S&P Global Ratings expects that most U.S. small businesses' revenues will suffer a severe near-term negative impact from the COVID-19 pandemic and the related social distancing measures, as well as the overall adverse effect on the economy and reduced household spending. Accordingly, we expect collateral delinquencies to rise in our rated small business asset-backed securities (ABS) universe due to payment deferrals and lower prepayments. In turn, this could increase liquidity risk and obligor defaults, which could also affect longer-term credit risk. Looking ahead, we believe the ultimate result on collateral and ratings will depend on the severity and duration of the coronavirus crisis, as well as the speed and scale of government stimulus intended to benefit this sector.

Scenario Analysis

The impact of government-mandated closures on nonessential businesses has been harsh and immediate. Business owners in some of the hardest hit industries (e.g., restaurants, bars, salons, gyms, and lodging) are frequent loan borrowers. To gauge the extent to which the 30 small business ABS transactions we rate can withstand liquidity and credit stresses, we broke down our rated universe into post-2008 transactions (which are typically backed by Small Business Administration [SBA] 7(a) loans' unguaranteed interests) and pre-2008 transactions.

Post-2008 transactions

In our opinion, the post-2008 transactions' SBA 7(a) loans will likely be able to withstand incremental liquidity stress in the short term because of recent government assistance programs, including the Coronavirus Aid, Relief, and Economic Security Act (CARES Act), which provide for the payment of principal, interest, and fees of current SBA 7(a) loans from April to September 2020. In addition, many of the SBA 7(a) borrowers may also be eligible for the Paycheck Protection Program, the Economic Injury Disaster Loan program, and other various state-level programs. While we believe these benefits will contribute to the relatively stable performance among small business ABS in the short term, we did not consider the current government support in our additional scenario analysis that gauges the pandemic's impact on credit or liquidity risk. These transactions have interest reserve accounts typically between 1%-2% of the original pool balance.

Pre-2008 transactions

The transactions issued prior to 2008 have low pool factors ranging from approximately 2%-15% and are predominantly backed by conventional, non-SBA 7(a) loans--primarily commercial real estate loans. While approximately half of the transactions have liquidity reserve accounts funded at or near the target balance, the remainder have depleted reserve amounts. Many of these transactions breached their performance triggers years ago and the notes are now paid in sequential order, with the senior-most outstanding tranche receiving all available principal. While the subordinated notes have a greater disadvantage compared with senior notes in deals with sequential pay, we recognize that diminished cash flows from high delinquencies could jeopardize an issuer's ability to pay timely interest on all notes.

Similar to borrowers of SBA 7(a) loans, many borrowers of these conventional loans may also be eligible for the Paycheck Protection Program, Economic Injury Disaster Loan program, and applicable state-level programs, which could provide some near-term relief.

Short-Term Liquidity Risk

In addition to the federal and state government stimulus that will benefit the small business ABS sector, there are also structural features embedded in these transactions that can mitigate liquidity risk.

We tested the adequacy of liquidity over a six-month period based on our economists macroeconomic forecast. For the pre-2008 transactions, which do not benefit from the same government support as the SBA 7(a) loans, we tested those transactions without a reserve account assuming there are zero prepayments and a 50% decrease in collections for the first six months. In the subsequent six months, we assume that payments are made according to schedule but there are zero prepayments. The non-investment-grade classes of the pre-2008 transactions, without reserve accounts, were not able to withstand a 50% increase in delinquency rates. These notes are mainly in the 'CCC' rated category. 'CCC' rated obligations are dependent upon favorable business, financial, and economic conditions for the obligor to meet its financial commitment on the obligation (see "Criteria For Assigning 'CCC+', 'CCC', 'CCC-', And 'CC' Ratings," published Oct. 1, 2012). For transactions with reserve accounts, available amounts are sufficient to cover senior interest on the notes (as required under the transaction documents) and fees assuming a 50% reduction in cash flows for six months. The risk associated with the pre-2008 transactions may be partially mitigated by the relatively low note balances and low monthly interest and senior fee obligations.

Almost all the post-2008 transactions are backed by SBA 7(a) loans. Under the CARES Act, SBA 7(a) loan amounts are expected to be paid by the government for six months beginning in April 2020. To determine whether available liquidity would be sufficient to cover senior fees and interest on the notes, we assume a 50% decline in available collections and determined that the remaining amounts would generally be sufficient to pay senior fees and interest on the notes.

We will continue to test liquidity coverage beyond the six-month period as we monitor delinquencies, the efficacy of government stimulus programs, and macroeconomic trends.

Structural Features That Can Mitigate Liquidity Risk

Typical small business transactions include the following structural features, which we believe mitigate certain risks.

Liquidity reserve account

Small business loan securitizations issued post-2008 typically benefit from an interest reserve account sized to a minimum required amount of 1%-2% of the original collateral balance, which increases as a percentage of the aggregate transaction as the tranches amortize. We estimate that this account typically covers interest and senior fee obligations for three months to over 12 months, in certain cases. As mentioned above, there are a limited number of pre-2008 transactions with no reserve balance, which could be vulnerable to liquidity risk if pools experience higher than expected delinquency levels, especially the subordinate classes.

Principal collections to cover interest

Most transactions permit the use of principal and interest collections on the loans to pay interest on the bonds.

Triggers to redirect cash flow

Pro rata pay transactions generally include performance triggers that redirect cash flows to pay down the notes in sequential order in the event of a breach in a performance trigger or event of default.

Servicer advances

Although our analysis does not give credit to servicer or optional advances, we acknowledge the benefit of structures that permit servicers to advance interest shortfalls. However, given the increased volume of deferrals and/or delinquencies (see "U.S. Commercial Mortgage Servicers Preparing For Impact From COVID-19," published April 3, 2020), we believe the servicers' abilities to advance amounts may be limited. Servicers in the post-2008 transactions are generally unrated non-bank licensed SBA 7(a) lenders..

Longer-Term Credit Risks

We are applying incremental scenario testing to the transactions we rate and are considering the following: the level of exposure to industries we deem adversely affected by the crisis, geographic concentration, seasoning of the loans, our opinion of the strength of the obligors, and the credit quality of the related collateral and recovery rates, if applicable.

Historically, in our analysis--considering that most small business loans are not rated--we use industry-specific default probabilities that are part of the SBA's 7(a) historical loan default data, which, in our view, capture a small business' historical default risk.

The SBA's 7(a) data capture small businesses' loan performance across many different business sectors and geographical locations over the past 30 years and across multiple U.S. economic cycles. These data are the largest publicly available source of small business loan performance information. We plot the distribution of cumulative default rates by different Standard Industrial Classification (SIC) codes based on the SBA's 7(a) program data in chart 1. The data show that small business risk varies, depending on the business sector or the SIC code.

Chart 1

In our analysis, the probability of defaults associated to each industry are derived using a vintage approach by following the default behavior of the loans originated in a particular cohort over different time horizons. For instance, the probability of default for a specific SIC code is the average of all available SBA 7(a) loan vintages of that particular SIC code. This means that very diversified SBA transactions with default rates associated with moderate stress (which we equate to a 'BBB' rating level) would generally equate to the cumulative default rates observed in the worst-performing vintages, which range between the mid- to high-30% range (see chart 2). This compares to default levels in the mid-60% range under extreme economic stress for those transactions rated 'AAA'.

Chart 2

We believe that pools with higher concentrations in industries hit by travel and consumption restrictions are more vulnerable to obligor defaults and could potentially exceed the highest historical levels. In our scenario testing, we may include higher cumulative default rates for certain affected industries.

Industry concentrations

Across our rated outstanding transactions backed by SBA 7(a) loans, industry exposure is generally diverse. The top five industry concentrations at the time of initial rating generally represent approximately 15%-50% of the overall pool, with the exception of one transaction that is almost exclusively concentrated in the hotel sector. Hotels and restaurants are among the top exposures in a number of our recently rated transactions.

As we noted earlier, collateral pools with higher concentrations in industries hit by travel and consumer consumption restrictions are more vulnerable to obligor defaults. We are in the process of obtaining the most up-to-date loan-level data, and will analyze the data to determine whether industry concentrations have implications on long-term credit.

Geographic concentrations

Currently, in general, our rated transactions are fairly geographically diverse across the U.S., with the exception of California, which has a disproportionately high exposure in certain transactions. At the time of our initial rating, a transaction's exposure to the top three states typically accounts for approximately 25%-85% of the total loan pool, while the largest exposure to any one state is approximately 75%. In our analysis, we applied a 30% intrastate and 15% interstate correlation coefficient, as local conditions have historically proven to play a significant role in default correlation.

While some areas may be more significantly affected than others, we expect defaults to be more heavily correlated across all states given the unique circumstances of the current recessionary period.

Our methodology accounts for regional correlation, such that joint defaults are simulated assuming a correlation structure between obligors in the same state and obligors in different states. The intrastate correlation parameter reflects the joint performance of two obligors in the same state. The interstate correlation parameter captures the performance of a pair of obligors in different states. The data used during the calibration of our analysis showed that intrastate correlation is generally higher than interstate correlation, and that correlation levels rise during periods of high defaults. For stress levels higher than historically observed, we believe that under extreme economic stress, the correlation level would also be much higher than historically observed. These correlation levels are higher than the sector correlation assumptions we use in our corporate collateralized loan obligation (CLO) methodology and were consistent with our view that small business defaults are more geographically correlated than corporate loan defaults.

The shelter-in-place orders as a result of the COVID-19 pandemic, however, challenge the assumption that geographic location rather than industry is a stronger indicator of correlated behavior. For example, certain industries have been more severely impacted than others during this economic downturn, regardless of location. In light of this concern, we may consider increasing default rates for certain more-affected industries in our analysis.

Seasoning

The impact of loan seasoning is an important factor in assessing small business loan defaults. We have observed that a loan is unlikely to default during the first year after origination. The probability of default then increases during the next few years before declining during the remaining years of the loan's life (see chart 3). For example, for loans collateralized by real estate, an improving loan-to-value (LTV) ratio may cause this behavior. During this downturn, given the severity and duration of the pandemic-related shutdown and the pace of the recovery, seasoning may not provide the buffer to defaults that it has historically. However, in the highly seasoned, pre-2008 pools, equity has likely built up for the borrower in the underlying real estate collateral, which should provide a mitigant to the increased rate of default for seasoned loans. As a result, in our analysis we may consider lowering the seasoning credit for transactions whose underlying collateral may have not reached low LTV levels and where the obligor may be deterred from defaulting.

Chart 3

Collateral

As part of our review, we look at the originators' or servicers' workout policies to better understand how these entities handle problematic credit situations. Historical loan recoveries are the key element in this analysis. In our analysis, we only apply recovery credit to first-lien commercial real estate and equipment. The methodology tiers recovery assumptions based on the level of stress associated with different rating stress scenarios. We apply a higher recovery level for lower rating stress scenarios and a lower recovery level for higher rating stress scenarios. Tiering such assumptions is consistent with our observation that recoveries are often lower during periods of greater economic stress. For similar reasons, we also assume that it takes longer to recover losses during those times.

Most of the loans in the small business loan securitizations that S&P Global Ratings rates are fully amortizing, collateralized by first liens on commercial real estate for which we generally base recovery assumptions on our U.S. commercial mortgage-backed securities criteria. However, the levels are generally lower to reflect the perceived lower quality of the properties backing the small business loans.

We believe that recoveries on loans backed by collateral other than commercial real estate may be stressed beyond our assumptions. Additionally, commercial real estate recovery-rate assumptions that we apply in the analysis of lower-rated notes may come under stress, as actual recoveries will depend on the depth and duration of the current recession and may be impacted differently by industry-specific concerns.

Servicer risk

The strength of the servicer is critical to ensure high-quality underwriting and loan servicing. During a period of economic stress, servicers' capabilities are often tested as collection efforts may be more intensive, resulting in an uptick in the foreclosure and repossession of collateral. In our transaction analysis, we consider the servicer's performance history and underwriting quality. Based on the outcome of our evaluation, we may increase or decrease the probability of default for an SBA 7(a) loan pool by applying a scaling factor of 75%-125% (see table 1).

Table 1

Ongoing Surveillance

As we develop more clarity on the expected size and duration of reductions in the transactions' securitized cash flows, we will evaluate whether adjustments to our assumptions are appropriate. If longer-term effects emerge that further reshape the economy or industry, we may revise our assessment of obligor default estimates, which, could put pressure on our current ratings. As we receive more issuer-specific and industry-level data--and learn more about the volume and speed at which government actions are benefiting the transactions, as well as what actions issuers re taking to mitigate the macroeconomic crisis--we will assess these transactions to determine whether rating reviews are warranted.

S&P Global Ratings acknowledges a high degree of uncertainty about the rate of spread and peak of the coronavirus outbreak. Some government authorities estimate the pandemic will peak about midyear, and we are using this assumption in assessing the economic and credit implications. We believe the measures adopted to contain COVID-19 have pushed the global economy into recession (see our macroeconomic and credit updates at www.spglobal.com/ratings). As the situation evolves, we will update our assumptions and estimates accordingly.

Appendix

We use our SBP Evaluator to determine a portfolio's loan default risk. Our analysis utilizes SBA 7(a) historical loan default data, which captures loan-level data across different sectors and geographic locations over the past 30 years, to help determine default probability at the aggregate pool level. Below is a list of key factors analyzed to determine the probability of a pool's default.

Table 2