Measuring Microcredit Delinquency: Ratios Can Be Harmful to Your Health
This paper explores a problem that haunts many microfinance institutions (MFIs), that of accurately expressing the ratios of loan delinquency, i.e. the ways of measuring the health of an MFI in terms of loans repaid to loans defaulted on or delayed. The author is writing primarily for practitioners but his clear and even humorous style make this paper broadly useful for non-specialist MFI staff as well.
He suggests that any mention of a delinquency ratio should include a precise description of the ratio’s numerator and denominator – otherwise the ratio cannot be inter¬preted meaningfully and may well suggest an un¬duly optimistic impression of portfolio quality. MFIs should track multiple delinquency indi¬cators, because no single indicator answers all the relevant questions. An MFI’s outstanding portfolio tends to be roughly one half of the original disbursed amount of its loans. Collection rates, which divide amounts paid by amounts falling due during some period, are useful indicators but are subject to drastic misinterpreta¬tion: an MFI can have a 97 percent collection rate and still be losing a third of its portfolio every year. Hundreds of real MFIs are deceived by high-sounding collection rates into thinking that their portfolios are solid.
The most useful collection rate for day-to-day port¬folio management is often an on-time collection rate that tracks success in collecting payments when they first fall due, supplemented by a clean-up report that tracks collection of late payments. MFIs should avoid using the Asian collection rate, which includes past-due amounts from prior peri¬ods in the denominator of the ratio. Prepayments and late payments can create fluctua¬tions that limit the usefulness of collection rates other than the on-time collection rate for measuring per¬formance over a short period. Frequent renegotiation – rescheduling or refinanc¬ing – of problem loans makes it hard for an MFI to track and measure its repayment risk. Renegotiated loans should always be flagged and seg¬regated from normal loans in a delinquency report.
MFIs should usually not use arrears rates, which divide the amount of late payments by some mea¬sure of total portfolio or loan volume, because such measures tend to understate risk. Almost all MFIs should follow international bank¬ing standards by tracking and reporting portfolio at risk (PAR): this measure analyzes outstanding bal¬ances of late loans as a percentage of total outstand¬ing portfolio. MFIs with weak information systems may wish to use a simplified PAR based on the number of loan accounts rather than the amount of account balances. When tracking PAR it is useful to age the portfolio: loans are broken down by degree of lateness, using time intervals that correspond to the MFI’s payment period and loan management process. Any PAR re¬port should specify the time interval(s) being used.
PAR information, supplemented by analysis of his¬torical portfolio performance, can generate a so¬phisticated estimate of probable loan losses. In judging an MFI’s portfolio quality, PAR infor¬mation needs to be interpreted in light of the MFI’s write-off policy and experience. PAR and arrears rates understate risk when a portfolio is growing very rapidly, or when there are long grace pe¬riods, unless loans on which no payment has yet fallen due are excluded from the denominator of the ratio. To the extent possible MFIs should disaggregate their delinquency measurement and reporting by loan product, region, branch, loan officer, and per¬haps client characteristics.
This paper, though densely written, is lively and gives a common sense series of explanations. The author’s arguments are logical and concise, and the overall effect is of a useful and professional piece which combines theory with practice, to the benefit of both. An annex summarises formulae for converting collection rates into annual loan loss rates.