Why use muscle? Big data can collect debt
The credit reporting institution’s survey shows 9% of industry participants expect investment in software to be a significant expense over the next year or two
image for illustrative purpose
The world's cheapest data prices have made smartphones ubiquitous
A 2021 remake of Get Shorty might not have much use for John Travolta's Miami mobster character. The debt collector role could go to a bot. You can test this hypothesis in a most unlikely place to roll out a new technology: the Indian countryside. The setting is perhaps not as odd as it seems, with about 5 per cent to 10 per cent of the country's farmers not repaying their tractor loans on time. The explanations for tardiness range from failed crops to medical emergencies and strategic defaults in anticipation of State-mandated debt waivers, a regular feature of the political economy.
But delinquency often stems from more mundane reasons: Borrowers forget their due dates, or fail to withdraw cash to pay the nonbank financiers who provide the bulk of loans for farm equipment purchases. Like in most emerging markets, these last-mile hurdles pose a frustratingly complex challenge to India's creditors. They also increase the overall risk premium for rural advances.
Of late, three things have changed. First, the world's cheapest data prices have made smartphones ubiquitous. Second, a strong push for financial inclusion has seen more than 400 million no-frills savings accounts opened in the last seven years. Finally, banks are now on a nationwide mobile payment network that is fast, convenient and supports apps like Google Pay and Walmart Inc.'s PhonePe. Google has even recommended the architecture to the US Federal Reserve.
Yet for all the help from technology, collections are still hard in villages due to barriers of language, education and entrenched cash use. Farmers simply don't know how to use the new digital tools. Or they may miss a deadline because of a temporary cash flow mismatch. When creditors respond by handing over borrowers to third-party recovery agents, scandals emerge. It's a universal problem. In Indonesia, the 2011 death of a cash-strapped small businessman in Jakarta after alleged harassment by collectors led to a two-year ban on Citigroup Inc. from acquiring new credit card customers in the country. Even when things don't go to such extremes, unpleasantness and soured relationships usually follow.
"The best way for a bank or a financial institution to lose a customer is to give the account to a collection agency," says Sumeet Srivastava, the chief executive of a five-year-old startup that aims to boost collections without human contact. "You can do collections without collectors."
Getting paid on time
Before setting up the Mumbai-based Spocto Solutions, Srivastava put in stints at General Electric Co. and Monsanto Co., the seeds and agrichemicals giant later acquired by Bayer AG. His Kisan Pay - Hindi for Farmer Pay - is not an app, but an automated voice call, which lets farmers choose when they can repay loans and how. The selections trigger text messages with online payment links. The bot stays on the call to help borrowers navigate the unfamiliar world of online money transfers.
It sounds fairly straightforward, until you consider that in a large, multilingual country like India, with more than 146 million operational land holdings, the service has to be offered in 100-plus dialects. Spocto is working on 1.2 trillion rupees ($16 billion) of loan portfolios spread across some of India's largest banks and nonbank financiers. Its recovery rate in cases where borrowers have missed one instalment is 85 per cent, compared with 70 per cent for traditional channels. The cost savings for creditors are in the range of 30 per cent to 40 per cent, Srivastava says.
The value to lenders comes from the behavioural clues Spocto picks up in the process of nudging customers to pay on time. By training his algorithms to mine the data, Srivastava is able to predict which customers will most likely resume paying after missing a couple of instalments, and who will probably default. Turning over a much smaller set of problem loans to collection agencies means good customers don't get put off by strong-arm tactics. For creditors, bad loan ratios and loss provisions fall; profitability improves.
Still low-tech
From remittances and working capital financing to micro insurance and buy now, pay later, fintech is disrupting conventional finance. However, collection technology has received far less attention and investment. Expect that to change as big-data analysis of repayment kinks and quirks complements the yes/no decisions of underwriting algorithms. The two can even combine - to screen good customers and retain them.
Ten years ago, there were almost 10,000 collection firms in the US Now, their number has dwindled below 7,000. More than 70 per cent have fewer than five employees, and it's these small agencies that are consolidating. "While letters and phone calls are still the most common across the industry, text messages, artificial intelligence-informed chatbots and other automated tools have been adopted by larger companies," TransUnion said in its 2020 collections report. The credit reporting institution's survey shows 9 per cent of industry participants expect investment in software to be a significant expense over the next year or two.
That's what is attracting the husband-wife team of Sumeet and Puja Srivastava to test the waters with credit card and other unsecured debt on the East Coast of the US Back in India, Srivastava is considering putting his money where his mouth is. He wants to finance overdue installments where Spocto's analysis shows delinquency is likely to be temporary. Taking on balance-sheet risk may be a leap for a bootstrapped startup, so Srivastava might partner with a bank. But it tells you where debt collection is heading - toward less muscle and more intelligence. Of the artificial kind. (Bloomberg)