Collateral performance in the U.S. auto loan asset-backed securities (ABS) sector continued its traditional seasonal weakness in July 2019. Losses, delinquencies, and recoveries worsened month-over-month for both the prime and subprime segments, though subprime recoveries improved annually.

Seasonal Losses Increased For Prime And Subprime

U.S. prime credit losses increased to 0.55% in July from 0.48% in June. They also increased year-over-year from 0.50% in July 2018 (see table 1 and chart 1). One of the key factors in the prime index's annual uptick is the index composition because higher-loss prime issuers make up a larger share of the index. Also, certain large issuers are reporting higher year-over-year losses, with some of this stemming from lower recoveries. However, most prime issuers continue to perform within our original loss expectations.

Subprime losses increased 128 basis points (bps) to 8.21% in July from 6.93% in June and 40 bps from 7.81% in July 2018. Santander Drive Auto Receivables Trust (SDART) and Drive Auto Receivables Trust (DRIVE) losses increased considerably month-over-month and year-over-year. These two platforms make up approximately 45% of the subprime index. After netting out three deep subprime issuers (including DRIVE), modified subprime losses increased by a lesser extent--91 bps month-over-month to 6.37% in July. On an annual basis, the modified index increased slightly year-over-year from 6.18% in July 2018.

Table 1

Chart 1

Seasonal Recovery Rate Declines Continued And Subprime Recoveries Improved Annually

Prime recoveries declined to 58.50% in July from 60.18% in June and from 59.71% in July 2018 (see table 2 and chart 2). Subprime recoveries decreased to 42.26% in July from 44.10% in June and increased from 40.33% in July 2018. Subprime modified recoveries decreased to 42.52% in July from 44.41% in June and improved from 41.23% in July 2018.

The monthly downtick in recoveries across the board in July is a seasonal feature. The annual improvement in the subprime space is led by improved performance of the DRIVE and SDART transactions.

Table 2

Chart 2

Delinquencies Increased Month-Over-Month

The prime 60-plus-day delinquency rate was nearly stable at 0.41% in July compared to 0.39% in June and 0.40% in July 2018 (see table 3 and chart 3).

The subprime 60-plus day delinquency rate increased by 25 bps both month-over-month and year-over-year to 5.25% in July. This is the highest July level since we started tracking this statistic, and it is largely due to the greater concentration of deep subprime loans being securitized. On a modified basis, after netting out three deep subprime lenders, the subprime modified 60-plus day delinquency rate increased to 3.75% in July from 3.60% in June and 3.56% in June 2018.

Table 3

Chart 3

Revised Loss Expectations

In August 2019, we revised our loss expectations for the following transactions:

• One Avid Automobile Receivables Trust transaction backed by subprime auto loan receivables.
• Two United Auto Credit Securitization Trust transactions backed by subprime retail auto loans.
• Six Toyota Auto Receivables Owner Trust transactions backed by prime retail auto loans.

Of the nine transactions we reviewed, we lowered our expected cumulative net losses (CNLs) for seven and maintained them for two.

Table 4

Table 5

Table 6

Historical Ratings Activity

Through August 2019, our reviews resulted in 212 upgrades and one downgrade (see "Honor Automobile Trust Securitization 2016-1 Class B Notes Downgraded," published Feb 15, 2019). In June, we placed the lowest most tranches from four CPS Auto Receivables Trust transactions (series 2016-B, 2016-C, 2016-D, and 2017-A) on CreditWatch negative following our review in June 2019 (see "Various Rating Actions Taken On 11 CPS Auto Receivables Trust Transactions," published June 13, 2019).

We later downgraded the four CPS transactions (series 2016-B, 2016-C, 2016-D, and 2017-D) in September because they were performing worse than our initial expectations and the overcollateralization amount for each series was not at its requisite floor level. The series' class E notes were not providing an adequate level of coverage, therefore we lowered the ratings on the notes to 'B+ (sf)' from 'BB- (sf)' (see "Four CPS Auto Receivables Trust Ratings Lowered And Removed From CreditWatch Negative," published Sept. 9, 2019).

Table 7

Appendix I: Auto Tracker Methodology And Definitions--Frequently Asked Questions

Effective with our U.S. auto loan ABS tracker report, "U.S. Auto Loan ABS Tracker: Full-Year 2017 And December 2017 Performance," published Feb. 22, 2018, we modified our methodology for calculating the loss, delinquency, and recovery rates reported. Under the new methodology, we do not incorporate a transaction's performance into the composite results until it reaches its fourth month outstanding. We have applied the new methodology to transactions that closed in December 2005 and thereafter.

How do you define prime auto loan ABS?

We generally categorize prime auto loan ABS transactions as those backed by loan pools with initial expected CNLs of 3.0% or less, average FICO scores of 700 or higher, and annual percentage rates (APRs) of 0.0%-5.0%.

How do you define subprime auto loan ABS?

We generally categorize subprime auto loan ABS transactions as those backed by loan pools with initial expected CNLs of at least 7.5%, average FICO scores of less than 620, and APRs that exceed 14.0%.

How do you calculate the monthly net loss rate?

The monthly net loss rate is annualized. It equals each transaction's net loss rate weighted by the transaction's ending pool balance for the current month over the aggregate ending pool balance of all transactions included in the index.

We only allow a transaction to enter the composite starting in its fourth month outstanding. Transactions usually have zero or low losses during their first three months, which dilutes the composite figures.

How do you calculate the monthly recovery rate?

We calculate recoveries by taking the recovery amount reported (which typically includes all recoveries, including disposition proceeds, post-disposition proceeds, and any other reported recoveries) over the gross loss amount for the current month. We then weight each transaction's recovery percentage by the transaction's ending pool balance for the current month over the aggregate ending pool balance of all transactions included in the index.

We only allow a transaction to enter the index starting in its fourth month outstanding. During a transaction's first three months, unusually high or low recoveries are reported, leading to a spike in the composite figures. Previously, we often excluded recovery rates in the first two months of a deal's life because of negative recovery rates (resulting from recoveries exceeding gross losses).

How do you calculate the monthly 60-plus-day delinquency rate?

We calculate delinquencies by taking each transaction's 60-plus-day delinquency amount over the ending pool balance for the current month. We then weight each transaction's 60-plus-day delinquency percentage by the transaction's ending pool balance for the current month over the aggregate ending pool balance of all transactions included in the composite.

We only allow a transaction to enter the composite starting in its fourth month outstanding. During the transaction's first three months, zero or lower delinquencies are reported, which dilutes the composite figures.

What is the ALSI?

Our Auto Loan Static Index monitors the credit performance of securitizations that were originated in the same year on a weighted average basis. The number of months displayed for each vintage is generally determined by the last point for which all securitizations for that time period have a data point. We calculate the prime and subprime ALSI CNLs by taking the weighted average of the CNLs of the transactions that were completed in the same time period (generally a year). Each transaction's CNL is weighted by its initial pool balance over the aggregate initial pool balance of all the transactions included in the index for that period. In the subprime ALSI, transactions from Byrider Finance LLC (doing business as CarNow Acceptance Corp.), Credit Acceptance Corp., and DriveTime Automotive Group Inc. are excluded because they do not have the typical indirect auto loan business model.

Which transactions are included in the prime, subprime, and modified subprime composites and indices?

For a list of the transactions included in our prime, subprime, and modified subprime composites and indices, see "U.S. Auto Loan ABS Tracker: March 2019," published May 23, 2019. However, note that we subsequently added S&P Global Ratings-rated transactions that have since closed, most prime transactions that closed and were not rated by S&P Global Ratings from 2016 through the present, and the Santander transactions S&P Global Ratings did not rate.

Related Research

• Speed Bump Ahead: As Auto Loans Accelerate Toward 84 Months, Caution Is Warranted, Sept. 18, 2019
• Four CPS Auto Receivables Trust Ratings Lowered And Removed From CreditWatch Negative, Sept. 9, 2019
• U.S. Prime Auto Loan ABS Are Seeing More Back-Loaded Losses As Loan Terms Lengthen, July 30, 2019
• Various Rating Actions Taken On 11 CPS Auto Receivables Trust Transactions, June 13, 2019
• U.S. Auto Loan ABS Tracker: March 2019, May 23, 2019
• Subprime Auto Loan ABS Tracker: Losses Have Stabilized, But Renewed Growth Bears Watching, April 29, 2019
• The Severity Of Subprime Auto Loan Delinquencies Is In The Eye Of The Beholder, March 18, 2019
• 10-Year Retrospective: Changes In U.S. Auto ABS In The Decade Since The Great Recession, Feb. 15, 2019
• Is There Extension Tension In U.S. Subprime Auto Loan ABS? Nov. 29, 2018