Foreign Exchange and Disaster Risk Management in Microfinance Institutions

"We should be managing risks instead of managing crises!" 1

Dr. Aberra Deressa,

Minister for Agriculture and Rural Development of Ethiopia

¹ Rojers, P.J. (2007) quoted after Dr. Aberra Deressa

TABLE OF CONTENT  

TABLE OF CONTENT  

TABLE OF FIGURES  

TABLE OF ABBREVIATIONS  

1 INTRODUCTION - 1 -  

1.1 ECONOMICS OF MICROFINANCE INSTITUTIONS - 1 -  

1.2 OVERVIEW ABOUT THE FOLLOWING CHAPTERS - 4 -  

2 RISK IDENTIFICATION AND ASSESSMENT - 5 -  

2.1 OVERALL RISK MAPPING IN MFIS - 5 -  

2.2 ANALYSIS OF DISASTER-RELATED RISKS - 7 -  

2.2.1 INTRODUCTION TO DISASTER RISKS - 7 -  

2.2.2 DISASTER-RELATED RISK MAP - 9 -  

2.2.3 ASSESSMENT OF DISASTER-RELATED RISKS - 10 -  

2.3 ANALYSIS OF FX-RELATED RISKS - 13 -  

2.3.1 INTRODUCTION TO FX RISK - 13 -  

2.3.2 FX-RELATED RISK MAP - 14 -  

2.3.3 ASSESSMENT OF THE RISKS IDENTIFIED - 15 -  

2.3.4 EXCHANGE RATE SYSTEMS AND DOLLARIZATION - 17 -  

2.4 MOST COMMON TREATMENT FOR FX AND DISASTER RISK - 19 -  

3 DISASTER RISK MANAGEMENT - 21 -  

3.1 INSTITUTIONAL DISASTER PREPAREDNESS - 21 -  

3.2 TRIGGER CONCEPTS AND PREREQUISITES FOR THE RISK TRANSFER - 23 -  

3.3 EVALUATION OF EXISTING FINANCIAL INSTRUMENTS - 26 -  

3.3.1 CATASTROPHE BONDS - 26 -  

3.3.2 WEATHER DERIVATIVES - 27 -  

3.3.3 CONTINGENT CAPITAL / CONTINGENT CREDIT - 31 -  

3.4 TACKLING DISASTER RISK ON HIGHER OR LOWER LEVELS - 32 -  

3.4.1 DISASTER-MICROINSURANCE - 32 -  

3.4.2 DISASTER LOAN FUNDS - 35 -  

3.4.3 PUBLIC-PRIVATE-PARTNERSHIPS - 36 -  

4 FOREIGN EXCHANGE RISK MANAGEMENT - 40 -  

4.1 SUSTAINABLE RISK ACCEPTANCE - 40 -  

4.2 RISK AVOIDANCE STRATEGIES - 41 -  

4.3 RISK MITIGATION STRATEGIES - 43 -  

4.3.1 CURRENT PRACTICES: OPERATIONAL HEDGES - 43 -  

4.3.2 EVALUATION OF FINANCIAL INSTRUMENTS - 47 -  

4.4 INNOVATIVE CONCEPTS - 51 -  

5 CONCLUSION - 55 -  

APPENDIX I Value at Risk ZAR-USD 5-year analysis APPENDIX II Degrees of Dollarization (1996 – 2001) APPENDIX III Memo Trueb, J., August 19, 2007 APPENDIX IV Overview about FX risk management strategies APPENDIX V Memo Zuidberg, J., September 19, 2007

BIBLIOGRAPHY

A) Non digital sources B) Digital sources

TABLE OF INTERVIEWS

TABLE OF ABBREVIATIONS

TABLE OF  FIGURES  

  Figure 1: Average loan size GNI per capita  

  Source: Microfinance Information Exchange Inc (MIX) based on  

  the 2005 Benchmarking  

  Page:  2

  Figure 2: Breakdown of specialised MFIs  

  Source: Meehan J (2004) p 3 (modified)  

  Page:  3

  Figure 3: MFI Risk Map  

  Source: self-provided  

  Page:  5

  Figure 4: Great Natural Catastrophes 1950  2003

  Source: http: www.ourworldfoundation org.uk (05 08 2007)  

  Page:  7

  Figure 5: Indian Climatic Disaster Risk Map  

  Source: http: commons.wikimedia org (05 08 2007)  

  Page:  9

  Figure 6: Hazard modeling (left) and Loss estimation (right)  

  Source: N.p World Bank (2006) pp 36  37

  Page:  10

  Figure 7: Liquidity flow during a disaster  

  Source: self-provided  

  Page:  11

Figure 8: Liquidity needs distribution

Source: N.p., World Bank (2006), p.37 (modified)

Page: 11

Figure 9: Value at Risk of the South-African Rand

Source: self-provided

Data : http://www.oanda.com/convert/fxhistory (03.08.2007)

Page: 14

Figure 10: Payout calculation for multiple hazard events

Source: N.p., World Bank (2006), p.42

Page: 27

Figure 11: Application ranges

Source: N.p., USAID-OAS (1999), Relationship of Return Period to

Annual Probability Distribution of Extreme Wind (modified)

Page: 34

Figure 12: Layering instruments

Source: Self-provided on the basis of:

N.p., World Bank (2006), p.37

Page: 35

Figure 13: Cost of debt with diversified local currency funding

Source: Zuidberg, J. (2007), p.6

Page: 48

Figure 14: Overview about FX risk management strategies

Source: self-provided

Page: APPENDIX IV

TABLE OF ABBREVIATIONS

TABLE OF ABBREVIATIONS  

ART Alternative Risk Transfer  

CAT Catastrophe  

CBOT Chicago Board of Trade  

CCRIF Caribbean Catastrophe Risk Insurance Facility  

CGAP Consultative Group to Assist the Poor  

CME Chicago Mercantile Exchange  

Cp. Compare  

DLF Disaster Loan Fund  

e.g. for example (lat.: exempli gratia)  

ECA Eastern Europe and Central Asia  

et al. and others (lat.: et altera)  

EUR Euro  

FAQ Frequently asked questions  

FLDG first-loss default guarantee  

FMO Nederlandse Financierings-Maatschappij voor  

Ontwikkelingslanden N.V.  

FX Foreign exchange  

GNI Gross National Income  

GOLF Guaranteed Offshore Liquidity Facility  

IFC International Finance Corporation  

ISDA The International Swaps and Derivatives Association  

L/C Letter of Credit  

LAC Latin America and the Caribbean  

LIBOR London Interbank Offered Rate  

MBP Microenterprise Best Pracices  

MENA Middle East and North Africa  

MFI Microfinance Institution  

MIGA Multilateral Investment Guarantee Agency  

Mil million  

MIX Microfinance Information Exchange, Inc.  

MXP Mexican Peso

n.d. no date

N.p. No publisher

NDF Non-Deliverable Forward

NGO Non-Governmental Organisation

No. Number

OFDA Office of U.S. Foreign Disaster Assistance

OTC Over-the-counter

p. page

PAR Portfolio at Risk

pp. pages

PPP Public-private-partnership

SBLC Standby letter of credit

SPV Special Purpose Vehicle

TCIP Turkish Catastrophe Insurance Pool

TCX The Currency Exchange Fund

U.S. United States

U.S.A. United States of America

UNCDF United Nations Capital Development Fund

USAID U.S. Agency for International Development

USAID-OAS U.S. Agency for International Development -

Organisation of American States

USD US Dollar ($)

VAM Vulnerability Analysis and Mapping

VAR Value-at-risk

WFP World Food Programme

WWB Women's World Banking

yr year

ZAR South-African Rand

1 Introduction

1.1 Economics of Microfinance Institutions

Microfinance institutions (MFIs) have been largely regarded as instruments of donor associations for the altruistic distribution of money in developing countries. In fact MFIs are commercial lending institutions to be found all over the world – in developing and developed countries likewise – that have partly been co-financed by donors as their business model supports some charitable goals like the reduction of poverty. Since MFIs become more and more profitable and their portfolio sizes as well as their numbers of borrowers are growing by up to more than 50% annually, MFIs increasingly seek for additional commercial funding sources – both locally and internationally. ² This increasingly enforces their self-responsibility for economical sustainability including a prudential treatment of existing and emerging risks.

The basic principle in credit risk management is that a loan has to be secured by collateral. This initiates a vicious circle that has often been interpreted as “the more you got the more you get” with the consequence that people who got nothing at all will not get any start-up capital to change this state. When Muhammad Yunus, the Nobel Peace Prize laureate of 2006, founded the first microfinance institution in 1976 in Bangladesh, his idea was to overcome this basic rule in banking with a leap of faith that initiated ongoing client relationships.

Today there are around 10,000 MFIs providing micro-loans to poor people and small enterprises reaching repayment rates up to 100%. ³ The global loan portfolio is estimated 7 billion US-Dollar (USD) in outstanding loans, serving around 13 million clients, and generating repayment rates of 97% in average. Since loans were formerly provided to potential micro-entrepreneurs to set up small businesses, they created assets broadening the demand for additional services. To date

² Cp. Coppoolse, M. (2007), p.2

³ Cp. in the following N.p., UNCDF (2007)

- 1 -

the range in some MFIs comprises additional services as checking and saving accounts, insurances, transfer services, remittances and even leasing contracts, while the majority still focuses on loans and deposits. The diversity of organisations, products, methodologies, clients and geographical locations exacerbates a clear definition of MFIs. The Microfinance Information Exchange Inc. (MIX) defines MFIs according to their size as financial institutions whose average balance of services is not greater than 250% of the average income per person in the country. 4

Based on this definition MIX conducted a survey in 2005 analysing 446 MFIs in the Middle East and North Africa (MENA), Asia, Latin America and the Caribbean (LAC), Eastern Europe and Central Asia (ECA) and Africa finding that the average

loan size ranges from 15% to 90% of the GNI per capita. (See figure 1 on the right) MFIs in this segment often provide micro-loans between only 10 and 100 USD of value. Interest rates for these loans range from

30% to 80% plus commissions and fees to provide loans on a cost recovery level and survive in the long-term because MFIs are subject to significant higher costs for transaction, risk cover and refinancing than traditional banks in other sectors. ⁵ The provision of micro-credit is very time-consuming and resource intensive since the customer base is multiple times broader and borrowers make repayments monthly, or even more frequently. Thus according to the ⁴ Cp. in the following N.p., Microfinance Information Exchange (2007)

⁵ Cp. Fernando, N.A. (2006), pp.1-3 as well as N.p., MicroCapital (2006), p.1

- 2 -

MicroBanking Bulletin 2006 the average expense ratio of MFIs around the globe is 24.9%. ⁶ Nevertheless micro-entrepreneurs are able to afford these high interest rates as they are only a small percentage of their total returns. Studies in India, Kenya and the Philippines computed an annual return on investment by micro-businesses between 117% and 847%, which are seen as commonplace. ⁷ Compared with the alternative of informal moneylenders in the black market which are charging monthly interest rates up to 100%, the official rates in MFIs are even comparably low.

Today’s MFIs have diverse legal states from Non-Governmental Organisations (NGOs) to credit unions, regulated financial companies and specialised banks strongly determining their product ranges and refinancing abilities. ⁸ The biggest differences result from the institution’s size and level of

development.

some Compartamos in Mexico or ASA in Bangladesh are institutions portfolios of outstanding loan amounts of up to 255 million USD, other

MFIs emerged out of neighbourly help and are still in an early stage of development. ⁹ Figure 2 illustrates an estimation of the market participant’s distribution concerning their different stages of development. Apart from the leading group of mature institutions MFIs are considered to be small with gross loan portfolio

⁶ Cp. N.p., MicroBanking Bulletin (2006), p.33

⁷ Cp. N.p., International year of microcredit 2005 (2007), p.1

⁸ Cp. Bruett, T. (2004), p.1

⁹ Cp. Coppoolse, M. (2007), p.2

- 3 -

amounts of less than 2 million USD and big if the value exceeds 15 million USD. ¹⁰ Profitability is likewise correlated with stage of development and size. “[…] 63 of the world’s top MFIs had an average rate of return of about 2.5% of total assets, after adjusting for inflation and after taking out subsidies programs might have received.” 11

1.2 Overview about the following chapters

Chapter 2 will enter the subject with an overall risk mapping in MFIs. In the following “disaster” as well as “foreign exchange” (FX) risk will be introduced separately. The concrete impacts from their different peculiarities will be pointed out and analysed in an exemplarily risk assessment. The chapter ends with a short insight in the common treatment of such “high impact – low frequency”-risks in developing countries and the aspect of the Samaritan’s dilemma to underline the importance of rethinking about current practices. To handle the overwhelming diversity of the examined field, this work will focus on the usual business model of MFIs offering loans and saving accounts to their clients and the financial dimension of the concerned risks. The third chapter shortly describes the initiation of institutional disaster preparedness as essential prerequisite for all following activities concerning disaster risk management. General basics for the transfer of disaster risks will be discussed isolating the most appropriate instruments. Chapter 3.3 introduces three financial instruments from the group of alternative risk financing and discusses their applicability for MFIs individually. To overcome the identified hurdles, three alternatives will be pointed out enabling the usage of the instruments described before.

¹⁰ Cp. N.p., MicroBanking Bulletin (2006), p.29

¹¹ Cp. N.p., The MIXmarket (2007)

- 4 -

Chapter 4 explains the concept of faithful risk acceptance as a framework for risk management considering influenceable risks as those arising from foreign liabilities. In a second step the different strategies of risk avoidance will be presented. Chapter 4.2 discusses instruments and strategies to minimise or transfer foreign exchange risk from international liabilities concerning applicability, effectiveness and efficiency in the microfinance context. This includes operational hedges and bilateral agreements as well as financial instruments as they are formerly used in more developed economies. Finally innovative concepts as “The Currency Exchange Fund” will be presented as new opportunities for MFI’s risk management of foreign exchange risks in developing countries.

Even though MFIs settled as well in developed countries like Germany, this work will focus on the vast majority of MFIs in developing countries to take regional specifics like underdeveloped financial markets and volatile exchange rates more into account.

The thesis ends in chapter 5 with a summary of the key findings and an evaluation of current possibilities. General constraints to effective risk management in MFIs will be highlighted describing possible future prospects.

2 Risk Identification and Assessment

2.1 Overall Risk Mapping in MFIs

The first step in the risk management process is to become aware of existing and emerging risks. All risks have to be identified and understood to assess their impact on the institution. The business model of MFIs is very similar to that of commercial banks, hence is creating an akin risk map.

- 5 -

Since for many MFIs lending is the core business function, default risks from the inability to collect interest and/or principals from the microentrepreneurs are the most frequently discussed risks in this context. Many successful strategies like group lending and family loans have been developed over the past years reducing this risk in well-managed institutions up to 100%. ¹² Reputational risk is important because many clients have few experiences with financial institutions, and a negative image can influence the client’s confidence particular in reference to savings and deposits. Liquidity risk is becoming more and more crucial as the majority of maturing MFIs start to offer saving and even checking accounts to their clients. The fact that they are predominant located in emerging countries which are heavily exposed to natural disasters and economic instabilities, highlights the operational risk from external events and stresses the liquidity risk additionally. Out of the market risks especially interest rate and foreign exchange risks are growing due to the fact that MFIs often borrow at floating rates and increasingly

¹² Cp. N.p. responsibility (2005), p.2

Krauss and Walter [Krauss, N., Walter I. (2006)] calculated a median default rate for 283

MFIs of 2.4%.

- 6 -

in foreign currencies. Their loans are largely offered at fixed interest rates in domestic currencies – this leads to a dangerous asset-liabilitymismatch. The increasing refinancing in hard currencies in absence of proper risk management practices is currently considered to be one of the biggest threats to microfinance institutions on the way to an independent business establishment.

2.2 Analysis of Disaster-related Risks

2.2.1 Introduction to Disaster Risks

The United States Department of Commerce defines a disaster as “a crisis event that surpasses the ability of an individual, community, or society to control or recover from its consequences.” ¹³ Accordingly, the classification as disaster is dependent on the magnitude of an event and the individual vulnerability of the MFI and its clients.

The causes of disasters can be differentiated in natural and man-made hazards with significant negative impact on the society or the environment. Frequency and severity of natural disasters worldwide increased exponentially in recent years. Due to multiple reasons areas, that faced infrequent natural disasters in the past, may experience more frequent floods, droughts and hurricanes in the future as can be seen in figure 4. (The thin black line is demonstrating the overall trend in natural disaster’s frequency.)

- 7 -

Moreover MFIs operate predominant in developing countries which are exceptional prone to natural disasters such as earthquakes, floods, tsunamis, droughts, hail, frosts, volcanic eruptions or hurricanes. “Between 1990 and 1998, 94 percent of the world’s 568 major natural disasters and more than 97 percent of all natural disaster-related deaths were in developing countries”. ¹⁴ A self-conducted analysis including all disasters worldwide with more than 10,000 people affected between 1997 and 2007 confirmed this result concerning the disaster locations. ¹⁵ Furthermore, the vulnerability of MFI’s clients in developing countries is exorbitant high due to their establishment in cheaper less disaster-protected areas, simple accommodation facilities and the strong dependence on agricultural business. The latter is seen as engine for a big share of the micro-businesses because even nonagricultural jobs are closely related to agricultural production and have little capacity to grow on their own. ¹⁶ As a matter of fact even less catastrophic events as heavy rainfall or drought can severely affect a MFI’s client base and result in a disaster according to the afore mentioned definition.

The eventualities of man-made disasters are infinite, ranging from technological disasters (e.g. engineering failures, transport accidents) to sociological disasters (e.g. crime, terrorism, riots or war), and can hardly be foreseen or managed from the point of affected organisations or companies.

Economic or financial crises can result from man-made events as overwhelming public debt or from natural disasters and their consequences. Investigations in the impact of macroeconomic distress showed weak correlation between the macroeconomy and the net operating income of MFIs. This can be explained with the circumstance,

¹⁴ N.p., World Bank (2000), p.170

Note: A disaster is classified as “major” if it caused more than 50 deaths or affected more than 100,000 people. (Source: USAID, OFDA 1999) ¹⁵ The analysis included 1184 events worldwide whereof 1114 took place in developing countries (=94%). Source of data: EM-DAT: The OFDA/CRED International Disaster Database - www.em-dat.net - Université Catholique de Louvain - Brussels - Belgium ¹⁶ Cp. Skees, J. et al. (2002), p.2 and 3

- 8 -

that MFI’s clients “may be less integrated into the formal sector of the economy”. ¹⁷ The Bank Rakyat Indonesia, for instance, was forced to write off 100% of its corporate portfolio and 50% of its middle market loans during the Asian financial crisis – in its microfinance portfolio on time repayment slid only one percent to 97.5%. ¹⁸ Beside the marginal impact on client’s businesses, a financial crisis can have profound consequences on a MFI’s refinancing activities that will be discussed in chapter 4 as a dimension of the foreign exchange risks.

Further remarks of this chapter will concentrate on the most likely derivation of disastrous situations for MFIs, which are severe weather events and natural disasters with a special focus on developing countries.

2.2.2 Disaster-related Risk Map

A natural disaster, regardless its nature, can affect MFIs in three major risk categories.

Operational risk arises from the direct interruption of operational activities through physical damages, e.g. the destruction of offices, equipment or information systems, or through indirect implications like the inability to reach offices and clients or damages to the overall infrastructure. Reputational risk can arise from an institution’s inability to fulfil its contractual obligations in the aftermath of a disaster, e.g. to disburse deposits to savers. ¹⁹ Additionally, the risk of fraud and theft is often designated to be increased in crisis situations.

Credit risk increases significantly by death or disability of micro-borrowers. Even physical unscathed borrowers may be inhibited to attain earnings and pay interest or principles because of destroyed livelihoods or injured family members. This can have severe impact on the liquidity situation of a MFI.

¹⁷ Cp. Krauss, N., Walter I. (2006), pp.12 and 13

¹⁸ Cp. Meehan, J. (2004), p.15

¹⁹ Cp. in the following Pantoja, E. (2002), pp.13 - 15

- 9 -

Liquidity risk is furthermore tightened by missing savings and immediate withdrawals of existing deposits to replace lost livestock or afford medical care. Those who are not in possession of deposits will probably demand emergency loans, whether they are already existing clients or non-clients.

2.2.3 Assessment of Disaster-related Risks

Assessing disaster risks, the possible hazards have to be identified for

every area the MFI operates in. Historical data about frequency and severity of past events has collected governments insurance Figure 5 shows a risk map of climate-related natural exemplarily for India, exhibiting hazard-prone areas for cyclones, floods, droughts and hot or frigid deserts. Hazard modelling transforms the data into hazard probability distributions for every possible threat. ²⁰ This can be done using historical or Monte Carlo simulations. The probability is often described in terms of “1 in X years” instead of percentages as can be seen in figure 6. Since damage and loss increase exponentially with the intensity of a disaster, the loss estimation shows the relation of the severity of the hazard and the expected damage. The inherent damage

²⁰ Cp. in the following N.P., World Bank (2006), p.36

- 10 -