State of Debt Collection in Fintech in India

Why Paytm and Lendingkart have performed well whereas Zestmoney failed? The contrasting performance can be attributed to the collection of debt on these platforms. The below table presents a better understanding:
 

Increasing non-performing assets (NPA) in the fintech industry is a concerning trend. According to the Reserve Bank of India (RBI), the NPA ratio for fintech lenders has been steadily rising. In 2021, the NPA ratio for fintech lenders reached an average of 8-10%, while traditional banks maintained an average NPA ratio of around 4-5% (source: RBI Annual Report 2020-21).

In consumer lending, which includes personal loans and buy-now-pay-later schemes, the NPA ratio can range from 10-15% due to the high volume of individual borrowers and diverse risk profiles.

In SME lending, fintech platforms that provide loans to small and medium-sized enterprises, the NPA ratio can vary from 8-12%. These lenders face challenges in assessing the creditworthiness of businesses with limited financial history or inconsistent cash flows.

P2P lending, where individuals lend directly to borrowers through online platforms, has also witnessed an increase in NPAs. The NPA ratio in P2P lending platforms can range from 12-18%, highlighting the higher default risk associated with this form of lending.

Factors Leading to High NPA

• Surge of BNPL (buy-now-pay-later) loans being unsecured increases the default risk for lenders
• Economic slowdown has led to an increase in unemployment and a decrease in disposable income
• Fintech industry has less regulation than traditional banks, giving more lending freedom to fintech companies, including high-risk borrowers

RBI Guidelines on Debt Collection in the Fintech Industry

• If harassed by a debt collector, borrowers can file a complaint with the RBI
• Fintech lenders must disclose the details of authorized recovery agents to borrowers during loan sanctioning and when there is a change in the agent
• Lenders should provide necessary guidance to recovery agents to ensure responsible conduct and compliance with instructions
• Debt collectors should not contact borrowers' family members or friends

Why Is Collection More Important than Ever?

With a compound annual growth rate (CAGR) of 39.5%, the digital lending market in India has experienced significant expansion. In 2022, the market was valued at approximately INR 19.8 lac crore, and it is projected to reach INR 25.6 lac crore by 2023, as per a report from Experian. With rising digital lending, the collection has become more important in the digital journey without any manual touchpoint.

Collection refers to the process of recovering outstanding payments from borrowers. This can be done through a variety of methods, including:

• Phone calls: Debt collectors contact borrowers to remind them of debts and negotiate repayment plans
• Letters: Debt collectors send reminders and warn of legal action if debts are not repaid
• Visits: Debt collectors may personally visit borrowers to collect debts
• Legal Notice: Digital lending platforms send legal notices to defaulting borrowers
• Litigation Support: Platforms provide support to lenders for legal action against defaulters

The Catch in Collection Management

According to a report by CreditVidya, the average ticket size of digital loans in India is around INR 25,000 to INR 50,000. On the other hand, the cost of the collection can vary depending on various factors such as manpower, technology, infrastructure and legal procedures. Reports suggest that the cost of collection for digital loans can range from 15% to 30% of the outstanding loan amount. This means that the earning on these loans in processing fees or interest may not be enough to recover the collection cost, let alone the profitability.

Considering these numbers, the challenge lies in ensuring that the cost of the collection does not exceed the potential recovery amount. For smaller loans, the cost of the collection can become a significant percentage of the loan value, making it economically unviable to employ a collection agency.

What Fintechs Can Do More to Increase the Collection?

Propensity-to-Pay Scoring of Clients

Propensity-to-pay scoring assigns a score to each client, indicating the likelihood that they will pay their bill. This score can combine like-behavior customers, prioritize collections efforts and tailor payment plans to each client's needs.

Propensity-to-pay scoring models are typically built using a variety of data points, including:

• Demographics: Age, gender, income and household size
• Payment history: Past payment performance, including late payments and defaults
• Account information: Account balance, outstanding charges and interest rate
• External data: Information from third-party sources, such as credit reports and social media activity

Customer Segmentation

It involves a lot of analysis of different types of customer data. It is also essential to separate the noise from the data; else it can result in analysis paralysis. Below are specific categories and data types that can help achieve customer segmentation:

• Economic and monetary data: Expense vs Income, Net expense balance, Credit card utilization, Occupation and its nature
• Behavioural and social data: Regular vs irregular bill payment, Travel, fuel, food and shopping data, Social sites subscription and usage
• Credit propensity data: Credit score, History of late payments, Amount of debt, Current economic conditions
• Online behavior: Different marketplaces vs preferred modes of payment, Loyalty vs offer
• Digital savviness: Amount spent online vs offline, UPI usage, Number of financial apps installed

Customer Segments

Using AI and ML, the customer data can be used to create a grouping of customers basis similar behavior, expense patterns, repayment pattern, etc. Below is the customer segmentation:

• First-time defaulters: They may be new to the credit market or have recently experienced financial hardship
• Lazy customers: These customers are just lazy to pay or clear their dues at the time
• Self-cure customers: These customers are those in the collection who with no or minimal collection effort return to their current collection cycle
• Bad debt: These customers have used their complete credit line and are in and out of delinquency
• Point of no return customers: Those customers who are most likely to turn bankrupt and not going to pay anything despite any collection effort

Based on customer segments, the companies can take action accordingly to increase collection efficiency.

Use AI and ML Along With Offline Strategy to Optimise Debt Collection

Lending platforms can efficiently use AI and ML with their offline strategy to determine the optimized number of collection executives and optimized routes for collecting cash, especially in tier 2+ cities. AI and ML can optimize debt collection by automating tasks, predicting borrower behavior and personalizing communication.

• Automating tasks can free up collection agents to focus on more complex tasks, such as resolving customer issues.
• AI can be used to send reminders, track payments and manage disputes automatically
• Predicting borrower behavior can help collection agents identify borrowers most likely to default on their loans.
• Personalizing communication can help to improve the effectiveness of collection efforts

Collection Personalisation

Most AI applications with real-world business significance for debt collection today seem to be in personalizing communications to customers and identifying clusters of similar debtor profiles. Collection personalization is the process of tailoring debt collection efforts to the individual needs of each debtor. This can be done by considering various factors, such as the debtor's financial situation, their communication preferences like SMS, in-app notification, IVR, etc, their past payment history and their preferred time of receiving the communication.

Personalization can result in increased and quicker repayment rates as it improves the overall customer experience. Debtors are more likely to respond positively when receiving relevant and considerate communication that aligns with their circumstances.

Level of success is measured by assessing specific indicators, such as:

• Did the debtor open the email?
• Did they visit the landing page?
• Did the individual make any payment attempts in the past month?

Paytm: Rethinking Debt Collection

Paytm saw a 250% jump in the total loan amount from a year ago in the quarter ended March 2023. It distributed nearly 12 million loans worth over Rs 12,550 crore ($1.5 billion). This growth pushed the fintech's revenue up by 61% to Rs 7,990 crore (~$966 million) in the year ended March 2023.

One of the main reasons driving the profitability of Paytm in lending is its collection efficiency. In 2022, the company reported collecting over ₹10 trillion in payments.

What Paytm Is Doing Differently

Paytm's success in the post-FLDG era can be attributed to its proposal of collection services in its agreements with lending partners. By committing to achieve a certain level of collections on the loans it distributes, Paytm reassured its partners.

Without FLDGs, lenders have reduced fintech’s processing fees to mitigate risk, impacting intermediaries' margins. However, Paytm has preserved its margins by incorporating collection services.

The lenders now rely on Paytm to onboard borrowers and collect from them. Paytm was able to unlock this collection opportunity after investing in the collection platform CreditMate in 2017. It fully acquired the company in 2021. Through CreditMate, Paytm offers the entire gamut of collection activities, from using diallers to call people about late payments to app notifications to employing collection agencies to do door-to-door collections.

Paytm makes the collection proposition more compelling by structuring what it earns from this as a "collection-performance bonus." For instance, if a lender expects a 2% loss on a portfolio, Paytm tells them to account for only 1.5%. If Paytm recovers the other 50 basis points (bps), it gets a share. If Paytm doesn't stem the credit loss at 1.5%, not only does it lose out on the revenue share, but lenders don't have to pay even a part of the sourcing fee.

Using AI and ML

Here are some specific examples of how Paytm uses AI and ML in increasing collection:

• Predicting customer behavior: Paytm utilizes AI and ML to predict loan default likelihood. This enables targeted early intervention efforts, such as reminders and offers, for at-risk customers
• Personalizing collection messages: Paytm uses AI and ML to personalize collection messages for each customer. For example, Paytm uses AI to identify the best time of day to contact customers about their payments. Customers who are contacted at the right time are more likely to respond and make a payment
• Automating collection tasks: Paytm uses AI and ML to automate many of the tasks involved in the collection, such as sending reminders, tracking payments and managing disputes freeing up the collection agents to focus on resolving complex customer issues

Overcoming Debt Burdens

Debt or loan has become part of regular life to fulfill dreams. However, reckless or habitual borrowing can land an individual into a vicious cycle of the debt trap. It is essential to keep your financial obligations in check. But sometimes it can become challenging for an individual to maintain his debt. This is the time when a borrower should reach out to a debt resolution platform.

When to Reach Out to a Debt Resolution Platform

There is no golden rule here. It is up to borrowers when to seek the help of a debt resolution platform. But below are some points to help the borrower decide on reaching these platforms:

• Facing difficulties in repaying EMIs or equated monthly installments due to a decrease in income or financial challenges
• Attempts to manage debt have been exhausted
• Creditors are harassing and causing distress
• High-cost debts, such as credit cards or loans with steep interest rates, are becoming unmanageable

How Does Debt Resolution Work?

The debt resolution companies assist in debt settlement, consolidation and elimination. Debt settlement involves setting aside funds for loan repayment in a trust account and negotiating with creditors to reach mutually agreed-upon settlements.

Debt consolidation focuses on combining multiple high-cost loans into a single loan, offering reduced interest rates and extended repayment periods. On the other hand, taking help from a debt counselor can be helpful for habitual borrowers. Debt counselors play a crucial role in helping individuals achieve a debt-free life by providing educational resources, budgeting guidance and tools to reduce and eliminate debt

‍