1
Performance and governance in microfinance institutions
Roy Mersland
a
, R. Øystein Strøm
b,*
a
Agder University, Kristiansand, Norway
b
Østfold University College, Halden, Norway.
This version: 1011-2008
Abstract
We examine the relationship between firm performance and corporate governance in
microfinance institutions (MFI) using a self-constructed global dataset on MFIs collected from
third-party rating agencies. Using random effects panel data estimations, we study the effects of
board and CEO characteristics, firm ownership type, customer-firm relationship, and
competition and regulation on an MFI’s financial performance and outreach to poor clients. We
find that financial performance improves with local rather than international directors, an
internal board auditor, and a female CEO. The number of credit clients increase with
CEO/chairman duality. Outreach is lower in the case of lending to individuals than in the case
of group lending. We find no difference between non-profit organisations and shareholder firms
in financial performance and outreach, and we find that bank regulation has no effect. The
results underline the need for an industry specific approach to MFI governance.
JEL classification: G30; G32; J23
Keywords: Microfinance; Governance; Performance; Boards; Ownership
*Corresponding author. Tel.: +4769215287.
e-mail addresses: mersland@online.no (R. Mersland), reidar.o.strom@hiof.no (R.Ø. Strøm)
Acknowledgements. We thank participants at the 8th Corporate Governance Workshop
(Cambridge, UK, 24-25 March 2007), participants at the 33
rd
EIBA Annual Conference
(Catania, Italy, 12-14 December 2007), and two anonymous referees for their comments. This
paper was originally submitted to Professor Giorgio Szegö on June 20
th
, 2007 and was revised
after submission through EES.
2
1. Introduction
In this paper, we investigate the impact of governance mechanisms on microfinance
institutions’ (MFIs) dual missions of financial sustainability and providing banking services to
micro-enterprises and low-income families. We identify three dimensions to this problem: a
vertical dimension between owners and management, a horizontal dimension between the MFI
and its customers, and an external governance dimension. Recommendations for better
governance are made primarily for the first and third dimensions. For example, Rock et al.
(1998), Otero and Chu (2002), and Helms (2006) suggest importing best practices in
governance from developed countries, such as board independence and shareholder ownership.
Van Greuning et al. (1999) and Hardy et al. (2003) argue for better MFI regulation.
However, problems of credit risk assessment and repayment
1
make governance of firm-
customer interactions potentially more important in banking than in other industries (Adams
and Mehran, 2003b). This is the focus in the present study. For example, if an MFI has a high
percentage of female loan clients, should it adjust its governance accordingly? Group lending is
seen as a method to ensure repayment (Armendariz de Aghion and Morduch, 2005). Is an
MFI’s financial performance enhanced when the MFI supplies its customers primarily with
group loans?
1
Two factors make an MFI’s loan portfolio different from that of a bank. First, it is generally
semi- or uncollateralised. Second, repayment time is generally short. Thus, an MFI faces the
risk of steep deterioration of its portfolio in a matter of only a few weeks.
3
We use recently released data from third-party rating agencies, yielding a unique dataset of 278
MFIs from 60 countries between 1998 and 2007. Thus, we respond to Morduch (1999) and
Hartarska’s (2005) requests for the use of better data in the analysis of microfinance questions.
Microfinance is high on the public agenda after the UN Year of Microcredit in 2005 and the
awarding of the Nobel Peace Prize to Mohammad Yunus and the Grameen Bank in 2006.
Nevertheless, microfinance still reaches only a fraction of the world’s poor (Robinson, 2001;
Christen et al., 2004). Helms (2006) and the Consultative Group to Assist the Poor (C-GAP
2004, 2006) consider the lack of strong MFIs to be a major constraint on the further
development of the microfinance industry, and CSFI (2008) identifies governance as a major
obstacle to MFI growth.
Few studies have been published on corporate governance in MFIs. Hartarska (2005)
investigates the relationship between governance mechanisms and financial and outreach
performance, using three surveys of rated and unrated Eastern European MFIs between 1998
and 2002. Governance mechanisms include board characteristics, CEO compensation, and
ownership type.
2
Hartarska (2005) includes several institutional and firm control variables and
finds that a more independent board gives a better return on assets (ROA). However, a board
with employee directors results in lower financial performance and outreach. The difference in
financial performance and outreach between various ownership types is negligible.
2
Ownership type refers to the various legal incorporations found in MFIs, ranging from
shareholder-owned firms to cooperatives.
4
Cull et al. (2007) also investigate MFI financial performance and outreach by focusing on
lending methodology.
3
They use data from 124 MFIs around the world and find that financial
performance improves, up to a point, with individual loans, and that MFIs concentrate more on
individual loans. Governance variables, such as board characteristics or ownership type, are not
considered.
Our study is therefore justified by the neglect of the MFI-customer dimension, the limited
number of academic studies available, our large and comprehensive global dataset, and the fact
that some governance mechanisms, like competition and internal board auditor, remain
unexplored in the literature.
Our findings indicate that most corporate governance mechanisms have little impact on MFIs’
financial and outreach performance. However, results show that financial performance
improves when the board is informed by an internal auditor and has local directors, and when
the CEO is a woman. For outreach, measured by the number of credit customers and average
loan amount, CEO/chairman duality increases the number of credit clients. Outreach is reduced
with individual lending. Generally, there is no difference between non-profit organisations and
shareholder firms in either financial performance or outreach. Similarly, bank regulation also
does not seem to have an impact on financial and outreach performance.
This paper proceeds as follows. Section 2 develops our hypotheses. Section 3 presents an
overview of the data sources and estimation method, and descriptive evidence is reported in
Section 4. Section 5 presents econometric evidence. Section 6 concludes and proposes a new
research agenda.
3
Lending methodology refers to the way loans are given. The categories used are individual
loans, group loans, and village banks, which are larger groups of approximately 20 members.
5
2. Governance and performance in MFIs
Governance is about achieving corporate goals. The first goal of MFIs is to reach more clients
in the poorer strata of the population, and the second goal is financial sustainability. We analyse
the relationship between governance mechanisms and both outreach and financial performance.
Financial performance is assessed in terms of overall profitability, through such measures as
return on assets (ROA
4
), operational self-sufficiency (OSS
5
), revenues (portfolio yield), and
operational costs (Christen, 2000). Using these measures enables us to pinpoint more clearly
under what conditions a particular governance mechanism is effective. The outreach measures
are the MFI’s average outstanding loan and the number of credit clients served (Schreiner,
2002). Table 1 summarises the dependent variables.
Table 1
The table confirms the high (nominal) portfolio yield usually experienced in MFIs. Thus, an
average of nearly 40% is not surprising in these markets. All returns in the regression analysis
are adjusted for inflation. Thus, we use real rates for ROA [(ROA – inflation) / (1 + inflation)]
and portfolio yield. The average loan reflects the “micro” in microfinance. The lowest loan
amount is US $2.22, the average loan amount is US $788, and the median is US $441. The
maximum amount of approximately US $25,000 is an extreme case, which is twice the amount
of the next largest loan. We filter out the extreme cases above US $10,000 and adjust the
remaining loans to purchasing power parity GDP (World Economic Outlook, IMF).
4
Debt/equity levels differ considerably between MFIs. Hence, ROA is more appropriate than
ROE (return on equity) when measuring financial results across different institutions. ROA is
calculated based on operational profits before donations and taxes.
5
OSS is a widely used proxy for institutional sustainability. Table 1 gives its definition.
