## Skinning the APR Cat

How important are compulsory deposits, insurance, etc. to calculating Annual Percentage Rates (APRs)? I played around some more with the MFTransparency dataset on India I wrote my last post on to answer this question, and it seems the answer is, “it depends”.

The Pricing Data Report looks at APR including deposits and APR including insurance, but in addition to this, the web portal offers APR figures that include neither, and that include both. It then becomes a trivial exercise to figure out how much compulsory deposits and insurance contribute to APR.

#### Intuitively Speaking …

The graph below is shows what happens when you calculate APR by including neither of, one of or both of the two constituents – deposits and insurance:

The red and blue columnar bars were presented before; the grey one is kind of a baseline without neither deposits or insurance, and the green one includes both. The error bars are the maximum and minimum.

By seeing how the red and blue bars change compared to the grey, we get an intuitive sense of what is going on. For co-ops, there is a small change of about 1% from the base 23% when you include insurance, but when you include deposits, it shoots up to 35%. On the other hand, for for-profit public MFIs, there is no difference from the base value of ~30% when you include deposits, but on including insurance, this moves up about 3%.

#### More Formally …

Fortunately, the MFT site allows you to download all the data and come up with precise numbers for what the differentials with respect to deposits and insurance. The graph above is built off of columns A, B, C and D from the table below. Columns E and F give us the differentials.

This tells us that:

- Co-ops charge an additional 1000+ basis points on top of interest and fees by taking
**compulsory savings and deposits** - For profit public MFIs charge an additional 250+ basis points on top of interest and fees by charging for
**insurance** - Other types of MFIs charge between ~100 and ~250 additional basis points on top of interest and fees through insurance and deposits

How’s that for “it depends”?

This also goes to show why proper pricing is so important – if one were to look at the top line numbers reported for interest rates, one would very easily conclude that Co-ops were about 10% cheaper than other MFIs as a source of funds.

By the way, the “Unexplained” in the last column is essentially the difference between the kitchen-sink APR as reported by MFT, and the kitchen-sink APR that we get when we reconstruct it from its constituent parts. That all the numbers are less than 1% is a good sign – we’re not missing some big chunk of data when we do the differentials.

#### Market Leveling

While there is a fair degree of divergence on what constitutes APR, it turns out that the averages for each type of organization is within a narrow band of **32.3% and 36.7%**. Not surprising, since the Indian microfinance sector has many players and is increasingly saturated, leading to a leveling of the marketplace when it comes what a financial institution can charge a for a fairly similar service to a fairly similar market segment.

It would be interesting to see how the reported interest rates diverge from this APR, and hopefully that will be the topic of the next post.

#### Skin Your Own Cat

I’ll briefly outline how to get the data I used to pull together the tables and graphs above, since it’s a little involved:

- Go to: http://www.mftransparency.org/data/countries/in/data/
- Under
**Filter Graph Results**on the right, select the kind of**Calculation Method**you want to use (there are four types – stick to one) - Also under Filter Graph Results on the right, select the kind of
**Institution Type**you want to use (there are six types) - Make sure the
**Loan Size**and**Number of Clients**boxes are empty (the site fills them up with some defaults – this may filter your results. Also, do not press “All” button under the graph – it reverts to some kind of default …) - Hit
**Filter Graph** - From the page that comes up, select the table only, and paste it into an Excel worksheet
**As Text** - Go back to step 3, choose another kind of institution and follow steps to #6, pasting in a new worksheet
- Once all institution types are done, go back to step 2, choose another calculation method, then go through #3 to #7
- Once all calculation methods are done, throw in your formulae in Excel

Yes, you will have to pull in 24 feeds from the site to get all the data, so take care as you consolidate data. Wouldn’t it be nice if you could have one extra column for institution type, and then a column each for each of the types of APR calculations.. Probably wouldn’t fit on the website, but that being available even as a CSV would have made life so much easier.

I should also note that the numbers I got from the raw data are slightly different than the average, maximum and minimum values that are in the report, but I couldn’t tell you why.