropped 4-7 points at 1-year after Katrina
Auto
Auto loans increased marginally and were skewed towards people who experienced at least 2 feet of flooding
Debt Balances
While auto and card balances increased, overall debt decreased. This is led primarily by home loan balance decreases of $8,500
The Louisiana Attorney General opened an investigation into reports of mortgage companies withholding insurance money intended for home repairs
Mortgage
Mortgage originations dropped to around 400 a month through the 2008 Great Recession from a pre-Katrina average of 1,200 a month
Lending transitioned from being predominately non-local lenders prior to Katrina to predominantly local lenders afterwards
Migration
Between 2-12 bps of residents left Katrina impacted areas for at least 3 years depending on severity of flooding
No studies are available yet to quantify migration impacts but anecdotal evidence suggests support of COVID-19 related migration from high-impact areas
Summary
If we assume the growth rate of unemployment claims for Hurricane Katrina serve as an economic indicator, we can predict several key outcomes during the COVID-19 pandemic. Without tools like the Term Asset-Backed Securities Loan Facility (TALF), it is likely that banks will begin to tighten credit policies for consumer financial products. Key segments of existing customers with outstanding exposure will utilize unused balances while other segments will be unlikely to utilize unsecured debt. While applications will increase, key risk indicators such as 90+ DQ rates and existing credit scores may not be as accurate while banks provide forbearance programs.
Here at AQN, we will be watching Louisiana unemployment numbers closely over the next few weeks. We will be sure to keep our blog updated as unemployment curves either align or diverge from those of Hurricane Katrina. As always, we will keep using unique data sources to help search for answers in these trying times.
With the lack of reliable data due to decreased bureau reporting, lenders will need to think proactively about how to incorporate new data sources, like unemployment data, to build confidence when lending. If you are interested in discussing these opportunities for your lending business, please reach out to: Joshua Suen - jsuen@aqnstrategies.com