Debt Restructuring in the Shadow of Moral Hazard: A Field Experiment with Small Businesses

The experiment focuses on a set of small firms (“borrowers”) that have borrowed from the lender and are currently in arrears. The small firms in our setting are minibus taxi owners who borrow from the lender using a debt contract, with the loan secured by the minibus. These business owners generate revenue by driving the minibus and are responsible for making fixed monthly payments, determined at loan origination, to the lender. The company decided to run the experiment on 3,186 borrowers who were delinquent at the time of the experiment.

We worked with the lender to randomize with stratification on eight groups. These eight groups are created by dividing the sample at the median across three variables: (a) the size of the effective payment reduction (%) the firm would receive if in the payment reduction group; (b) the size of the effective principal reduction (%) the firm would receive if in the principal reduction group; (c) the debt to income ratio. Our randomization aims to obtain three equally split samples across the control and two treatment conditions.

The lender will manage the communication with the customers. The experiment is structured in a way that customers are automatically enrolled in the new program, but they are allowed to opt out of this option at no cost. Customers will first receive the communication about their enrollment into the offer via an SMS. The message will contain a link to the letter detailing the new contractual terms. If the borrower decides to not take-up the offer, then she needs to opt-out within five business days. The company will follow up with phone calls in case the SMS is undeliverable. After the opt-out period, the lender will connect with borrowers explaining to them the terms of their new contract using follow-up phone calls.

2023-11-13

2024-03-31

1. Index of Loan performance
2. Index of Labor Supply

See the description below and more details are available in the Pre-Analysis Plan

1. Index of Loan performance: we will examine an index of loan performance, which combines information from three of the lender’s key metrics to assess performance. To be precise, the index is constructed as the (equally weighted) average of three metrics defined below, after they are standardized to have zero mean and standard deviation one. The components are:

i. Indicator variable of whether the account is not current.
Definition: An account is not current if Total arrears < 100 ZAR in the month, with unmet obligations > 30 days late considered arrears.

ii. Total Arrears Amount.
Definition: Total Arrears Amount reported by the lender.

iii. Total Arrears Amount, Scaled
Definition: Total Arrears Amount divided by Monthly Owed Installment.

Monthly loan performance: measured as a snapshot taken each monthly reporting date.
12-month loan performance: measured as a snapshot taken on the 12th-month reporting date.

Dealing with unbalanced panel of loan performance. If a loan leaves the books, we will impute that month and subsequent months as follows:
• Current and zero arrears, for a loan paid in full
• Not current and the last observed arrears amount, for a loan not repaid in full

2. Index of Labor Supply: we examine an index of labor supply, which combines information from four key metrics constructed using the GPS data on borrowers’ driving. As before, the index is constructed as the (equally weighted) average of the metrics defined below, after they are standardized to have zero mean and standard deviation one. The components of the index are:

i. Distance driven (in kms).
Definition: Sum of the total distance covered by the vehicle.

ii. Number of hours driven.
Definition: Sum of the total time the vehicle was active.

iii. Number of days worked.
Definition: Sum of total number of days worked in the month. A working day is defined as the day in which non-zero trips were made.

iv. Number of hours on the job:
Definition: The total duration between vehicle first trip of the day and the end of the last trip of a day.

Monthly labor supply measures: Total for a given month.
12-month labor supply measures: Summed across all months.

Dealing with unbalanced panel of labor supply measures: We will only have data for drivers while their loan is on the books. We do not plan to impute missing values.

Details in Pre-Analysis Plan

Details in Pre-Analysis Plan

Each borrower is randomly assigned to one of three experiment arms. Every borrower, including the ones in the control arm, receives a loan transformation. Borrowers in the control arm are then used as a benchmark to assess the size of the payment and principal reduction. The three offers received are the following:

(1) Control group: firms in the control group are presented with a new loan, featuring the same principal amount as their original outstanding loan, and with a monthly payment resembling that of the initial loan. Accepting this offer can still be advantageous for borrowers, as it helps them avoid accumulating late fees (in the form of interest on unpaid interest) from their delinquent loan. This method serves the purpose of bringing all borrowers in the sample to a "good standing" and also facilitates a more straightforward evaluation of the benefits associated with each of the other intervention methods. The modification allows the borrowers to avoid a negative flag on their credit report.

(2) Principal write-off group: the restructured loan for firms in this treatment arm has the same monthly payment as their original loan. However, these borrowers receive on average a 17% reduction in the total size of the loan they owe to the lender at the time of the experiment. As a result, the new maturity of the restructured loan is lower than the control group.

(3) Payment reduction group: the restructured loan for firms in this treatment arm matches the total amount paid in the control group. However, these firms receive on average a 13% reduction in monthly payments. This reduction allows borrowers to effectively extend the maturity of their loan, accommodating the smaller monthly payments.

Randomization done in office by a computer

The unit of randomization is the loan account (borrower of the minibus taxi). The firms will be randomized into one of the three groups, i.e., principal write-off, payment reduction, and the control group.

No

0

3,186 borrowers

1,061 borrowers (payment reduction treatment); 1,062 borrowers (principal write-off treatment); 1,063 borrowers (control)

Pre Analysis Plan