PPP Lenders Face Myriad Risks Responding to Reeling Businesses

Challenge #1 – Application Data Entry

SBA’s E-Tran system requires applications to be inputted manually and has frequently crashed, causing applications to be restarted from the beginning. Recently, amid lenders’ significant challenges accessing E-Tran, the SBA announced it is prohibiting the use of robotic process automation, or RPA, to submit loan applications directly into the system. The SBA has, however, allowed one-time bulk submissions of 5,000 applications or more via XML transfers, and has also opened up a more automated Application Programming Interface (API) submission process to all lenders, building from a pilot that a smaller group of lenders had tested between when the first and second funding pools were established.

Short-Term Fixes

Notwithstanding the difficulties, RPA data entry bots were used initially to cut the input time for each application drastically, down to two minutes or less in certain cases, compared to 20 minutes or more using manual processes. With the SBA’s recent decision to prohibit direct bot access to E-Tran, lenders will need to submit forms through the SBA’s new API. We are helping clients migrate to the SBA’s API submission tool, including assisting with converting existing RPA processes to connect to the API. Additionally, we are helping clients utilize RPA for supporting processes, such as data quality checks, account creation and documentation review – activities which are even more critical now. RPA significantly enhances the efficiency of teams entering data into the E-Tran system. Leveraging dedicated workflow tools for PPP, like those developed by vendors such as Nintex, can also enhance these processes. Overall, our solutions are helping clients reduce the number of failed applications due to crashes and increase the accuracy of data entered into E-Tran. In addition, lenders, particularly those who continue to submit applications manually, should also be performing manual quality control, “maker/checker” reviews.

Long-Term Solutions

The SBA’s pilot API tool is expected to allow higher volumes of applications to be submitted in a much more automated fashion. Concurrent with efforts to implement the fully automated solution, lenders should also work to create additional automated exception reporting to help manage data integrity. For example, in the up-to-ten-day period between when PPP loans are approved and funded, lenders should have developed and be able to run exception reports that compare and identify any differences between customer data maintained in the bank’s core system on existing relationships against data submitted on borrower applications, and then further compare that information to data submitted to E-Tran. Any discrepancies should be investigated and corrected as appropriate before disbursement.

Challenge #2 – Borrower Documentation Verification

Although the PPP Interim Final Rule and related FAQ guidance issued by the SBA and Treasury provide broad rights for lenders to rely on information submitted by borrowers, the FAQ states that “lenders are expected to perform a good faith review, in a reasonable time, of the borrower’s calculations and supporting documents concerning average monthly payroll cost.” We have observed significant differences across the industry regarding the types of documents each lender is requiring borrowers to submit and how this information is being reviewed and validated. Lenders are concerned about the possibility of SBA guarantees being invalidated in the future should loans default and the SBA later determine the lender missed obvious signs that a borrower was not eligible for the program at all or the loan amount received exceeded the maximum allowable, or that outright fraud existed.

Lenders should establish a basic risk rating process to define different categories of applicants and document the specific information that must be gathered and how that information will be verified for each category.

Short-Term Fixes

Lenders should establish a basic risk rating process to define different categories of applicants and document the specific information that must be gathered and how that information will be verified for each category. For example, categories might be based on some combination of things like:

• The existing customer risk rating established for anti-money laundering/know your customer (AML/KYC) purposes, the extent of KYC documentation already maintained, and/or how long the borrower has been a customer of the lender.
• The nature of the borrower’s business (e.g., brick and mortar retail or manufacturing versus online and/or services-based businesses).
• Whether and how recently the lender issued other loans to the applicant, the extent of information that was collected and verified when the earlier loan was originated and/or during subsequent credit risk reviews.
• Whether and how recently a site visit was performed on the applicant, and the extent of information collected during that visit.
• The quality and independent verifiability of documentation initially submitted with the application. For example, financial statements audited or at least prepared by a reputable CPA firm would be considered lower risk than spreadsheets prepared by the applicant itself. Similarly, payroll data produced by third-party providers, such as ADP or Paychex, would be considered more reliable than handwritten paystubs produced by the applicant.

Based on the risk level identified and the initial review of the documentation received for each applicant, lenders should be able to flag obvious discrepancies between application information and what the lender believes it already knows about the client. For example, borrowers that submit applications for loans that would cover $20,000 per month in payroll but also recently applied for another loan from the bank and claimed $10,000 per month in revenue in connection with the prior loan should warrant further review.

Also, and similar to the recommendations above related to data entry verification, lenders should establish quality control processes to make sure the required documentation is collected and consistently verified per procedures for each application between loan approval and funding.

Long-Term Solutions

Given the short amount of time the PPP is expected to be active, opportunities for significant automation of the verification process are limited. Lenders should consider creating automated feeds from existing source systems to generate simple worksheets containing the key KYC and financial statement data the bank already has about each customer. This makes it easier to compare that information to application data without taking the time to manually query multiple source systems. Additionally, lenders that have more advanced, artificial intelligence-driven decisioning models that rely on third party data sources for their normal course of business lending activities are using these tools to help spot red flags within PPP applications as well.

Lenders should establish quality control processes to make sure the required documentation is collected and consistently verified per procedures for each application between loan approval and funding.

Challenge #3 – Application Pipeline Management

The short amount of time the PPP will be offered coupled with demand that significantly exceeds the funding pool available has created a literal “race to the bank,” with applicants and lenders rushing to ensure they are eligible to participate in the program. Failing to timely communicate to customers about an application and to process it effectively, will likely, at a minimum, result in a lost customer relationship, or, at worst, the failure of the customer’s business and potential litigation against the lender.

Short-Term Fixes

Lenders should create dedicated processing teams aligned with the borrower categories described above in Challenge #2 to be able to quickly identify and consistently and accurately communicate to each borrower in that category the information that will be required. Lenders should also confirm the borrower’s preferred communication channels while the application is in flight and consider providing daily status updates via those channels. Where possible, e-mail addresses should be collected, as well, so that lenders can maintain an audit trail to support written communications provided about any missing information on a timely basis.

Lenders should also establish effective interim manual processes to aggregate the status of the application pipeline on at least a daily basis. This information should be used both to help make sure any applications stalled in the process are identified and actioned, and also to measure cycle time per application at each stage and help make sure the lender is dedicating adequate resources to the stages that create the greatest risk of slowing down the process.

Many lenders have quickly adopted dedicated PPP software solutions offered by platform providers, such as FIS, to help automate the management and reporting of their PPP pipelines.

Long-Term Solutions

Many lenders have quickly adopted dedicated PPP software solutions offered by platform providers, such as FIS, to help automate the management and reporting of their PPP pipelines. Alternatively, common commercial software tools such as ServiceNow can be used to build simple workflows that meet many of the same objectives. These types of tools can also be used to manage and retain documentation to support that the required edit checks and QC reviews were performed for each loan prior to funding.

Challenge #4 – Loan Forgiveness and Borrower Credit Risk

PPP borrowers that maintain or restore the required levels of employee payroll eight weeks after their loan is disbursed qualify to have their loans forgiven. Between this provision and the fact that even loans not forgiven are fully guaranteed by the SBA, lenders do not directly bear any credit risk related to PPP loans themselves. Additionally, there is no provision in the Interim Final Rule or the SBA’s FAQ guide that contemplates lenders performing traditional underwriting of PPP applicants’ ability to repay the loan along with any other outstanding borrowings. Therefore, in the event a PPP borrower does not eventually qualify for their PPP loan to be forgiven, this additional extension of credit may make it more difficult for borrowers to repay existing loans held by the same creditor, and/or introduce a competing preference claim if the borrower eventually declares bankruptcy.

Short-Term Fixes

Considering the pressure lenders face to originate PPP loans as quickly as possible, and that it is difficult if not impossible for most borrowers to find other lenders willing to originate a PPP loan if their primary banks refuse to do so, we don’t believe it is practical or advisable for lenders to decide on a case by case basis not to accept a PPP application due to creditworthiness considerations. That said, there are several steps lenders can take to mitigate this risk, including:

• Developing clear and detailed borrower disclosures, including things like payroll calculation worksheets, to help make sure borrowers fully understand the actions they must take to qualify for loan forgiveness.  
• Periodically and proactively communicating with borrowers throughout the eight-week forgiveness period to monitor their progress towards meeting the forgiveness standards.
• Beginning to collect required forgiveness documentation by week six of the eight-week forgiveness period in order to identify and give the borrower the chance to correct any issues prior to the deadline, in order to maximize the amount eligible for forgiveness.
• Conducting more frequent and detailed credit risk reviews of borrowers that hold a PPP loan along with other extensions of credit.

If appropriate and in line with safe and sound lending practices, considering providing additional funding for businesses that will need a working capital infusion on top of their PPP loan to restart operations and rehire employees by the end of the eight-week forgiveness period. As an example, certain borrowers may be eligible for a term loan under the Federal Reserve’s Main Street Lending Program, which provides benefits both to the lender (e.g. government guarantee for 95% of principal) and the borrower (principal and interest deferred for a year).

We expect more SBA scrutiny of, and potentially more subsequent disputes and litigation related to, the loan forgiveness stage as compared to the loan origination process.

Long-Term Solutions

Unlike the PPP application process, the eight-week tail for loan forgiveness provides lenders enough time to establish more efficient and better controlled processes and infrastructure for this stage. Correspondingly, however, we expect more SBA scrutiny of, and potentially more subsequent disputes and litigation related to, the forgiveness stage as compared to the origination process. As a result, lenders should consider:

• Establishing an automated forgiveness application workflow to help guide borrowers through the documentation they will need to provide during this process.
• Putting in place forgiveness quality control and/or audit processes within the next 30 days, including calculation logic for determining eligibility for full versus partial forgiveness.
• Piloting quality control processes early in connection with a sample of the earliest borrowers likely to qualify for loan forgiveness. The Federal Reserve Board and the SBA recently clarified that banks may make PPP loans to related parties of their directors and shareholders under certain circumstances, so this might represent a willing population with which to pilot the forgiveness process for lenders that originated any such loans. Based on the results of that exercise, lenders should track the time required to compile and submit the forgiveness application and ensure adequate staffing is in place for the expected peak period of forgiveness requests.
• If not already done during the application process, developing RPA tools to increase the speed and accuracy of the process for submitting loan repurchase and forgiveness applications to the SBA.