• We have serious questions about the extent to which “best practice” matches with “in practice.”
• PAR vs. what we’ve been calling “real repayment rate” seems like a different conversation from “one day late” vs. “several days late.” An equivalent “real repayment rate” could be constructed based on the % of loans paid within 30 days of coming due, for example.
• What is important about the contrast between measures, to us, is that while PAR may be of use in understanding an MFI’s financial exposure and risk, the “real repayment rate” seems closer to what people have in mind when they promote repayment rates as evidence that microfinance is “working.”

More at our recent post on the subject.

Best practice is to treat all loans that have been rescheduled as PAR.

In addition, it’s unclear to me why paying back loans on time is of critical interest. If a borrower is a day late, does that mean they aren’t doing so well? Not really. That’s why PAR, aged into 30 day buckets, is used to assess MFIs’ financial performance but also the ability of borrowers to pay back within reasonable amounts of time. At many MFIs, borrowers go delinquent 1-30 days but are able to catch up quickly because their cash flow doesn’t map to loan products’ repayment schedules perfectly.

I think this is more of a case of choosing to make assumptions about the meaning of rates instead of actual problems with the rates themselves.

Ben

The main thrust of this post, though, is that “real repayment rate” seems the most straightforward and intuitive interpretation of a term like “repayment rate” or “repayment ratio” – yet this term turns out to mean something extremely different, not just technically but conceptually. When I hear “99% repayment ratio,” I think “Loans are almost always getting paid back on time,” yet that turns out not to be necessarily the case at all.

I recognize that taken together, the write-off ratio and portfolio-at-risk figures can give a meaningful picture, although there is one more piece of important information - % loans renegotiated, don’t always seems to be counted either as “at risk” or as “written off” and which we have had trouble getting from applicants.

Ben’s comment describes well how PAR ratios can provide meaningful information at a single point in time on the potential for future losses. Write-offs tell you about realized losses over a period of time from the past. Both tell you something about borrower behavior, and about the institution’s own practices. It does not take much for a donor to analyze both ratios jointly should they so choose. Reporting requirements are not cost-free and need careful consideration to meet the needs of multiple stakeholders efficiently.

A high repayment rate is often taken as an indicator that clients are participating in a healthy and successful way (not just that the institution is in solid financial shape). I’d guess that most who cite high repayment rates as evidence that “microfinance is working” have this meaning in mind (it wouldn’t make much sense for them to claim that “microfinance is working” based on low PARs, which in isolation simply indicate that a given institution’s position is unlikely to worsen significantly in the short term).

On time repayment rate is easy to understand but conceptually offers little to investors or donors. Let’s say all clients make payments 1 day late. In this case, on time repayment would be 0% but writeoffs/default would be 0% as well. Portfolio at Risk (PAR) is used to capture the total portfolio that may default in the future (indicated by some delinquency already on the loan). This allows for sensitivity analysis on the hit to the balance sheet and allows donors and investors to age the arrears and portfolio at risk to see how bad the delinquency is. Depending on the underlying characteristics of the loan (type, repayment frequency, etc.), different aging is acceptable.

Perhaps all MFIs are not “transparent” about this metric because no one uses it and some fairly intelligent people have created appropriate indicators to measure the core underlying issue: delinquency.