Impact of Adverse Selection on Digital Lending: Heads-up to Community FI's

Introduction?

Adverse selection poses a significant hurdle for Community Financial Institutions (FIs) as they embark on their digital transformation journey. This phenomenon, rooted in information asymmetry and sub-optimal decisioning occurs when institutions attract a disproportionate share of high-risk customers relative to their competitors. To successfully navigate this challenge, Community FIs must leverage real-time decisioning technology with sophisticated analytics to ensure they can expand their digital presence while maintaining the quality of their customer base.

Let's dig deeper to understand the core reasons for adverse selection.??

Impact of sub-optimal decisioning on Adverse Select: Why is this compounded in the Digital Credit Era??

Context: Digitization is a paradigm shift in banking, which is becoming prevalent across all banking products, both on the asset and liability side of the balance sheet. It started with unsecured credit like installment loans, and credit cards, and then the pandemic accelerated its adoption exponentially.? This trend has now moved to the digitization of traditional credit products like auto loans and even mortgages.?

For the first time, consumers or businesses have the ability to easily compare offers and services from multiple institutions and FinTechs, for both deposit and credit products. The key shift is the CHOICES available to well-informed customers and low switching costs at the start or during the customer relationship.

Let’s take a scenario to understand the impact of adverse selection due to suboptimal decisioning:?

Scenario: A consumer researches online for an installment loan and if needed gets help from online intermediaries to learn about the various choices they have, applies simultaneously to multiple banks and Fintech lenders online, gets instantaneous offers, and chooses the best one in real-time.? i.e. accepts one out of 4 (as an example).??

From a bank's perspective, this is an ‘Acceptance Rate” of 25%.?

This concept is NEW for all banks/credit providers. Why? Banks have been fixated on optimizing approval rates for loans in the past with the implicit assumption that all approved loans will be accepted. That is no longer the case in the digital marketplace where competition is brutal to acquire the same customer.?

The most important question is, how does choice and acceptance rate impact the portfolio a bank holds? ?In all cases it increases the risk.?

If a bank won a customer - is it because no other bank thought they could make a profit from the loan at lower rates or is it because the winning bank had more knowledge and did a better assessment of risk-return trade-offs??

In all cases due to competition, the profit margin is squeezed and if the decision turns out to be wrong then the portfolio could turn unprofitable rapidly.?

In summary, the key differentiators of digital lending versus non-digital lending are:?

Note, that the impact on credit portfolios is subdued in low credit loss benign environments but gets exaggerated when credit losses climb. The potential concentration of defaults is even higher if a bank freezes in a recession given that refinancing or moving products has a very low digital hurdle and the most profitable segment will usually be sought after by the digital competition!?

The extent of the problem mainly depends on a bank's level of sophistication to use all relevant information and make the right trade-offs between risk and return.? In today's digital marketplace, every bank is competing with sophisticated and experienced players (both banks and FinTechs) and losses will be more volatile, and economic stress may make segments of the portfolio that are sub-optimal to spike.??

So, if you are a bank who has created a front-end digital experience to compete in digital credit and have not worked hard to transform the sophistication of back-end analytics, and decision management systems – you are particularly at a high risk in a downturn.

The below infographic visually shows the pre and in-digital scenarios where sub-optimal decisioning can lead to a higher concentration of bad customers in digital credit portfolios.?

The Impact of Numerator & Denominator Effect on Loan Loss Rates?

Let's use a simple example to illustrate the numerator and denominator effect in the context of loan portfolios.

Consider two banks, Bank A and Bank B, both of which have issued loans to their customers. We'll look at the performance of their loan portfolios over two different periods: Year 1 and Year 2.

Year 1:

Bank A: Total loans issued = 1,000. Number of bad loans (defaults) = 10. Loan loss rate: 1%

Bank B: Total loans issued = 500. Number of bad loans (defaults) = 5. Loan loss rate: 1%

Both banks have a bad loan ratio of 1% in Year 1 based on the number of loans issued.

Year 2:

Bank A: Total loans issued = 1,500. Number of bad loans (defaults) = 15. Loan loss rate: 1%

Bank B: Total loans issued = 400. Number of bad loans (defaults) = 10. Loan loss rate: 2.5%

In Year 2, you can see that Bank A's bad loan ratio remained the same at 1%, even though the number of bad loans increased. This is an example of the numerator effect. The numerator (the number of bad loans) increased, but because the denominator (the total loans) also increased, the loan loss rate remained constant.?

On the other hand, Bank B's loan loss rate increased from 1% in Year 1 to 2.5% in Year 2, a 2.5X increase. This is an example of the denominator effect. Even though the number of bad loans (the numerator) increased only slightly, the decrease in the total loan volume (the denominator) caused the loan loss rate to rise significantly.

In the numerator effect, banks take an offensive stance to enhance profitability through strategies like aggressive lending by expanding the risk box. In the denominator effect, they adopt a defensive approach, prioritizing customer retention, risk mitigation, and regulatory compliance to safeguard against negative outcomes and loss of good customers.

Understanding the numerator and denominator effect can help you analyze whether the deterioration is due to an actual increase in defaults or a change in the composition of the loan portfolio (the mix of good versus potentially bad customers).

Let's look deeper into how the denominator effect can influence the concentration of high-risk customers.

Concentration of High-Risk Customers in a Recessionary Environment?

If a bank decides to tighten its risk box during a recessionary period, it may choose to do so by reducing lending to riskier customers who are more likely to default on their loans.

This action can result in a concentration of bad customers within the remaining loan portfolio. This effect can be further compounded negatively if good customers are poached by competitors.?

For example, let's consider Bank C:

Year 1: Total loans issued = 1,000. Number of bad loans (defaults) = 20. Loan loss rate? = 2%.

Year 2: Total loans issued = 500. Number of bad loans (defaults) = 15. Loan loss rate? = 3%.

In Year 2, Bank C reduced its total loans issued. To do so, it might have scaled back lending to riskier customers, which caused the loan loss rate to increase (denominator effect). As the denominator shrinks further due to poaching by competitors of your good customers (say due to unfavorable pricing terms) , this can further spike your loss rates.?

To mitigate the impact of the denominator effect and retain good customers, banks need to implement improved or sophisticated decisioning systems. These systems can help identify and offer favorable terms to creditworthy borrowers, thus promoting customer retention while maintaining a healthy loan portfolio. Otherwise, they might find themselves losing out to the bigger institutions with best-in-class decisioning systems.

Impact of Information Asymmetry on Adverse Selection?

In this scenario, adverse selection typically arises when one party in a transaction possesses more comprehensive information regarding a product or service compared to the other party. In most cases, the party with greater knowledge is the seller or provider. However, in banking, adverse selection often arises when the borrower has more information about their own financial situation than the lender.??

Let's look into how this plays out.

When an individual applies for a personal loan at a bank, they may possess knowledge of their own financial difficulties, including past instances of late payments on loans. However, when completing the loan application, they choose not to disclose this history. Consequently, the bank lacks access to this negative financial background and proceeds to approve the loan based on the information provided by the borrower.

Note, that this information asymmetry is further exacerbated in a highly volatile environment where bureau and 3rd party credit proxy data is lagging behind the actual financial situation of the customer.?

As a result of this information gap, the bank approves the loan without a comprehensive understanding of the borrower's risk profile.? This can lead to adverse selection because the bank unknowingly takes on a riskier borrower than they intended per their risk appetite, which may result in higher default rates and financial losses. In essence, this information asymmetry allows the borrower to withhold critical risk-related information, resulting in potentially unfavorable outcomes for the lender.

To Recap :?

In the next blog, we will delve into how FIs can address these issues by building best-in-class decisioning capabilities.?