Sacrificing Microcredit for Unrealistic Goals

You are here

Jan 1, 2011
Adrian Gonzalez

In a recent New York Times opinion piece, Muhammad Yunus proposed government regulation to enforce an interest rate cap of 10 - 15 percentage points over cost of funds for microfinance institutions.  The intention of this cap is to prevent profit-maximizing businesses from extracting megaprofits from poor borrowers. Unfortunately, the proposal misses its goal in two ways.  First, operating costs, not profits, drive microcredit interest rates for the vast majority of MFIs.  Second, the proposed caps would affect mostly MFIs working with poorer clientele and would impact the majority of MFIs in most countries of the world.

This methodology for categorizing microcredit interest rates is not new, as it was previously presented in Yunus’ 2007 book, Creating a World Without Poverty.  The methodology defines three zones:

  • Green Zone: (Interest Rate – Cost of Funds) ≤ 10 percentage points.  According to Professor Yunus, these are “poverty-focused” microcredit programs (p. 69).
  • Yellow Zone: (Interest Rate – Cost of Funds) ≤ 15 percentage points.
  • Red Zone: (Interest Rate – Cost of Funds) > 15 percentage points.  Yunus labels institutions operating in this zone as “profit-maximizing” MFIs, adding that these programs are “commercial enterprises whose main objective appears to be earning large profits for shareholders or other investors.” Yunus even refers to this as the zone of the “moneylenders” and “loan sharks”[1].

Figure 1 shows the percent of borrowers in the ‘red zone’ by country. We can easily see that in most of both Africa and Latin America, and much of Asia, the majority of borrowers are served by MFIs in the ‘red zone’. Only a few countries, often with large state-run institutions, slip through this metric.

Figure 1: Percentage of Borrowers in the Red Zone by Country


Last year, MIX published a review of this methodology using data for 2008.  Given the return to these metrics in the recent editorial, we’ve updated these figures using data for 2009 from 1027 microfinance institutions (MFIs), covering 88 million borrowers. The updated figures confirm the conclusions from last year’s paper:

  • Three out of four microfinance institutions worldwide fall into the ‘red zone’.
  • The categorization can almost entirely be explained by operating expenses, rather than profits, since operating expenses represent 62 percent of all the expenses that need to be covered by the average yield[2] and 80 percent of expenses covered by the premium, as defined in the methodology.
  • Looking across the broad universe of MFIs, there is no evidence that institutions in any ‘zone’ are taking supernormal profits. Removing all profits from all MFIs would not substantively change the distribution of MFIs into green, yellow and red zones.
  • Most MFIs that have low average loan sizes (suggesting they reach poorer clients) are being mislabeled as in the ‘red zone’.
  • Not-for-profit NGO MFIs are more likely to be in the red zone than for-profit MFIs (like banks) and credit unions.

The updated metrics reinforce the point that the main challenge to reduce microcredit interest rates is to improve efficiency. Overall, megaprofits are an exception and not widespread among MFIs around the world.  As shown in Figure 2, operating expenses contribute close to two-thirds of the costs that interest rates covered in 2009, while profits represented less than 10 percent of all costs covered.  In other words, from every $100 collected by MFIs from borrowers as interest rates and fees in 2009, $63 dollars covered operating expenses, $21 dollars financial expenses, $7 dollars portfolio losses, $3 taxes, and only $7 were left for profits.

Figure 2: Spread and Yield Components, Profitable MFIs (ROA>0), 2009


Potential Effects on Microcredit Outreach of the Proposed Interest Rate Caps.

In his op-ed, Yunus argues that “The maximum interest rate should not exceed the cost of the fund … plus 15 percent of the fund … The ideal ‘spread’ between the cost of the fund and the lending rate should be close to 10 percent.”  In addition, he proposes that to “enforce such a cap, every country where microloans are made needs a microcredit regulatory authority.”

Given these specific recommendations, what could happen to microcredit outreach if local regulatory authorities apply these interest rate caps? For now we look at the actual distribution of borrowers in the three zones using 2009 data from 1027 MFIs covering 88 million borrowers. Figure 3 shows that close to two-thirds of all borrowers in 2009 were served by MFIs in the red zone, those potentially affected by an interest rate cap of 15 percentage points over the cost of funds.  Less than one quarter of all borrowers were served by MFIs in the green zone, those that will not be affected by an interest rate cap of 10 percentage points over the cost of funds.  Consistent with last year’s analysis, the regions most affected by the proposed interest rate caps would be those with the highest operating expenses:  Sub-Saharan Africa, Latin America and the Caribbean, and Middle East and North Africa.

Figure 3: Percentage of Borrowers by Zone in 2009










Perhaps surprisingly, as shown in Figure 4, the proposed interest rate caps will also affect NGO providers the most. These institutions provide the smallest loans (as indicated by average loan balance values shown in the chart), and thus likely serve poorer borrowers than other types of institutions.

Figure 4: Percentage of Borrowers by Zone and Legal Status in 2009


A better way to look at the analysis is to focus on the total number of borrowers.  Of the 88 million borrowers served in 2009, only 19 million are served by MFIs that would not be affected by a (hypothetical) 10 percent cap on spreads, and only 14 million more (or 33 million total) are served by MFIs in the green and yellow zones. 

Figure 5: Simulated Impact of the Caps on Number of Borrowers by Region in 2009


The database used for this analysis is available here, and we invite readers to simulate the impact of the proposed interest rate caps under alternative scenarios.[3]


Finding ways to reduce costs for poor borrowers is among the most important goals for microcredit.  However, sometimes we focus too much attention on prices and forget the other factors that influence how and why the poor decide to borrow: quality of service, appropriate product terms and loan amounts, speed of delivery and access to future loans.

However, the main reason microcredit interest rates are higher than those of other financial institutions is the high operating cost necessary to deliver small loans - paying administrative and staff expenses.  Making 1,000 loans of $100 each requires far greater staff expenses than making a single loan of $100,000.  Because of this well-known relationship between loan size and cost, the MFIs most likely to be affected by the proposed interest rate caps are those making the smallest loans, and most likely, reaching the poorest microborrowers.

Some MFIs do obtain very high profits from their microcredit operations; however, this is not the main reason why interest rates are higher in microfinance in comparison with other credit markets.  The previous analysis shows that reducing all profits to zero for all MFIs still leaves 63 percent of borrowers worldwide in the ‘red zone.’  From the point of view of benchmarking, a global evaluation of interest rates needs to first take into account differences in operating expenses worldwide.

This article is also available in Spanish and French.


[1] Yunus, Muhammad (2007), Creating a World Without Poverty, Public Affairs, New York, p69.

[2] Throughout the article, we use yields as a proxy for interest rates. Yields reflect all revenue from interest and fees actually received by the MFI over the course of the year for their loan products. While yields and interest rates will differ when borrowers don’t make payments or if fees or other charges are not fully disclosed, they should closely track each other.

[3] A future update of this analysis will include alternative scenarios like assuming a 3% target for ROA.