The COVID-19 crisis is the largest public health crisis the world has ever faced. As citizens we must do our part to help ‘flatten the curve’.
As bankers, we must provide financial relief to our clients and understand the financial impact of the crisis on our financial institutions (FIs). While the recently passed CARES Act provides significant financial relief to many, it also presents new financial challenges to FIs.
Economically, this is a very uncertain time. The range of possible impacts of COVID-19 on our financial well-being is greater than any previous event on record. A natural disaster that starts as suddenly as a hurricane yet keeps growing in force over weeks or months is unprecedented. No one can know with any real level of certainty what the rest of 2020 will hold, so we must plan for the unknown.
The recent CNN article “How quickly can the US economy bounce back? That depends on the virus” discusses the “shape” of a recovery and offers three possible views – V-shaped, U-shaped and L-shaped. While we all hope for a V-shaped recession (with a quick bounce back after a dramatic downturn), we also need to consider the possibility of the U-shaped or L-shaped recovery, which have much longer recovery processes.
Early economic and disease data suggest that a V-shaped recession is unlikely, so we need to plan for a much longer-lasting event. This range of possible economic scenarios makes it ESSENTIAL to stress test your portfolio using various economic scenarios. Simply hoping for a speedy recovery is not a prudent strategy for FI management.
The Range of Possibilities
Stress testing is the equivalent of health insurance for FIs and if you aren’t already doing it, you need to. A scenario of a short wave of COVID-19 infections in the US and a corresponding V-shaped recession is ideal, but the $2 trillion rescue package recently passed by Congress confirms that the Mild scenario is unlikely at this point.
A more likely outcome is the combination of shelter-in-place requirements and dramatically reduced business activities cause a significant reduction in GDP through the Spring and Summer before stabilizing (U-shaped).
Unemployment, which grew in March faster than ever seen before, would reach recessionary levels approaching 10% after adjusting for government programs to encourage businesses to retain workers. Although the onset of the recession would be at record speed, the duration would be short, with a very strong rebound by the end of the year. The net impact of all this volatility would look like half of the 2009 recession, although with the stresses coming in different industries and loan types from the last recession.
If we fail to control the spread of the virus in the short term with corresponding restrictions lasting into the fall, we are likely to experience a U- or L-shaped recovery (i.e. a prolonged recession), possibly reducing US GDP by 5% to 6% for 2020 and the GDP growth stabilizing in early 2021 at a lower level. Even with government supports, unemployment could top 15% and remain there for many months.
The greatest risk in the prolonged recession scenario is that it triggers a wave of defaults in C&I, CRE, and Small Business lending. With the dramatic, leveraged loan growth in this section, this is the real risk of a replay of the Great Recession. So far, the Federal Reserve appears ready to buy as much corporate debt as necessary to prevent that from happening.
Just as we have never seen a recession as sudden and severe as with COVID-19, we have never witnessed such an immediate government response of such magnitude.
As we move toward more severe economic scenarios, we expect additional waves of government support. In the Hong Kong SARS Recession, we learned that the best way to model a recession from an epidemic was to look at consumer response to economic conditions rather than trying to model the number of people who were sick.
The US government response is so intense this time, that loss forecast models looking at macroeconomic factors will over-predict the consequences to consumers, so neither the number of confirmed cases nor the level of unemployment will truly express the net impact on consumers.
With no historic data on a comparable pandemic, our only path is to run our models against macroeconomic data and incorporate qualitative adjustments (Q-factors) to dial the results back some amount to compensate for government assistance. The Coronavirus Aid, Relief & Economic Security (CARES) Act is a $2 trillion attempt to prevent exactly what our models would predict. But no government program will be able to erase all the effects, so reality will be somewhat less than the unadjusted model forecasts, but not negligible.
Examiners and auditors often dislike negative Q-factors, or management judgment reducing loss estimates from models, but this is a clearly valid exceptional case. Perhaps the best approach is to assume that the COVID-19 Severe Recession will occur but assume that portfolio losses will look more like the COVID-19 Adverse (median) case.
“Good” outcomes for lenders will not be achieved without aggressive action and some amount of pain. This is where lenders must plan for a marathon. If otherwise good borrowers will be distressed for 6 to 12 months, but we are convinced that a strong rebound will happen eventually, how do we get them there? Hardship programs, loan rewrites, and other support programs must consider that the stressful period could be lengthy.
From a collections perspective, expect the payment hierarchy to shift. With large numbers of people out of work or working from home, the vehicle may be the least useful asset. Payment and foreclosure moratoria on mortgages mean that consumers might stop paying those early as well. In a home-based economy, credit cards and lines of credit may be the greatest asset to both consumers and small businesses.
Consider also how to lend through the crisis. Governments are encouraging lending and are offering substantial support to do so, but this still must be thoughtful. The crisis is a very industry-specific event. While many people are suddenly without work, others in critical infrastructure industries are working harder than ever to support society. Both groups may continue to be good credit risks, but what kind of loans does each group need? Those 0% teaser rate credit cards, targeted to the good consumers, look like an ideal way to help borrowers across this chasm and obtain longer-term goodwill.
This is a different kind of crisis, and it creates an opportunity to go beyond marketing clichés and really forge partnerships between lenders and borrowers. Our society needs both to bounce back stronger, the way we all hope.
The COVID-19 scenarios described above are available to Profitstars CECL customers from ScenarioAI.com at 50% off the regular price. Contact us for how to use these in your loss forecasting.
This post was co-written by Joseph L. Breeden, PhD from Deep Future Analytics and Brad Dahlman from Profitstars.