Drop out rates
A drop out rate or a retention rate measures the percentage of clients that leave a program in a specified period of time. Drop out rates are measured using different formulas and definitions by different microfinance institutions. In order to understand an institution's drop out rate and how it compares to other institutions' rates, it is necessary to know:
- How a "drop out" is defined. For example, a microfinance institution may define a drop out as a client who does not immediately take out another loan after repaying the previous one or as a client who does not take out another loan within a certain number of months. Other institutions may consider a client as retained if s/he continues to access services such as savings or insurance even if s/he does not continue to borrow. Institutions may also have different ways of accounting for clients who are not issued a new loan due to default.
- What formula is used to calculate the rate. From what we've seen, there is not a standardized formula for the drop out rate. Different specificiations of what quantities are used in the numerator and denominator of the formula can result in different quoted rates for the same underlying rate of drop out. Examples:
- The Small Enterprise Foundation defines its drop out rate as the number of clients who completed a loan in the six months prior to a given date and did not take out a subsequent loan in that period or within one month following the end of the six month period divided by the total number of clients who completed a loan in the six months prior to the end of the period.
- Chamroeun defines its drop out rate in terms of loans, rather than clients, and uses a time period of 12 months. It calculates this rate as the "number of loans taken out by clients who previously had a loan divided by the number of loans closed in the last 12 months." Since this rate is measured in terms of loans (of which a client may have more than one per year), it will result in a lower drop out rate than if clients were the unit of analysis.
- Another formula we've seen used is: 1 - [(Clients at the end of the period) / (Clients at the start) + (Clients added during the period)]. Clients whose loans do not become due during the period, and thus, do not reach a "decision point" of whether to drop out or not, are included in both the numerator and denominator of the formula, deflating the drop out rate especially over short periods.
We specify the formula used by each charity when reporting the drop out rate.
For more, see "The Challenges of Measuring Client Retention" (PDF), a report on this issue by the SEEP Network, and "Estimating Client Exit Rate" (PDF), a note by the microfinance rating organization M-CRIL.
From the blog:
We discuss microfinance interest rates in-depth and provide an example on our blog. In short, to fully understand the cost of a loan to a borrower, it is necessary to know:
- The stated interest rate and the length of time for which this is calculated.
- The interest rate method. "Flat" means that interest is calculated on the full amount of the loan, while "declining balance" means that interest is calculated on the outstanding balance of the loan. The true cost of a flat rate will always be at least as high as the same rate quoted as declining balance.
- Frequency and size of repayments, length of loan, and whether there is a grace period (i.e. time after loan is disbursed when no payments are due).
- Size and timing of fees. Fees may be a set amount for each loan or may be based on the size of the loan and they may be charged up-front or over the course of the loan.
- Savings requirements. Some institutions require clients to set aside part of the loan as "compulsory savings" that serve as collateral to protect the institution somewhat from losses due to default. Institutions vary on what percentage of the loan they require in forced savings, whether and how much interest they pay on savings, and when they allow clients to withdrawn this money.
We present interest rates in three forms: monthly rate, Annual Percentage Rate (APR), and Effective Interest Rate (EIR). The APR is equal to 12 times the monthly interest rate, which does not incorporate interest compounding. The EIR fully incorporates "compounding," whose relevance to microloans is debatable. Using information provided to us by microfinance institutions, we estimate the cash flow schedule for a loan and calculate the internal rate of return (IRR) per period (i.e. the time between payments), incorporating all fees and savings requirements. APR is calculated as the IRR multiplied by the number of periods per year. EIR is calculated as (1+IRR)^(# of periods per year) - 1.
We have found the website MFTransparency (http://www.mftransparency.org/ ) helpful for understanding microfinance interest rates and comparing rates across institutions. See also, a report on "Microcredit Interest Rates" (PDF) from CGAP.
From the blog:
There are a number of ways that a microfinance institution may calculate and report its repayment history. A CGAP paper, "Measuring Microcredit Delinquency: Ratios Can Be Harmful to Your Health" (PDF) identifies the following measures:
- Collection rates. "measure amounts actually paid against amounts that have fallen due." There are a number of variations of the collection rate, including:
- On-time collection rate. "tracks success in collecting payments when they first become due."
- Asian collection rate. "divides all payments received during a period by all amounts due during that period, including past-due amounts from prior periods."
- Current collection rate. "divides cash received during a period by cash that first fell due during that same period."
- Cumulative collection rate "covers payments received and payments due over the entire life of the MFI."
- Arrears rates. "measure overdue amounts against total loan amounts."
- Portfolio at risk (PAR) rates. "measure the outstanding balance of loans that are not being paid on time against the outstanding balance of total loans."
For the purpose of understanding whether clients have consistently repaid loans in the past, we prefer an on-time, current, or cumulative collection rate as these have straightforward interpretations, and, when calculated correctly, are not susceptible to an institution's decision to write-off or reschedule loans.
In cases where an organization is able to provide details on value of loans disbursed, value of loans written off (i.e. loans that have been deemed uncollectable and been removed from the loan portfolio), rescheduled (i.e. loans whose terms, especially payment size and due date, have been changed), and in arrears (i.e. loans with overdue payments) and the length of the longest loan term, we have used this data to calculate a conservative (or lower-bound) collection rate.1
From the blog:
Standard of living
We seek evidence that microfinance institutions (MFIs) are generally serving people who have low incomes. Information on clients' standard of living may come from surveys of entering clients or surveys of a sample of all clients. We look for evidence that surveyed clients are representative of all new or all clients. To ensure representativeness, clients generally must be selected randomly from the full population under study or all (or nearly all) clients within a population under study must be surveyed. We also look for evidence that surveys were designed to yield credible information: that there were not incentives for clients to lie (perhaps because answers were linked to loan approval) and that the questions asked were easy for clients to answer. We prefer questions such as "do you own a TV?" to ones such as "what was your income last year?"
We do not have a single standard for what qualifies as poor. We use subjective judgements in incorporating this information into our overall rating of a microfinance organization, while presenting as full a picture as possible in our reviews so that donors can come to their own conclusions.
Some MFIs use 'poverty scorecards' to estimate their clients' poverty levels. Such scorecards use clients answers to questions such as "do you have electricity," "how many meals per day do you eat," and "what is your roof made of" to estimate clients' poverty level. The Grameen Foundation's Progress Out of Poverty Index2 is an example of this type of measurement system. We find such analyses useful for rough comparisons across organizations, but prefer to see the disaggregated responses to each of the questions along with the aggregated poverty level data, as we believe that this provides a clearer (though necessarily incomplete) picture of what life is like for an MFI's clients.
In cases where MFIs provide information on client incomes and these incomes are not adjusted for purchasing power parity (PPP), we perform this adjustment. We use price level data from the World Bank's International Comparison Program.3
Lower bound collection rate during time t = (D – W – R - A) / D where
- D is the total value of loans disbursed during time period t
- W is the value of all written off loans during t + m where m is the length of the longest loan term after t. If t is 2008 and 2009 and the longest loan term is 6 months, then t + m is January 1, 2008 to June 30, 2010.
- R is the stock of rescheduled loans at the end of t + m (i.e. on June 30, 2010 in our example).
- A is the stock of loans in arrears at the end of t + m (i.e. on June 30, 2010 in our example).
For an example of how we've done this for an MFI, see our review of the Small Enterprise Foundation .
For an example, see our review of the Microloan Foundation
Category: Payday loans