DueDiligence TechGuide ENG

7/31/2019 DueDiligence TechGuide ENG

1/58

Due Diligence

Guidelines for theReview of MicrocreditLoan Portfolios

A Tiered Approach

Robert Peck Christen and Mark Flaming

2009


2/58

These guidelines were irst developed by Robert Christen, with assistance romPatricia Mwangi, both o CGAP. Mark Flaming dra ted the current versiona ter testing the methodology. Richard Rosenberg o CGAP wrote the annexon yield gap analysis. Lynn Curran and Deborah Drake o ACCIN, Todd

Farrington o MicroRate, and Mark Schreiner o Washington Universitycontributed to earlier versions o a related document.

The authors and CGAP would be very grateful for feedback from people or organizations that have used this tool in testing the quality

of a microloan portfolio. Please send any comments on your experiencewith it to [email protected]. Your observations will

help us improve future editions of the tool.

2009, Consultative Group to Assist the Poor/The World BankAll rights reserved.

CGAP1818 H Street, N.W., MSN P3-300Washington, DC 20433 USA

[email protected]


3/58

ABSTRACT

Micro inance institutions (MFIs) operate with risks that investors need tobe concerned about. Un ortunately, external audits, ratings, evaluations, andeven supervision too o ten ail to identi y the primary riskinaccurate repre-sentation o port olio quality. These due diligence guidelines are designed tohelp an analyst evaluate, with statistical precision, the accuracy o accountingand per ormance reports about the port olio, as well as an MFIs compli-ance with its own port olio management policies. The document guides the

analyst through three increasingly robust levels o appraisal. This allows theanalyst to adapt the exercise to a wide range or uses and requirements aboutlevels o con idence. In the third level, the use o statistical sampling to vali-date conclusions rom conventional appraisal techniques in the prior levelsgenerates uniquely robust results that should satis y the most demandingcommercial investor or regulator.

Tier I is a ive person-day review o the MFIs basic policies, procedures,and systems or managing and reporting on the per ormance o its loanport olio. This review should be required as a part o any credible external

assessment, appraisal, or ratings exercise.Tier II is a 1014 person-day assessment o whether operational practices

are consistent with the MFIs policies and procedures and with standardso best practice in microcredit port olio management. This normally wouldbe a minimum requirement or donors, investors, and ratings agencies thatevaluate and assess MFIs or potential investment.

Tier III is a 34 person-week exercise that includes an audit to con irmthe port olio quality through a sampling o loan iles, accounting iles, andthe loan tracking so tware. Tier III due diligence is recommended or inves-

tors or regulators who need a quantitative and statistically certain measureo port olio per ormance.

These due diligence guidelines provide a comprehensive checklist o whatto analyze, some instruction about appraisal technique, and very little guid-ance about how to interpret the results. The Tier I review can be carried outby someone with general knowledge o micro inance but Tiers II and III relyheavily on the judgment o an analyst with extensive experience in the evalu-ation o MFIs and their loan port olios.

For another microcredit portfolio testing tool, produced by MicroSave,

see http://microsave.org/relateddownloads.asp?id=14&cat_id=313&title=Loan+Portfolio+Audit


4/58

Due Diligence Guidelines for the Review of Microcredit Loan Portfolios2


5/58

INTRODUCTION

Micro inance institutions (MFIs) operate with risks that investors need tobe concerned about. Un ortunately, external audits, ratings, evaluations, andeven supervision too o ten ail to identi y the primary risk aulty represen-tation o port olio quality. These due diligence guidelines are developed tohelp investors, donors, and regulators veri y the real level o risk in an MFIsloan port olio.

Discrepancies between real and reported port olio quality are common inMFIs. In act, many well-known MFIs have experienced at least one signi i-cant port olio crisissustaining delinquency and de ault rates well abovewhat they reported to the public. Finansol, or example, was a ColombianMFI that was widely considered to be one o the top institutions in LatinAmerica during the early 1990s. In November 1997, the Colombian banksuperintendent required Finansol to inject several million dollars o capitalto shore up reserves or bad debt. The intervention came a ter inspectorsreclassi ied a signi icant number o Finansols loans as doubt ul. The addi-

tional reserve requirement put Finansol in technical insolvency, since theassociated expenses decreased capital below acceptable adequacy levels.

Finansols case was highly public, but certainly not unique. Because itwas a supervised inancial institution, its problems were a matter o publicrecord. But there have been many cases that did not receive publicity. Mostmicro inance network organizations have discoveredmore than oncethat one o their most important a iliates had a ar more serious problemwith loan delinquency than previously stated. These incidents have costmany millions o dollars and, in some cases, have orced donor agencies torecapitalize or shut down operations they supported or years.

The occasional ailure o MFI managers, donors, and investors to detectport olio ailures is simply part o the growing pains o a generally vibrantmicro inance industry. Fortunately, most MFIs have been able to weatherthese episodes because they were only moderately leveraged, their port olioproblems did not create an acute liquidity crisis, or they were able to indanother donor to recapitalize the institution. However, MFIs are integratinginto regulated inancial systems, taking deposits, and leveraging themselves

with commercial sources o unding. The Finansol case demonstrates thatpoor oversight can lead to severe consequences in these conditions.


6/58


The purpose o these due diligence guidelines is to provide bankingsuperintendents, donor agencies, and potential investors with a tool or amore accurate and deeper understanding o the underlying quality o anMFIs loan port olio than current methods o er. This tool is designed tocreate the levels o con idence required or MFIs to leverage their capitalwith commercial unding and deposits.

These guidelines provide a comprehensive checklist o what to analyze,some instruction about appraisal technique, and very little guidance abouthow to interpret the results.This is not a training manual for aspiringanalysts; it is a checklist for an analyst with extensive experience in evalu-ating MFIs. Earlier versions o this document attempted to annotate thechecklist with guidance about how to interpret the results o the appraisal,

but these proved to be unsatis actory and cumbersome in practice. Thisversion relies ultimately on the judgment o the analyst to use the checklistto sort out what matters most and to identi y the root causes o risk in theloan port olio.


7/58

THE DUE DILIGENCE PROCESS

Due diligence is the process o examining a companys per ormance. It istypically per ormed by or on behal o investors who are primarily interestedin veri ying that per ormance is accurately re lected in the inancial state-ments and company reports, and that management and operating systemsare robust enough to sustain good per ormance into the uture.

In the commercial inancial industry, investors rely heavily on external

appraisals o a inancial institution to analyze per ormance trends. Audits,ratings reports, stock analyst reports, supervisory reports, and even mediaarticles provide an array o appraisals o a inancial institutions per or-mance. This varied and ongoing measurement o per ormance providesincentive to the inancial institutions, creates consensus around best practicemetrics, and provides an evaluator with many sources o analysis. Thus thedue diligence analyst typically arrives at a inancial institution with consider-able in ormation.

The due diligence process itsel typically entails a review o the basic

management and operational systems and a care ul analysis o the inancialinstitutions business model.The important point to make about due dili-

gence methods in a commercial and regulated financial industry is that theanalysts typically rely heavily on the information provided by management and by the accounting and performance reporting systems. 1 They do thisbased on the assumption that an institution that operates success ully in suchan industry has robust systems and competent management, and that grossde iciencies tend to get weeded out by competition, audits, supervision, orexternal appraisals.

This assumption is not always valid in the micro inance industry, soconventional approaches to audits or rating appraisals can ail to capturethe peculiar weaknesses o MFIs. The micro inance industry has placed greatemphasis on adapting appraisal and evaluation tools, audit and inancialdisclosure guidelines, and per ormance standards. There are assessmenttools, such as ACCIN Internationals CAMEL, WOCCUs PEARLS, andCGAPs Technical Tool or appraising MFIs; assessments by rating agencies

1

In researching due diligence practices, the authors interviewed many representatives rom ratingsagencies, investment banks, and commercial unders who con irmed that in ormation risk is nottypically addressed as a primary concern in traditional due diligence practice.


8/58


(MicroRate, M-CRIL, Micro inanza, CRISIL, and PlaNet Finance); creditratings by credit rating agencies (Standard and Poors, Fitch, Apoyo, Class,Paci ic Credit Rating, and Equilibrium); guides or analysts, including theTechnical Guide o the Inter-American Development Bank, CGAPs externalaudit manual, and minimum disclosure guidelines or MFI inancial state-ments. There are also benchmarking projects, such asMicroBanking Bulletin and MIX Market. In addition, there are numerous local or national e orts todevelop many o the same tools and capacities to support network reportingrequirements, apex investment decisions, and inspection visits by bank super-intendents.

These appraisal tools may e ectively assess an MFIs overall inancialper ormance. But they typically do not test the veracity o the underlying

in ormation. Appraisal work plans typically ocus on loan port olio admin-istration without delving much into a detailed reconciliation o loan iledocumentation, accounting records, and loan tracking system reports. Theappraisals typically assume that the in ormation reported is correct, andmake that assumption based on the existence o an acceptable audit report.Conventional audits usually include some tests o these systems, but theyare seldom intensive enough to justi y much con idence in the accuracy o system reports, particularly loan tracking reports.

The ollowing example demonstrates this point. In a typical audit, lettersare sent out to a random sample o loan clients asking them to veri y theloan they received, their outstanding balance, their latest payment, and anyother pertinent in ormation. Clients are asked to send the letter back to theauditor i they disagree with any o the in ormation provided in the letter.The auditor assumes that i no letters come back, there are no disagreements.The laws in this assumption are obvious to anyone with experience with thetypical clientele o an MFI.Due diligence in an MFI must go urther than conventional methods to

veri y per ormance precisely in the areas that typically challenge MFIs. Thesedue diligence2 guidelines are designed or that purpose.

A THREE-TIERED APPROACH This document guides the analyst through a three-tiered approach orassessing how well an MFIs loan port olio is per orming and how well it ismanaged. The tiers build on each other, with each tier using more sophis-ticated analytical procedures and more evaluation sta resources than the

2

These due diligence guidelines deal speci ically with port olio quality and management. A completedue diligence process in an MFI would involve the many other areas that are typically included in aull due diligence exercise.


9/58

Due Diligence Guidelines for the Review of Microcredit Loan Portfolios 7

previous tier. A due diligence process that incorporates all three tiers o ersthe greatest precision and highest degree o certainty about the actual levelo loan port olio risk.

Tier I. This is a ive-day review o the MFIs basic policies, procedures,and systems or managing and reporting on its loan port olio. It is both ahigh-level analytical review o port olio trends and a stock-taking exercise.In the stock-taking exercise, the analyst collects the MFIs ull inventoryo credit policies, procedural manuals, key accounting policies, and inan-cial and per ormance reports. The Tier I assessment is based largely onthe completeness and coherence o this inventory and on interviews withmanagement about port olio trends and the policies, procedures, and systemsor managing the loan port olio. The collection o documents also provides

the basic in ormation or Tier II analysis.Tier II. This is a 1014 day assessment o whether MFI operational

practices are consistent with policies and procedures and with standardso best practice in microcredit port olio management. The analyst veri iesthat accounting policies and reporting ormats are consistent and based onsound practices; that credit department sta conduct operations accordingto policy and procedural guidelines; and that the MFI does not regularlyengage in practices that generate unmeasured credit risk, such as reclassi-ying loans, re inancing, and granting parallel loans. The Tier II assessmentis based primarily on how closely the MFIs practices match internal policiesand how reports match industry standards o best practice.

Tier III. This is a 34 week exercise that includes detailed testing o transactions to con irm the port olio quality through a sampling o loaniles, accounting iles, and the loan tracking management in ormation system(MIS). The Tier III audit uses statistical sampling methods that supportconclusions, within acceptable levels o con idence and error, about realarrears levels and, more broadly, the risks related to accounting practices

and credit policy. The Tier III conclusions are supported by the results o the auditnamely, whether port olio per ormance is accurately re lected inaccounting and per ormance monitoring reports.

The Tier I review should be required as a part o any credible externalassessment, appraisal, or ratings exercise. It will give the analyst an overviewo recent company trends that may a ect the loan port olio and a basic butcomprehensive look at the MFIs policies, procedures, and reports. The eval-uation incorporates enough discussion, cross-re erences o key per ormanceindicators, and accounting and MIS review to provide a sense o whether

the MFI at least has the basic tools to manage a port olio. The analyst willmake qualitative and subjective judgments in a Tier I review, so the value o


10/58


11/58


Contracting or a Tier I exercise should be airly straight orward. Thescope o work is already explicit in these guidelines and will not varysigni icantly.

The Tier II guidelines also provide a comprehensive scope o work.However, the Tier II exercise is more dynamic in that the analyst must be ableto direct e orts toward problem areas as they are discovered. The sponsorand analyst should develop a workplan that combines the Tier I and Tier IIinstructions and that anticipates mid-course adaptation at key points. Theirst checkpoint will be just be ore the ield visit, when the analyst knowsthe condition o the MFIs basic operating documents. The second will bea ter the analyst has interviewed headquarters management sta . At thispoint, both the sponsor and the analyst should expect to review and adapt

the remaining workplan based on initial indings.The Tier III exercise includes care ully selecting the analyst team and

de ining the scope o work. These elements will vary depending on the objec-tives o the sponsor and the characteristics o the MFI. The lead analyst oranalysts must have the capacity to organize the Tier III procedures to testthe conclusions o the Tier II assessment. This requires extensive experiencewith MFI management and evaluation. The Tier III team must also havethe requisite skills in inancial and statistical analysis, spreadsheet use, andauditing procedures. The most easible way to construct such a team is tohire a lead analyst, who in turn hires a supporting analyst and/or a localauditing irm that is amiliar with the micro inance industry and with statis-tical sampling methods. As in the Tier II exercise, the scope o work shouldanticipate review and adaptation o the workplan at the key checkpoints. InTier III, this is especially important as the analyst makes decisions about howto sample the port olio in the audit. The analyst, audit sponsor, and the MFIare encouraged to design the sample method with clear expectations aboutthe levels o certainty in the results.

Due diligence sponsors should make a special e ort to clari y all expec-tations when working with audit irms. The three-tiered approach is moreextensive and more detailed than the scope o conventional audits. Thecontract with the auditor should speci y both the procedures and the audi-tors responsibility or ormulating conclusions on the results. Many auditorsare disinclined to use any but the standard procedures employed by theirirm. Even irms that agree to conduct the procedures may not be willing toissue an opinion on their results. In these circumstances, sponsors will dowell to hire a lead analyst who is responsible or ormulating inal conclu-

sions and or subcontracting and managing the audit irm.


12/58


13/58


in ormation as possible rom sources both inside and outside the MFI. Theollowing is a summary o the required documents. (See Section 1 tables inAnnex I.)

Financial reports Audited financial statements (balance sheet, pro it and loss, and

cash low) or the past two complete iscal years. The audited inan-cials serve as a benchmark or the other sources o in ormationin the desk review, as they represent an outside validation o thein ormation provided by MFI management. The audited inancialstatements report should consist o an auditors opinion, inancialstatements, and notes to the inancial statements. The analyst should

ask or the complete report, including the management letter i theauditor provided one, and the MFI managements response to theauditors report.4

Current year-to-date financial statements if more than one quarterhas elapsed

Supporting internal financial reports for the same periods thatprovide su icient detail o the loan port olio (classi ication by risk/ reserve categories), reserves, income, and expenses

Current liquidity and asset and liability management report

Accounting policies. This will consist o an accounting policy manual andmemos related to the ollowing transactions:

Accounting of loan balances by arrears classification Rescheduled loans Loan loss provisions, 5 loan loss provision expenses, loan write-o s,

and recoveries (it is particularly important to get a clear statemento the MFIs policies in these areas)

Fee and interest income on loans, including the accrual of interest Allocation of client payments to outstanding balances Liquidation of collateral Any noncash methods of loan repayment (e.g., in-kind, refinancing

old loans with new ones)

4 For urther detail on the audit o MFIs, seeExternal Audits of Microfinance Institutions: AHandbook, CGAP Technical Tool No. 3 (Washington, D.C.: CGAP, December 1998), especially

pages 2223.5 This is sometimes re erred to as impairment allowance, impairment expense, or impairmentloss.


14/58


Credit policies and procedures. The MFI should have documents thatprovide clear and comprehensive guidelines or port olio and delinquencymanagement. These documents should include any credit-related policies

and procedures manuals or memos to sta .

Portfolio reports. The loan port olio reports are typically produced bythe MFIs management in ormation system (MIS)6 and provide detailedin ormation about lending activities. The MIS should be able to produce theollowing reports or the analyst:

Portfolio classification of loans and amounts by number of daysdelinquent

Comparison of actual loan repayments to scheduled repayments

(during the past year or quarter) Breakdown of portfolio by loan product (lending methodology),

including number o active loans, total port olio, and port olioquality or each category

Portfolio reports by branch and/or loan officer (including portfoliosize, details on products, port olio quality)

Detailed reports on restructured and refinanced loans Summary portfolio performance reports for branch managers,

senior management, and board o directors

Agent banking policies and procedures. Some MFIs contract with third-party agents who execute some aspects o the port olio management on theMFIs behal or instance, ormation and management o borrower groups.In such cases, the analyst will have to assess the policies and procedures thatthe MFI employs to manage the agent, and the risks associated with theagents participation. MFIs may use agents or any combination o servicesand there ore the analyst must be able to identi y the policies and procedures

that are necessary in each case. The analyst should be able to gain a basicunderstanding o the role o the agent rom the ollowing documents:

Contracts or other agreements with agents MFI guidelines for the agent (e.g., client eligibility requirements,

reporting requirements, etc) MFI internal policies and procedures for management of the agents

services

6

For urther in ormation on the development o quality MIS or MFIs, seeManagement InformationSystems for Microfinance Institutions, CGAP Technical Tool No. 1 (Washington, D.C.: CGAP,February 1998).


15/58


Agent and MFI reports on agent activity

Assessments, appraisals, and external evaluations . The MFI should send

the analyst the report generated by any assessment, appraisal, or externalevaluation that has been done during the two previous years. Thisin ormation is generally con idential and should be treated as such by theanalyst. Increasing numbers o these reports are being prepared by ratingsagencies or a iliate networks, donors, and investors.

1.2. Desk ReviewThe analyst should prepare or the site visit with a desk review organizedaround three objectives. One objective is or the analyst to gain an under-standing o the business model and operations o the MFI, including themost recent growth, provisioning, and write-o trends. This will guide theanalysts interviews with senior management. The analyst should also discussany major changes in the market or in MFI operations and determine howthese might a ect port olio management. A second objective is to determinei the MFI has adequately documented policies, procedural manuals, andinancial and port olio per ormance reporting. The ability o the MFI toprovide these documents is the irst indication o the management capacity

and pro essionalism o the institution. Anything short o immediate deliveryo a complete set o documents and reports should guide the analyst towardareas that require special attention during the ield visit. And the third objec-tive o the desk review is to identi y policies or per ormance that deviatesubstantially rom best practices o similar MFIs.

The Section 1 tables in Annex I serve as an appraisal guide or the analyst.The main components o the analysis are summarized as ollows:

The MFI Operating Environment. Most o this in ormation will be gath-ered rom interviews with senior management during the ield visit. However,the analyst should gather in ormation rom www.themix.org to compare theMFI to competitors in the market (or similar MFIs elsewhere) and to generalmarket trends.

Financial Reporting and Performance. The analyst should assess the overallinancial per ormance o the MFI as well as the quality o the underlyingin ormation.

Accounting policies and financial reports


16/58


It is very important to con irm that accounting policies and reportormats use accepted standards or the key variables that re lect port-olio per ormance (e.g., interest accrual, arrears classi ication, loan loss

provisioning, write-o s, etc.). Reporting consistency

It is important to veri y that the amounts are consistent among reports,especially the port olio balances reported in the loan tracking systemand on the balance sheet. Discrepancies are common because data o tencome rom di erent sources or data lows. The analyst should discussdiscrepancies with MFI management to determine whether managementactively reconciles discrepancies and whether the discrepancies representa material de iciency in the reporting system.

Financial PerformanceAssetsThe inancial analysis should ocus on the evolution o key indicatorsover time, looking or abrupt changes or per ormance that deviates romnormal standards. Key indicators include the ollowing:

Growth rates o loan disbursements and port olio balance Port olio in arrears

Loan-loss provisions Write-o s

Financial PerformanceIncome and Expenses Yield on port olio, including a yield gap analysis that compares

actual with expected income7

Cost o unds Provisioning and write-o expenses as a percentage o the

average loan port olio Operating expenses

The analyst will be guided to a great extent by his or her own judgmentabout the indicators that best reveal port olio quality in each particular case.

7 This quick and very e ective check or hidden arrears in the port olio is discussed in Annex II.


17/58


There may be good reason to make use o additional indicators o inancialper ormance.8

The analyst will also want to consider the MFIs per ormance against

peer institutions. MFIs di er widely in the clients they serve, their methodso managing port olio risk and delinquency, their institutional structure, andthe legal and regulatory environments in which they operate. In addition,contextual actors, such as geographic location, local market conditions, andage o the institution, in luence per ormance indicators.

In recent years,MicroBanking Bulletin (MBB) has begun to address theproblems o peer group analysis. MBB tracks the inancial per ormanceo over 700 MFIs worldwide (www.mixmbb.org). Although individualprogram results are con idential, MBB displays results or peer groups o

similar institutions that the analyst can use as a guide to determine where theper ormance o given institutions all. In addition, a iliate networks, suchas ACCIN International, FINCA, CASHPOR, MFN, and SEEP, may haveinternal databases on which to draw peer comparisons. Finally, the analystcan look at the per ormance o other individual MFIs in the country at MIXMarket (www.mixmarket.org).

Credit Policies and Procedures. A clearly documented credit policy is anecessary condition o prudent lending. The MFIs lending policies shouldestablish the necessary authority, rules, and ramework to operate andadminister its loan port olio and manage risk. The policies should be appro-priate to the size o the institution and the nature o its activities, and theyshould be consistent with prudent banking practices and relevant regula-tory requirements. The policies should be broad and not overly restrictive insetting guidelines to maintain sound credit appraisal and approval standardsand control, manage risk, and deal with problem loans.9

The desk review o credit policies and procedures should ocus on ourkey questions:

Are the eligibility criteria and appraisal and approval proceduresadequate and clearly documented to ensure consistent loan analysis?

8 Several publications explain the indicators commonly used in the micro inance industry: SeeTillman Bruett, ed.,Measuring Performance of Micro-finance Institutions: A Framework forReporting, Analysis, and Monitoring, SEEP and ACT (Washington, D.C.: USAID/AcceleratedMicro inance Advancement Project, orthcoming); CGAP,Selected Definitions of Financial Terms,Ratios and Adjustments for Microfinance , Micro inance Consensus Guidelines (Washington, D.C.:CGAP, 2003); Joanna Ledgerwood, Sustainable Banking with the Poor inMicrofinance Handbook:An Institutional and Financial Perspective (Washington, D.C.: IBRD/The World Bank, 1998); CGAP,

Measuring Microcredit Delinquency, Occasional Paper 3. (Washington, D.C.: CGAP, 1999.9 Board o Governors o the Federal Reserve System,Commercial Bank Examination Manual.Division o Supervision and Regulation (Washington, D.C., March 1994).


18/58


19/58


even i they think they have given clear orders or how they should be calcu-lated. It is worth stating again that the analyst may learn the most about theMFI rom mangers ability to produce a complete set o these documents andreports and explain how they are used.

The ield visit should begin with a meeting with the MFIs seniormanagement team. The meeting should include discussions o the economy,regulatory environment, the state o the micro inance market in the country,and the institutions clients. Through these discussions, the analyst shouldgain an understanding o the context in which the MFI operates anddisburses credit.

The analyst should then hold subsequent meetings with members o the management team to discuss the MFIs loan port olio per ormance and

management. These ace-to- ace meetings should cover all o the policiesand practices o the institution that are listed in the Section 1 tables inAnnex I. This process is iterative in that the analyst will ask similar ques-tions o the general manager, chie inancial o icer, credit manager, andcredit o icers o the head o ice. Answers to these questions should beconsistent among all members o the team and should correspond to thestated policies o the institution.

Analysis o credit policy should be grounded in discussions with bothmanagement and credit department sta and with re erence to the creditmanuals. The analyst must look at the principles underlying an MFIs creditmethodology to answer this essential question: Are there clearly de inedand consistently applied credit policies? There is requently a di erencebetween what is printed and what is practiced. For this reason, whatmanagement and sta say and do is always more important than what isprinted in the manual.

1.4. Final AnalysisThe analyst needs to orm an overall impression o the MFIs port olio risk.Much o this impression depends on the analysts con idence about the trust-worthiness o the in ormation and the senior management team. The analystshould consider the ollowing:

Transparency of data Ease with which data are provided Top managers familiarity with loan portfolio quality indicators and

reports Consistency of information across several MIS reports

Staff commitment to ironing out inconsistencies


20/58


General environment of orderliness and internal control Presence of actual ex-post control mechanisms Clarity in all aspects of the MFIs operation

I the analyst concludes the Tier I exercise with an uncom ortable eeling,chances are high that the results o Tiers II and III exercises will besubstantially worse. Rarely do microcredit port olios look better on closerexamination than they do at the outset.

2. TIER II DUE DILIGENCETier II due diligence builds on the Tier I exercise.11 A ter completing all o thetasks or Tier I, the analyst completes ive additional areas o analysis:

Integrity and completeness of the MIS Verification of loan classification, provisioning, and write-off

procedures Actual performance of credit operations, especially in branch

o ices Degree to which the work of individual loan officers reflects MFI

policy Integrity and completeness of individual loan files

The essential question o Tier II is whether or not the policies and procedurespresented by senior management and in MFI documents is re lected inpractice at the operational level and whether those operational practicesare sound. In Tier II, the analyst engages sta in the credit, accounting,and in ormation technology departments and, most importantly, the brancho ices. The process is iterative, sometimes requiring the analyst to consultwith several departments to con irm how transactions are recorded, howreports are generated, and how all loan management procedures are actually

carried out.A Tier II exercise may be more e ective i conducted by a team o two

analysts, or example one with inancial analysis skills and the other withcredit operations experience, to bring a wider range o experience to thetask. The analysts are likely to bene it rom comparing perspectives anddiscussing indings.

11 For the sake o simplicity, the remainder o this section will re er only to the additional workrequired in a Tier II exercise. It is also important to clari y that a Tier I analysis conducted in a Tier

II exercise will usually be more robust than i carried out alone. Although the analyst asks the samequestions in both cases, in a Tier II exercise the analyst has more opportunities to collect, veri y, andanalyze in ormation.


21/58


2.1. MIS ReviewIn Tier II due diligence, the analyst will ocus on veri ying the accuracyo reports and the capabilities o the accounting and port olio tracking

systems. The Tier I review will likely guide the analyst toward the areas o these systems that merit urther study. A lack o clarity around accountingpractices, or discrepancies among reports, provide important clues orurther investigation.

The most e ective method or testing accounting practices and reportingaccuracy is to audit the transactions on a sample o loans, ollowing eachtransaction through the port olio tracking and accounting system. Theanalyst will need to irst develop a spreadsheet template that replicates all o the transactions associated with a single loan: disbursements, ee assessments,

interest calculation, distribution o principal, interest and ee payments,arrears calculation, and so orth. The ormulas must precisely re lect thepolicies and procedures or these transactions. The analyst then selects asample o loans rom the accounting system, collects the correspondingcontracts, and prints a report rom the port olio tracking system that hasa ull transaction history or each loan. The analyst enters the disburse-ment and payment transaction in ormation rom the loan report manuallyinto the spreadsheet template so that the spreadsheet ormulas calculate therunning balances, principal, interest and ee allocation, and arrears status.The analyst compares the result o the spreadsheet template to reports gener-ated by the port olio tracking system.

The analyst con irms that the transactions are passed accurately intothe accounting system. The analyst should also inspect the physical receiptsassociated with the disbursements and client loan payments. This completesa ull audit o the transactions that are re lected in the balances that appearin the inancial statements and port olio reports.

In this exercise, the analyst will con irm whether the port olio tracking

system (i) re lects the exact terms and conditions o the individual loancontracts, (ii) accurately calculates and reports principal, interest, and eedistributions, (iii) accurately calculates the arrears classi ication o the loans,(iv) produces accurate reports, and (v) passes the in ormation accurately tothe accounting system.

The analyst should conduct the transaction audit on a sample o loansthat includes examples o all the MFIs lending products. This sampling tech-nique makes no attempt at statistical signi icance. It simply helps the analystunderstand the details o the loan products, ocuses discussion with sta ,

and detects systematic deviations rom stated policies.


22/58


The analyst should also comment on whether the accounting and loantracking system are capable o expanding with the MFI. The challenges andpotential o these systems vary signi icantly. Custom-designed so tware, o -the-shel so tware, and systems purchased rom other MFIs require di erentskill sets to operate and maintain. The analyst should assess whether theMFI has the resources, or access to the resources, to maintain the system itis using.

2.2. Loan Classification, Provisioning, and Write-offsThe Tier II analysis should evaluate the MFIs method or classi ying itsport olio, estimating and provisioning or loan losses, and writing o andrecovering bad loans. The evaluation should compare real losses to provi-

sioning levels. For most MFIs, the risk classi ication system should be basedon repayment per ormance as well as historical patterns associated withactors such as speci ic lending products, guarantee coverage, economicsectors, geographic location, and branch o ice. MFIs with inely tuned clas-si ication criteria use classi ications not only or provisioning but also to setdi erential risk-based interest rates.12

2.3. Branch Office ManagementThis part o the Tier II exercise ocuses on how well the MFI manages creditoperations through the branch o ices. Most o the in ormation or thisanalysis is gathered through interviews with credit department managementand branch o ice sta . The analyst should structure the analysis aroundive basic areas.

Distribution of authority and responsibility. The analyst should map howthe key credit activities and decisions are organized between headquartersand the branch o ices. Ultimately, the analyst will be looking or an appro-priate balance o operational e iciency and control in the MFIs management

o lending targets, loan approvals, cash management, accounting, andreporting.

Performance management. The analyst should pay special attention tohow branch o ice per ormance is managed. The mapping exercise shouldreveal how responsibility is distributed or setting per ormance targets, gath-ering per ormance indicators, per ormance reporting, per ormance review,and management response to per ormance achievements. The analyst should

12 The ACCIN CAMEL, Technical Note, Microenterprise Best Practices, p. 42.


23/58


expect to ind robust systems or measuring per ormance and decisivemanagement response to both good and poor results.

Core operational policies. The operations most directly related to port-olio quality are loan analysis and delinquency management. The interviewsshould reveal sta understanding and competency in these areas.

Consistency between policies and practice. The analyst should pay specialattention to the consistency between how things are done at the branch andhow headquarters management thinks things are done, in all areas o creditoperation. Discrepancies re lect lack o management control.

Branch office morale. Sta morale is an important indicator o current anduture port olio per ormance. Low morale is o ten a product o sta rustra-tion over ine ective policies, inadequate management, poor compensation,

or chronic poor port olio per ormance. Sustained port olio per ormance isunlikely with demoralized sta .

2.4. Loan Officer PerformanceLoan o icers are one o the most important sources o in ormation aboutlending operations. They carry out most o the procedures, and they areclosest to the clients. The analysts should expect to spend at least threeull days with the loan o icers, away rom senior management, to capturethe reality o ield operations. The analyst should use in ormal conversa-tions with loan o icers to cover the topics o port olio quality, how closelylending policies are ollowed, how lexible they are, routine contact withcurrent clients, loan rescheduling/re inancing, and delinquent loan proce-dures. Analyst should be care ul how they question loan o icers to avoidintimidating ield sta and impeding the discussions.

Analysts should also accompany loan o icers on client visits to witnessirst-hand how procedures are ollowed in the ield. On these visits, analystsshould be able to get a sense o the clients history with the MFI, level o

satis action, relationship with loan o icer, and so orth. Discussions withclients should be brie and unstructured with a ew open-ended questions.The analyst should also review the MFIs training program or credit

o icers and any incentive systems to encourage good per ormance. Incentivesystems should be based primarily on port olio quality. Incentive systemsthat reward port olio growth without quality will undermine rigor in loanselection and eventually port olio per ormance.


24/58


25/58


o opportunities to evaluate whether the MFI conducts business accordingto its policies as well as to observe signs o success or ailure. Ultimately,the analyst should be able to present a well- ormed opinion about the core

per ormance areas, substantiated by observation and analysis.

3. TIER 3 DUE DILIGENCETier III due diligence aims to achievea quantifiable level of certainty regardingthe quality o the loan port olio that would satis y a commercial investor.The exercise uses statistical methods to veri y individual account balances,measure the level o compliance with lending policies and procedures, andcheck or material discrepancies and raud.

The exercise will enable the analyst to answer three basic questions withstatistical certainty:

Do the portfolio tracking and accounting systems accurately reflectloan balances, transaction activity, and port olio risk?

Does the MFI rigorously follow sound lending policies and proce -dures?

Does the MFI have adequate systems of internal control to ensurerobust per ormance monitoring and compliance with policies andprocedures?

A Tier III analysis is appropriate or a third party that requires a higher levelo certainty about port olio per ormance than is normally available roman audit or ratings report. This certainty may be necessary or commercialinvestors, bank regulators, or a strategic investor, any o whom may want toestablish clear benchmarks with MFI management about improvements tolending policies and operating systems.

3.1. Audit Scope

Analysts de ine the scope o the Tier III audit based on their assessmento the Tier II results. Speci ically, analysts must use judgment to determinewhich reports, accounting policies, and loan management procedures havesigni icant impact on the risk o port olio per ormance, and then identi y thespeci ic data to audit. These decisions will likely require some modi icationsto the scope and organization o the MIS and procedural audit templatesused in Tier II.

At a minimum, the Tier III MIS audit will go urther than the Tier IIexercise by con irming the payment vouchers and general ledger entries asso-


26/58


ciated with each o the disbursements and payments that appear in the loantracking system.

For each sample loan, compare the disbursementsand payments thatappear on the loan summary report with the accounting vouchersthat support actual disbursements and payments. Take note espe-cially o how payments are made (cash, check, post-dated check,new loan, repossession o collateral, other).

Follow each of these payments through to the general ledger, andcheck compliance with the MFIs policies or recording interest andprincipal amounts, and policies or noncash payments.

The Tier III procedural audit should include an audit o second or third loans

to the same client. The audit should con irm that the previous loans werepaid o according to MFI policies.

Determine whether the clients prior loan was repaid by a noncashor partial cash payment (through rescheduling, re inancing, post-dated or third-party checks, or receipt o collateral or equipment).This is done by reviewing the inal two or three payments o theclients prior loan to determine repayment behavior. The inalpayment voucher o the prior loan should be checked to determineactual means o repayment.

Determine whether the increase in loan and payment amounts isconsistent with MFI policy, and generally reasonable.

The analyst will also need to create a sample speci ically or making ieldvisits to the clients. This is especially important i loan o icers manage cashpayments in the ield. The MFI should help introduce the analyst to theclients by presenting these visits as a client-service evaluation.

The main objective of the visit is to compare the clients records or

recollection with the repayment history and arrears status in theloan summary report. Compare their understanding o their repay-ment status with the loan port olio delinquency report.

If a sample client is unavailable, the analyst should visit the sampleclients neighborhood to ask other clients in that vicinity to con irmthe sample clients existence and business activity.

Determine how cash payments are actually made to the MFI. Forexample, are payments made to loan o icers in the ield? Do groupleaders make the groups payment to the MFI o ice or to a bank?

Determine what documentation is provided to clients or the cash


27/58


payments. It is particularly important to ocus on the time betweenthe actual repayment and when that repayment is entered into theMIS o the MFI.

Determine the veracity of the basic information on the loan applica -tion and loan report.

Look for evidence of fraud, such as loans granted to nonexistentgroups, kickbacks o any o the loan amount to the originating loano icer, bribery, or other orms o mal easance.

The analyst should not be accompanied by the clients loan o icer.

In the Tier III audit, the analyst will need to take special care to organize theresults o the audits so that the in ormation can be analyzed quantitatively.

Two organizational methods are recommended. First, the results o eachloan audit should be entered into a separate worksheet template. Second,each template should have a summary table that quanti ies the audit results.In the MIS audit, or example, the summary table should capture the loanamount, type, actual versus reported days in arrears, and an indicator or thevalidation o payment vouchers and accounting entries. In the proceduralaudit, the summary table should capture the type and number o materialdeviations rom policy. The analyst will want to inalize the audit templatesa ter care ully planning which data and calculations will be required orthe analysis.

3.2. The Tier III Sampling MethodThe Tier III sampling methodology is designed to provide levels o certaintythat are adequate or most investors or regulators to make decisions, withoutincurring the enormous costs required to achieve high levels o certainty aboutthe precise value o the port olio. The primary objective o this samplingmethod is to validate the indings o the Tier II exercise. Speci ically, the

analyst will weight the sample to test the areas o high risk detected duringthe Tier II analysis. This added level o statistical certainty makes Tier IIIconclusions uniquely robust. The Tier III results can, or example, supportthe ollowing statement:

In our opinion, the MFIs practices comply in all material respectswith its policies and procedures for loan portfolio management;internal control systems are sound, adequate, and effective; and,

accounting and portfolio reports can be relied upon to generateaccurate component balances.


28/58


OR

In our opinion, the MFIs practices deviate substantially from its

policies and procedures for loan portfolio management; internal control procedures are inadequate to guarantee compliance; and,accounting and portfolio reports exaggerate actual operating incomeand underreport the portfolio in arrears by as much as 20 percent.

Investors should be aware that the discretionary sampling in Tier III limitsto some degree the robustness o in erences about per ormance levels inthe entire port olio. A ocused sample can be very e ective at detectingproblems and estimating their dimensions, but because it is not statisticallyrepresentative o the entire population the results cannot be extrapolated toquanti y a value adjustment to the balances that appear on the MFI reports.That level o precision is possible with larger sampling methods, but at acost that exceeds the marginal bene its or most investors or regulators. Formost investors, the results o the discretionary sampling will be adequate.For example, a sample o a speci ic loan product that constitutes 65 percento the entire port olio could show that the MFIs reports are inaccurate inover 20 percent o the loans, that over 15 percent loan iles show material

deviations rom lending policies, and that the analyst detects widespreadindications o management de iciencies. In such a case, an investor orregulator has su icient in ormation to take actiondeclining investment orlaunching an aggressive corrective intervention.

3.3. Defining the SampleThe Tier III analyst makes two basic decisions about how to sample theloan port olio. One decision is about what part o the port olio to sampleand the other is how many loans to include in the sample. This means thatthe analyst uses his or her judgment to select the types o loan to sampleand then uses statistical models to create the sample. So or example, theanalyst may decide to sample the loans o a speci ic branch o ice and thenuse statistical tables to determine the sample size and a statistical programto randomly select the sample loans.14

14 For random selection o loans, the analyst can import a list o loans generated by the MFIs

port olio tracking system into EXCEL, SPSS, or similar o -the-shel so tware. When using EXCELwith the Analysis Toolpak add-in, click on Tools>Data Analysis>Sampling. In SPSS, go toData>Select Cases and use the option Random.


29/58


The analyst makes the irst decision based on the indings o the TierII exercise. Port olio per ormance de iciencies are typically concentrated incertain branch o ices, types o loans, credit o icers, or accounting methods.A key advantage o the Tier III sampling method is precisely that it allowsthe analyst to base the sample on an in ormed Tier II opinion about likelyproblem areas. The general guiding principle is to oversample the loansthat are likely to reveal practices associated with poor per ormance. Typicalproblem areas are

larger loans delinquent loans loan types with a history of problems loans to clients who were delinquent on previous loans

loans made by loan officers with higher than average loan write-o s

loans with peculiar amortization schedules that may be miscalcu -lated in the port olio tracking system

loans that are originated or managed by third-party agents

Typically, the analyst will want to design the sample so that it represents theentire port olio as well as ocuses on probable problem areas. For example,the analyst might randomly select well-per orming branch o ices in additionto branch o ices selected speci ically or their historic poor per ormance.Likewise, the global sample o loans might include a random selection romeach loan type in addition to a more robust sample o a loan type withknown problems. It is important that the analyst justi y the rationale or thediscretional sampling and its extrapolation to the port olio.


30/58


The ollowing is an illustrative example o Tier III sampling.

MFI Micronova has three loan products. Business loans were Micronovasirst loan product and still make up 45 percent o the port olio. Theyare collateral-based loans and signi icantly larger than any other loanproduct. Microcredit loans are the newest and make up only 10 percento the port olio. They were developed with the help o internationallyexperienced technical partners and are based on a cash low analysis o the clients household. Salary loans are 30 percent o the port olio.Micronova has 12 branches. Five are at least our years old, and togetherthey have 70 percent o the entire port olio. The other seven brancheshave been in operation less than 18 months.

The analyst is particularly concerned about the business loans, whichhave higher reported arrears rates than the other loans, especially inthe ive oldest and largest branches. There ore, the analyst decides tosample three o the ive large branches and two o the seven newerbranches. Within each group, the branches are selected randomly. Ineach branch, the analyst decides to sample each loan product separately.The analyst is also concerned about ollow-up loans and there oredecides to create a sample or each loan product that is hal irst-time

loans and hal ollow-up loans. Within these categories, the loans areselected randomly.

3.4. Sample SizesThe choice o sample size depends on the desired level o certainty in inter-preting the indings.15 The relationship o sample size to con idence levelsis strictly mathematical: larger sample sizes will produce higher con idencelevels and smaller margins o error in extrapolating the sample results to

the port olio. Nevertheless, the analyst will use judgment to determine howmuch certainty is required. This manual recommends a standard sizingstrategy that will probably be su icient or most situations, but the analystcan adjust it as needed.

The Tier III exercise uses an incremental approach to sampling that allowsthe analyst to stop when the results produce the desired level o certainty. Aninitial sample set o 50 loans may return results that are su icient or somepopulations. For example, i a sample o 50 loans contains 10 cases o unre-

15 Re er to the statistical tables in the annex to determine the margins o error associated withdi erent combinations o sample size and proportions.


31/58


ported loan rescheduling, the analyst may in er, with 95 percent certainty,that the incidence in the entire port olio is between 10 percent and 33.7percent, or a proportion o about 20% with a margin o error o roughly11% on either side o that value. The range o this estimate is indeed large.However, the analyst may decide that even the lower bound is unaccept-able and that the results are su icient to con irm the initial hypothesis thatserious rescheduling problems exist. The analyst has the option o pullinganother set o 50 loans or a combined sample o 100. I the analyst were toind 22 cases in the sample o 100, the range would be 12.227.8 percent, ora proportion o 22 percent with a lower margin o error o about 8 percent.The analyst may continue to enlarge the sample size and urther reduce themargin o error. However, in many cases where problems are detected in the

irst 50 loans, the slightly smaller margin o error that could be achievedwith a larger sample size will not be worth the cost.

Larger sample sizes may be desirable when the analyst wants to gainmore precision or good results in the initial sample. For example, i theanalyst only inds one problem case in a sample o 50 loans, the 2 percentinding can be extrapolated to the port olio with a margin o error thatplaces the incidence between 0.1 percent and 10.7 percent. A sample o 300with six cases would render a smaller margin o error with the incidence inthe entire port olio between 0.7 percent and 4.3 percent. The analyst can usethe tables in the Annex I to select the appropriate sample size.

3.5. Quantifying the ResultsThe inal output o the audit should be a set o spreadsheets that quanti y theresults or each loan. The analyst needs to compile the results o the sampleand extrapolate the results to the port olio. The results need to be expressedin two variables. The irst is the incidence o error. The analyst can calculatethe proportion o negative indings in the sample and then consult the tables

in Annex I to determine the margin o error or extrapolating the sampleindings to the entire port olio.The second variable is the magnitude o the error. Here again, the analyst

needs to use judgment to determine the appropriate metric. As an example,the magnitude o discrepancies in the arrears status o a loan should prob-ably be measured in days (e.g.,the arrears status of loans is, on average,underreported by 17 days ). The increment in loan sizes between irst, second,and third loans may be measured as a percentage o the previous loan, or innominal amounts, depending on how the MFI measures the increments in

its own policies. In any case, when calculating the average error, the analyst


32/58


should consider the standard deviation in the data set or clues to the under-lying cause o the error.16

The incidence and magnitude variables enable the analyst to quanti ythe discrepancies. For example, the MIS audit might show that the port-olio tracking system misstates the arrears status in 15 percent o all loansand that the average error is 34 days (arrears is typically underreported).Furthermore, a small standard deviation in the errors (meaning that theerrors are airly consistent) might also lead the analyst to investigate whetherthe MFI is consistently per orming an adjustment that produces the under-statement o arrears.

In the procedural audit, the results might show that 35 percent o loaniles re lect an average o three material incidents o ailure to comply with

MFI policies. I the analyst has segmented the procedures into operationalcategories (e.g., loan analysis, approval, disbursement, ile management,etc.), he or she can determine whether the compliance ailure is concen-trated in speci ic procedures or spread randomly throughout all phases o loan management.

3.6. Final Discussion with ManagementAt this stage o the Tier III exercise, the analyst will ormulate preliminaryconclusions about the overall per ormance o the MFI, loan management,the integrity o policies and procedures, and the accuracy o the port olioreporting and accounting systems. The inal step in the investigation phase isto present these tentative conclusions to MFI management or discussion.

The inal consultation with management serves at least three purposes.First, it engages management in the inal interpretation and this helpsgenerate management buy-in to the inal report and recommendations.Second, it gives management an opportunity to correct misconceptions andprovide perspective that adds important nuance to the inal interpretation.

Finally, the discussion gives the analyst a inal look at management capacityand attitude.

3.7. Analysis and Write upThe quality o Tier III due diligence depends ultimately on the ability o theanalyst to interpret audit results together with the more subjective discov-eries o the Tier I and Tier II exercises. The analyst should be able to quanti y

16 Excel (or SPSS or any statistical audit package) will calculate the median or mean and the standard

deviation. In Excel, use the appropriate @ unctions, or in the Analysis Toolpak Tools > DataAnalysis > Descriptive Statistics; in SPSS, use Analyze > Descriptive Statistics > Explore.


33/58


material deviations and trace them to their root cause in de iciencies o management capacity, policies and procedures, and systems.

The Tier III report needs to be explicit about the objective o the due

diligence sponsor and the sampling method. The rationale or discretionalsampling and extrapolation should be clearly supported by the Tier I andII indings. And the precision and con idence levels o statistical samplingshould be clearly explained.

4. GENERAL COMMENTS ON PORTOFOLIO DUEDILIGENCEIn many MFIs, the due diligence exercise will reveal that actual loan port olioper ormance is weaker than it appears in o icial reports. The micro inance

industry is still relatively young, and ew investors or regulators have devel-oped robust systems or oversight. MFIs, just like banks, have a strongincentive to put their best ace orward when it comes to the quality o theirassets. And sometimes, their donors or investors or regulators have strongincentives or not questioning strong per ormance reports. Ultimately, theanalyst has to determine whether the weaknesses or discrepancies creatematerial risks that threaten the interests o the MFIs donors, creditors,investors, regulators, or depositors. The analysts biggest challenge is how

to present the indings in a constructive manner. The ollowing observationsare meant to provide general guidelines: While there is a cure for ignorance, deceit is cancerous. Most

MFIs are truly ignorant about the degree or consequences o theirpoor MIS per ormance, but once trained, they can become ardentdisciples o transparency. In contrast, MFIs that deliberately hideport olio per ormance e ectively eliminate the possibility o accu-rate measurement and undermine their own credibility. In practice,an analyst can usually distinguish between those MFIs that think

transparency is good management practice and those that thinktransparency threatens their access to unds.

A weak MIS will inevitably lead to deterioration of loan portfolioper ormance. There should be no doubt about this. Anytime thereis a basic ailure to reconcile the loan tracking system results withthe accounting system, or loan delinquency reports do not re lecttransactions in a timely or accurate manner, quality will eventuallydegrade, no matter i current quality looks high.

The way MFI management deals with a weak MIS says much aboutits commitment to accurate port olio reporting. Ad hoc measures


34/58


35/58


about the key actors that in luence the inal recommendation to aninvestor or regulator.

Some potential donors or investors may want to engage in a two-step approach to the loan port olio due diligence process. The irststep examines practice in an initial visit, discusses negative ind-ings with management, and allows management time to resolveunhealthy practices. In the second step, the analyst can delve intodeeper due diligence once basic issues have been addressed. Theappraisal process o can be instructive to MFI management, showingit better ways to manage its organization.


36/58


37/58

Table 1.1

Tier I Review: MFI Operating EnvironmentKEY ISSUES APPRAISAL GUIDELINESWhat is the growth potential o the MFIstarget market?

Pro le the size of the target market

Are the MFIs lending products appropriateor the target market?

Pro le the lending methodology, includingloan conditions or the client (size o loanrelative to GDP per capita and e ectiveinterest rate)

Review drop out rates

How is the MFI managing competition in itsmarket?

Account for other MFIs offering similarservices

Identify competitive advantage of MFIservices vis--vis competitors (pay specialattention to scope and price o services,unding and operational costs o the MFI,and long-term unding capacity)

Assess management strategies for dealing

with competitionDo local MFIs generally comply with stan-dards o best practice?

Focus on nancial performance, arrearslevels, operating cost margins, and adop-tion o institutional orms capable o und-ing sustained growth

ANNEX I: Appraisal Tables


38/58


Table 1.2

Tier I Review: Financial Reports and Accounting PoliciesKEY ISSUES APPRAISAL GUIDELINESAnnual nancial statements are audited bya reputable rm and published in a timelymanner.

Anything less than publication of the fullauditors report onwww.themix.org andthe MFIs website requires urther investi-gation into management/owner commit-ment to transparency and disclosure.

In addition to audited summary nancialstatements, the MFI has monthly nancial re -ports that provide adequate detail on incomesources, expenses, portfolio classi cationand provisioning, loan products, and branchoperations.

The level of detail in the nancial reportsrefects the MFIs ability to manage theloan portfolio and measure nancialper ormance by operational departments,lending and unding products.

The MFI produces daily or weekly reportsor liquidity positions, asset and liabilitymanagement, unding, and any other time-sensitive operations.

This level of reporting will depend on thenancial structure of the MFI. The analystshould also assess whether managementmakes adequate use o these reports in therespective areas.

The balances o the detailed monthly reportsreconciles precisely with the balances in theaudited nancial statements.

Any discrepancy in the balances of detailednancial reports vs. audited nancialstatements is an indication o integrationproblems between components o theaccounting system. Management shouldbe able to explain and reconcile any di er-ences.

The port olio balances in the loan trackingreports reconcile precisely with the balancesthat appear on the balance sheet.

Any discrepancies here warrant carefulscrutiny to ensure consistent and veri ableaccounting o loan per ormance rom theloan registers all the way to the balancesheet.

The portfolio report classi es the outstandingbalance o every loan by arrears categories(this is a portfolio-at-risk classi cation of theentire port olio).

PAR reports are necessary for an accurateaccounting and management o loans inarrears.

The arrears classi cation categories andcorresponding loss provisions must beconsistent with international standards orthe MFIs documented historical experi-ence.

Re nanced or rescheduled loans are clearlysegmented in the portfolio classi cation.

These loans should be classi ed separatelyand subject to speci c provisioning poli -cies.

Accounting or loan loss provisions, provi-sion expenses, write-o s, write-o recover-ies, and interest accrual on non-per ormingloans are clearly presented in the nancialstatements.

These policies and procedures should beclearly documented and their applicationreadily veri ed in the nancial statements.A comprehensive set o policies will enablethe analyst to calculate all key indicatorsrelated to PAR and loan provisioning/lossexpenses over several accounting periods.Policies that obscure clear accounting inthis area are a sign o inadequate port olio

management.


39/58


Tier I Review: Financial Reports and Accounting PoliciesKEY ISSUES APPRAISAL GUIDELINESWhat are the accounting entries related to anindividual loan? Are disbursement, interestaccrual and repayment transactions preciseenough to ensure accurate individual loantracking and aggregate port olio reporting?

A robust accounting system will berefected in port olio reports that reconcileindividual account detail with the globalamounts that appear on the nancial state -ments. Anything less is a limit on manage-ments ability to supervise the per ormanceof loan of cers and their individual loans.

Is accrued interest income reversed whena loan alls into arrears? How is interestincome accrued? What is the treatment o accruals?

Policies governing reversal of accrued in -come should be consistent with the natureand tenor o the loan. As a general rule,accrued interest should be reversed a terno more than 90 days o arrears.

How is cash related to loans paid out, re-ceived, and controlled?

Is unrecorded cash in the hands of loanof cers for any amount of time betweenthe point o payment and recording in theloan tracking system? (Controls shouldexist to ensure that cash is accounted orimmediately and accurately.)

Can loans be paid o through in-kindpayments, receipt o collateral guarantees,post-dated checks, rescheduled loans, or anyother non-cash means?

Widespread practice of booking loanrepayments be ore the cash is received cansigni cantly exaggerate actual income. Theaccounting o these transactions shouldprovide a clear trail to the actual cashreceipt or write-down.


40/58


Table 1.3

Tier I Review: Financial PerformanceKEY ISSUES APPRAISAL GUIDELINESOverall nancial performance is consistentwith industry standards or similar MFIs

Key per ormance indicators are: Portfolio yield

PAR

Provisioning and loan loss expenses

Operating cost ratio, or cost per borroweras a percent o per capita GDP

ROA

Port olio quality is good over time, as mea-sured by port olio at risk (PAR).

Persistent arrears are a sign of manage -ment complacency and poor credit culture.

Portfolio trends should be consistent. Theanalyst should watch or signs o abruptchanges in per ormance.

As a general rule, MFIs that have totalloans more than 30 days in arrearsthat are greater than 10 percent o theport olio, or annual loan losses greaterthan 5 percent, are unlikely to maintaindelinquency at sustainable levels.

The e ective port olio yield is consistent withexpected yields.

Yield gap analysis is a quick check on hid -den non-per orming loans in the port olio.

Is the institution growing at a sustainablerate?

Rapid growth inevitably leads to stress onsystems and collection per ormance.Growth stress is typically evident in theseindicators:

Portfolio quality

Return on portfolio

Number of clients or disbursements

Number of clients per credit of cer

Staff turnover

Client turnover

Is the MFI acing liquidity constraints thata ect its ability to disburse loans? Signi cant restriction in credit disburse -ments breaks implicit contracts and isusually associated with deterioration o credit discipline.


41/58


Table 1.4

Tier I Review: Credit Policies and ProceduresKEY ISSUES APPRAISAL GUIDELINESAre loan appraisal and approval criteria andprocedures clearly documented and adequateto ensure consistent loan analysis?

Guidelines may need to be speci c to dif -erent loan products.

Do the application and approval orms refectthe loan analysis and approval procedures?

Forms should be comprehensive and orga -nized to facilitate ef cient loan processing.

Is the approval process structured with ad-equate levels o authority and control?

The analyst should be able to diagram thedi erent levels o approval authority andreadily identi y the control mechanisms.

Authority levels should be appropriate toloan amounts and related risk exposure.

The system should reinforce the account -ability of the credit of cer who recom -mends loans or approval.

How are loan restructuring and renegotiationde ned and treated?

The de nition should include all loans thathave had their original terms altered eitherin amount, interest rate, or amortizationperiod.

Check to see if the MFI allows new loansto a client whose prior delinquent loanshave not been paid o .

What is the institutions policy regardingmultiple, concurrent loans to an individual

client?

Multiple lending that is not clearlyreported to show the total borrowing by

an individual client is a recipe or highde ault.Does the MFI have clear policies about thesize o loans and loan payments relative tothe clients ability to pay?

Lack of analysis of client repayment capac -ity will typically lead to eventual repay-ment problems.

What is the rate o increase in loan amountsand average size o payments or repeat bor-rowers?

Increases should be incremental. The his -tory o arrears will indicate whether theincrements are appropriate.

How big are any obligatory savings require-ments relative to loan amounts?

Obligatory savings reduce the effectiveloan amount and at some level lead tohigh client turnover.


42/58


43/58


Table 1.6

Tier I Review: Portfolio Monitoring and ControlKEY ISSUES APPRAISAL GUIDELINESDoes senior management set targets andmonitor loan delinquency on a routine basis?

This will be refected in widespread under-standing o the targets, requent delinquen-cy reports, and well-de ned action plansor the di erent levels o sta involved incredit operations.

Can senior management clearly explain thecauses and cycles o arrears?

The analyst should make the subjectiveassessment o whether managementsunderstanding o loan per ormance is sup-ported by port olio reports and the viewso operational sta .

Do the port olio reports show per ormanceby loan product, branch, and credit of cer?

It is important that management and sta can track port olio per ormance by thesekey actors. Without this breakdown,management is not able to identi y thesources o strength or weakness in creditoperations.

Does the institution track payment per or-mance in ways that help it understand clientbehavior?

This capacity will also be refected in loanappraisal methods and lending productsthat are speci c to client groups.

Do sta incentives include loan port olioquality as a primary per ormance marker.

Sta who are not held accountable willnot produce optimal results over time.Incentive systems that reward growthwithout quality will encourage poor loan

selection and lead to poor per ormance.What are theex-post internal controls em-ployed by the institution, independent o theoperations department, that review loan ap-provals, scrutinize problem credits, evaluateprovision coverage, look at trends, and citedocumentation exceptions?

In a maturing MFI, internal control sys-tems must be robust and well supported bymanagement and the board o directors.

Who handles collection o late payments, theoriginal loan of cers or a specialist?

The analyst needs to determine i the creditof cer has suf cient incentive to ensurethat his or her loans are per orming.

At what point is a loan sent or collectionthrough the legal system? How e ective is

this remedy and how long does it usuallytake to collect?

The legal remedy must be ef cient andcost-e ective. I not, problems will likely

accumulate over time and result in largeperiodic write-o s.Has raud played a role in losses? Fraud control and remedies should be

robust. The analyst should ask about pastcases.


44/58


Table 1.7

Agent Banking Policies and ProceduresIn cases where the MFI or its sta have outsourced some aspects o port oliomanagement to a third-party agent or agents, the analyst is advised to ollowa three step appraisal process.

1. Identi y the precise role o the agent.A simple mapping exercise can be use ul to clari y the respective roles o the MFI and the agent in the di erent levels o port olio management:

Client identification and recruitment;

Loan appraisal; Loan approval; Loan disbursement and payment collection; Post disbursement client management; Portfolio information management.

2. Identi y the risks associated with the agents role.The analyst will need to consider how any o the ollowing risks mightmaterialize in the agents role:

Poor agent performance; Inadequate MFI oversight of agent performance; Fraud; Customer abuse, including collecting unauthorized commissions; Lack of customer loyalty to the MFI; and, Cash handling risks (theft, loss, miscounting, etc).

3. Assess the mechanisms the MFI employs to manage the risks. The ol-lowing table summarizes the typical mechanisms.


45/58


MECHANISM APPRAISAL GUIDELINESDoes the MFI have a clear legal contract withthe agent?

The MFI should have a contract or someother agreement document that de nes allimportant aspects o the agents role, respon-sibilities, and compensation.Does the arrangement generate appropriateincentives or all three parties the clients,the agent and the MFI to act in the MFIsinterest?

Are measures in place to ensure that the cli-ent repays and cannot access uture servicesin case o de ault?Does the agent have incentive to select goodclients and keep those clients loyal to theMFI?Do MFI sta have adequate incentive tomonitor agent per ormance?

Is the MFI able to de ne clear operationalguidelines or the agent role?

The MFI must have an appropriate levelo say over the operating procedures o theagent (e.g., eligibility screening, appraisalcriteria, collection procedures, etc.)

Does the MFI have adequate control mecha-nisms to audit agent per ormance?

This will vary depending on the agentsrole. I the agent is conducting appraisals ormanaging clients, then the MFI should beable to review les and verify the existence ofthe clients.The MFI needs ways to check whether itssta are using external agents ( or instance,group organizers) without its knowledge orin violation o its policiesThe MFI needs to be able to determinewhether its agents are also managing clientsor competing lenders, a situation posing seri-ous confict-o -interest risk.

Does the MFI have adequate reports to moni-tor agent activities?

The MFI should have access to reports onkey agent activities. It may also be releventor the MFI to generate special port olio re-ports or all loans associated with the agent.


46/58


Table 2.1

Tier II Review: MIS Review

Note: This table guides the comparison o the results rom the analysts MISspreadsheet template and the MIS loan report.

KEY ISSUES APPRAISAL GUIDELINESDoes the MIS accurately classi y individualloan principal balances by arrears status?

This is primarily to a test to see i the MISinternal calculations on arrears are accurate.

Does the MIS accurately assign individualloans to the correct provisioning category

Check to see that the classi cation of indi -vidual loans is consistent with MFI policy.

Do individual MIS loan accounts accuratelyrefect the contractual conditions o the loan

contract?

Using the spreadsheet template, the analystwill veri y consistency in the ollowing loan

conditions:loan amortizationinterest calculationeesdistribution o payments between principal,interest and ees

Do individual MIS loan accounts accuratelyrefect MFI lending guidelines and accountingpolicies?

Does the system generate a schedule or eachclient that compares actual payment datesand amounts to the original amortizationschedule? And does the system generate aconsolidated actual vs. scheduled repaymentreport?

These reports are necessary or trackingclient and port olio per ormance against theexpected repayment schedules.This in ormation is necessary or per orming

the yield gap analysis.Are actual disbursements and payments ac-curately entered into the MIS?

This comparison requires a con rmationo the dates on the physical payment anddisbursement vouchers.

Table 2.2

Tier II Review: Loan ClassificationKEY ISSUES APPRAISAL GUIDELINESDoes the MFI have clearly de ned criteria forassigning loans to risk categories?

Some MFIs may consider other actors inaddition to arrears.

Are risk categories applied consistently toindividual loans?

The analyst can test or this on the small TierII sample o loans; robust assessment willrequire Tier III analysis.

Are provisioning levels consistent with his-toric write-o activity

This will involve calculating write-o s as apercentage o port olio and tracking the evo-lution o risk category balances over time.I the MFI has not done such an analysis, areprovisioning levels appropriately conservativein relation to international benchmarks?


47/58


Table 2.3

Tier II Review: Branch Office ManagementKEY ISSUES APPRAISAL GUIDELINESDoes the distribution o decision makingauthority and operational responsibilitybetween headquarters and the branch of cesfacilitate ef cient and controlled operations?

Speci c areas of importance:

determination o lending targetsloan approvalsdisbursement and payment cash managementdata entry into loan tracking systemaccounting reconciliation

Does the MFI have a robust system ormanaging the performance of branch of ceoperations?

Speci c areas of importance:

determination o lending and delinquencytargets

per ormance data collectionper ormance data reportingper ormance reviewollow up actions or good and poor per or-mance

Does the MFI adequately manage the coreunctions related to port olio quality?

The analyst should assess management andsta understanding and commitment toloan analysisdelinquency management

Are branch of ce practices consistent withheadquarters policy and expectations? Check or discrepancies or disagreementsaround policies

Are there signi cant differences betweenbranch of ces?

Is sta morale positive? Check branch of ce staff attitudes aboutheadquarter policies and management;Note level o sta enthusiasm or reachingper ormance targets


48/58


Table 2.4

Tier II Review: Loan OfficersKEY ISSUES APPRAISAL GUIDELINES

Are loan of cers adequately trained? What is the typical education and employ-ment background of a credit of cer?

Are the MFI training programs adequate?Do loan of cers have an accurate under -standing o policies related to each lendingproduct?

Lack o clarity on lending products shouldalert the analyst to the possibility o wide-spread problems in the port olio.

Do loan of cers speak knowledgably aboutthe credit risk o their clientele and how toevaluate and control those risks?

This can be observed in discussions abouthow the credit of cer uses personal judg -ment during loan analysis and dealing withdelinquent clients.

Do loan of cers receive frequent reports ontheir port olio per ormance?

Report requency should be daily or weeklyto keep the credit of cer up to date on pend -ing loan applications, arrears status, andpriority cases.

Do loan of cers follow vigorous proceduresor dealing with delinquent loans?

Credit of cers should be very clear aboutprocedures and remedies and show dedica-tion to their implementation.Policies should speci y how quickly a delin-quent borrower must be contacted, and bywhom.

How often do credit of cers see clients, andis this in the of ce or at their place of busi -

ness?

In general, credit of cers should be spendingthe majority o their time with clients at their

place o business, or at their group meetingsite. Any deviation rom this norm should bewell justi ed.


49/58


Table 2.5

Tier II Review: Loan Files

Note: This table provides general guidelines or interpreting the results o theprocedural audit on the loan iles.

KEY ISSUES APPRAISAL GUIDELINES

Does the MFIs loan le template adequatelydocument the essential components o loanappraisal, approval and management?

This is ultimately a judgment about whetherthe les have the appropriate level of docu -mentation.

Are the loan les complete? This is a question of whether the loan lesactually contain the documents they aresupposed to. A complete assessment is onlypossible in Tier III.

Do deviations from policy or incomplete lesconstitute a material de ciency?

Typically, the level o compliance will be con-sistent with the analysts other ndings abouthow well the MFI conducts business accord-ing to

Documents

DueDiligence TechGuide ENG