3 Steps Banks Must Take to Ensure PPP Loans Are Forgiven
- mayo 14, 2020
In times of crisis, quick, decisive action is required to help prevent a bad situation from getting worse. That strategy recognizes that longer-term problems are accepted to provide more immediate benefit. The Paycheck Protection Program (PPP) is a great example of rapid response that presents long-term challenges. According to the SBA, more than 1.6 million PPP loans were given out under round one of PPP. With an additional $310 billion funding a second round, it is safe to assume that the program will see another 1.6 million loans. The speed at which the program was implemented, and the volume of loans requested, presents unprecedented servicing challenges banks will have to overcome to maintain regulatory compliance, deliver on the promise of the program to help small businesses, and avoid public criticism.
There is a lot being said about the potential shape and timing of the economic recovery. Amid all that uncertainty and speculation, one certainty emerges: banks and other lenders will have to deal with servicing these loans for the next two years or more. “Unprecedented” is a word applied to many elements of this crises and this servicing challenge is no exception. Lenders must deal with what is essentially a new loan product with features, processes, and volumes unlike products they are more comfortable dealing with. Here are three steps lenders can take to begin to scale to meet the challenge.
Step 1: Validate legitimacy
For most lending, the underwriting process takes time. This time is needed to collect and analyze documents to support the claims made by the borrower. The PPP dramatically abbreviated this process in favor of speed, delaying the banks’ need to perform these typical checks and balances until after the loans are made. No bank wants to be party to inappropriately originating loans. So, a first step in tackling the service challenge is to review the loans they have originated. Start by validating the applicant’s responses to eligibility related to criminal activity. It doesn’t take a risk management expert to know that an applicant who has been indicted or convicted of a felony might lie on the application. Lenders need to look beyond just the applicant, as the program requires evaluating the record of any individual owning 20% or more of the equity of the applicant’s business. If a lender’s KYC process for determining and validating beneficial ownership is not up to the challenge, they may be subject to KYC penalties, just as they would for all other loans. Next, the delinquency and default questions should be addressed. If the applicant is delinquent on any federal loan program, they are ineligible for a PPP loan. One would reasonably expect the SBA to validate delinquencies and defaults for the SBA, but testing if the applicant is delinquent with any other federal agency is a bit tougher. Again, a prudent risk manager would not choose to depend on efficient and effective cooperation between government agencies when looking for delinquencies.
Data management and robotic process automation (RPA) are the keys to efficiently meeting this challenge. Financial services organizations can use these technologies to query the relevant sources and systems, many of which live outside the bank’s firewall, to quickly collect the information necessary to confirm the truth of the applicant statements. It is not a perfect solution; nothing guarantees 100% accuracy, but from a risk management approach, being able to demonstrate that reasonable processes were followed for all applications, without slowing the application process, puts the bank in a defensible position when regulators look at the appropriateness of a given lender’s approval decisions. With manual processes disrupted by the need for employees to work remotely, the best way to properly work through surges in volume like the one created by the PPP is through technology.
Step 2: Validate the need and loan amount
After addressing the legitimacy of the application, it’s time to look at the appropriateness of the application and the amount of the loan. The first challenge is dealing with the subjective assessment of an applicant’s need for the loan. There are multiple examples of large companies with significant financial reserves applying for PPP loans. Some have been shamed in the media and have returned the funds, but the question is, “How many others are out there?” Regulators have already signaled an intention to review all large loans under the program, and banks need to prepare for this review. If an individual borrower is found to have not needed the loan, it will reflect on the borrower. However, if a lender shows a pattern of lending to applicants that didn’t need the loans, that will reflect on the lender and expose them to potential fines or public criticism. In addition to providing proof of their need for PPP funds, borrowers must submit documents supporting their operating dates and payroll. Given that 75% of the funds must be used for specific purposes, (such as paying employee salaries), for the loan to be forgivable, lenders will need to collect documentation defending the loan amount.
There is good news here as banks look to meet this challenge. As with the prior challenge, success at scale depends on automation, specifically automated document processing and data capture. Until recently, OCR and other document processing technologies have been more of a promise than a reality. There are now a small number of vendors that have developed a new approach to applying AI to the data capture challenge and are delivering on the promise. This technology delivers higher levels of accuracy with much lower training times. The algorithm does not need to be trained on each variation of a template. Paystubs, for example, all contain the same basic information, but in thousands of different formats. This new technology needs to be trained on only a handful of versions before it will work on other versions.
Step 3: Help your clients obtain forgiveness
After confirming the loans are legitimate and appropriate, the most important step is ensuring they are forgiven. If the loans are not used for approved purposes, the loan will not be forgiven and becomes a two-year loan at 1%. Worse is the fact that instead of helping small businesses, the PPP may have just increased overall indebtedness. This is one area of the program where the details are particularly subject to change. Lenders need to be flexible enough to be able to respond to the changes and scalable enough to work with all borrowers to maximize forgiveness.
Automation and data capture are critical to success. Borrowers will need to submit additional payroll, rent payment information and other documents to support requests for forgiveness. There may even be a new request form required. Through use of data capture and RPA, documentation can be processed, and the information it contains assessed, to confirm that the loan meets the requirements for forgiveness. This will allow lenders to be “exception-based” and focus on serving the needs of those clients having a harder time qualifying for forgiveness while borrowers with proper and complete documentation can move quickly through the process.
A question of confidence
I regularly meet with lenders of all sizes to discuss challenges like those presented by the PPP. I regularly hear clients tell me, “We’ve got this.” Sometimes their solutions are to apply extra resources to manual processes. Sometimes they just assume processes work properly. My question is always the same: “How confident are you that you are right?” The answer too often is that organizations don’t really trust their data. As the PPP moves from mass origination into the servicing phase, confidence in data and the ability to process both structured and unstructured data from forms and documents is critical to effectively serving those 3.2+ million borrowers looking to come through the crisis and return to success.
Find out more how NTT DATA can help you get the confidence that your customers’ loans will be forgiven.
Subscribe to our blog
