I just got the annual report from Compartamos in Mexico.  Putting data there together with some MIX data, here are their interest rates:

 Interest income (almost all from portfolio) divided by average gross loan portfolio:

    2005   88%
 2006   82
 2007   85
 2008   71

I asked Compartamos management whether the steep decline in 2008  was due to competition. They told me that there is considerable competition in the Mexican market, but that the competition is not driving prices much, yet at least.  Last year’s substantial reduction in Compartamos’ interest income was due to implementation of a longer-term pricing policy, including a feature that gives lower rates to groups that repay well and have more members, both of which lower Compartamos’ costs.

Compartamos says that their own price drop hasn’t driven down competitors’ prices as much as they expected.  Note that this is in line with classical competition theory as described in a Focus Note David Porteous wrote for CGAP  a couple of years ago:  in the early phases, as a market is becoming saturated, competition first focuses on product features rather than price.

I see that their admin costs as % of portfolio dropped by about 5% last year. Analysis by Adrian Gonzalez of the MIX shows that the learning curve is probably the biggest factor pushing down costs:  mere age drives costs down, even after controlling for things like loan size, portfolio volume or numbers, etc.  Because Adrian had no statistical variable to represent competition, he could not control for competition in his regression analysis, but most microfinance markets are not competitive yet, so the strong cost reductions we’re seeing in the MIX data are probably mainly learning curve rather than competition.

Because Compartamos’ costs didn’t drop as fast as their interest income did, their profits were down last year, but still plenty fat:  15% return on assets (down from 20% in 2007).  They’re still not very leveraged, so that translates into a 39% return on equity (down from 46% in 2007).  

Last week Compartamos announced that due to the recession its past-due loans will to the highest level in their history.  But that will be about 2.5% (PAR30?) by the end of 2009–still darned low.  They’re slowing new branch openings (20 this year, down from 60 last year), and will stop hiring new loan officers in the cities for a while.

The investors who paid monster prices to buy Compartamos shares in the mid-2007 initial public offering have gotten clobbered.  The share price was over $6 shortly after the IPO. By November of last year it had crashed to $1.33, suffering along with almost all other bank stocks.  Its latest price was $2.86–which made it one of the best-performing stocks in Mexico for the first half of 2009.

Here’s my narrative….  The 2007 public offering produced an “irrationally exhuberant” buzz, and new investors bought shares at ridiculous price/earnings multiples.  By June of 2008, sanity had returned and the trading price settled to a more defensible level ($3.85).  Then the universal panic about bank shares drove their price down to the low of $1.33 near the end of last year.  But eventually the market figured out that Compartamos was largely immune from the asset quality problems affecting other banks, and the price has climbed back to nearly $3–still far below the price right after the IPO.  Given how strong earnings have remained, one would expect to see further price gains as (or if ?) investors get less nervous about banks generally, and it becomes clearer that Compartamos is unlikely to have a serious default problem.  (If their loan portfolio were seriously shaky, I think it would have shown up by now, since most of the loans are short-term.)

Note that this price evolution affects, not only new purchasers of shares, but also the insiders who annoyed a lot of people by getting rich off of the IPO.  They still own a big chunk of the company.  As shareholders, they’re still sitting pretty, but not nearly as pretty as they were in the fall of 2007.