Synopsis: In today’s article, we will discuss the differences in the models employed by lenders to determine collection risk and the underwriting risk of loans and how these processes can be improved further through analytics.

Individuals, companies, and governments have been raising capital through loans to fund their requirements, ranging from purchasing equipment to building factories, for a long time. While the reasons for applying for a business loan have remained pretty much static through the decades, the process for determining the suitability of borrowers has undergone a transformation. 

Financial lenders, including banks and NBFCs, typically conduct credit analysis to determine the risks entailed in underwriting a loan to a particular borrower. But the story of a loan disbursal doesn’t end here. Lenders need to continue to monitor their loan portfolios continuously for collection risks as well.

Let us first understand underwriting risk, which is the primary factor that banks account for while deciding on a loan sanction.

What is Underwriting Risk?

The very first step in a loan sanction is to determine whether the borrower is creditworthy. This process of risk evaluation is called underwriting. Thus, underwriting risk denotes the financial risk of an individual or business. 

Conducting an underwriting analysis is essential as it helps in assessing the degree of risk of the individual/business/government. As a result, lenders are in a position to fairly price their loans after risk determination. Underwriting, therefore, enables lenders to take a binary decision: whether to lend or to refuse.

How Does Underwriting work?

Traditionally lenders have primarily relied on credit scores to evaluate a borrower’s suitability for a loan. The process of underwriting would begin with an evaluation of the borrower’s income and current financial position. If a loan is being sought against property, then underwriting would also involve the collateral’s appraisal.

The most important determinant in risk analysis is the borrower’s credit score as it helps in understanding whether the borrower is reliable in the repayment of his loans. This requirement for a thick credit file has resulted in a huge credit gap in the economy. As a result, many financial startups have been coming up with tech-driven proprietary models to bridge this gap. 

The burgeoning of FinTechs has caused a transformation in the way lenders underwrite their loans. These financial companies utilize big data and alternate data to get a holistic picture of the borrowers’ financial standing to determine their creditworthiness. In contrast to banks that rely on 20-50 data points, FinTechs generate over 500 data points by aggregating alternate data into machine learning algorithms to decide on a loan sanction.

Thus, the process of underwriting is informed by thorough research on risk assessment and has been vastly improved with analytics. However, as mentioned before, the loan life cycle doesn’t end at sanctioning of the loan. Instead, the lenders also need to account for collection risk. But what is collection risk, and how do lenders determine it? Let’s find out.

What is Collection Risk?

Collection risk refers to the probability of a debtor failing to make his interest or principal payments as per the terms of the contract. This does not necessarily imply that the debtor is in a state of bankruptcy. Instead, it indicates that the borrower’s financial position or a business’ performance has deteriorated.

Traditionally, the collection process has been labor-intensive and focused on finding customers who can pay. Lenders generally employ manual data processing methods, which are inefficient in terms of categorizing and assessing borrowers as per their repayment patterns. Many lenders are still reliant on traditional collection practices of automated calling by aggressive agents – methods that are not favored by customers who now prefer digital-first contact.

Underwriting Risk vs Collection Risk

While lending institutions have graduated to employing models for determining collections risk, these collection models share similar parameters with the underwriting models, which may not be appropriate under all circumstances.

This is because the focus during the collection period is to zero in on the customer who is at the most risk of defaulting on payments but is also amenable to payment if better terms are offered; the focus is not on collecting from every defaulter. Thus, analytics are employed to ensure better segmentation of customers and to tailor collection strategies accordingly.

Additionally, the data requirements increase manifold during the tenure of the loan. Depending on the loan term, repayment abilities may be impaired by the business cycles, interest rate environment, and global headwinds. So, instead of averaging out the future income potential, as is done in underwriting models, the collection models will have to project for specific monthly cycles. This requires close monitoring and evaluation of the entire loan cycle to ensure repayments are made on time and provisions can be created in case of defaults.

Another difference is that a collection risk analysis needs to account for the reasons for delinquency. To illustrate, a borrower may fall back on payments due to forgetfulness or due to temporary blocking of funds. Such loan accounts will require a different strategy compared to a borrower who has been skipping payments in succession. This nitty-gritty isn’t dealt with in an underwriting risk model.

Collecting Digitally

Traditional risk models are becoming less effective in predicting collection risk over time. Many upcoming financial lenders are instead focusing on developing consolidated data portals to improve their collections and recovery processes. Such integrated portals enable banks to set financial triggers in the system that can highlight the borrowers that are likely to fall into delinquency.

Lenders are now increasingly employing analytics based on machine learning (ML) to calculate the behavior scores of a borrower. As a result, lenders can identify high-risk accounts and segment their loan portfolio into broad categories based on borrowers’ risk profiles. 

To illustrate, if out of 100 borrowers, 8 are highly leveraged versus the remaining applicants, then lenders will prioritize these 8 borrowers as they denote high risk. Thus, lenders can employ different collection strategies when they are armed with such value-at-risk-based classification.

Furthermore, behavioral segmentation helps identify a borrower’s behavioral clues and receptivity towards different contact approaches, including SMS, automated calls, or emails. Thus, lenders can employ an omnichannel contact strategy that is customized as per the borrower’s profile, thus reducing any unnecessary spamming. This strategy ensures that the right contact channels are used in the right sequence for the right customers. 

Bottom Line

Real-time data and advanced analytics will continue to power the lending environment. Armed with insights into borrowers’ financial behavior, credit events, and default probabilities, lenders are in a better position to manage underwriting and collection risks. This will help in reducing delinquency losses – an important factor that can make or break a business.

We, at Protium, have created superior proprietary models that can holistically capture an applicant’s creditworthiness. Apply for a collateral-free loan today, even if you don’t have a formal credit history. We offer business loans at attractive interest rates with minimal documentation. Call 8828827800 to know more!