Ghana - Non-Bank Financial Institutions Assistance Project

Documentof The World Bank FOROFF ICIAL USEON LY Report No: 25213 IMPLEMENTATION COMPLETION REPORT (IDA-27920) ONA CREDIT INTH E AMOUNT OF SDR1 6.OMIL LION (US$ 23.9 MILLIONEQ UIVALENT) TO THE REPUBLIC OFGHANA FOR A NON-BANK FINANCIAL INSTITUTIONSASS ISTANCEPRO JECT DECEMBER 15, 2002 Financial SectorD ivision AfricaR egion This docunenth as a restricted distribution and mnayb e usedb y recipients onlyi nth e performance of their officiald uties. I ts contents mayn ot otherwise bed isclosed withoutW orld Banka uthorization.  CURRENCYEQ UIVALENTS (Exchange RateE ffective June 30, 2002) Currency Unit = Cedi GHC80 50 = US$ 1:00 US$ 0.0001242 = GHC 1:00 FISCAL YEAR Januaryl December3 1 ABBREVIATIONS ANDA CRONYMS ARB Association of Rural Banks BOG Bank of Ghana CAS Country AssistanceStr ategy CSD Clearing,Sett lement & DepositorySy stem FINSAC FinancialSe ctor Adjustment Credit GHAMFIN Ghana Micro-Finance Institutions Network GSE Ghana Stock Exchange GRC Ghana Reinsurance Corporation HFC Home Finance Company Limited ICAG Instituteof Cha rteredA ccountants Ghana IDA Intemational Development Association MDPI Management Developmenta ndPr oductivityl nstitute MIS Management Information System MOF Ministryof Finance MEI Micro-Financel nstitution NBFI Non-Bank FinancialI nstitution NIC Nationall nsurance Commission PT PrivatizationT rust PCT Project Coordination Team RTGS Real Time Gross Settlement SEC Securities & ExchangeCo mmission SIC Statel nsurance Company SOA Schoolof Administration SOE StateO wned Enterprise SSNIT SocialSecurity& Nationall nsuranceT rust Vice President: Callisto E. Madavo Country Manager/Director: MatsKa rlsson SectorMa nager/Director: Gerard A. Byam Task TeamLe ader/TaskMa nager: Kofi Boateng Agyen  GHANA NON-BANK FINANCIAL INSTITUTIONSASSIST ANCEPR OJECT CONTENTS PageN o. 1. Project Data 1 2 P rincipalPe rformance Ratings 1 3. Assessmentof Development Objectiveand Design,and of Quality atE ntry 2 4 Achievement of Objective and Outputs 6 5. Major Factors Affecting Implementation and Outcome 15 6. Sustainability 17 7. Banka ndBo rrower Performance 18 8.Less ons Learned 20 9. Partner Comments 21 10 Additional Information 21 Annexl. Key Performance Indicators/LogFr ame Matnx 22 Annex 2. Project Costsa nd Financing 31 Annex3. Economic Costs and Benefits 33 Annex 4. Bankl nputs 34 Annex5. RatingsforAch ievement ofObjectives/OutputsofCo mponents 36 Annex6. Ratings of Bank and BorrowerPer forrnance 37 Annex 7. L ist of Supporting Documents 38 Annex 8. Sum mary of GOG'scom pletion report 39   Project ID: P000943 Project Name NON-BANKIFI NINS AST Team Leader. Kofi-BoatengAg yen TL Unit AFTPS ICR Type: Core ICR Report Date: December 23, 2002 1. Project Data Name: NON-BANKFI NINS AST L/C/TFNumnber. IDA-27920 Country/Department: GHANA Region* AfricaRe gionalOffi ce Sector/subsector- General financesec tor (63%); Central government administration (28%);Comp ulsory pensiona nd unemployment insurance (6%); Tertiary education (3%) KEY DATES Original Revised/Actnal PCD: 03/22/1995 Effective: 05/26/1996 08/16/1996 Appraisal: 11/06/1995 MTR. 07/31/1998 11/11/1998 Approval- 12/05/1995 Closing: 06/30/2000 06/30/2002 Borrower/lmplementing Agency: Government of Ghana/Ministry of Finance Other Partners: STAFF Current At Appraisal Vice President: CallistoE. Madavo Edward Jaycox Country Manager: Mats Karlsson K. K.Fr amji Sector Manager: GerardA. Byam Peter Watson Team Leader at ICR: Kofi-BoatengAg yen AnilC handramani ICR Primary Author: Ravi J.Ru parel 2. Principal Performance Ratings (HS=Highly Satisfactory,S= Satisfactory,U =Unsatisfactory,H L=Highly Likely,L =Likely, UN=Unlikely,H UN=Highly Unlikely, HU=Highly Unsatisfactory,H= High, SU=Substantial, M=Modest, N=Negligible) Outcome: S Sustainability: L Institutional Development Impact: SU Bank Performance S Borrower Perfornmance. S QAG (if available) ICR Quality at Entry: U Project at Risk at Any Time: Yes  3. Assessment of Development Objective and Design, and of Quality at Entry 3.1 Original Objective: The Ghana Non-Bank Financial Institutions Project or "NBFI Project" (Credit No. 2792-GH) was approved by the Board on December 5, 1995. It was designed to build on previous projects in the financial sector namely the two Financial Sector Adjustment Credits (FINSAC 1, 1988 and FINSAC 11, 1991) and the Rural Finance Project (1989). FINSAC I and 11 had supported the development of the banking structure by improving the regulatory framework, building capacity at the central bank (Bank of Ghana), restructuring distressed banks, and divesting public ownership in banks. The Rural Finance project had supported the financial restructunng of rural banks, strengthening the Association of Rural Banks, and strengthening the Rural Finance Inspection Department (RFID) at the Bank of Ghana. Despite the achievements of the earlier projects, much remained to be done in terms of strengthening and implementing the regulatory framework to ensure that the regulations were applied rigorously enough to secure the ongoing soundness and improving efficiency of the Financial Sector. Along with this came key investments in financial infrastrure, most notably the envisioned improvements to the National Payments System. The NBFI Project was consistent with the Country Assistance Strategy (CAS) discussed by the Board in May 1995. The project was intended to directly support two of the five key elements of the Bank Group's strategy outlined in the CAS: private sector development and capacity building. Each component of the project was targeted at supporting greater pnvate participation in the development of the non-bank financial sector and in building local capacityin the sector. The design of theNBF IPro jectwaslarg ely influenced byth e Ghana Financial Sector Report (GH-I 3423 of Decemberl994) whichhadcon cluded that: * Ghanahadno t achieved muchsuccess in themo bilization and allocation of domestic and foreignprivate capital, whicha re key factorsi n stimulatingprivate sec tor-led growth; * One factor that had contributed significantly to the low formal savings and investment rates was poor financial intermediation in the economy; * Effective financial intermediationwou ldr equire:(a ) the promotion of greater competitiona ndmo re innovation int he sector( including divestiture and wellsup ervisedentr y and exit);( b)t he strengthening ofthe financialandl egal infrastructure( including human resources and paymentsystems) ;a nd (c) the integrationof t hef ormal financial and informal financialsec tors; * While previous IDA operations had gone a long way in addressing problems in the banking sector, issues related to the non-bank financial sector hadnotbeen addressed. TheNB Flcre dith ad two objectives: 1. To promote the growth of an efficient, competitive, well regulated, non-bank financial sector and an efficient payments systemin orderto: a. Increase the mobilization of domestic savings inbo thf ormal and informalsectors; b. Facilitate the flow of internationalc apital into Ghanat hroughbet ter financial intermediations;a nd c. Supply financialse rvices competitively to the economy. 2. Tocr eate anenv ironment conducivetogreater privatesecto ractivi tyand supportthesu ccessful implementation ofthe divestiture program for state-owned enterprises (SOE) and the three largest state-owned commercial banks. Given that different institutions in the non-bank financial sector were at different stages of development, specific interventions were designed in each sub-sector. This approach was reflected in GOG's strategy for the development of the non-bank financial sector which was detailed in a letter of sector strategy prepared by the MOF in October 1995. - -2 -  For eachben eficiary, thepro ject was intended to support one ormo re of thefol lowingactiv ities: a. Legislative reforms b. Regulatory reforms andstr engthening of regulatory capacity c. Organization structuring andi nstitution building d. Building local capacity through training e. Marketa ndprodu ct development f. Divestitureof GOG shareholding 3.2 Revised Objective: Thepro ject became effective on August 17, 1996af tera delayof 3 months.Theinit ial duration ofthecreditwas 4 years withanexp ected completiondate ofJu ne30 ,20 00. Theprojectwentthrou gh two major changes between l9 96 and20 00: * Duringth emi dter m review, in Novemberl 9 98,performan ce indicatorsf or most ofthecom ponentswere more clearly defined, someco mponents were scaled back, ando nene w component was added. * In the periodf rom March- July 2000,three componentsw ere cancelled andthe decision was taken to extend thepro jectfo r another two years for some of theco mponents. During both these times of change, componentob jectiveswere revised andou tcome indicators weresp ecifiedor revised, in eachcase to better align projectactiv itiesto the primary development objective, tobet ter takeacco unto f complementary activities in otherprojrec ts, most notably divestituturebeing carried out under PEPTA, and in response to thec omplex polictical environment which mitigated against implementationof some ofthe original activities The second objective had been stated as: "To createan environmentco nducive to greaterpriv atesec toractiv ity and support the successfulimplem entation of thediv estiture programf orstate-owned enterprises (SOE) andthe three largest state-owned commercial banks." Thediv estiture aspectof thisob jective becamer edundantear ly in the projectas the Government decidedtochan nel its restructuring and divestitureef forts throughthe PEPTApr oject. The log framematri x developedduring the mid term review workshop makes no mention of this objective andit does not appear that this objective was ever seriously pursued. Itseem s thata decision wastake n to supportdiv estitureof SOEs andstateown ed commercial banks with another Bankc redit - theP ublicE nterpnses and Privatization Technical Assistance (PEPTA) credit which wasapp roved in 1996. The firstob jectivehad been stated as: "To promote the growthofanef ficient, competitive,wellr egulated, non-bank financialsector and an efficient payments system in orderto: a. Increase the mobilization of domestic savings inbo thf ormal and informalsectors; b. Facilitate the flow of internationalc apital into Ghanat hroughbet ter financial intermediations;a nd c Supply financial services competitively toth e economy." This objectivewasthepredo minanton e andf ormedthebas is of the logf ramematrix atthemid termr eview workshop.Howev er, during the course of projectimp lementation, there was a continuouspro cessof re-aligning priorities andproject activities to best supportth e achievement of this objective. The FSAP mission of July 2000 provided clearguid anceonf inancail sector priorities. Boththe Government and Bank teams used this input from theFSAP to focusth e project towards progressin these key priority areasov erth e lasttwo years of thepro ject. 3 3 Original Components: Initially the project had 14 components. Each componenthadsever alsub -componentswit h ther esult thatthe technicalannex of the appraisal documents defined 45separate sub components. Thel4 components were tobeimp lementedby 11 beneficiaries andth e MOF. Thel I beneficiariesincl udedth e - 3-  Bank of Ghana (BOG), the National Insurance Commission (NIC), the Securities and ExchangeComm ission (SEC), the Ghana StockE xchange (GSE),the Home Finance Company (HFC), theSocial Security and National InsuranceT rust( SSNIT), theScho ol ofAdmin istration( SOA) at theUniversityof Ghana, the Instituteof Chartered Accountantsof Ghana (ICAG), the Management Developmentand Productivityl nstitute (MDPI), the Statel nsurance Company (SIC), and the Ghana Reinsurance Company( GRC). The MOF was designatedthe implementing agency for the project. APr oject Coordinating and Implementation Team (PCT)cons isting ofa n advisora ndon esup portstaf f wasestablish ed atM OFund erthe chairmanship ofthe Deputy Minister of Finance. While each beneficiaryhad the primary responsibility for implementation of itsown components, the PCTw asr esponsible for coordinating and monitoringthe implementation ofthe project. The different componentsa ndt he proposed areasof technicala ssistance are indicated in Tablel. Tablel: Original Project ComponentsandTechni calAssis tancearea s Component Legislative Strengthening Building Building Market Divestiture Reform regulatory& Institutional local and supervisory Capacity capacity Product capacity Devt. 1 BOG - Payment Systems X X x 2 BOG- NBFI Supervision X X X 3 NIC X X X X 4 SEC X X X 5 GSE X X X 6 Micro / Rural Finance X X x 7 HFC X X 8 SSNIT X X 9 SOA X 10 ICAG X 11 MDPI X 12 SIC X X 13 GRC X X 14 Pnvatization Trust X 3.4 Revised Components: As mentionede arlier, theprojec tw asr estructured during the mid term reviewi n 1998a nd before the extensionin 2000. MidT erm Review: At the time ofthe mid term reviewon ly$4 .6mill ion (I 8.6%)of the total $24.9millio nhadbeen disbursedor was committed. During the mid-term review thebud getof themicro and rural finance componentwa s reduced from $3.75 millionto$80 0,000. Furthermore, asper the GOG's request, $1.47 million was reallocated to BOG to finance some of the activities which were originally scheduled tobe funded under thePnvate Enterprise and Export Development Project (PEED) which was closed in March 1998. Wh ile these funds wereus edtosu pport various departments in BOG,on ly the activitiesr elated tothe BankingSup ervision Department (BSD) were treated as a separate component. Project Extension: The periodbet ween January - June 2000 was a crucialon e for thepro ject Thefoll owing events tookpl ace: * Three nonper forming components - SIC, GRC, andthe PrivatizationT rust - were cancelled; * Therewa s partial extensionof three components (GSE,M icro/RuralFinance and HFC) inorder tof acilitate the orderly completionof ongoing activities and contract commitments; -4 -  * Thepro ject was refocusedwith an emphasiso nstreng theningr egulatorycapacity a nd improvingtheeff iciency of thenat ional paymentssy stem. A decision wasma deto continue NBFIsu pportfortwo years (up to June 2002)forth e four regulatorycom ponents(BOG/ BSD, BOG/NBFI, SEC,an dNIC) and the paymentsy stems component. Eachof the institutions wereasked to prepare impact-oriented businessplan s asa conditionof extension; * There had beendelay s in the procurement of theauto mated clearing,settlemen t anddep ositorysys tem( CSD) forth e GSE. A decision was made to also extendthi s component to allowf or furtherrevie w ofth isiss ue; * Atthetim e of projectexte nsion the Bank suggested to theon going beneficiaries of the project to includea program of HIVW AlDsawar eness in theirwork programs; * The transitionto asmaller, extendedprog ram of activities reduced the need for ac ontracted project management team. The MOFdecided to mainstream projectmanagemen twi thin theM OF at thee xpiration of the Project Coordinator'sc ontractin August 2000. FSAP Recommendations: Anotherkey event in the project wastheFinan cial SectorAssess ment Program( FSAP)miss ion in July20 00.T he FSAP waspart of a joint World Bank-IMF program of studies that aims at assessing vulnerabilities off inancial systems, their developmentneeds, and implementationof r elevant financial standards TheFSAPcov ereda numberof institutions/activities supportedby the NBFI project. And althoughthefinalFSAP report wasno t issuedun til May 2001, recommendationsf romther eportwere discussed with the beneficiaries dunng the March20 01 supervision mission. T he beneficianes for the four components affected( BOG/BSD, BOG/NBFI, SEC, and NIC) were asked to commenton the FSAPrecom mendations and provide their proposed actions toimplem ent ther ecommendations. Followupon ther ecommendationsthenbecame an integral part of the remaining supervision missions. 3.5 Quality at Entry: Quality of Entryforth is project is assessed as unsatisfactory. There are three main reasons fort hisa ssessment * Inconsistency between project objectives andco mponents:As previously discussed the secondob Jective didno t have anyc omponents relatedtoit. Furthermorethere were noperf ormanceind icatorsspecifiedfor the sub-objectivesfo r thefirs t component. * ComplicatedDesig n: The large number of componentsmean t thattheprojectwas unable to go asdeep as it could have, or should have,i n providing support to the key components and institutions. Furthermore having a design with a number ofsm all unrelated componentsmean t that there waslikely to bedifficultiesinproject coordination andmonit onng. * Incomplete Design Team:T heprojec tdesign included a largemicr o andr ural finance componentwi th a budget of $3.75 million. However none ofthe preparation or appraisal teams includedindiv iduals with expertisei n this area. I n factthiscom ponentw asno t designed until thesecon dsup ervisionmiss ioninM ay 1997. -5-  4. Achievement of Objective and Outputs 4.1 Outcome/achievement of objective: Assessment of outcomesan d achievement of objective issi ightly hindered by theweak nesses inth ede finition of objectives asdis cussede arlier. Takingacc ount of the fact that thepriv atization agendawas handledby aseperate project(PEP TA), the key focus ofth ep roject becme the other Develeopment Objective, i.e. "To promotethe growth of an efficient,c ompetitive, well regulated, non-bankfinancial sector and an efficientpaym entssy stem." Whenass essed against this objective, theo verall projecto utcomeis satisfactory.Thep rojecth asclearly contributedto thestr engtheningof regulatory agencies for both thebankin g and non-banking financialsec tor. The project has also contributed toimp roving the efficiencyof various institutions andhasprov ided the infrastructure for an efficientpay ments system Mos tofth e components have metth eir relevant objectivesan dare expectedto achieve satisfactoryd evelopmentres ults as discussed below. It isalsoworthno ting thatdu ringthe last two years of theprojectthe Ghanaianecon omy wasr ecovering from a majormac roeconomic shock (discussedmo re in section 5.1). Therefore thefact thatth e project wasab le to achieve many key outcomes,pro vides further supportf or a satisfactory rating. The keyou tcomes were achievedpart lyasa resultofthecan cellation of three components and the focusing of time, energies,a ndf inancialr esources tothe other components. The fact thatthe Bank and the GOG were ableto implementt his cancellation and restructuring andt her esulting achievement of outcomes( at a lower cost than budgeted)f urthersup portsthe satisfactoryr ating. 4.2 Outputs by components: Theratin gs ofou tcomes bycom ponentare summarized in Table 2 and discussed in detail below 1.BO C-Pay mentsSy stem The objective of thisco mponent was theim provement of the financialin frastructure. The keyexp ectedben efits werethe reductionof cash dominance inthee conomyand animpro vement inthe clearingtimesf orchecks. The key outcomes of this component include: * TheRealT ime Gross Settlement System (RTGS)for High Value FundsT ransfer was launchedon Octoberl I, 2002. At the time of ICR preparation, the RTGS, knowna s the Ghanal nterbank System, (GIS) hadbeen operatingf or 30 days without problems. * Computerized settlement of interbankc learing istakin gplace for the Accrar egion. BOG decided notto undertake computerizationf or otherregio ns because of the low volume of transactions. * Major improvements weremad einthe cleanng times for payments. At theinceptio nof theprojectc learing times were four days for local clearingand upto2 I daysfor outstation clearing. The originaltargets outcomes were2day s forlo calpay mentsand 4day sforou tstation payments. Thesetarg etswere fou ndto be unrealistic and were revised to 3 daysf or localpay ments ( within inner zones); 5 days within thesam eou ter zone and betweeninn er zones; and8day sfor remotelo cations. Asu rvey of clearingtimes in Jan20 02,un dertakenby BOG,sho wedthat 96%of paymentsc learedwithin thedefined standards. * There duction of cash dominanceintheecon omy wastobemeas ured by changesin Cash/M2. The target was to reduce this ratio from 38%to 20%. During the project the Cash/M2 ratiof ell to 24% by September19 99 but then went up to 36% by December 2000. Since then ithas been gradually fallingandr eached24 % in September20 02. Theov erallou tcomeof this component is rated satisfactory. Whilepro gress wasslo win theearly ye arsofth e projectand was then adversely affectd by the 2000macro economic cnsis,thestatu s at closing was encouraging - 6 -  The primary unfinished activity islegislativ er eform. At the timeof thel CR mission, the BillsandCheck s Bill and the Payment Systems billhad beendrafted and reviewedand were expectedtobe put before parliament during December 2002.The two bills, once approved, should addressmo stof the remainingcon cerns raised in the FSAP report. Table2: Revised ProjectCo mponents andRa ting End Date Rating before Rating May Rating MTR 2000 Final I BOG - Payment Systems June 2002 S S S 2 BOG- NBFI Supervision June 2002 U S PS 3 BOG - Bank Supervision June 2002 n/a n/a S 4 N1C June 2002 S S S 5 SEC June 2002 U S S 6 GSE June 2002 S_ S S 7 HIV/AIDs awareness June 2002 n/a n/a S 7 MicroandRura I Fmance December 2000 U S S 8 HFC December 2000 S S S 9 SSNIT June 2000 S U 10 SOA June 2000 S S 11 ICAG June 2000 S S 12 MDPI June 2000 S n/a 13 SIC Cancelled U n/a 14 GRC Cancelled U n/a 15 Pnvatization Trust Cancelled n/a n/a 2.Ba nk of Ghana - NBFISup ervision The NBFI sector is made upof 10 categories of institutionssu pervised by BoG, some of which are regulated under theFinancial Institutions( Non-Banking) Law of 1993.Cr editUnio nsa ndt he mortgage company,H omeFi nance Company (HFC ) are regulatedby separate laws. Leasing companies aresu bjectto the Finance Lease Lawof 1993. Onlysavin gs and loans institutions, building societies,andcr editunio ns are authorizedto takedepo sits fromthe public.The other NBFIsseektheir funding from banksand other wholesale marketparties.T he NBFI sectoris relatively small, with total assets of under5 percentof total commercial bank assets. At the timeof projectapp raisal the BOG wasintheproces s of establishingtheNBFI supervisiondep artment to regulate and supervisein stitutions coveredby thelaw. Key outcomes of this componentin clude: * TheNBFI supervision departmentis fullyop erational with a staff of 51 anda functional MIS. * BOG established two types of Business Rules for NBFIs:on e set for Deposit Takinglns titutions andano ther for Non-DepositT akingl nstitutions. * The department hassho wn significantimprov ement in its supervision role. 29 operational NBFIs (100% ofa II operational NBFIs) were subjectedtoon -siteins pection during 2001 and 9 were found to be in non-compliance.Using the newbu siness rules, the Department began to levy penaltiesf or non-compliance. The NBFldepartmen t hassho wna modestincrea seinenforce mentmeasures.Dunng 2001 a total of25f ines were levied on NBFIs for non-compliance withr eportingr equirements. TheFS AP report madea numb er of recommendations regarding licensing rules, and investment and lending restrictions Thenew businessr uleshav e takenint o accountmo stof the recommtendationsmad e byth e FSAP report, except int wo areas: (i) defining a time table fori nsolventinstitutionsto raise fresh andstr ict capital; and (ii) strengthening enforcement of prudentialr egulations. - 7 -  Theov erallou tcomeo f thisact ivityis partally satisfactory. T he last implementation supportmiss ion had indicatedthat this rating wouldbe upgradedto " satisfactory" depending on the department'sprogress on applying sanctions against non-complianti nstitutions in a timely manner.F rom interviews dunng the ICRpr eparation, it appears that the BOG isst ill hesitant about strengtheninge nforcement of prudential regulationsa nd applying sanctions. 3 out of8S avingsandL oan companies were stillnotcomp liant withminim umca pitalr equirements.O f the 2 building societies thathadbeen previously given until June 30, 2002 tor estructure andr ecapitalize,onehad presenteda restructuring plan while the other had been given a revised deadline of Dec 2002. 3. Bank of Ghana - Banking Supervision Thisco mponent was addedju stbe foreth e mid -term review. So meacti vities which were scheduled tobefun ded under thePr ivateEnter priseand ExportDevelop mentCre dit weretr ansferred to theNBFI project atthe GOG's request. Keypro ject outcomes of this componentin clude. * A new Bankof G hanaBi 11 (which seeks to, among other things,enhance a nd strengthen the Central Bank's Independence)was approved in February 2002. * All 17 banks and 114 rural banks were inspectedduring 2001 .A sa tN ovember 2002 the annual inspection for alll7banksa nd 69 rural banks had been completed. * There was a marked improvement inoff sitesup ervision following theprov ision of IT equipment andtr aining financed under thec redit. * BOG has9 "relieving managers" whoar e sentt o rural bankst o assist managers,str engthen boards etc. * BoGhas established new minimum capital requirementsof 25billion Cedis for Ghanaian Banks, 50 billion Cedisf or ForeignBan ksand 70billio n Cedisf or Development Banks. These requirements are forlicens ing of new banks and arei ncluded inthe revisedBankin gBill. BOGhas givenall the existingbank s 3 years to attain the minimum capitalr equirements. The lastsupervisionmissionhad ratedthis componenta spar tially satisfactory. It had indicated that achievinga "satisfactory" rating would dependto alar ge extenton theability of theBOG's Banking Supervision Department (BSD) to improve its trackr ecord ofenfo rcingprud entialr egulations, andthepassag e ofpen dinglegis lation by Parliament. The overall ratingof this component at the timeof theICR is satisfactory. T he BOG is making strong efforts to meet itssup ervision targets. The BankingBillhasbeen draftedand reviewed and is thepro cess of being forwarded to Parliament. 4. National Insurance Commission When the project began,N IC already existed and hadr esponsibility for regulation, supervision, andl icensingi n the insuranceind ustry. Howev erN IC had not beenable to undertake its dutiesdu e to shortageof financial resources, office space, and qualifiedpro fessional staff. Whilein dustrypro vided the officesp ace, thepro ject providedthecom puters and training. Key outcomes of thiscom ponent include: * NIC is fully operationalwitha staff of 34, a functionalM ISand trained staff. * Draft manuals forins pectionhav e been preparedand havebeen usedfor inspections.NIC expects to have the manualsf inalized by December 2002. * In20 01,NIC completed offsite inspections for 16 of the 18ins urance companies. * For 2002,asat November, NIC hadcom pletedoffs ite reportsforl6 companies. 7 companies had been subject - 8-  to onsiteins pections. For 6 of these7 companies reports hadbeen completedand discussed with management andthe companies had beena sked to prepare remedial action plans. The NIC alsomnheri ted another activitythat had originallybeen designedas part oft he SIC component. T he existing SIC training center was tobe converted intoa I nsurance Industry Training Center( IITC) for the entire insuranceind ustry. However other private insurance companies wereno t keen on an industryc enter beingown ed and runby ani nsurance company. Therefore the decision wasta ken to have theN ICc onstructa ndownthe IITC. Theprojectpro vided70 % of thefun dingf or the IITC. Thebu ilding has beencom pleted andN IC expectsthe center to be operationalbyM arch 2003. NIC is planning togive a management contract forthe trainingc enter to apr ivate management training group. The lastsupervisionmissionhad ratedthis componenta spar tially satisfactory. It had indicated that achievinga "satisfactory" rating would depend to a large extent on theappro val of the pending legislation. The overall rating oftheNI C component at the time ofthel CR is satisfactory.Th eNIC h ass uccessfully implemented most of the activities it wassup posedto implement. Asf orthe outstandingl nsuranceBill and Insurance Legislations, theN IC andM OF haveunder takent hea ctivities underthei r control andthe le gislation is expected to bepu t forwardbeforePar liament shortly. The NICis alsomak ing strong effortstomeetits annual inspection targets. 5.Se curitiesan dEx changeCo mmission The objectiveof this component wastost rengthen the regulatoryan dsu pervisorycapa city of theSecu nties RegulatoryCo mmission (asit was then called). Itwas believed thata well supervised securities industry would improve the confidence and increasethe number of savers and investors. At project inceptionthe SRC had been legislated to supervise and regulate the securitiesmarket (by the Securities IndustryL aw, 1993) bu t hadno t yet beenest ablished asa separate agency The Govemor ofthe BOG wasex ercising the functions ofth e'SRC on a temporarybas is. Under the NBFI Project, the SEC wassetup asa separateagen cy andanind ependent Commissionerwas appointed. Key outcomes of this component include: * TheSEC is fully operationalwitha staff of 20 staff.The new IT system that allowsSE C to tracktradin g activity became operational in June 2002 * The Securities Industry (Amendment)A ct 590 was passed in 2000. The Regulations on Mutual Fundsa nd UnitTr usts( LI 1695) wereapp roved by Parliament in December20 01. T he Secutitiesand Exchange Regulations whichw illsp ellou t requirementsr elating to licensing, disclosure, transaction levies etc.we re to be presented toParliam entbefo reth e end of 2002. * SEC has prepared a compliancemanu al which is usedf orbo thon site and offsiteins pections. * Inboth 2001a nd2O 02 all14 entities wereinsp ected throughr eview of offsiter etums. During200 lon site inspections were carried out onl I of14 firms( 78%). During20 02 (up to November), onsiteinsp ection had been carried out on 8 entities. SEC's target for completion by year endwas 1Oenti ties ( 71%). The overallr ating ofthisc omponentis satisfactory. 6. GhanaStockEx change The objectivesof this component included the strengthening oftG SE's institutional capacity and improvement in its efficiency through the establishmentof a clearinga ndsettle ment system, and MIS, and links withoth er stock exchanges. T hec omponenta Iso aimedto buildl ocalc apacitybystr engthening GSE'str aining program and expandingtheown ership of securities by individuals. -9-  The keyou tcomeso f this componentare: * Efficiency improved as a result of installation of a newM IS with related hardware. Informationdissem ination improved through intemetconnections. * Settlementperio d wasredu ced to T+5 (from T+10); Failed transactions wereredu ced to below 1% (from 14%). * Formal links were established with 3 exchanges( Johannesburg, Nigeria, and Nairobi) The GSE hadi tsf irst cross-borderlisting inN ovember20 02, when a bank from The Gambialisted itshares on the GSE. * Thenu mber of individualsecurity holdersincreased to27 0,000f romthep re-project estimate of5, 000(lar gely due to thesuccess ful listingof Ashanti Goldfields and Ghana Commercial Bank) * The number of participants on securities training coursesa ndprofessio naldevelo pment workshops has more than doubled. In 2001 income fromtr aining courses andseminars comprised 13% of GSE'soperat ing income. The shortcomings of this componentinclud e: * The number oflisting s only increasedf rom 18 to 23 (farlower than thetar get of 40+) * The automatedcle anng, settlementand depositary systemwa s not installed. Theoriginalbud get for this systemwas $1.4m and the procurement processbeg an in 1999. Howevertherewere a number ofpro curement delayswith the result that byMay 2000 the procurementhad notbeen completed. Aspartof the deliberations regardingthe extensionof the project,t he Bank teamr e-assessedthisc omponent. The Bankte am initially questioned the appropriateness of thepar ticular CSD package beingp roposed forGS E,in v iew ofis sues related toe conomic viability and regional integration. The Bankte amf urther recommended thatt heCSD part oftheGSE com ponent be included undertheexten ded NBFI. Upon further discussions with GOG and subsequentr eviewofth e proposedpac kage (withassi stance fromth e Securities MarketsSpec ialist onth e FSAP team), the Bank team advisedt hat, inview of the slump int rading values andvolu mes,t he prevailing macroeconomic conditions, the financial position of GSE, and other technical issues, there was no longer an economicju stification for making an investmentofthetyp e and scalepro posed.T herefore the Bank teamwas unable to provide ano -objection fortheparti cularCSD package proposed. The Bank, however, continued dialogue with GSE and assisted theGSE to visit markets ofa simi lar make-up as Ghana's Subs equentto thesevisits,the GSEr evised their CSD plans and identifieda more appropnate package withco stest imates of aboutUS$ 400,000 Unfortunately by the time GSE completed theirrevi sed proposal the extendedpro ject had come to an endand GSE wasun able to accessNBF Ifun ds. The management of GSE are stillkeentoimp lementanauto matedCSD system and are in the process of identifying financingoptio ns.Withthe encouragement ofthe Bank,the GSE isa Iso in discussions withthe BOG aboutth e possibility ofco nfiguring a CSD system aroundth e RTGSsy stem procured by BOGun der this Credit. Thisc omponenti s ratedas saisfactory. 7. HIV/ AIDS Awareness This component wasadd ed atthe time ofprojectexten sion. The Bank encouraged BOG,NI C, and SECto consider developinga HIV/AIDs awareness campaigna nd tobu dget forthis activityin their revisedworkpro grams. The BOGun dertooka veryinte nsiveand successful campaign aimed atthe2, 600 staffof the BOG. BetweenApril 2001 and June20 02 theyimp lemented the following activities: - 10-  * Defined the objective for the component: toincreas e awareness and to preventth e spread of HIV/AIDs among the staff of the Bank by providing education and promoting safer sex. * Appointed a steering committee, and a implementationsec retariat.T heste ering committee included representatives from vanous staff associations. * Trained 133 Peer counselors to disseminate information and provide advice. * Provided information and educationin thef orm of lectures, seminars, workshops, filmsho ws, and drama. The secretariat alsocon tracted an outsidef inn to conducta baselinesu rvey of 11 00staff and a finalsu rvey of 800 staff. The key outcomes of thisco mponent are: * A large number of staffpart icipatedin theac tivities. 72% of ther espondents in the final survey had takenpar t in atleas t one activity. * Knowledge about HIV/AIDs, its mode oftrans mission, and preventionincreas edsig nificantly. * Voluntarytestin gwasenco uraged and 74%of thestaff who knew theirHIV status, gotto know their status duringt he intervention period. * Therewasso me success increatin ga moresu pportive environmentf or infected bankstaff without fear of stigmatizationor discrimination. * Staff reported satisfactionw ith theaw arenessca mpaign. 81%of ther espondentsin the final survey said that the program had influencedt heir sexualbehavior .A nd 59%of t her espondents said thatt hepr ogram had increasedtheir awarenessabo utHIV- related issues. UnfortunatelyBOG defined theirtarg etsb eforedo ing thebas elinesu rvey. Therefore an assessmentofou tcomes against targetsi s difficult. The NIC undertook a similar program, albeit on a smaller scale, from March-June 2002. All staff were involved in one or more activities. Outcome indicators were notdefined and were notmeasu red The SEC held a workshop on this activity. The overall outcomeofthi s activity is rated satisfactory. Both BOGa ndNI Cimp lemented thorough, well organized and committed camnpaignsat relatively lowcost. 8. Microand Rural Finance This componentwas not designedat the appraisalstag e. The originalob jective wasdefined vaguely asa diagnostic reviewtoexam ine expansionof financialserv ices andtheorig inal target outcomeof thiscom ponentwas to have an established pilot schemet argeting small saversa nd borrowers inur ban andr urala reas. This component went througha number of revisions andcanbestbeun derstood intwoparts:theGHAM FiNsub component andthe ARBAp exBa nk component. GHAMFIN: In Mayl99 7,du ring thesecond supervision mission, the microa nd rural financec omponent was better defined andi nvolved working with existing institutions T her evised progratmhad two main thrusts: * Institutional strengtheninga nd innovation-:thi si ncluded provision of trainingf or ruralmic rofinance institutions (RMFIs),best practice studies and creationof aninno vation fund. * Coordination and PolicyFormu lation: building consensus among stakeholders anddev elopmentof astrate gic framework.  A Microfinance Coordinator was appointed att he MOFa nd theprogramn wasl aunched.A ta roundthe samet ime the Bankhad been supporting theM icrofinanceAc tion Research Network undera trust funded program. This network wascon verted into the Ghana Micro Financelnst itutions Network (GHAMFIN). The NBFIPr oject provided funds to GHAMFIN toallowit to register asa professionalasso ciationand to establish a secretariatand equipa permanentoffic e.GHAMFIN'sactiv itiesdu ring thatperio d formso me of the key outcomes for this sub-component. They include: * A trainingneedsa ssessment, designof trainingpr ogram and provisionof trainingser vices to approximately 80 member MFls; * Design ofa t rainingf und and a capacitybuildin g fund; * Establishment of a resource center asa on e-stop information sourcef orM Fls,w hich the micro-finance unit in the Ministry ofFi nance has taken over itsopera tion. ARB ApexBank: This subcomponent was notpartof theorig inal design. Howevera detailed proposal and budget submittedby the Association of RuralBan ks (ARB)wasapp roved duringthe midterm review and the ARB became a beneficiary. The primary objectiveofthi scom ponent was tosetup an Apex Bankfor theruralba nks. The key outcomesofth is sub-componentwer ethattheoffice of theApex technicalsecretariat were set up andsen sitizationsem inarswer e conducted for Rural Bank managers.O perationalmanu als for the Apex Bank wereprod ucedanda new law for establishing thebank was drafted. The micro and ruralf inance component was partiallyexte ndedin 2000 inor der to fund the existing contractof the microfinancecoo rdinator. Eventually the component became partof theRural FinancialSer vices Project which becameeffe ctive in 2001. The overall rating for thiscomp onentis satisfactory. Althoughits design wasdelayed and implementationwas erratic it provided strong supportf orthe establishmentof GHAMFIN as an umbrellaorganization formicro and ruralf inance institutions. It also provided thekeyinitial support for the ARB Apex Bank which was granted a banking license in June 2001 and isnowf ully operational. 9.H omeFin anceCo mpany Theob jectiveofth isco mponentwas to improveHFC 's efficiency andim prove itscap acityto ori ginate new mortgages. Assistance provided included anew integrated MIS systemand awidear eanetwo rk tolink HFC's branchesto the head office. HFCwas also able toget assistance fromABN Amro undera twinn ing arrangement funded by the project. During the project the number of mortgages increased by1 1 7% (from 1, 676to3,6 39) far exceeding the target of 50%. Thisc omponent isr ated satisfactory. 10. SocialSecur ityan d National Insurance Trust The objectiveofth is componentwas to expand social security toth e informalse ctor by developing newpro ducts. Thetar get outcome was that5% (40,000) of the working population intheinfo rmal sector(80 0,000) wouldbe covered by June 2000. This componentisrated unsatisfactory.SS NIT did not develop any newp roducts for the informalsec tor and attempted tomar keti tse xisting products with ther esult thatonly500 0 individuals in the informal sector were covered. Theman agement of SSNIT lost interest in this effort andclos ed down the informal sectordepar tment. It isw orth notingthat although thecom ponentdid not meetits objectives, the Bank continued tousethe project and thepo ssibility ofr estored funding to carry on a dialogue, anchored inFSAP findings, encouraging SSNIT management to initiate seriousr estructuring efforts. As ar esultSSN IT prepared a Memorandumof Understandingbetweentheir Board andSe niorM anagement and subsequentlyGhanawas selected as oneof nine - 12 -  Sub-Saharan Afncan countrnes to be included ina WorldBankpensio n reforminitiati ve. 11. School of Administration,Univ ersity of Ghana The objective of this component was toexp andan dim prove the SOA'sexec utived evelopment program The project providedassi stance toun dertakeanin dustry assessment andtode velop new study materials. A portfolio of 30sp eciallydes igned courseswasd evelopedu nder thepro ject.Partic ipation inth e executive programin creased by 78% (from 42 inl9 98to 79 in 2000) exceeding the target of 60%. The assistance appears to haver esulted in sustainedbenefits as the number of participants increased further to 130 in 2001 andis expected to reach 350 for 2002. This component isr ated satisfactory. 12. Institute of Chartered Accountants The objective of this componentw as toexpan da ndimpro ve thel nstitute'st rainingprograms inorder to increase the number of qualified accountants. With the assistance provided under the projectthe ICAG wasabletodev elop new study modules Thenu mber of qualified accountants increased byl 2 % (from 770 in 1998 to 870 in 2000) exceeding the targetoflO %. Theimp rovementapp earsto besu stainable as the number furtherincreasedto95 Oby the end of200 1. This component is rated satisfactory. 13. Management Developmentan d Productivity Institute This small component comprised$50 ,000 to financea trainingsp ecialist to conduct acom prehensiver eview ofthe activitiesof the MDPI. T he review wasinten ded to startinNov ember 1998.T he mid term review mission team visited theM DPI and foundou t that the MDPI wasin the process of implementinga multi-year expanslonprog ram involving on-going massive investments in infrastructure. Based on those findings, the mission indicatedt hata comprehensiverev iewwo uld not provide any value added. 14. State Insurance Company The objectiveo f this component wasto create alev elp layingfield and a competitivean deffici ent industry structure byth e removal of SIC's monopolyon allstat e-ownedbu sinessesan dth e divestitureofSlCto private investors. T hec redital located $2.625 milliont o restructure and modernize SICpri ort o privatization. T he technicalass istancewassu spended becauseGOG decided to place SICo nthe marketim mediately followingthe effectivenesso f the NBFI project.Th erefore the componentwas canceled. TheGO Gproceededwith the planned privatization and bids were receivedan drev iewed.Ho wever theGOGfeltth at none of thebi ds were acceptable and theprivatiza tionprocess wasc anceled. Discussions with SIC indicatethat as at November 2002: * Although SIC stilllegally has the monopoly on allG OGbu siness, SICis choosing not to use this regulation andiswelco ming competitionon a"lev el playingfield ". NeverthelessSIC believes itsti 11 controls about 60% of the GOG and SOE business. * SIC has beenund ergoing a major restructuringto reduce its contingent liabilities andto improve its financial positionun der a new managementteam. This restructuringis expected tobe completedby theend of 2004 at which timeSI C should berea dy to restart the divestiture process. 15. GhanaRein suranceCo rporation The mainob jective of thisco mponent was therem oval of GRC's monopolyo nrein surancebu siness.Thi s wasto be doneby thee limination of the legalr equirement for 20%cessio n for allins urance policies. TheGRC had been initiallya llocatedU S$ 1.05million under the NBFI credit tosupp ort its effortsin maintaining its viability after it lost its monopoly position. GRCus ed part of this amount to develop a newMIS. In addition the credit provided financialsup port to prepare GRC for privatization. In November19 98theM oF - 13-  informed theBank that funds had already been provided toG RC for itsprivatizati onund erthe PEPTAproject and NBFIfu nds ($ 250,000) wouldno t be needed Furtherdiscussion s after that mdicated that theG OG was unlikely toel iminatet hele gair equirement for 20% cessionof all insurancepo licies to GRC. Discussions between the Bank and GOG stalled over this issueand furtherac tivitiesun derthiscomp onent wereputonho Id.I n20 00the GOG eventuallyr equested thatthis component be canceled Fromdiscuss ionswi th industrypar ticipants,the indicationsa re thatt he 20% cession rule hasbeenr evised to benefit allGhan aianreinsu rancecom panies and notjustGRC. However GRC still gets about 60-80% of the local reinsurancebu siness because it isf eltthat the only other domestic competitor,M ainstream Reinsurance,is not adequately capitalized. Thel argest partof the reinsurance business still goes to internatonalc ompanies such as MunichReand Swiss Re. 16. PrivatizationTrus t Theorig inal objective of thiscom ponentwas to createa privatization trust to foster publicplacem ent ofenterp rises to be divestedun derthediv estiture program. Thedes ign was that the GOGwou Id place blocksof minority shares ofth e publicen terpnsesbein g divested witha trustwhi chwou Id have a GOG nominated board but wouldbe managedby a private portfoliomanag er.The shares wereto be held in the trustf ora penod note xceeding three years and werethento be sold to( a) theperson nel of the public enterprise being privatized;( b)the Ghanaian public; and( c) institutional investors. An initial study was done and the recommendationswer er eviewedby the mid term reviewmis sionteam and discussed with the MOF. During the discussions the MOF put forward anotherpropo sal thatdiff ered in design fromthe originalcon cept. T he MOF did notwishthe trust to bemanag edbya pnvatepo rtfolio managerbut rather bythe MOF.Furth ermore the MOF wanted to create unittr ustswhich would beman aged by the PrivatizationT rust. T he Bank adviseda gainstt his approach andt her esult of thedisc ussionsw as that the Privatization Trustcomp onentw as dropped from theproject andthe GOGr equested thatit be canceled. 4.3 Net Present Value/Economic rate of return: Notap plicable. 4.4 Financial rate of return: Notap plicable 4.5 Instituttional developnment impact: The overall institutionaldevelopm ent impactisr ated as substantial, asa num ber ofkey institutions inthe financial sector werestr engthened. T hese include theSE C, NIC, GSE, HFC, and BOG. I n each ofthe institutions the project had an impact in twoway s: * by improving efficiencythrou gh thedesig nand installation offiew managementinformatio nsys tems,the project enhanced the institution'sabil ity to makeeff ective use of its human resources; * by providingop portunities for training and skills transfer, the project enhanced theskills of eachi nstitution's humanr esources. In the absenceofa project suchasNBFI, institutional development wouldpro bablyhav etake npla ce in BOG (NBFI Department), the GSEan dHFC. However it would probablyh ave been at a much slower pace. BO G has become considerablymo re rigorousinitssu pervisionand follow-upof bothban ks and NBFIs Ontheoth er hand it isun likely that the SEC and NIC would have gotveryf ar withoutsup port from theNBFI project. At the beginning of the proejct, neither NIC nor SECwer e functioning as active supervisors. Both have now initiated on-siteand off-site inspectionreg imes. - 14 -  Oncethe pendingl egislation isapproved,G hanashou Id havea mod ema nd strengthenedl egal and regulatory frameworkf or banking, insurance, and capital markets. T he stncter regulations combinedw ithst rengthened supervisoryc apacity atSEC, NIC, and BOGsho uld have alasting positive impact ontheinstitu tional structureof theG hanaianf inancial system. 5. Major Factors Affecting Implementation and Outcome 5.1 Factors outside the control of government or implementing agency: The two main factors werethe macroeconomic situation in the countryand Bank performance: Macroeconomic situation: Thesevere terms of trade shocks (a dropin cocoa prices and an increase in oil prices) led to a very sharpdep reciation (from26 50 cedis perdo liar at the endof 1999to55 00 cedis perdo liar at the endof 2000). Inthe aftermath oft his depreciation thestockof medium term financial assets almost vanished. I nterest rateshadr isen to 46 % by June20 00andlittle mo neyw aspu t into long-term assetsother than realestate Stock market activity deceased significantly and commercial banks found it much easierto invest in T-bills rathert han real sectorle ndingopp ortunities. BankP erformance: The Bank performance during supervision affectedthep roject bothn egativelyand po sitively as discussedbelo w inSection 7.2.Altho ugh the changesinT TLs and lack of IT expertise ledtoprocu rement and implementation delays, the useof FSAP recommendationsand refocusingof the project at the extensionph ase helped lead to satisfactory outcomes. 5.2 Factors generally subject to governnient control Thetwomain factors were the change in government and lackof support forsom e components: Changei n government: Aftertheelection s in December 2000there wasa newgo vemmentwhich inheritedman y projects. T henewg ovemmentwan ted to reviewth e portfolioof donor funded projects to ensureth at thepro jects were consistentwithth e new government's development agenda Therefore they chose to reexamine the NBFI project in great detail and according to theT TL,thisled to considerable delays inso me components Lack of support: Altho ugh itwas not an explicit partof theNBFI project, implementationof the govemment's divestiturep rogramwasex pected to supportth e growtho fth e stockmark et. Because of theGOG'sde cision not to use the stock market (only 6 or7of the 300div estitureswer edo ne through the GSE),theGSE component was adverselyaff ected. In addition, the quality ofandinter estof headsof beneficiaryagencies had a significant impact oni mplementation 5.3 Factors generally subject to implementing agency control. The main factorsthat are subject to implementing agency controla re project coordination, procurement, and financial management. Project Coordination: T he PCT was initiallystaffed bya Project Coordinator who washired extemally ona contract. Wh ilethe beneficiaries did notr eportanyissu es withproject coordination duringhistenure,Bankstaf f felt thathis projectcoo rdinationwas not fullyef fective. Atthetimeof project extension,theBan kand GOGhad been discussing the conceptof mainstreaming the coordinationof allBan k funded projectsintheir respective implementingagencies. Therefore inli ne with this thinking, theM OF appointeda full timeM OF employee as Project Coordinator andproject coordination and monitoringimp roved substantially. Procurement: Mostofthe beneficianes complained aboutpro curement problemsr elatedto delays ingett ing "no objections". While some of thesesit uations were truly problems on theBan k's side (and aredi scussed below under Bankperforman ce), a number of issueswer er elated to poorplan ning and executionby theben eficiaries. I tis not unreasonable to expectth at theseissu es could havebeen avoided ifthef irst ProjectCo ordinator had provided stronger assistance to the beneficianes. An ex-postr eview of procurement during 1999 and 2000 revealed that therew asnoe vidence of procurement planning andmon itonng being done at thePCT . - 15 -  FinancialMana gement Som e beneficiaries felt that in thee arly partof the project there hadbeen delays in paymentsbecau se the amountallocated to the Special Account ($200,000)was notadeq uate. This point hada Iso been raised by the GOGand theBan k hadsu ggesteda two-foldsolu tion:( a) to makethe requestfor replenishment moreoften; and( b)to request anincrea se inthe SpecialAc count to$ 400,000. The MOF decidedthat itpre ferred tomain tain theSpecial Account atthe lower limit. Anoth erconcernr aised by thebeneficiar ieswa s thatthere wer e delays caused-by theGOG's decision tom oveth e Special Accounts fromco mmercialb anks toth e BOG. It appears that BOGwa s not preparedf or this activity and the change was notimp lemented smoothly. Beneficiary Commitment. Given the natureof the project-with so many beneficiaries - implementation was also affectedby facto rs subject toth e beneficiary agenciesco ntrol. Fo r example the unsatisfactoryou tcome on the SSNITc omponent was directlyattr ibutable to the factthat senior management at theagency had changedsince the inception ofthe project and the newman agement had little desire to develop new products for the informal sector 5.4 Costs andfinancing: Table 3 shows the revisedpro jectco mponents, revisedbu dgetsand actual costs by component. Obtainingth e necessary informationfor this tablewas made difficultby thefa ctthat neither the Bank nor the MOF had accurate figuresabo ut the revised budgetallo cation by component. Furthermore the ICRmi ssion observedth at the MOF did not track changes in thebu dgetbycomp onent and does not have a regularmon itoring process for comparinga ctual costs to budgeted costs Table 3: RevisedP roject Components, Budgets and Actual Costs (aUlfigures in U S$ million) Original Revised Actual Cost Budget Budget I BOG - Payment Systems 4.05 7.07 2.75 2&3 BOG- NBFIan d Bank Supervision 0.93 1.05 2.35 4 NIC 1.27 1.65 1.95 5 SEC 1.25 0.92 0.54 6 GSE 3.12 2.97 2.22 7 HIV / AIDS 0 0.15 0 8 Micro /Rural Finance 3.75 0.75 1.32 9 HFC 1.00 0.81 0.93 10 SSNIT 1.25 0.79 0.99 11 SOA 0.60 0.51 0.65 12 ICAG 0.08 0.07 0.17 13 MDPI 0.05 0.04 0.04 14 SIC 3.20 0 0 15 GRC 1.05 0.39 0.41 16 Privatization Trust 0.45 0.03 0 17 ProjectM anagement 0.60 0.60 0.76 18 Unallocated 2.25 0.84 0 subtotal 24.90 18.64 15.08 ExpectedGOG/ Ben. Contribution 1.00 GOGrequ ested cancellation 2.66 Exch.Rate Adjustment (US$ 2.60 _ _ XDR)__ _ _ __ _ _ _ Total T 24.90 -16 -  Notesto Table3: * The revised budget figures are based on estimates included in the May 2000 Aide Memoire Actual costsare as per GOG/MOF records * Althoughthe BOG supervision was treated as a separate component for monitonng results, a separate cost center was not estabhshedby MOF, and all additionalBOG costs were recorded under the NBFI component * Although there was an estimate for a separatebu dgetfor theHl V/AID)S prograrn,the recording of the costs was done under the existing cost centers. * The GOG and beneficiaryco ntnbution was not broken downby component in the onginalbu dget The original cost of the project atprepar ation wasestimatedat US$ 24.90million, includingan IDA credit of US$ 23.90 million. Of this amount $2.66 million was cancelled in March200 0. The project closingdate wa s extended fromJun e30 , 2000 to June30, 2002. At closing,disb ursedf unds amountedto $ 15.08million which represents 63 % of the original budget of $23.9millio nand 81 % of the revisedbu dget of$ 18.64 million. Theactualdisb ursementon the project was significantlylower thanthere vised budget, even after the additionof newcom ponents suchas the HIV/AIDSawa reness program and the BOGBank ing Supervision component. The primaryr easons forthis are savings in somepro curements( such astheBOG payments system) andlackof implementationof somemajo r subcomponents (such as the GSECSDsy stem). The GOG's contnbution toto talpro ject cost was estimated to be$50 0,000wh ileth e beneficiaries wereex pected to contribute asimilar amount. This totalof $1.0millio n represented4%of thetotal projectcost.At closingthe GOGcon tribution was estimated by the MOFat $60,000 (about 0 03% of the disbursed amount).MOF was unable to provide estimates for the beneficiariescon tnbution. 6. Sustainability 6 1 Rationalefor sustainability rating. This ICR benefits fromth e fact thata majority of project components overtwo years the ICR wasdo ne. An "End-of-Term"Revi ew wascar ried outwithall projectben eficiaries dunng the final supervisionwhere the contnbutionofth e projectwas assessed by eachofhe beenficiaries. Withth ispe rspective, sustainability of the project, inits restructuredf orm, isr ated as likely. A s mentioned earlier,once thependin g legislationi s approved, Ghana should have a modem andstr engthened moder n legal and regulatory framework for banking,insura nce, andca pital markets. The stricter regulations combined with strengthened supervisory capacity at SEC, NIC, and BOG shouldhave a lastingpositive impactonthe institutionalstr ucture of theG hanaianf inancial system. The key challenges for the sustainabilityof thechan ges introduced in the NBFI sector can be analyzedby looking at four broad areas- regulatory agencies,pay mentsys tems, savings providers, and localtr aining providers. a Regulatory agencies. BOG, SEC, and NIC nowa llhav ethe capacity toun dertake their regulatorydu ties. However the main factor thatwill impact on the sustainability oftheir efforts ispend ing legislation.Until the pendingla wsandr egulations arepassed,the regulatory agenciesw illbe hamperedin their activities both in terms oflicensin g and supervision.The GOGand MOF together with theBan k will need to constantly follow upon this issue Furth enmore to ensurethatthe agencies carry outtheir duties effectively and in asus tainable way, these agencies will require GOGa nd MOF attention andsup port inthe following areas: * Monitoring ofsup ervisiont argets: Dunng thelast 2year sof the project( and dunngt he ICR) the Bank missions emphasized the need to set targets forsup ervision activityand to monitorpro gress. I n the absenceof future missions theon us will be on the Boardsof these agencies and the MOF toens ure that supervision (especially onsite) is done on a regularbas is. * Taking action:Even if onsite supervision visits are undertaken regularly and areef fectivelydo ne by the inspectionofficer s, the onus ison theSenio r Management of the agenciesto take action against the subject institutions. As thishasno t been done to date, the agencies will need to develop the "willpower" and "muscle"t o take the correct action, againstall institu tionsr egardlessof size or ownership - 17 -  * Continued skills development: The inspection staff at thethree a gencies will need to continue improving their skills Interviews withpr ivate sector participants me achof the threesec tors( banking, capitalmar kets, insurance) revealed that theinstitutio ns beinginsp ecteda re oft he opinion that the inspection staffdono t have thenecessar y skills and experienceto challenge them. * Roledef inition: SEC andNI Chave a dual role of bothr egulating the industrya nd facilitatinggr owth inthe industry. Both institutions will needM OFassistan ce intryin g to find the right balance between these two roles b Payment Systems: Now that BOG has implementeditsReal Time Gross SettlementSys tem, it is in a good position to makef urther additions and improvements tothe system (such asl inking allits branches and allowing for settlement offhardc urrencytr ansactions) T he Board of theBOG,tog ether with the GOG andM OFwill need to monitorth isactiv ity to ensure thatthesy stem con tinues tobeeff ectiveand ef ficientand thatth e necessary expansionsa nd improvementsa re made on a timely basis Furthermore there hasbeen discussiont hata n alternative totheGSE having its own automated CSD system, is for GSE to onlyhav e a depositary system and for the cleanng and settlement tobe done through theBankof Ghana's GIS. Ifthis option ispursu ed theM OF will havetoens ure thatthereis effective collaboration between the BOG and theGSE. c. Savings Mobilization. The savings mobilization challengesti ll remains. Although the GSEc omponent registered success inincr easing the numberof individual shareholders, there isun likelyto beanysus tamed increaseif the GOG continues to chooseno t to use theGSE as a divestiture vehicle. And on the insurance front, whilene w efforts by SSNIT may eventuallybesu ccessful in attracting informal sector savers, there isa need for substantial reform in order to stimulate private pension plansand increasedsavings d. Local TrainingP rovaders. TheN IC is in thepr ocess of awarding a contract to an international training provider to operatethel ITC.N ICwill get revenue in the formof rentand a share in theprofitsof the center.T he successof the IITCas acenter ofexcellen ce for the insurance industry and as a sustainable source of revenue will depend largely on the quality of training provided by thet raining firm and the financialt ermsof the agreement between NICa nd the training firm.T he improvement intr aining capacity at the SOAa nd the ICAG should besusta inable ift het woor ganizations are able to strengthen theirmar keting efforts while at the same time maintaining quality. As both these institutions would liketo improve their training facilitiesthey have been encouraged to consider approachingthe NICf or possible use ofspace in the newl ITC, atleast on asho rt term basis. 6.2 Transition arrangenent to regular operations: As most ofthe projectcom ponentshav e been completedor are sufficiently wellestab lished, no formal transition arrangementsa re necessary. SEC,NI C,GSE ,and the relevantdepartmentof BOG are expectedtoc ontinue the work initiated underth e project. Throughout the last3 years ofthepro ject the Bank supervisionteamwer every focused on mainstreaming activitiesin thevariou s institutions.T his efforthas reduced theneedf orspecial transitionalar rangementsa ndhasc ontributed positively tosus tainability. TheBankwill carryouta Post-FSAP Stock-takingmis sioninthef irst halfof20 03 whichwill providea n appropriate vehicles foron going policy dialogue onou tstanding issues. In addition, theGhan a InvestorsAdv isoryCo uncilisactivel ylob byingf or passage of thel egislation outstanding under the project. The follow-on activitiesof the micro and rural finance componenta re being undertakenund er the Rural Financial ServicesProject whichiswork ingwith both theARBAp exBan k, GHAMFIN and the micro-finance unit ofth e Ministry ofFinan ce. 7. Bank and Borrower Performance Bank 7. 1 Lending: TheBa nk'spe rformancedur ing identification, preparation, andappr aisalw as unsatisfactory. The project design was complex and unwieldy with a largenu mber of small components.And while the project objectiveswere in line witht he Bank's CAS,the projectc omponents did not match the objectives. Further more thele nding teamdid not - 18 -  have thenecessaryexp ertisef oron e of the major components (micro-rural finance) and thereis little evidence that thedesign team obtained supportf rom theta sk teamof thethenong oing RuralFinance Project. 7.2 Supervision The Bank's supervision efforts wereled byfou r different TTLs over the 6 year life of theproject. The appraisal TTLsu pervisedthepro ject until Februaryl 9 98.T he secondT TL tookthepro ject through the midterm review stage. The third TTL took over inJune 1999 andl edthe restructuring ande xtension. The fourth TTL, who is locally basedand had beeninv olved in supervision support from the start,too kov er the projectin July 2000 The third TTL remained involvedin thepr oject witha t echnical role. The MOF andmany ofthe beneficiariesar eof theop inion that the "frequentchang es in TTLs"wa s a negative factorthat adversely affected the project. When asked forsp ecificstheben eficiaries indicated thattheseco nd and thirdT TLs had different viewsof how theproject was run, that thehand over led todelays,a ndt hat the third TTL was "tougher" and questioned someactivities( including overseas training) that had already been agreed toby the second TTL. The ICR teamis oftheop inion that, while the number of TTLchan gesw ere more than what would bedesirabl e, each change actuallyledto improvementsinsu pervisionan dult imatelytoward sachiev ingofou tcomes. Thisvie w ise xplainedf urther below insom e detail as italsoillustratessom e lessons learned. * Supervision under thef irst TTLwa s inadequate.T here wasnof ormal project launchandthere was a gap ofl 2 monthsbetweenthef irstand second supervision missions. T he first missiontoo kplace in March 1996, two months before theexp ectedef fectiveness date.T he secondmis sion did not take place until April 19 97,eigh t months after theactualeff ectiveness date.And while theseco nd supervision mission tackled some implementation issues,a large part of themiss ion'sactivitiesw ere focusedon thedesig n of themicro-r ural finance component. * The second TTLIed the project through the mid term review and effectivelyapp lied the log frame matrix in a MTRwo rkshopto better definepe rformance indicatorsan dex pected outcomes. Theseco nd TTL alsoactiv ely pushed for GOGdeci sions onth e divestiture components. * The third TTL refocused theproject on five components with additionalperf ormance indicators and required specific annual businessplans fromthe beneficiaries. T his TTLalsoinco rporatedthe FSAPr ecommendations intoth e project and also made a successful casef or extending the project. * The fourth TTL, being locally based, playeda n instrumental role insupp orting the second and third TTLs. Whenth is individualbecam e the TTL, he wasabletospeed upthepro curement process byr educing the time takenf orgiv ingno -objections Thisw asa significantimp rovementinthe procurement aspectas under the third TTL there had been delays in issuing no-objections because oftheind ividual's heavy travel schedule Supervision efforts wereno taid ed byth e fact thatth ere were a large number of beneficiaries, eachofwho expected attention dunng a supervision mission, with ther esult that each mission could only pay cursory attentionto each component. Neverthelesslargel yas aresu It of theT TL changes discussed above, the focusing onou tcomes,and thesh iftinsup ervisionto the field, theov erallsup ervisionef fortimp roved significantly and is rated as satisfactory. 7.3 Overall Bank performance. Overall theBan kper forinance is satisfactory,esp ecially in lightofth e positivefina I outcomeofthe project. Although the projectwas notwell designed anddid notgetoff to ago od start, the Bankwork ed hardtor escuethe projectaf terth e midtermr eview. The decision to extendthepro ject was correct andprov idedth e necessary time tocomp letethe crucialcom ponents related tostre ngtheningthe regulatoryagencies and implementing thenew payments system. The stronger supervision inthelastth reeyears (discussed above) greatlycon tributed tothe project'so utcome. - 19-  Borrower 7.4 Preparation: Theov erallbo rrower performancea tp reparationw as unsatisfactory.Thego vemmentapp earstoh avef ollowed the Bank's lead in thedesig nof the projectwithlittle independent analysis. This isevid ent from theGOG' sLe tter of Sector Strategy, whichis datedl3 October, 1995 -a nd wasw ritten duringproject appraisal. Furthermore, discussionswithben eficiaries leads to the impression that the GOG madelittle attempttoestim ate the costsof various components and simply accepted whatevertheBan kteamsu ggested. 7 5 Government implementation performance: The govemment'sov erall performancedu ring implementation was satisfactorygiventher esults. Inthee arly stages ofthepro ject,th e govemment wasr eticent to take decisions necssary toexp edite certain parts of the project, particularly those relatingto the divestiturecomp onents. Although thec hangein govemment led tosom edelays, the new govemmentwasableto createa supportive environmenttoim plement some of thecom ponents of the project. Inpartic ular the govemment'sdecision toinv itethe IMF/Bank to undertake anFSA Passessm ent contributedstron gly to the implementationof the NBFI project as well as settingthestagef or futuref inancial sector reforms. 7.6 Implementing Agency: Like the Bank, the implementing agencystartedoff po orlybu t finished strongly. The mainstreaining ofproject coordination in the MOF after the extensionof the project greatly improved projectmo nitoring and relations with the Bank. Furthermore therew erenomaj orpr oblems with procurement andf inancialmanagement and the MOF's performance is rateda s satisfactory. 7.7 Overall Borrower performance: Although theBor rower's performance at the preparationa nd initialstages of implementationw ase rratic,the Govemment's desire to pushahead with the projectand to invitetheFSAPr eview,leadthel CRteamtocon clude thatth eo verall borrowerp erfommance is satisfactory. 8. Lessons Learned * Project design should be simplea ndha ve a few componentswith sufficientdep th. Thep roject wouldh ave been easiertoimp lement andwou ldhav e had stronger outcomesif it hadf ocusedon two main arease.g. strengthening regulatoryagenciesa ndimprov ing paymentsystem s. T his wouldhaver esulted in a manageable number of components and beneficiaries (fouro r five).It would haveals oallo wedfora better definitionof project objectives and their related component objectives. * Having weH defined,mea surable, understood and accepted outcome indicatorsiscritica 1. T he refinement of outcome indicatorsdu ring the mid termr eview, together withthe new indicators introduced at project extension, enabledbo ththeBan kand Borrower to monitorpro gress. The value ofthes e indicators needs to be discussed indetail with thebeneficiar iestoe nsure that theind icators areac cepted. Theind icators need tobe used over a periodof timeas a management toolandmad epart ofthe mainstream monitonng ofa ctivities within an institution to ensure that they are notjust seena sa Ba nk requirement and aredr opped after the end of a project. * Having well definedand measurable indicators would also have insulated the project frominterferen ce at the political level which, in tum, resulted ini mplementation delays during the first half of thepr oject. An understanding ofthe outcomeind icatorswou ldhav e mitigated theinev itablef ocus on inputs( vehicles, computers, trips)especiall y atthepo litical level. * TheBa nk needs to developa n approach tosup portt he reviewan dap proval of large,co mplex IT procurements. A large aspect of this project was thepro curement ofIT systems for the various beneficianes. When thetimecameto review the procurementr equest andprov ide "no objection" the Banksup ervision team struggled tof ind the necessary expertise within the Bank. T heAf rica region'sOper ationalSupp ortand Procurementteam could use thispro ject as a casestu dy to develop an approach that wouldprev entsu ch - 20 -  situations onf uture projects * Procurement delays can cause irreparableha rm to projectimp lementation. Oneo f the two main causes of delaysinthispro ject was thetimetaken to pass legislation( largelyou tsidethecon trol of theBan k,the implementing agency, or the beneficiaries). Theoth erwa s thetime takenf orprocu rement of goods and services( very muchun der thecon trol ofthe Bank, theimp lementinga gency, and the beneficiaries). * FSAP assessments and reportsca n be veryus eful toolsinpro jectdes igna ndimp lementation. There can be substantial benefits when there is close integration betweenFSAPsand projects. The ideal situation would be where projectdes ign followsand is builtup on aFSAP,thu sensu ring that the projectis designed to address theke yweak nessesid entified in the FSAP. However,as the NBFI experience shows, even TTLs of ongoing projects canus e the results of aFSAPtorefocu sth e project's direction andim prove developmentim pact. * Thereis a needf or flexibilityfor projects whichare creatingimpo rtant agencies and systems. Although the decision toexten dtheproj ect wasmad e reluctantly bytheCo untry Director, in hind sight it proved to be theco rrectd ecision. Intensive supervisiono f a morefoc used project allowedforach ievement of outcomes Furthermore the extension probablyallo wed forthesu rvival of two important regulatory agencies, SECand NIC and theimp lementation of ac ritical payment system. * Project Coordination should be mainstreamed withinimp lementingag encies. During2000, the Bank and the GOG agreed that the project coordination should bemainstreamedw ith the respective implementing agencies.T his decision wasinlin e with the Afticaregio n's policywhic h is based on years ofexp enence from vanousprojects. IntheNBFI project, project monitoring andr elationswiththeBank improvedsign ificantly afterthe change. And the benefits ofimp roved skills andgreater ef ficientsho uld continue as the NBFI ProjectCo ordinator isno w the Coordinator of the Micro-finance componentof the Rural Financial Services Project. * The Bank can stay engaged inpo licy dialogue,bey ond the immediatecom ponent objectives. In the caseo f theSSNIT com ponent, fundingwas stopped because ofthe institution's lack ofdesire to implementthe initial activities in the project design. NeverthelesstheBan kcon tinued to useth e projectand the possibilityof restored fundingto carry on a dialogue,a nchored in FSAP findings.T he result of thesee fforts is that Ghana has been selected as one ofnin e Sub-Saharan Afncan countries tobe included in a new World Bankpens ion reforminitia tive. 9. Partner Comments (a) Borrower/limplementing agency: TheGov ernmenthasr eviewed and commented on thedraft version of thel CR.T hesecom mentshav e been considered and incorporated into the final version. Further, the Government hassubm itted its reportw ith considerations about theimp lementationa ndcom pletion oftheproject. This reportcon firms mostof thef indings of the ICRtasktea m. A summary of the Government's reportis included in thel CR asAnn ex 8. (b) Cofinanciers N/A (c) Other partners (NGOs/private sector): N/A 10. Additional Information N/A - 21 -  Annex 1. Key Performance Indicators/Log Frame Matrix Atthe time ofappraisa 1, whileGhanahad successfully undertakenthe first major stepsin itsFinancia I Sector reform program, its financial institutions were still weak and financial infrastructure was a severe hinderance tofurther economic development. Despite having to weather some severe economic shocks during project implementation, most notably inthe period around 2000, the project has been instrumental in assisting the Ghanaianauthori ties put the sector on much stronger footing. Inevitably in a dynamic situation prioritieschan ge.Res ponsibility for divestiture transferred to another IDA-financed project (PEPTA). Suppo rt services suchas trainingof accountants and executive education were closed on time as the project focused onthe key areas ofpote ntiafly greatest impact. The FSAP in 2000 provideda thorough review of the sector and gavecle ar direction to the authorities on which areas were of highest priority. The project has been instrumental infacilitatin g progress inmost of theseareas. Whereas previously none of thereg ulatory bodieshad had much experience of operating in afra mework of prudential regulation,bythe end of the projectall ofthe m were operating uwithin such a frameworkand weremo re actively forcing compliance. The number of Financial Institutions meetingsolv ency requirements has improved considerably over theproject timeframe in Banking,NBFIs, Securities and Insurance. Thepayme nts system infrastructurehas been redesigned and restructured withheav y investment support from the project and has thereby enhancedits ability to ensure timely andrel iable transfersoffun ds around the country- a marked improvedove r the unacceptablepos ition atthe startof the project. Progress as measured against the FSAPreco mmendationsoverthe last two yearsof the project demonstratessubstantialperform ance gains in several critical areas. Inaddition,the avallabilityof critcal support servicesalso increasedsignifica ntly induding the number of qualified accountants and access to executive business and professional training. Spedfic details are provided inthe tables that follow. Component: Bankof Ghana -Pa yment Systems Target Indicators Actual /Latest Estimate Original- Appraisal Legislative reform: Passingan den forcement of The Bills and Checks Bill and Payment Systems Bill were Integrated Negotiablelns truments Act and preparedby BOG. The billsare with MOF for onward Electronic Funds TransferLaw transmissiontoCab met for approvalbef ore beingsub mittedto Parliament. BOG is in the process of organizing a seminar for parliamentarians. Installationof Real-timeGro ssSet tlement System The RTGS known as the Ghana InterbankSy stem(G IS) was (RTGS)for high value finds transfer operational launchedon October 11. At the time of ICRpr eparationthe system capacity ofl 0,000 highva luetra nsactions perday had been operating for 30 days withoutpro blems. The system was on real-timebasis financed by the NBFI project Computerized settlementofinter bank cleanng Computenzed settlement of interbankclea ringis taking place for andin tercitych eckcle aring the Accra region. BOG decided not to undertake computenzation foroth er regions because of thelo w volume oftra nsactions.Th e systemw as purchasedby BOG. Training was financed by the NBFI project. BOGha s also upgraded its softwarefor the cleanngof credii nstruments. Theon ginal targets were: Local checkscle ared A surveyofcle aring timesinJan 2002, undertaken by BOG, showed D+l (in 2 days), Outstation checkscle ared D+3 that 96% of payments cleared within the definedstanda rds.An (in 4 days); These targets were revisedto D+2 (3 independent survey commissioned by MOFsho wed similar results. days) for local checks (withininn er zones); D+4 (5 days) within the same outer zone and between innerzo nes;an d D+7 (8da ys) for remote - 22 -  locations. Enhanced shulls/knowledgeofBO G staff BOGstaf f attended several intemal payments system conferences andtraining courses at the Center forCentralBankingStudies at the Bank of England. Cash/M2 dropped to 20%h from 38 % Cash/M2 rato fell to24W bySept ember 1999 but then wentup to 36%by December200 0 Since then ithas beengrad uallyfall ing andreached24% inSeptembe r 2002. Additional- afterMTR, Extension orFSAP None - 23 -  Component: BankofGh ana - NBFI Supervision Target Indicators Actual /Latest Estimate Original- Appraisal Operational NBFIDep artmentwith fully btained NBFI supervision department is fully operationalwit h a staff of staff 51. Established Ongoing regulations,hcen sing and BOG established two types of Busmess Rulesfo r NBFIs:on e set for supervision procedures for each lindof institution Deposit Takingln stitutions and another forNo n-Deposit Taking in the non-bank financialsec tor Institutions. BOG also established astanda rd formatfor Returns which NBFIs have to submit to BOG. Additional- after MTR,Extension or FSAP IndustryRe ports Completed % of NBFis supervised as measuredbyons ite & 29 operational NBFIs (I 00