Using time series of Northern Arc portfolio data

Debt markets in India have witnessed a series of event-based shocks in recent years, and borrowers at the bottom of the pyramid are particularly vulnerable to such shocks. Northern Arc, an India-based impact debt platform, has a large repository of loan data spanning over a decade, with around 200 live pools of securitizations, resulting in millions of repayment observations every month. This is anchored on Nimbus, an in-house technology platform. The data enable a dynamic understanding of credit behavior over time across geographies, originators, loan sizes, loan cycles, demographics, and credit bureau scores. This blog summarizes our findings on the aggregate performance of microfinance loans in India after events that have impacted the repayment capacity of borrowers and assesses the time taken to return to normalcy.

Base data and assumptions

We consider three events — demonetization, the Kerala floods, and cyclone Fani in Odisha. Demonetization was a nationwide event impacting all microfinance institutions. The Kerala floods and cyclone Fani were localized events severely affecting specific districts.1 The following key metrics were used to assess recovery patterns:

  • Periodic collection efficiency: the percentage of demand/dues collected in a month. Although the mean collection efficiency provides a good point-in-time assessment of recovery, the volatility of collections provides insight into recovery behavior over time and is a more nuanced reflection of what happens on the ground after a shock event.
  • Portfolio at risk trends (PAR 0, PAR 30 or PAR 90): the loan portfolio outstanding for 0, 30 or 90 days, measured as a proportion of the overall portfolio.
  • Recovery rates: the percentages of borrowers with at least one installment paid in a month, grouped into days-past-due (dpd) buckets to differentiate the behaviors of regular and overdue customers.
Psssssst :  You asked: How do bank construction loans work?

Performance of microfinance institution loans post–macroeconomic shocks

For the three events, a trend analysis was conducted of the key metrics to derive insights into post-event portfolio repayment behavior. The loans in the portfolio were cherry picked using proprietary algorithms, so there could be some selection bias. However, given the size and granularity of the portfolio, long time series, and high level of diversification across originators, states, and districts, we consider it a fair representation of the behavior of the sector in India.

Event 1: Demonetization, November 2016

We considered loan data for 3.7 million customers in more than 400 districts in India (the country has 700 plus districts). The key findings are as follows:

Source :

Back to top button