Since the COVID-19 pandemic emerged, S&P Global Ratings has generally implemented higher base-case cumulative net loss assumptions, additional sensitivity scenarios, and adjusted certain cash flow assumptions when applying its criteria in its analysis of U.S. auto loan asset-back securities (ABS) transactions (see "The Potential Effects Of COVID-19 On U.S. Auto Loan ABS," March 26, 2020). These adjustments reflected our view of the negative impact that the COVID-19 induced economic dislocation would have on wages, unemployment, and, ultimately, the ability of borrowers to continue making payments on their auto loans.

We have now observed the COVID-19 impact through 10 months of surveillance reports following the start of the pandemic in March 2020. In addition, we have received feedback from servicers and originators regarding performance trends, including extension data and any changes in their underwriting and operational processes due to the pandemic. In many cases, we have augmented this data with loan-level performance information that demonstrates the degree of success lenders have had with extended loans resuming normal payments. We have also updated our macroeconomic forecasts, including our most current unemployment forecast, to 6.4% for 2021 from our prior forecast of 7.2% in June of last year. We have incorporated this additional information regarding the effect of COVID-19, both into our ratings for new issuance transactions and in our surveillance of outstanding ratings.

COVID Performance And Its Impact On Our Analysis

The generally better-than-expected loss results, improved employment outlook, and wide variances in extension rates across issuers led us to reduce the fixed incremental adjustment to base-case cumulative net losses that we applied to our auto loan ABS transactions last year beginning in late March. In the application of our criteria, we are now factoring into our analysis issuer-specific performance data since the start of the pandemic, among other factors, to determine the degree of the COVID-19-related loss adjustment for a given transaction.

Amid the pandemic, we also ran additional liquidity sensitivity runs according to our criteria to test potential cash flow reductions early in each transaction's life due to higher extensions, forbearance, and delinquencies. We discontinued performing these additional runs in the summer after we observed a declining trend in extensions and determined that collections would have to decline by an extreme amount before subordinated classes would experience a default on timely interest. This is partly due to the ABS structure's ability to use all collections, including principal and excess spread, to pay bond interest expense.

At the onset of the COVD-19 outbreak, we reflected our expectation of a decrease in collections due to higher levels of unemployment and higher extension rates in our assumptions for the money market Rule 2a-7-eligible classes by testing zero prepayments or applying other haircuts to expected cash flows (previously we had typically ascribed a 0.25%-0.50% voluntary ABS credit in the sizing of such classes). This was in addition to our already conservative modeling assumptions in the sizing of the money market class (i.e., we generally assume only 11 months of scheduled payments are made over a 12-month period and give no credit to excess spread or possible reserve account draws to pay the class). Unlike subordinated classes, the money market class has a short tenor (generally 12-13 months) and is the senior-most class in the capital structure. This makes the class more susceptible to payment shortfalls if collections should decrease. As a result, the use of zero prepayments and the employment of other haircuts on an issuer-specific basis may remain in some cases to reflect the risk of a decrease in collections should delinquencies or extensions begin to materially increase again. However, we will continue to monitor issuer-specific prepayment data (to the extent they are available) and reflect them in our assumptions as warranted.

For non-money market classes, we similarly began incorporating additional scenarios in our analysis after the pandemic started, or required more months when testing the legal final maturity to accommodate lenders liberalizing their extension policies. These adjustments, in our view, helped to account for the potential impact on transaction cash flows resulting from previously extended loans in the pools and issuers' current extension policies, and we may continue to use them in our analysis on a transaction-by-transaction basis, as appropriate.

High Extension Rates, But Muted Impact On Credit Performance

As we expected at the onset of the COVID-19 pandemic in March 2020, extension rates increased significantly as servicers sought to provide payment relief for borrowers that were unemployed due to business closures from the statewide mandated shutdowns. Monthly extension rates for U.S. prime and subprime auto loan ABS peaked in April, but subsequently trended downward as statewide closures eased in May and through the summer months (see chart 1).

Chart 1

Extended loan status

Despite the reduction in extension rates from their peak last spring, we have yet to see delinquencies and losses increase to typical pre-COVID levels (see charts 2 and 3), although they are starting to trend upward. Strong delinquency and loss performance has been due to a confluence of factors that have occurred since the start of the pandemic. These included federal aid from the Coronavirus Aid, Relief, and Economic Security (CARES) Act, which provided enhanced unemployment benefits and stimulus checks; the unprecedented levels of extensions granted by auto loan servicers; and soaring used vehicle market values. The federal aid from the CARES act has significantly improved the liquidity and savings of many borrowers across the U.S., although it has had an uneven impact, as demonstrated by higher extension levels in subprime.

Chart 2

Chart 3

Since March, the cumulative number of charge-offs on previously extended loans in public prime and subprime pools remained low through November, at 1.37% and 3.87%, respectively (see chart 4 and "Higher COVID-19 Infection Rates Played A Role In Higher November U.S. Auto Loan ABS Extensions In Some States," Jan. 25, 2021).

Chart 4

Keeping An Eye On The Road

We are continuing to monitor how COVID-19 is affecting outstanding transactions and considering how it will impact future ones. As the pandemic has intensified in recent months, we expect monthly extension rates to continue to rise, but not to the levels we saw in in the spring of 2020. The high rates of infection have led to reduced business activity, which, in turn, has cut working hours and salaries for many of those still employed, and have kept first-time unemployment claims high (at nearly 780,000 for the week ended Jan. 30, 2021--about 3.1x pre-pandemic levels). Still, unemployment levels remain substantially below where they were in April (peaking at 14.7%) and we expect them to average 6.4% this year. Also, a potential third round of stimulus checks is currently being discussed by President Biden's administration that, along with the stimulus check received in January and the upcoming tax refunds, could reduce the need for extensions and limit the potential increase in losses. We also expect used vehicle supply to remain constrained throughout most of 2021, which could continue to support strong used vehicle values and assist with decreasing the severity of losses on charge-offs.

S&P Global Ratings believes there remains high, albeit moderating, uncertainty about the evolution of the coronavirus pandemic and its economic effects. Vaccine production is ramping up and rollouts are gathering pace around the world. Widespread immunization, which will help pave the way for a return to more normal levels of social and economic activity, looks to be achievable by most developed economies by the end of the third quarter. However, some emerging markets may only be able to achieve widespread immunization by year-end or later. We use these assumptions about vaccine timing in assessing the economic and credit implications associated with the pandemic (see our research here: www.spglobal.com/ratings). As the situation evolves, we will update our assumptions and estimates accordingly.

Related Research

• U.S.Biweekly Economic Roundup: Jobs Are Off To A Bad Start, Feb. 5, 2021
• Higher COVID-19 Infection Rates Played A Role In Higher November U.S. Auto Loan ABS Extensions In Some States, Jan. 25, 2021
• U.S. Auto Loan ABS Tracker: November 2020 Performance, Jan. 14, 2021
• The Potential Effects Of COVID-19 On U.S. Auto Loan ABS, March 26, 2020