Peru - Fifth Agricultural Credit Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 3471 PROJECT PERFORMANCE AUDIT REPORT PERU FIFTH AGRICULTURAL CREDIT PROJECT (LOAN 933-PE) May 21, 1981 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. ABBREVIATIONS AU Animal Unit ARE Agricultural Reform Enterprises BAP Banco Agrario del Peru CAP Agricultural Production Cooperatives IBRD International Bank for Reconstruction and Development IDB Inter-American Development Bank MAA Ministry of Agriculture and Food OED Operations Evaluation Department PCR Project Completion Report PPAM Project Performance Audit Memorandum FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT PERU FIFTH AGRICULTURAL CREDIT PROJECT (LOAN 933-PE) TABLE OF CONTENTS Page No. Preface ................................................................. i Basic Data Sheet ...................................................... ii Highlights ............................................................ iv PROJECT PERFORMANCE AUDIT MEMORANDUM I. SUMMARY .1................................... ....... 1 II. MAIN ISSUES ............................................ 4 A. Sub-Loan Appraisal ............................... 4 B. Technical Assistance to Sub-borrowers .............. 4 C. The Relevance of Farm Models ....................... 5 D. Farm Mechanization ............................... 6 Attachment: Comments from the Borrower ........................... 8 PROJECT COMPLETION REPORT I. The Agricultural Sector ...................................... 9 II. Project Formulation ........................................ 13 III. Project Execution .......................................... 18 IV. Project Impact ................................................ 27 V. Institutional Development .................................. 31 VI. Bank Performance ........................................... 36 VII. Conclusions and Lessons Learned ............................ 37 Tables 1-23 Map This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  PROJECT PERFORMANCE AUDIT REPORT PERU FIFTH AGRICULTURAL CREDIT PROJECT (LOAN 933-PE) PREFACE This is a performance audit of the Fifth Agricultural Credit Project in Peru for which Loan 933-PE was approved on August 14, 1973 in the sum of US$25 million. The loan was fully disbursed: the final disbursement was made on October 31, 1978. The audit report consists of an audit memorandum prepared by the Operations Evaluation Department and a Project Completion Report (PCR) dated June 4, 1980. The PCR was prepared by the Latin America and the Caribbean Regional Office on the basis of a country visit in February/March 1980. The audit memorandum is based on a review of the Appraisal Report (No. 55a-PE) dated June 20, 1973, the President's Report (P-1191-PE) dated July 9, 1973, the Loan Agreement (Ln 933-PE) dated September 12, 1973, the PCR and inter- views with Bank staff associated with the project. Internal Bank memoranda on project issues as contained in relevant Bank files have also been consulted. An OED mission visited Peru in December 1980. The mission held dis- cussions with officials of the Banco Agrario del Peru (BAP). The information obtained during that mission was used to test the validity of the conclusions of the PCR. A copy of the draft report was sent to the Borrower on February 4, 1981. Comments were received from BAP: they were incorporated in the report. The full comments are attached as Annex 1 to the PPAM. The audit finds that the PCR covers adequately the project's salient features and the audit generally agrees with the conclusions. In addition to summarizing the objectives and results of the project, the PPAM expands upon the discussion of the effects of the loans made for farm machinery purchases. The valuable assistance provided by the Banco Agrario del Peru-s staff met during the preparation of this report is gratefully acknowledged.  - ii - PROJECT PERFORMANCE AUDIT REPORT BASIC DATA SHEET PERU FIFTH AGRICULTURAL CREDIT PROJECT (LOAN 933-PE) KEY PROJECT DATA Appraisal Actual or Item Expectation Current Estimate Total Project Cost (US$ million) 41.7 51.2 Overrun - 23 Loan Amount (US$ million) 25 25 Disbursed )- 25 Cancelled ) Repaid to November 30, 1980 - 3.2 Outstanding to )- 21.8 Exchange Adjustment - 3.7 Borrower-s Obligation - 25.5 Date Physical Components Completed 12/31/77 10/31/78 Proportion Completed by Target Date (%) - 83 Delay (% of Time Overrun) - 21 Economic Rate of Return (%) 26 30 Cumulative Estimated and Actual Disbursements (US$ millions) FY74 FY75 FY76 FY77. FY78 FY79 Appraisal Estimate 0.6 5.0 13.0 22.0 25.0 - Actual - 3.4 10.1 17.7 23.7 25.0 Actual as % of Estimate n.a. 68 78 80 95 n.a. OTHER PROJECT DATA Original Actual or Item Plan Revisions Est. Actual First Mention in Files - 12/07/66 Negotiations - 05/73 Board Approval - - 08/14/73 Effectiveness Date 12/12/73 - 12/12/73 Closing Date 12/31/77 06/30/78 12/31/78 Borrower Banco de Fomento Agropecario del Peru (changed name to Banco Agrario del Peru in 1975) Executing Agency Banco de Fomento Agropecario del Peru Follow-on Project None  - iii - MISSION DATA Sent Month/ No. of No. of Man- Date of Item by Year Weeks Persons Weeks Report Identification Bank 02/67 1 1 1 02/27/67 First Appraisal Bank/FAO 06/67 3 4 7 01/68 Post-Appraisal Bank 04/68 1 2 2 04/12/68 Preparation I Bank 05/71 3 2 1 06/11/71 Preparation II Bank 02/72 1 2 3 03/23/72 Second Appraisal Bank 10/72 4 5 20 06/20/73 Total 34 Supervision I Bank 08/73 1 1 1 09/07/73 Supervision II Bank 03/74 1 1 1 04/08/74 Supervision III Bank 05/74 2 1 2 06/20/74 Supervision IV Bank 03/75 2 2 4 04/10/75 Supervision V Bank 10/75 2 1 2 11/17/75 Supervision VI Bank 03/76 2 1 2 05/11/76 Supervision VII Bank 10/76 1 1 1 11/11/76 Supervision VIII Bank 05/77 2 2 4 06/01/77 Supervision IX Bank 03/78 1 2 2 05/20/78 Supervision X Bank 09/78 4 2 8 10/17/78 Supervision XI/ Completion Bank 03/79 2 1 2 04/04/79 Completion Bank/Cons. 03/80 2 2 4 06/06/80 Total 33 CURRENCY EXCHANGE RATE Currency: Soles (S/.) Years: Exchange Rate: 1973 and 1974 US$1 = S/. 38.70 1975 US$1 = S/. 40.10 1976 US$1 = S/. 55.70 1977 US$1 = S/. 88.40 1978.La US$1 = $/. 155.40 WEIGHTS AND MEASURES 1 kilogram (kg) = 2.205 pounds 1 ton (t) = 2,205 pounds 1 hectare (ha) 2.47 acres 1 liter (1) = 0.264 gallons /a After 1978, the rate of currency devaluation has accelerated rapidly, until in December 1980, US$1 was equal to about S/. 340.  - iv - PROJECT PERFORMANCE AUDIT REPORT PERU FIFTH AGRICULTURAL CREDIT PROJECT (LOAN 933-PE) HIGHLIGHTS The Bank started to support medium-term agricultural lending in Peru in 1954, when its first US$5 million loan to the agrarian bank (now called Banco Agrario del Peru - BAP) was approved. Subsequent loans were made in 1957 and 1960 (each for US$5 m equivalent) and in 1965, for US$15 m equiva- lent.!/ The US$25 million loan reviewed in this audit was the fifth in the series. The most significant change in Peru-s agricultural sector since 1965 was the land reform, which, between 1969 and 1975, transferred ownership of most of the large estates, formerly individually owned, to cooperatively operated Agricultural Reform Enterprises (ARE). These transfers were com- pleted in 1980. The Government of Peru had, understandably, strong vested interests in the success of the AREs and therefore requested that at least half the sub-loans be directed to them. The other half was allocated to small farmers and livestock owners whose holdings were below the limit that would bring them under the purview of the land reform. A little over half (55%) of all loans were granted to support purchases of farm machinery. Most of the borrowers requiring assistance for machinery purchases were AREs. Of the remaining project funds, 34% supported the livestock indus- try: dairying in the traditional milk producing areas and beef production, much of the latter in the Selva (Amazonian) region of the country. Of the balance, 10% financed the production of permanent crops and 1% supported agro-industries, like rice milling. The project's economic rate of return of 30% is above the appraisal estimate of 26%. Despite this seemingly satisfactory rate of return, in the macro-economic sense the project was not an unqualified success. Its imple- mentation coincided with a period when, owing to the Government's preoccupa- tion with land reform, it did not deem it prudent also to introduce policy changes in agricultural credit practices that could have ensured a self- sustaining operation by BAP. The Bank, in responding to the challenge to work with the "reformist" Government, after a lending gap of some six years in Peru, had largely to forego its attempt to impose sound credit policies: had it persisted, it may have had to withdraw from the project (PCR para 7.01 and PPAM para 4). Nevertheless, the fact that the dialogue was maintained with the Government contributed to their adopting to some extent more realistic policies in non-agricultural lending and eventually (1980) even for BAP. 1/ Loans 105-PE, 162-PE, 257-PE and 415-PE.  The Government's policies made BAP virtually the sole lending agency for agriculture, for short-term credit as well as medium- and long-term devel- opment loans. This situation resulted in: (i) a preference of, - compounded by pressured on, - BAP to concentrate on short-term lending (PPAM para 3 and PCR para 7.02) and (ii) all economic signals reaching the farmer on input and output markets coming from only one source (BAP, in effect Government) and therefore the farming community did not have access to a "second opinion" properly to evaluate and judge entrepreneurial prospects (PCR para 7.02). The following points are of special interest: - sub-loan appraisals which were expected to serve as technical and managerial guidelines for the farms did not fulfill expec- tations (PPAM paras 9 and 12); - despite virtual cessation of the agricultural advisory services during project execution, most of the farmers, due to their resilience, were able to reach production targets (PPAM paras 10 and 11); - the disproportionately high contribution of the farm mechaniza- tion sub-loans to compensate for the inevitable early disrup- tions of the land reforms in maintaining production levels (PPAM paras 16-18); and - the combined effects of 900% inflation and low interest rates, which resulted in the decapitalization of BAP and heavy subsi- dization of the farming sector (PPAM para 4 and PCR para 5.05).  PROJECT PERFORMANCE AUDIT REPORT PERU FIFTH AGRICULTURAL CREDIT PROJECT (LOAN 933-PE) I. SUMMARY 1. This project was the fifth in a series of Agricultural Credit Proj- ects supported by the Bank in Peru. The four previous loans were approved in 1954, 1957, 1960 and 196511. A follow-on project was appraised in 1967 but processing of the documents was halted because interest rates proposed for on-lending and the Government-s policy on prices of farm inputs and outputs could not be resolved. In addition, general economic problems and the circum- stances under which foreign properties were being expropriated at the time were deemed major issues to be solved before loan processing could proceed. 2. Eventually, after these issues were resolved in 1972, the project was re-appraised and the Board approved it in August 1973. As before, the Agricultural Bank [renamed Banco Agrario del Peru (BAP) in 1975] was the executing agency. It was expected that individual farmers and the Agricul- tural Reform Enterprises (ARE), the successors of the large individual hold- ings taken over under the 1969 Land Reform legislation, would borrow about equal amounts. The loans would be made for: dairy and livestock enterprises; annual crop production; establishing permanent crops; and agro-industries and secondary processing. Borrowers were to contribute 20% to project costs, BAP, out of its own resources, 20%; the Bank loan would cover the remaining 60%. On-lending rates were to be 10% for AREs and 13% for individual farmers; the lower rate for the larger enterprises being justified by the political and social need to ensure the ARE's success. Total project costs were estimated to be US$41.7; the Bank loan, accordingly, was determined to be US$25.0 million and would cover the estimated 46% foreign exchange costs and contrib- ute 14% to local costs. 3. BAP was slow to start the project. The reason was that they had other funds available which could be used to finance 100% of a proposed investment (under the Bank project, the sub-borrower had to provide 20% of investment costs - see para 2 above) and re-lent at 7% interest and therefore sub-borrowers obviously preferred them. About eight months after loan effectiveness the cheaper funds were exhausted and lending under the Bank project was started. However, appraisal of sub-loans was slow and processing was delayed because of the priority BAP attached to short-term loans at the expense of dealing with the medium and long-term loans required by this proj- ect. Performance was only marginally influenced by BAP-s "World Bank Project 1/ Ln 105-PE, US$5 million; Ln 162-PE, US$5 million; Ln 257-PE, US$5 mil- lion; and Ln 415-PE, US$15 million. - 2 - Office," set up to administer the loan. The project was completed about a year later than estimated. BAP in fact continued lending according to the terms of, and objectives set by, the project even after the loan was fully disbursed, expecting that a proposed sixth credit project would, retroac- tively, reimburse them for such lending. The sixth project, however, has not yet materialized; consequently, actual project costs under the fifth project have increased from US$41.7 million to US$51.2 million. The difference was covered by BAP funds. Partly because of the rapid changes of the socio- political environment between project appraisal and execution, and partly because accelerating inflation encouraged the use of loan funds to purchase durable items (cattle, machinery, building, and to a lesser extent, fence construction) rather than for short-term investments like fertilizers, the proportion of on-lending has changed from appraisal estimates: Appraisal Estimates Actual (a) Dairy and livestock enterprises 42% 34% (b) Annual crop production (mainly tractor 24% 55% and farm machinery purchases) (c) Permanent crops 12% 10% (d) Agro-industries 20% 1% However, due to the Bank-s insistence and the efforts of the supervision missions, AREs and individual farmers borrowed approximately the same sums, as expected at appraisal. 4. During project execution, BAP has suffered substantial losses due to the negative real interest rates that the Government of Peru insisted they charge. Although the Bank had foreseen this possibility, and made adequate provisions in the Loan Agreement to counter it, the relevant covenant was not enforced. Eventually, though interest rates had been raised to 24% and 27% for AREs and individual farmers respectively, this was only about half the inflation rate. Despite massive injections of funds by the Government, BAP's capital base between December 31, 1973 and December 31, 1979 decreased in real terms by about 3%. Its staff has increased by 46%, yet total short and long-term lending, adjusted to the Consumer Price Index, actually declined by 27% between 1974 and 1978.1/ Moreover, BAP's operations showed a sizeable shift to short-term loans, since they required less staff time to process and were therefore preferred by the branch managers. 5. It was estimated at appraisal that there would be approximately 2,500 sub-borrowers under the project, each taking loans of about US$17,000 equivalent. Eventually, BAP granted about 4,600 sub-loans but, because lending continued even after the Bank loan was fully committed, were able 1/ See PCR paras 5.03, 5.05 and 5.07. - 3 - to claim reimbursements only for about 4,200 sub-loans. Individual farmers borrowed an average US$6,400; AREs, US$50,600. The overall average sub-loan amount was about US$11,400 equivalent. 6. Since the loans granted by BAP under the project represented only about 4% of their total annual lending, they did not keep separate records of overdue payments attributable to the Bank-supported project. In 1980, about 15% of their total loan portfolio (including capital and interest) was overdue for more than 90 days and about 2% for less than 90 days. However, branch managers interviewed by the audit mission stated that repayments on the "annual crops" (farm machanization) component were completely up to date. BAP's headquarter staff are of the view that arrears on the IBRD-supported loans are under 5%. No documents were available to corroborate this claim. 7. Under the project, BAP was called upon to employ a Credit Specialist. The Specialist, appointed in 1974, was a recently retired former Assistant General Manager of BAP; he has in fact been recalled in 1980 and appointed General Manager. During the early years of project implementation, the Spe- cialist, notwithstanding some intermittent restrictions placed on his move- ments in 1974 and 1975, has played an important role in getting on-lending moving; senior officers in BAP are of the opinion that without his assistance lending in the remote areas, where the need was greatest, would have started much more slowly, and would not have reached the extent it did: more loans would have gone to easily accessible, relatively prosperous areas than was actually the case. The PCR (see para. 3.25) states that, because of movement restrictions placed on him, the Credit Specialist was not as effective as BAP staff suggest. The audit mission has further examined this issue by solicit- ing the views of staff members closely involved with the project but who no longer serve in Peru. Based on their views, as well as on the achievements and subsequent service record of the Credit Specialist, the audit mission is unable to support the judgement expressed in the PCR. As the project pro- gressed, the Credit Specialist's role became less important, as BAP staff got more accustomed to their task. 8. The economic rate of return of the entire project, estimated at appraisal and recalculated by the audit at project completion, is as follows: Appraisal Estimate La Actual Dairy development 26% 27%/b Beef cattle development 21% 13% Annual crop production 38% 387/c Total project 26% 30% /a See Appraisal Report, Annex 15, Table 12. T Supplementary PCR, para. 46. /c Estimate based on sample survey. Loan disbursed almost entirely for purchase of mechanical equipment. II. MAIN ISSUES A. Sub-Loan Appraisal 9. The project, as designed, expected BAPs staff to evaluate each loan application and, if feasible, to prepare a detailed appraisal, based on the sample farm models presented in the Appraisal Report. Even though a Credit Specialist was made available to train staff in this operation and to assist in implementation, the task has proved too much for BAP, for the following reasons: (a) The agricultural extension staff, already heavily committed to assistance with the implementation of the Land Reform legislation at the time the project was appraised, became involved with it full- time and therefore was unable to help would-be borrowers to submit plans for medium and long-term investments, or to assist in evaluat- ing them. Thus, the whole burden of loan preparation and appraisal fell on the limited number of BAP technicians, in consequence reduc- ing the total number of loans they could handle in a given period. (b) A corollary to the foregoing was the fact that specialist advice to evaluate finer technical points was not available, only the necessarily generalized knowledge of the few BAP technicians. (c) The same technicians were also called upon to cater to BAP's other business, which was mainly the processing of short-term production loans. With short-term loans, technical considerations are usually quite simple and loan processing largely depends on previous experi- ence gained with the borrower and his repayment record. Therefore, the technician-s apparent productivity was much higher. Conse- quently, both he and his branch manager favored short-term loans. The fact that most of the medium and long-term loans were nevertheless success- ful may be more a tribute to the planning ability of the borrowers than the time and attention that could be brought to bear by BAP staff. Inflation and low interest rates, which reduced loan repayments in real terms to a fraction of the actual value of the investments made, also played a significant role in the project's nominal financial performance. B. Technical Assistance to Sub-borrowers 10. With the virtual cessation of the agricultural extension service (see para. 9 above), farmers were compelled to undergo a sometimes painful learning process to exploit fully the benefits of their new investments. BAP did not change its established policy of not allowing its technicians to advise on farming techniques or to get involved in farm management decisions, even though such advice would have been important, for instance, to maximize benefits from machinery purchased. The technicians, when visiting farms, only - 5 - verified the use of funds (PCR, para. 3.25). In contrast, in the dairying areas, milk collecting centers offered some advice on animal husbandry and veterinarians were available to assist with brucellosis control (brucellosis being the principal cause for low conception rates in all cattle); conse- quently, technical performance was much better. 11. Adaptive research activities pertaining to pastures and field crops, previously the function of the extension service, were also severely cur- tailed because of the service's preoccupation with land reform. Consequently the generation of relevant new technical information for farmers almost ceased. The AREs were able to obtain what information became available despite the reduced activities of the services because they were run by full-time farm managers. As a result, they benefited more from the loans received.1 Since land reform, and involvement of the extension service, date back to 1969, the question can be asked why no provisions were ever made to organize a project technical service. The appraisal report prepared in 1967 had provided for expatriate technical advisers but the Bank subsequently decided to drop this project component because in 1973 the general atmosphere was not receptive to such a suggestion. Undoubtedly, benefits of the sub- loans were affected by this omission. Bank staff is aware of this fact and intends to include a component similar to the one they were forced to omit in 1973 in a proposed Sixth Agricultural Credit Project that would facilitate the flow of technical and farm management advice to sub-borrowers. C. The Relevance of Farm Models 12. Since it was recognized during project appraisal that little or no reliance could be placed on receiving full technical support from the agricul- tural extension service, the relevance and accuracy of the farm models, as a basis of sub-loan appraisal, assumed major importance. An evaluation of sub- loans made for dairy and beef cattle production, conducted by BAP and analyzed by the PCR mission (see PCR para. 4.07) revealed, however, that many expecta- tions on which farm models were based were not fulfilled. Milk yields of individual cows did not increase significantly under the project; on the other hand, yields obtained by participating farmers (3,100 liters per lactation) were already high compared to the national average (975 liters per lactation) (see PCR para. 4.03). Milk production per ha did increase approximately at the predicted rate; but before farmers took loans, their pastures were under- stocked. Loans have enabled farmers to increase output principally through enlarging their herds, thereby making better use of previously underutilized pastures. The fact that demand for additional fencing and grassland improve- ment was considerably less than could be expected from the farm models (see Appraisal Report, Annex 14) further underlines the fact that the major impedi- ment to increased milk production was shortage of cattle to utilize existing grazing. 1/ The audit considered the managers to be reasonably competent, but had to base this judgement on a 3% sample (PPAM para 18). The projects depart- ment concerned advises however that many of the AREs were not competently managed. - 6 - 13. The foregoing highlights the need for the preparation of meaningful farm models, which, in its turn, demands intimate knowledge of the environ- ment and an accurate assessment of its development potential. While the prime responsibility for the models' accuracy clearly rests with the project prepa- ration team, it is equally essential for Bank appraisal missions to verify the findings. In this case, the verification process could have been more careful. 14. The dairy component has, nevertheless, met its objectives in eco- nomic as well as financial terms, even though the nature of the investments, and the source of additional farm incomes, differed from the model envisaged in the appraisal report. It was unfortunate that the Bank agreed to drop the provision in the 1967 appraisal report to set up a comprehensive records system that would have provided both BAP and the supervision missions with evidence of changing trends. 15. At appraisal, the expectation was that beef cattle enterprises in the Sierras (high mountain plateaus) as well as the Selva (the tropical eastern slopes) would avail themselves to loan facilities. However, most of the sub-loans for beef cattle production went to the Selva. Loans for beef cattle production were expected to be used for the purchase of single-purpose beef animals; dual-purpose breeds were not eligible, despite a market for milk in the Selva which would have increased farmers' income. Moreover, the sub- borrowers profits were further impaired by the fact that individual animals, taken from the dry Sierras to the humid Selva did not adapt readily to the different climate and became affected by pyroplasmosis and anaplasmosis, which the thinly stretched veterinary services of the area were unable to control completely. However, adaptation problems are expected to be overcome in sub- sequent cattle generations. In the meantime the rate of return (13%) never- theless obtained was mainly due to increasing the herds and expanding grazing areas, thereby permitting better utilization of other fixed investments and management skill. D. Farm Mechanization 16. As noted, some 55% of the loan funds supported purchases of mechani- cal equipment: wheel and crawler-type tractors, as well as tillage imple- ments. Some combine harvesters were also purchased. Most of the loans for this purpose went to the AREs. According to the land reform legislation, the former owners of the large farms were to receive relatively little compensa- tion for their machinery. Therefore, since land reform proceedings took place over several years, maintenance had been minimal prior to takeover, and no replacements took place during the time; on the contrary, the owners endeav- ored to sell as much of the used equipment as possible. Thus, the AREs took over a much dilapidated machinery park which, never really adequate, was depleted even beyond its previous level. Repair and spare part facilities were at their lowest ebb in 1974-76. Therefore the AREs using the foreign exchange resources of this project, made a major effort to re-equip. Most of the loans were given to enterprises in the coastal areas, where sugarcane, cotton and later rice production has traditionally been at a high level and - 7 - pursued on modern, mechanized lines. The machinery purchased under the project contributed to increased farming intensities where water was available and to increased yields and production on all farms, due to more timely and thorough tillage operations and by enabling cropping pattern changes that would demand more, and more reliable, farm power. For instance, it became possible to grow rice on the heavier soils, where cotton brought less than 60% of what the more suitable soils yielded but where around 6 tons/ha of rice, a very fine achievement by any standard, could be obtained, thus increasing per ha net income from US$149 to US$280. Maize (net income $90/ha), was replaced by cotton (which, under the soil conditions of that particular enterprise, gave a net income of US$290/ha) due to the availability of additional farm power. Overall cropping intensities (where water was not a limiting factor) increased by nearly 30% (and not by 67%, as expected at project appraisal; see PCR, para. 2.09), mainly by enabling sorghum to be grown after rice. Net income of sorghum was estimated at US$150/hal. 17. A major issue associated with farm mechanization is the impact on employment and sometimes also the displacement of small and/or tenant farmers. Because of the land ownership pattern following land reform, farmer replace- ment could not occur under this project. On the contrary, some of the AREs expanded membership because higher cropping intensities were possible and most were forced as a consequence to increase their casual labor force, often turning casual laborers into full-time employees. Combine harvesters were mainly introduced into areas where labor shortage occurred at the peak harvesting seasons due to the conflicting demands of rice harvest and thresh- ing on the one hand and the spraying and picking of cotton on the other. Further, combines permitted the introduction of dwarf hybrid sorghum which created additional labor requirements for preplanting and after-cultivation and, to a lesser extent, for handling the harvested grain. Sorghum also provides grazing and forage for animals in the surrounding areas, including stock owned by the itinerant laborers. In the opinion of the audit, the social impact of the mechanization component in no way diminishes the economic impact; it enhances it. 18. Based on the small (3%) sample survey carried out by the audit mission, assisted by BAP staff, in an area where over 90% of machinery loans were granted, the economic rate of return of this component is estimated to be in excess of 60%. However, because of the small size of the sample, it is more prudent to accept the appraisal report estimate (38%). 1/ All values shown are 1980 figures, applying December 1980 rate of ex- change (US$1 = S/. 340).  -8- FORM NO. 788 (1-74) IBR) ANGUAGE SERVICES DISION CONTROL NO. F-1 F 9o/ IDATE: Mnr I PAI ORIGINAL LANGUAGEt Sganish (Peru) DEPT, OED ITRANotAToR: ENMcMe mk ANNEX 1 Banco Agrario del Peri Gerencia General (Office of the General Manager) Ref.: No. 63/GG-81 Lima, April 24, 1981 Mr. Shiv S. Kapur Director, Operations Evaluation Department The World Bank Washington, D.C. Subject: Project Performance Audit Report on Fifth Peru Agricultural Credit Project (Loan 933-PE) I am pleased to reply to your kind letter of March 18 asking for comments on the Project Performance Audit Report on the Fifth Peru Agricultural Credit Project (Loan 933-PE). In this regard I am pleased to tell you that the Report properly reflects the background to the Project, its implementation, and the results and experiences obtained. I should merely like to point out that the Summary states: "Despite its many years of experience in working with the Bank (IBRD), the Banco Agrario del Peri acted slowly in initiating the Project." In this context it should be noted that, when IBRD Program No. 933-PE started, BAP had alternative sources of financing available, such as the Programa de Pr6stamos en Fideicomiso (Trust Loan Program) with the Ministry of Agriculture, which offered similar loans for 100% of the investment at interest rates of 7%. Approximately eight months after the start of the Program, when the Trust Fund resources were exhausted, applications from farmers for loans under IBRD Program 933 picked up. I shall be grateful if this comment is included in page 2, paragraph 3, of the aforesaid Summary. Very truly yours, Banco Agrario del Peri /s/ Horacio Castro Medina General Manager  - 9- PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report I. THE AGRICULTURAL SECTOR 1/ General 1.01 Agriculture contributes about 13% to Peru's GDP, but it occupies more than 42% of the labor force and accounts for about 29% of total merchan- dise exports. Food processing employs an additional 4% of the labor force and contributes about 9% of the GDP. The value of food production at constant prices rose initially in the 1970s and then trended downward through 1978. The population, increasing at about 2.9% per annum, is now about 17 million; some 7.4 million persons (about 45%) live in rural areas. Per capita food production has dropped continuously during 1971-78; according to USDA statis- tics, by 1978 it was about 69% of the average per capita food production during 1961-65. Consequently, the country's dependence on imported food- stuffs has increased, mainly in dairy products, pulses, vegetable oils, and grains such as wheat and maize. Improvement of the performance of the agri- cultural sector has now become critical to the country's economic progress and to the well-being of the rural population. The Resource Base 1.02 In comparison to other Latin American countries, Peru has a relatively limited natural resource base. While the total area of the country is about 128 million ha, only about 3.7 million, or 2.9%, are cultivable, and another 22 million ha (17%) are natural pasture land. Most of the soils suitable for intensive agriculture are already cultivated. Additional land could be developed either in the Ceja de Selva (lower eastern slopes of the Andes), which would be expensive due to lack of access and basic infrastructure and services, or in the Coastal Zone, also requiring substantial investments to provide irrigation on presently marginal land. The ratio of cropland actually harvested to rural population is relatively low at 0.34 ha per person, but actual figures vary according to the three main ecological regions into which Peru can be divided -- the Coastal Zone, the Selva and the Sierra. 1.03 The Coastal Zone. The,Coastal Zone, which lies between the 3,000-km long Pacific Ocean coastline and the Andes, consists mainly of flat desert plains, sand dunes and the dry Andean foothills. It covers about 13 million ha (11% of the country's area) and has about 46% of the country's estimated 17 million population. As rainfall is extremely low, coastal agriculture is limited to the river valleys where irrigation is 1/ This section is based on the Agricultural Sector Updating Report issued in September 1977, whose findings are incorporated into the Basic Economic Report (No. 2204-PE) issued April 13, 1979. While the section may appear longer than typically the case for PCRs, this is considered necessary here in order to properly appreciate the difficult sectoral context in which the project was carried out. - 10 - possible; these valleys produce over 40% of the country's gross value of crop output. The crops in the Coastal Zone include sugar, rice, cotton, maize, potatoes, citrus, olives and grapes. The region enjoys relatively close proximity to the major urban markets, a good transportation network and other infrastructure, and is relatively prosperous. 1.04 The Selva. The Selva, or jungle region, which covers about 77 million ha (62% of Peru's area) and accounts for 10% of total population, is a humid tropical region. It contains two sub-zones, the low Selva and the Ceja de Selva. Fruits, coffee and tea are the main crops in the Ceja de Selva, and livestock production is being developed in some areas. Despite the relatively low soil quality, the Selva region has been gaining importance economically because of oil discoveries and development of two Trans-Andean Highway projects. It is likely that these projects will give impetus to further agricultural development in the region. 1.05 The Sierra. The Sierra, or Andean highland region, with about 44% of the population, covers over 33 million ha (27% of the total land area). It consists of steep mountains, reaching over 5,000 m, and high valleys located between the mountain ranges. The Sierra accounts for well over half of the arable cropland in Peru (about 2.3 million out of a total of some 3.7 million ha). However, this represents only 7% of the Sierra's total area. About half the remaining area (14.3 million ha) is used for grazing of sheep, cattle, alpaca and llama, mainly on natural pastures under marginal conditions because of steep slopes and high altitudes. Crop farming is generally dif- ficult and limited to one short growing season from about November to March. Accentuating the farming difficulties are: severe topography, erratic rain- fall, and extreme changes in temperature, including the risk of frosts during the greater part of the year. Approximately 55% of the country's population dependent on agriculture lives in the Sierra, with an average of about 1.9 ha of cropland per family. Agricultural techniques are primitive, fallow periods are long (three to four years for every year cropped), transport costs to market centers are high, and opportunities for off-farm employment are limited. This region is the major locus of rural poverty. While there is little scope for putting new land into production, there is significant potential for increasing crop and livestock productivity, raising output and improving farmers' living conditions. Policy Background 1.06 Some of the Government's very interventionist policies in the agricultural sector since 1968--such as those reviewed below--have slowed the expansion of output or contributed to its decline. Shortfalls of domestic production in recent years have resulted in considerable imports of basic foodstuffs--often distributed to consumers at highly subsidized prices-- queuing, meatless days, parallel markets, and such. The Government has recently initiated a series of policies which include a reduction of the earlier detrimental approach. Marketing and price control policies are being altered in order to promote increased production, reduce agricultural imports and improve producers' returns. This reorientation of sectoral policies is expected to resolve some of these problems, although it would be premature to undertake an assessment at this point. - 11 - 1.07 Agrarian Reform. In June 1969, the Government issued a new agrarian reform law which introduced significant changes into the agricultural land tonure pattern of the countryside. Large private landholdings were expropri- ated and property rights transferred to groups of workers, small farmers and Indian communities. The main target of the reform process was to expropriate some 14,500 farms covering 10.3 million ha (43% of the country's area in farms) by the end of 1977 and adjudicate 10.6 million ha (including 0.4 mil- lion ha of state land); 77% of the land area was adjudicated to--and 95% of the beneficiaries were grouped in--associative or cooperative operating units (Agrarian Reform Enterprises (AREs)). The expropriation and adjudication stages of the agrarian reform process were by and large completed by the end of 1978. This formidable task was accomplished relatively rapidly and with notable stress on the institutional capabilities in tbe sector; the Agricul- tural Ministry's trained manpower was concentrated on the land transfer process and other priority items were downgraded or neglected. In this process, the Ministry's extension services were severely reduced. Moreover, despite the concentration of credit and technical assistance on the newly established AREs, performance has been uneven. 1.08 The Government had great expectations that agrarian reform would alter the political and economic structures of the country, promote agricul- tural production and eliminate rural poverty. However, achievements have fallen short of expectations and a new set of serious problems is emerging from the reform sector. The more that cooperative farming changed the pre- reform patterns of production and the smaller the initially distributable ARE profits, the less willing members were to participate in attaining the AREs' goals. Member rejection has been particularly noticeable in the poor tenant- based haciendas of the Sierra, where the shift to cooperatives has been accom- panied by attempts to eliminate individual plots/herds. The rigid size cri- teria imposed on cooperatives have made many of the Sierra AREs uneconomically large. Basic conflicts of interest have arisen on some AREs between members/ workers (who want to maximize wages and reduce employment) and members of the surrounding peasant communities (who want to reduce permanent worker wages and maximize seasonal employment). At the same time, since individual profit sharing is small and the member has little control on the allocation of the reinvested profits, he pushes for high wages and low profits; since the wage and the job are both virtually guaranteed, there are few incentives toward labor effort and discipline. Similarly, the incentives encourage members to mechanize operations rather than expand non-member employment opportunities. These organizational problems have made it difficult to manage the AREs, especially when the managers and technicians hired by the enterprise are inexperienced and given very limited authority. Nevertheless, while most public sector resources in the past decade have been concentrated on imple- menting the reform, which involved about 43% of the area in farms and 25% of the rural families (and probably the relatively better-off families), the 40% of the farm families which are minifundistas and 12% which are landless laborers have benefited little or none from this costly process. 1.09 Marketing. As the Government considered that market forces were not allocating sufficient resources to the production of food, and that monopolistic practices resulted in excessive profit margins and inefficiencies in marketing, in 1975 it set up a complicated central planning system to - 12 - partially replace the market on production and pricing decisions. It also imposed controls in prices and marketing margins; took over all marketing of exports, imports and some domestically produced foodstuffs; and sought to regulate the physical movement of agricultural goods. As a result, the farm area in food production has tended to increase, but gross value of output per hectare has declined, and net foreign exchange savings appear to have been negative. Price subsidies on fertilizers helped the richer farmers and often resulted in excess use with little impact on production. The Government lacked manpower, money and storage facilities, while wholesale markets and other marketing infrastructure remain inadequate. The inefficiencies intro- duced by this intervention in the marketing system have become a serious bottleneck to agricultura development. The Government is presently changing its approach, moving over to a system of partial intervention to eliminate the possible monopolistic practices now exercised in some cases by middlemen and to ensure that marketing margins are as small as possible, benefitting both producers and consumers instead of trying to police the whole. In line with this new approach, in the last few years, the Government has started to relalx its controls on prices and to reduce consumer subsidies on basic foodstuffs. These new measures are intended to stimulate production and improve producer returns, lower imports and reduce the heavy burdens placed on the Government budget by consumer subsidies resulting from earlier policies. 1.10 Credit. Less than 20% of farms obtain institutional credit, at substantially negative real interest rates. The remainder operate with informal sources at high interest rates or obtain no credit at all. The bulk of credit--institutional or informal--is short-term; medium- and long-term lending averaged less than US$20 million annually from 1975 to 1978. The Government's credit policies discouraged the commercial banking system from lending for agriculture and at the same time promoted inefficiency and de- capitalization in the Peruvian Agricultural Bank (BAP). 1/ The BAP had traditionally adhered to credit policies established by the Government; nevertheless., at the start of the 1970s the Government reduced even further the BAP's already limited operational autonomy, largely mobilizing it to support the newly-established AREs. Borrowers came to depend on subsidized credit in a highly inflationary environment, rather than applying financial discipline in their farming operations. Previous Bank Involvement in the Sector 1.11 The Bank has made 12 agricultural loans to Peru. Five of them, totalling US$55.0 million, were for agricultural credit and were channeled through BAP (105-PE, 162-PE, 257-PE, 415-PE and 933-PE). Disbursements of the last loan (933-PE), which is the subject of this PCR, were completed in 1978. Two other loans (67-PE and 98-PE), totalling US$3 million, were for farm machinery imports and four (144-PE, 418-PE, 1403-PE and 1771-PE) were for irrigation. The first two of these loans for irrigation, totalling US$24.5 million, financed irrigation works and land settlement of the San Lorenzo 1/ The Banco de Fomento Agropecuario (BFA) was renamed Banco Agrario del Peru (BAP) in 1975. This latter abbreviation will be used throughout this Report. - 13 - scheme in the Piura Valley. The Second San Lorenzo Loan (418-PE) suffered managerial problems, inadequate counterpart funding and water shortages. While it took 11 years to disburse, it nevertheless achieved its basic goals and produced important benefits (OED Project Performance Audit Report SECM-48 of January 1978). The third irrigation project (Loan 1403-PE for US$25 mil- lion) finances the rehabilitation of irrigation and drainage systems in six southern coastal valleys, and the fourth (Loan 1771-PE for US$56 million) finances similar works in the Lower Piura River Valley. A Loan for US$15 million (1812-PE) financing the Puno Rural Development Project has recently been approved by the Bank. II. PROJECT FORMULATION Appraisal 2.01 In December 1966, with two-thirds of Loan 415-PE for the Fourth Agricultural Credit Project utilized, the BAP applied to the Bank for a US$20 million loan to cover 50% of a three-year crop and livestock lending program amounting to US$40 million equivalent. Through these previous projects the Bank had established a close working relationship with BAP and had sought to promote credit policies which would strengthen BAP and lead to the estab- lishment of a financially autonomous agricultural credit institution. The project had been prepared by BAP and represented a continuation of the sublending program being carried out by the on-going Fourth Project. A Bank mission to appraise the proposed project visited Peru in June-July, 1967. Following the working party review of the Yellow Cover Appraisal Report in February 1968, a post-appraisal mission visited Peru in April, 1968 to update the report. This mission reviewed the impact on BAP of economic measures applied subsequent to the initial appraisal and updated the appraisal report as regards BAP's financial situation. The issue of index-linking subloan repayments was raised, owing to an acceleration in the rate of inflation in the preceding twelve months (to about 20%). The revised Yellow Cover Appraisal Report was reviewed by the Working Party in June, 1968. Major unresolved issues were Government price subsidies or controls on inputs and output, negative real interest rates charged by BAP and the critical economic problems facing Peru. On this basis, a Green Cover Appraisal Report was issued in August, 1968, proposing a US$17 million loan to the BAP, financing 50% of the proposed sublending program. The documents for the proposed loan were never circulated to the Loan Committee, as the Government which came into power in October, 1968 did not satisfactorily resolve the outstanding country economic problems or an issue of expropriation of Foreign properties. Re-Appraisal 2.02 Preparation. In May 1971, a project preparation mission visited Peru to revive interest in a possible loan to the BAP for agricultural credit. With the rapid execution of the agrarian reform program, and other government policies, the sectoral context was notably changed from that of several years earlier. Initially, the Government sought a Bank Loan only for sublending to AREs, but private farmers were also included after conversations with the - 14 - mission. This mission brought back a BAP proposal for a US$28 million loan for a three-year project. The BAP undertook to prepare a project, focusing mostly on small- and medium-size private farmers and the sugar production cooperatives. A follow-up Preparation Mission visited Peru in February, 1972. According to that mission, the major issues were: (a) Instability and uncertainty caused by the agrarian reform program; (b) The level of arrears (35%) in the BAP portfolio; and (c) Government interest rate policies. The preparation mission recommended that the Bank favorably consider appraising the proposed project, with the caveat that it appeared "just barely" suitable for Bank financing. 2.03 The appraisal mission visited Peru during September/October 1972. 1/ The proposed project, costing about US$42 million equivalent, would be par- tially financed by a Bank loan of US$25 million. The major issues identified by the mission (and the recommendations) were: (a) BAP arrears (currently 27% of its portfolio), to be reduced to 10% within five years; (b) Subloan interest rates (the mission recommended 10% and 13% for AREs and individuals, respectively); and (c) Beneficiary contribution to project cost (20% as a minimum). Negotiations 2.04 During negotiations the following changes were introduced into the draft legal documents: (a) it was agreed that the introduction of the proposed 10% p.a. interest rate for sublending to AREs be delayed fifteen months to avoid further ARE protests such as had followed on the earlier increase from 7% p.a. to 9% p.a. It was also agreed that, should inflation surpass 9% p.a., sublending would be suspended until the Bank, BAP and the Government had agreed on a new level of interest rates; (b) A change in wording was agreed to regarding the ineligibility of sugar cooperatives for subloans; (c) A change in wording was agreed to in the standard text regard- ing audit, to enable the Controller General's office to appoint auditors acceptable to the Bank; I/ The mission did not include any members of the appraisal mission which prepared the August, 1968 report. - 15 - (d) At the request of the Peruvian Delegation, side letters were prepared stating that the Bank would (i) review, in special circumstances, subloans under which sub-borrowers would con- tribute less than 20% of subproject investment cost and (ii) review, from time to time, the desirability of including forms of ARE organization other than Agrarian Production Cooperatives (CAPs) as beneficiaries; and (e) the Loan Contract clause requiring the Government to issue import licenses for all machinery and equipment, including tractors and spare parts, as required, for investments under the Project, was replaced by a side letter which recognized that tractor imports would be restricted once domestic assembly starts and stipulated that the supply of tractors and spare parts would be adequate to satisfy demand under the Project and that prices charged for tractors assembled in Peru would not be substantially higher than those charged for imported tractors the year before domestic production started. This was intended to prevent a government-sanctioned tractor assembly monopoly from adversely affecting the supply and price of farm tractors. Studies of BAP's operating costs and subloan arrears situation were also provided for in Loan covenants. During negotiations the Bank had to reiterate its interest that sublending be equally divided between private farmers and AREs. Description 2.05 The project was to be part of the BAP's lending program, over a three-year commitment period. It would finance investments made by small and medium farmers and AREs in livestock (beef and dairy cattle, sheep and alpacas), on-farm development (including minor irrigation and drainage), storage facilities, mechanization, orchards, swine and poultry operation, and processing facilities. 1/ About 2,500 small- and medium-size farmers and AREs were expected to participate in the project. A large portion of project investments would be made in the intensive coastal agriculture, while much of the livestock investment would be in the Sierra and Selva. The project funds were earmarked 50% for individual farmers and 50% for AREs. The Loan was made to t'ie BAP and guaranteed by the Government. Total project cost was estimated at US$41.7 million equivalent (S/.1,614 million). The Bank's contribution of US$25 million would cover 60% of project costs (including US$19 million estimated foreign exchange cost) while the Govern- ment and the beneficiaries would each contribute 20% of project costs. Executing Agency 2.06 The borrower and executing agency for the project was the BAP, a successor to the Government-owned Agricultural Bank established in 1931. The BAP's predecessors had eKecuted four previous Bank Loans; the last one was 1/ Sugar cane AREs were not eligible for subloans for sugar cane produc- tion and processing. - 16 - fully disbursed in 1968. In addition to operating with its own resources, BAP managed -- for a fee -- Government trust funds for credit to the agricul- tural sector. It also managed several non-banking activities, the major ones being a large rubber plantation (Fundo Iberia), natural rubber imports and jute marketing. 1/ 2.07 BAP's financial position at the time of appraisal was a matter of concern to the Bank. Authorized capital on December 31, 1971, was US$258 million equivalent, of which US$43 million equivalent was paid-up and reserves were an additional US$3.2 million equivalent. Its spread of seven percentage points (between income as a percentage of its average portfolio and financial costs as a percentage of its portfolio) barely covered operating costs (5%) and provision for bad debts (2%). It was able to maintain low lending rates (i.e., 7%-13% p.a.) only by having access to low-cost financing. Loan recoveries had been a problem since the early 1960s. 2.08 BAP's geographical coverage of the country was relatively good, with 14 branches, 42 agencies and 46 inspectorates (these last were limited to receiving and forwarding loan applications for consideration by agencies and branches). Its staff of 2,186 included 247 professional agricultural spe- cialists. The appraisal mission noted that operations appeared to be highly centralized at the Lima head office, but anticipated that BAP's hiring of a Credit Specialist under the project would promote greater coordination and decentralization in the future, which would improve staff efficiency and lower operating costs as a percentage of loan portfolio. Targets and Goals 2.09 The objective of the project was to increase the supply of crop and livestock products, almost exclusively for sale in domestic markets, increase producer incomes and relieve the pressure of growing food demand on the trade balance. In addition, through studies of BAP's operating costs and subloan arrears situation, as well as the stipulation for interest rate adjustment if inflation rose above 9%, the Bank anticipated that BAP's financial condition would be strengthened. Five specific farm models for crop and livestock operations were presented in the Appraisal Report to provide a basis for assessing credit demand and project costs and returns: (a) Small dairies. A 15-ha farm with 10 animals would have 7.5 ha of pasture renewed in grass and clover and the remaining 7.5 ha improved (21% of investment cost). Fencing (11%) was provided to permit proper pasture rotation, as were structures (15%) for herd handling. Breeding animal purchase (46%) was foreseen.and stocking rates would increase to 1.5 AU per ha. Some machinery purchase (7%) was also anticipated. By full production (year 7), milk yield per cow would be 3,660 1 per year. Average investment per sub-borrower would be about US$4,000 equivalent. The financial rate of return was estimated at 24%. About 1,500 subloans of this type were expected under the project. 1/ The Bank had for some time encouraged BAP to divest itself of these non-banking activities. In compliance with the Fifth Project Guarantee Agreement (Section 3.03), BAP transferred the Fundo Iberia to the Govern- ment in 1974. - 17 - (b) Beef Production. A 110-ha farm with about 220 head of cattle would have 50 ha of new pasture and 60 ha of improved existing pasture (23% of investment). Provision was also made for breeding stock (32%), machinery (25%), fences and water supply (10%) and buildings (7%). Stocking rates would rise to 1.5 AU per ha, steers would be sold at 24 to 40 months age with 400 kg live weight and the calving rate would rise to 70%. Average investment per sub-borrower would be about US$16,000 equivalent. I/ The financial rate of return was estimated at 19%. About 400 such subloans were expected under the project. (c) Pig Production. It was envisaged that these subloans would be executed mainly by AREs. Investment components were buildings (37%), breeding stock (23%), feed (20%), labor and animal health (10%) and machinery (10%). the production unit would have a herd of 315 sows; annual production would be not less than 400 pigs averag- ing 100 kg liveweight. Investment per unit would be about US$330,000 and the financial rate of return would be 35%. (d) Annual Crop and Alfalfa Production. About 22,500 ha would be developed through investment in on-farm irrigation and drainage (15%) land levelling (18%) and mechanization (67%). Cropping intensity would increase from 91% before the investment to 166%. Crops produced would be mainly cotton, sorghum, maize, potatoes and alfalfa. Farmer net income (on a 450 ha operation) would increase from US$21,500 equivalent to US$103,100 equivalent, with investment cost for such an operation estimated at about US$189,000 equivalent. 2/ The financial rate of return would be 39%. About 50 subloans were expected under the project. (e) Perennial Crops. Subloans were anticipated for mango, banana, pineapple, avocado, citrus, olive, grape and asparagus, on units averaging 10 ha. Investment components would be land preparation (10%), planting (30%), labor (39%), fertilizer and pesticides (12%) and machinery and equipment (9%). About 6,000 ha would be incorporated. Investment costs would average about US$8,000 per unit and the financial rate of return was estimated at 26%. About 600 subloans were expected under the project. 2.10 No models were prepared for the agroindustrial investments expected to take place under the project. On the basis of information provided to the mission, subloans were expected to be for lemon oil and olive oil extraction plants, milk pasteurization plants, feed mills, sausage-making plants and cold and grain stores. About 30 such subloans were expected under the project. The estimation of the project's overall economic rate of return -- 26% -- was based only on the crop and livestock models outlined above. 1/ Relatively larger subloans were foreseen to AREs for beef, sheep, llama and alpaca operations. 2/ In the cases of small- and medium-size farmers, investment costs were estimated at about US$6,000 and US$20,000 equivalent, respectively. - 18 - III. PROJECT EXECUTION Effectiveness and Start-Up 3.01 Conditions of effectiveness for the loan were (a) BAP would employ a credit specialist to assist in improving its procedures for the technical and financial evaluation of investment subloans, and (b) the Government would make a capital contribution of S/. 100 million (approKimately US$2.6 million) to the project account. The latter condition was fulfilled on December 3, 1973. The first candidate proposed by BAP on October 19, 1973 for the credit specialist position was not accepted and on December 11 the retired Central Manager was proposed by BAP and accepted by the Bank. The loan was declared effective on December 12, 1973. Furthermore, at its own initiative, BAP established a "World Bank Project Office", staffed by three to four profes- sionals, to backstop the branch offices and agencies and train credit spe- cialists. The first subloan request under the project %as submitted to the Bank in April 1974. In part, the slow start was attributed to the restriction that subloans could not finance more than 80% of the costs of the investment plan (sub-borrowers were accustomed to subloans financing 100% of investment costs), the availability of competing funds at an interest rate below those for the project (BAP was reluctant to raise its general lending rate), inade- quate experience of BAP staff with subproject preparation and evaluation, and BAP's preference to work with AREs while not promoting investment subloans to private farmers. 3.02 In May 1974, BAP raised its overall interest rate structure to the level provided for under the project, ending the problem of competing funds. The Bank expressed concern that at this time that there was still a major effort required to ready BAP staff for preparing and processing subloans. Owing to the lack of progress in project execution the Government proposed, and the Bank concurred in, delaying the further S/.223 million (approximately US$ 6 million) capital contribution the Guarantee Agreement called for the Government to make by June 30, 1974. Progress 3.03 The first application for withdrawal from the Loan Account was submitted in August, 1974. By early 1975, the pace of project execution had picked up. About two-thirds of the lending program was going to the AREs. Main investment iteis were tractors and livestock. With the accelerated pace of project execution, it appeared timely for the Government to make its second and final capital contribution (S/.223 million). 1/ Bank supervision visits to branch offices noted the slow rate of subloan processing, lack of proper training for BAP staff in project subloan procedures, staff concentration on processing short-term subloans and a discrimination against small individual farmers in favor of politically powerful AREs. While the rate of subloan processing accelerated, supervision missions in 1976, 1977 and 1978 noted that 1/ This contribution was made in the second half of 1975. - 19 - the capacity of BAP field staff for preparing and processing investment sub- loans continued to be deficient. The "World Bank Project Office" staff undertook to train BAP field staff in project preparation and evaluation. This Project Office staff also made periodic visits to BAP field offices to clear up the backlog in subloan processing and accelerate the preparation and evaluation of subloans. In this respect it performed an important service and improved communications of the Head Office with the branches and agencies. 3.04 The delay in project initiation caused Bank Loan disbursements to lag behind the Appraisal estimate (Table 1). The pace of subloan approval increased markedly in 1975 (Table 2). The annual data on number and amounts of subloan approvals reflected the Bank's early concern over the emphasis given to agrarian reform enterprises (AREs). While the average subloan amount (in US$ equivalent) is not large, it fell considerably in the final two years of project execution, as the BAP accelerated lending to private farmers to comply with the undertaking that not more than 50% of the lending program be for AREs. With respect to total sublending (Table 3) this pro- vision was complied with, although a slightly higher percentage (58%) of Bank Loan disbursements were made for subloans to AREs rather than private farmers (Table 4). Project Cost 3.05 At the time of appraisal, project cost was estimated at S/. 1,614 million (US$41.7 million equivalent); 80% of this cost would be financed by the sublending program and the remainder by beneficiaries. 1/ Appraisal Actual S/ .M US$M % S/ .M US$M % Bank 968 25.0 60 1,848 25.0 49 BAP 323 8.4 20 1,170 15.8 31 Sub-borrowers 323 8.4 20 754 10.2 20 1,614 41.7 100 3,772 51.2 100 The increased project cost was absorbed by the BAP, whose share of project cost rose from 20% foreseen at appraisal to 31%. As we see below, the Bank disbursed against 91% of the subloans (4,199) and covered 65% of BAP's disbursements (S/. 2,837 million) on those subloans: 1/ A side letter permitted BAP to waive this latter requirement if it determined, and the Bank concurred, that the beneficiary could not make such contribution. There were few cases where this side letter was implemented. On the other hand, a spot check of subloan records suggests that there were cases of ex-post subloan increases where this 20% requirement on the increased subloan amount was not applied. Thus the figure labelled "Actual" is in reality an estimate. - 20 - Subloans on which partial Total BAP Bank Loan Bank Reimbursement Disbursement Project Disbursements Requested on such Subloans Total No. of Subloans 4,199 4,199 4,199 4,620 Amount (S/. M) 1,848 2,464 2,837 3,018 Thus the Bank Loan accounted for only 61% of total project subloan disburse- ments rather than 75% as established by the Loan Contract. This outcome owes in part to the problem BAP had in establishing a suitable reporting system which permitted the Project Unit to keep abreast of the pace of subloan approval, and in part because the Project Account was kept open as the Govern- ment anticipated--on the basis of communication with the Bank--that a bridging loan would be quickly available to continue the credit program shortly after the Fifth Loan was fully disbursed. 3.06 It should be noted that the average size of the 4,199 subloans for which the Bank at least partially reimbursed BAP was about S/. 675,000, while the average size of the 421 subloans not submitted to the Bank for reimburse- ment was about S/. 430,000. As smaller subloans tended to be concentrated near the end of the project execution period, and a greater effort was made to incorporate relatively more private farmers into the project, the final sublending program was evenly divided between private farmers and AREs. 3.07 On the basis of records reviewed in two branch offices (Piura and Huacho, which accounted for about 20% of the sublending under the project), the average size of subloans approved was smaller (exceptionally so in the case of Piura) than the original subloan applications. There also appears to have been a significant "weeding out" process, since only 62% and 42%, respectively, of the subloan requests in these two offices were approved. Nevertheless, any possible limitation of sublending to larger (i.e., most likely ARE) sub-borrowers under this Project did not necessarily result in denying them appropriate investment credit, as the BAP was concurrently executing IDB loans which were expressly intended to provide such subloans to AREs under similar terms and conditions. Credit Demand 3.08 At appraisal it was anticipated that the major demand for credit would be for beef and dairy production units; 43% of the Bank Loan was allocated for financing these activities. Secondly, investments associated with the production of annual crops and alfalfa were assigned 25% of Loan funds. Subloans for agroindustry, poultry, swine and to sugar cooperatives were expected to account for 20% of credit demand. Finally, lending for permanent crops was anticipated to account for about 12% of subloan demand. In each of these categories, Bank Loan financing was divided evenly between private farmers and AREs. The outcome of the lending program (Table 5) was that the overall demand for beef and dairy subloans was considerably less than anticipated among the AREs but more than expected among private farmers. The demand for investment subloans relating to annual - 21 - crops and alfalfa was exceptionally higher than anticipated on the part of AREs, but less than expected among private farmers. Subloan demand for permanent crops was less than expected among both types of sub-borrowers. Finally, the demand for agroindustrial subloans was far below anticipations, especially by private farmers. Overall, allocation of Loan proceeds did not depart too widely from the original even division between private farmers and AREs; the latter accounted for about 58% of Bank Loan disbursements. I/ 3.09 As anticipated at appraisal, the bulk of the sublending program was in the coastal area (Table 6), especially as regards the AREs. A rela- tively larger share of the lending to private farmers was in the Sierra and jungle regions. This outcome in part reflects the emphasis put by BAP on supporting the newly formed AREs in the relatively more developed irrigated coastal areas, which specialized in cotton and rice production. Of the various BAP branches, the highest level of activity in terms of both number and value of subloans was in Arequipa, an area specialized in small-scale dairy operations (Table 7). A similar pattern of sublending characterized the large number of subloans made by the Tacna branch. These two branches accounted for 58% of the number of subloans to private farmers and 52% of their value. On the other hand, the high level of sublending by the Huacho, Trujillo and Ica branches was concentrated on financing mechanization of the AREs in their respective areas. These three branches undertook half the number of subloans to AREs and 53% of their value. Overall, these five branches accounted for 68% of the number of subloans and 61% of their value under the project. The remainder was distributed-unevenly--among the thirteen other branches of BAP. 3.10 The size distribution of subloans (Table 8) varies widely between private farmers and AREs. 2/ Almost 80% of the subloans to the former were less than S/. 500,000 and accounted for 40% of the sublending to private farmers. With respect to subloans to AREs, 57% were in the range S/. 1,000,000 to S/. 5,000,000 and accounted for 47% of the sublending to such farms; about the same share was accounted for by the 15% of the ARE subloans over S/. 5,000,000. With the exception of dairy and perennial crop subloans, average subloans resulted notably smaller than appraisal anticipations (Table 9). The increased number of private farmer subloans mainly accounts for higher number of subloans than expected. 3.11 Dairy Production. 3/ Those 2,043 subloans (2,026 for private farmers and 17 for AREs) represented the largest single share of subloan numbers and accounted for 21% of project sublending. Most of the activity was centered in the Arequipa-Tacna region of southern Peru, where small-scale dairying was well established. Other major milksheds, especially Cajamarca and Huancayo, 1/ Of the S/. 3,018.6 million in subloans approved under the project, private farmers accounted for 50%; 88% of the 4,620 subloans were to private farmers. 2/ Owing to the acceleration of inflation during project execution, these data are difficult to interpret. 3/ A more detailed review of project sublending for dairy and beef production is presented in the report prepared by 0. Cordon (Cons) and filed under Loan 933-PE. - 22 - provided relatively little demand for such subloans. The main deviation between appraisal expectations (Table 10) and actual use of subloan funds (Table 11) is the considerably higher share (79%) actually devoted to cattle purchase (11,937 head) and the much lower share used to develop and improve pastures (7%). Few imported cattle were financed under the project. 3.12 Beef Production. There were 568 subloans for beef production; 554 of these were to private farmers (84% of the amount lent) and 14 subloans were to AREs (16% of the amount lent). The major use of subloan funds (53%) was to purchase 9,438 head of cattle, averaging 13 head per subloan to private farmers and 143 head per participating ARE (Table 12). Somewhat over half the cattle purchases were by sub-borrowers in the eastern lowland areas served by BAP's La Merced and Iquitos branches. Relatively little financing of pasture establishment and improvement was carried out; an important part of what is so labelled in reality merely consisted of clearing out jungle trees and bush. 3.13 Other Livestock and Poultry. Subloans for these purposes formed a relatively minor part of the project. Seven subloans (four to AREs and three to private farmers) were granted for sheep ranching (Table 13). These were principally to finance the purchase of 8,828 animals (86% of the amount sub- lent). The amount sublent (S/. 33.7 million) represents barely 1% of the lending program. Six subloans (three to AREs and three to private farmers), totalling S/. 8.9 million, financed purchase of alpaca breeding stock (4,758 head). Finally, eight subloans were granted for broiler and egg production. The total amount sublent for these purposes was S/. 53.3 million. No subloans were made for swine production. 3.14 Annual Crops. These 1,290 subloans (representing 28% of the number of subloans made under the Project) accounted for 53% of the sublending, more than double the share anticipated at appraisal. The demand for these subloans represented 29% of the amount sublent to private farmers (879 subloans), but 76% of the amount sublent to AREs (411 subloans). Purchases of machinery and implements represented a far higher share of subloan financing than was anticipated--90% rather than 67%, on the average, and reaching 96% in the case of AREs (Table 14). Tractor purchasing was the main item financed, accounting for 52% of the amounts sublent to AREs. 1/ 3.15 The main justification presented for this unbalanced financing program is the reportedly deteriorated condition of the farm machinery in- herited from the previous farm owners by the newly formed AREs. Furthermore, tractor imports were relatively low in 1968 and 1969, largely owing to the September, 1967 devaluation as well as uncertainty over the agrarian reform program. On the other hand, the desire of these AREs to rapidly mechanize was rational from their own point of view: 1/ 746 tractors were purchased under the project, nearly three times the number (250) foreseen at appraisal, representing about 12% of the esti- mated national tractor fleet (ca. 6,000) in 1972. - 23 - (a) They were anxious to reduce as much as possible the non-member regularly hired labor force, to avoid having to accept these laborers as ARE members and thus diluting member incomes; (b) Tractor prices were increasing at a higher rate than the general price level, thus providing a hedge against inflation (as was also the case with livestock); and (c) Financing was readily available on very soft terms, i.e., negative real interest rates and no index-linking of subloan principal. 3.16 When tractor imports were discontinued and only locally assembled tractors were available, complaints multiplied of the poor quality of these tractors and the unavailability of spare parts. 1/ While complaints regarding availability of spare parts appear to have been well-founded, it may also have been the case that the new AREs were not fully capable of undertaking proper tractor maintenance and tractor operators were not properly trained, thus the effective utilization of this machinery was hampered and breakdowns became more likely. 3.17 Subloans to private farmers under this category also included a relatively large farm machinery component (76%, of which half was accounted for by tractors), while the remaining 24% of the financing was for land level- ling and on-farm irrigation and drainage works. 2/ Subloans to both classes of sub-borrowers were not accompanied by related programs of technical assistance or extension services to ensure the timely adoption of more productive farming techniques. As the Ministry's agricultural extension services had virtually disappeared (absorbed into the execution of the land reform program), possible weakness in the sublending program arising from the absence of such support services could not be corrected. 3/ 3.18 Permanent Crops. 571 subloans were granted to finance 2,059 ha of permanent crops; 510 subloans were for private farmers and 61 were for AREs (Table 15). The amount for permanent crops represented 10% of the total sub- lending program, reasonably approximating the appraisal anticipation of 12%. Principal items financed were vineyards (22%), apple orchards (20%), orange groves and pepper plantations (14%). Subloans for fruit orchards and vine- yards were concentrated in the Central Zone (between Huacho and Ica), while the pepper plantations were located in the Selva near Pucallpa. No plantings of coffee, tea, cacao or sugarcane were financed under the project. 1/ The issue of tractors financing is also discussed in paragraph 3.24. 2/ These were concentrated in the northern coastal area specializing in rice production. 3/ The absence of effective extension services also appears to have hampered the success of other project sublending activities. - 24 - 3.19 Agroindustry. The agroindustrial component of the project fell far short of the appraisal anticipattons; only four subloans were made, two were for primary processing of coffee beans, one was for a rice mill in Chiclayo and one for a tea packing plant in Lima. The S/. 25 million sublent for these four projects represents less than 1% of the lending program. Several factors are largely responsible Eor this relative inactivity: (a) BAP staff was not adequately trained to prepare and evaluate such subloans; (b) BAP is permitted to lend only to farmers or groups of farmers for agroindustries related to their farm production; (c) The investment climate during much of the project ececiitton period, when e.g., price controls were maintained on most processed foodstuffs, was not propitious. Subsequent to the completion of the project, a trust fund (FRAI) partially financed by AID to rediscount agroindustrial investment loans was established in the Central Bank. While the commercial banking system has actively partici- pated in the FRAI program, the BAP has made only one subloan. Loan Covenants. 3.20 Staffing. The Loan Contract (Section 4.02) required that BAP assign or employ, in consultation with the Bank, a qualified and experienced crodit specialist for the duration of the project. This was a condition of effective- ness. BAP hired a former General Manager for this position. The Bank did not finance this position, or any other technical assistance under the Project, nor was financing provided for any training of BAP staff or equipment for BAP. 1/ 3.21 BAP Operating Costs. The Loan Contract also set out that BAP would (a) review its operational and administrative costs per department, branch office and field office, and express the aggregate of such costs in terms of a percentage of the aggregate principal of loans outstanding, (b) on the basis of this review, send to the Bank by June 30, 1974, a plan to improve its efficiency, and (c) after exchanging views with the Bank, carry out such a plan. An initial review of operating costs was sent to the Bank in October, 1974; it was not accompanied by a program to reduce costs, although the study indicated that 30 out of 48 branch and field offices studied were operating in the red. The Bank suggested that BAP initiate a program to increase staff and administrative efficiency and reduce operating costs over a twelve month period. BAP's response of its intention to reorganize the Huacho and Arequipa offices was not considered by the Bank to comprise a program of sufficient scope to adequately resolve the problem and the Bank continued to express its concern over BAP's inability to generate sufficient income to cover its rising costs. 2/ 1/ It appears that this Credit Specialist was of limited effectiveness (para 3.25). 2/ BAP's operating costs during the project period are further reviewed in paragraphs 5.05-5.08. - 25 - 3.22 BAP Subloan Arrears. The Loan Contract (Section 4.04) provided that the BAP would (a) by December 31, 1974, review its portfolio in arrears, (b) on the basis of the review send to the Bank a plan for writing off, rescheduling or other such appropriate measures in regards of such arrears, and (c) taking into account the Bank's comments on such plan, promptly carry it out. A review of the situation regarding the short-term portfolio was sent to the Bank in December, 1974, while that relating to the long-term portfolio was sent in April 1975. The Bank did not consider the program for correcting the arrears problem to be adequate and continued to communicate its concern to BAP in this respect. In the meantime, under the provisions of an IDB loan (497/SF-PE) BAP was concurrently undertaking a program to reduce its level of arrears. Progress in this regard was slow, although by 1976 the rate had fallen to 18.6% (from the level of 27.1% in mid-1972) and in 1978 fell further to 11.8%. 1/ 3.23 Interest Rates. According to the Loan Contract (Section 3.05), if the rate of inflation would in any fiscal year exceed 9% p.a., upon notifica- tion by the Bank to the Borrower, there would be no further sublending until the Government, the Bank and the Borrower had reviewed the effective interest rate on subloans and agreed upon an increase in these rates. 2/ The trend in inflation and interest rates was as follows: Year Inflation Rate Interest Rate (date implemented) (%) (%) 1973 14.2 9 1974 16.7 10 - 13 1975 21.1 10 - 13 1976 34.6 12 - 14 (7/76) 1977 39.1 14 - 16 (2/77) 1978 57.5 16 (12/77) 29 (8/78) 33 (11/78) 1979 67.0 33 Invariably, adjustments in the BAP lending rate were delayed and the new interest rate levels consistently failed to approach much more than half the rate of inflation. From 1975 onwards, these rates were notoriously negative in real terms; nevertheless, after internal Bank review of the sectoral and country situation and relations with Peru, the Bank chose not to invoke the 1/ Excludes renewals, which accounted for 8% of the portfolio in 1977 and 5% in 1978. In all likelihood, a substantial share of this achievement resulted from the impact of the high rate of inflation on BAP's accounts. 2/ Interest rates on project subloans were the same as those charged by BAP on its other operations. The original (August, 1968) appraisal report recommended that the interest rate for subloans under the project be not less than 16%, unless BAP proposes a system of index-linking satisfactory to the Bank. - 26 - previously cited sublending suspension provision in the Loan Contract. In this respect, it appears that the Government was not appropriately advised of the Bank's interest rate policy position. While the above-cited increases in the BAP lending rate were applied to outstanding as well as new loans, the resulting erosion in the real value of interest receipts and subloan recoveries caused--and is continuing to cause--serious decapitalization of the institi- rLon, even though in 1978 and 1979 its accounting records demonstrate nominal profits. 1/ 3.24 Procurement. As was foreseen, procurement was through existing com- mercial channels for both locally produced and imported goods. Performance was, in general, satisfactory, despite Government policies to progressively constrain imports. A special problem foreseen at appraisal related to the Government's plan to award a contract for a tractor assembly plant to be established in Peru. Once this plant was operating, the Government intended to prohibit imports of models Ln the horsepower range assembled in Peru. Therefore, a side letter was signed in which the Government undertook to ensure that the supply of tractors and spare parts would be adequate to satisfy demand under the Project and that prices charged for tractors assembled in Peru would not 'be substantially higher than those charged for imported tractors the year before domestic production started (para 2.04 (e) above). (All imported agricultural wheel tractors and implements were at that time fully exempt from customs duties.) The complications introduced by accelerating inflation and rapid exchange rate movements make it diffi- cult to assess Government compliance with this undertaking. The question arises as to whether or not the Bank should have accepted a case of virtual reserved procurement as regards tractors. 2/ The impact of this monopoly on the prices of locally-assembled tractors has been a persistent issue with the Government; prices are presently about 50% or more above list prices for similar imported tractors. Moreover, Farmers persistently complain of the shortage of tractor spare parts and of the poor quality of the locally assembled tractors. This monopoly market situation was raised as an issue with the Government during a 1976 pre-appraisal mission of the proposed Sixth Agricultural Credit Project. 3.25 Operating Policies and Procedures. In schedule 4 of the Loan Agree- ment, it was provided that, in lending to cooperatives (AREs), BAP would pay particular attention to ARE management capability and that BAP would supervise the use of subloans in accordance with appropriate banking and agricultural practices. Because the AREs were only recently established and the avail- ability of qualified experienced managers was limited, problems were to be expected. (A small number of subloans to AREs were prematured.) 1/ In a recent IDB report (PR-968-A of 25/9/79) it is reported that losses of BAP's capital and reserves amounted US$91.4 million equivalent during the three years 1976-1978. 2/ The import prohibition applied only to the locally assembled HP range, so farmers moved to duty-free importing of over-sized tractors. - 27 - BAP supervision normally involved a visit by a credit technician to a sub- borrower twice a year. During this visit, the technician verified that funds had been used as foreseen in the subloan. It is not BAP's standard practice that the technician would examine or advise on farm management, production techniques, etc. Unfortunately, the Ministry of Agriculture's extension service had been virtually dismantled at the inittation of the agrarian reform program. This situatton may have made it difficult for farmers to reach the higher output and productivity levels foreseen to be attainable under the investment credit program. In supervising subloans, BAP was also constrained by: (a) the relatively higher priority given to assigning staff to carry out the short-term lending program; (b) the Government's clear preference to promote lending to AREs rather than private farmers; and (c) the restriction on the Project Credit Advisor not to travel outside of Lima, for a period during 1974-75, as he (and the General Manager of BAP) was under indictment (and sub- sequently cleared) for involvement in mismanagement (as Board Member) of the State Marketing Agency (EPSA). IV. PROJECT IMPACT Incremental Output and Productivit.) 4.01 As pointed out in Chapter I, during the period of Project execution the agricultural sector was experiencing profound structural changes. In addition to a rapidly implemented agrarian reform, Government intervention in the pricing and marketing of agricultural inputs and output was notably heavy- handed. These policies tended to discourage the carrying out of investment programs whose gestation periods might be relatively long, while incentives for improved management and adoption of yield-increasing production techniques were depressed. On the other hand, investments in readily marketable, movable assets (e.g., animals and machinery) were encouraged. 4.02 On the basis of the highly aggregated presentation in the Appraisal Report, it is difficult to ascertain incremental output by product line. The with project yield assumptions used in the annual crop model (Table 16) appear not to be excessively high, and are in line with those presently observed on the typical, not exceptionally well-managed coastal medium and large farms. - 28 - 4.03 The Appraisal Report provides limited data on the technological coefficients and production parameters underlying the livestock models (Table 16). The assumptions made in the dairy model are compared below with the results observed on a sample survey carried out by BAP for the PCR mission: Appraisal BAP Sample Survey 1/ Without With Without With Item Project Project Project Project Stocking rate (AU/ha) 0.75 1.5 1.6 2.4 Milk production/cow/lactation (1/year) n.a. 3,660 3,186 3,152 Milk production/ha (1/year) n.a. 5,500 2,921 4,288 Milk production/cow/day (1) n.a. n.a. 11.0 10.9 4.04 These results indicate that (a) participating dairies appear to have entered the project at higher productivity levels than assumed in the Appraisal Report and (b) output increases appear to be almost wholly accounted for by the expansion of herds. 2/ Although a relatively small share of subloan financing was for pasture improvement, stocking rates increased by 50%, so it appears that before project carrying capacities were considerably above stocking rates and/or additional animal feed was subsequently purchased from off the farms. These results are from smallholder dairy operations; there was also a small number (17) of dairy subloans to AREs. Several of these were for sugarcane cooperatives and were primarily intended to produce milk for sale at subsidized prices to cooperative members, while several others were intended to diversify other specialized-crop AREs (e.g. cotton). These enterprises have occasionally not been well-managed or financially viable. 4.05 Technological parameters underlying the beef model, as set out in the Appraisal Report, are compared to the BAP sample survey results below: 1/ Carried out in Arequipa and Tacna, two areas accounting for a major share of dairy sub-borrowing. More detail is presented in Table 17. 2/ Almost 12,000 head of cattle were purchased under the 2,043 dairy subloans, averaging 5 head per private farmer subloan and 53 head per ARE subloan. - 29 - Appraisal BAP Sample Survey 1/ Without With Without With Item Project Project Project Project Stocking rate (AU/ha) n.a. 1.5 0.9 1.2 Calving rate (%) n.a. 70 67 57 Liveweight/head (kg) n.a. 400 380 380 Liveweight/ha (kg) n.a. 224 92 58 1/ Carried out in Pucallpa, Oxapampa and Tarapoto, three areas in the eastern lowlands accounting for an important share of beef production sub-borrowing. More detail is presented in Table 18. These sample survey results suggest that project subloans fell noticeably short of appraisal expectations as regards technological or managerial improvements. Technological Change 4.06 The Project, as designed, did not include a component for carrying out crop and livestock research or strengthening the institutional mechanisms for linking the transfer of more productive technologies and use of improved inputs with the provision of investment credits. Several well-established institutions existed at the time of appraisal which could have been provided some support: crop research had for some time been carried out at the National Agrarian University, while IVITA (Instituto Veterinario de Investigaciones Tropicales y de Altura) worked in livestock research. Unfortunately the Ministry of Agriculture's extension service had been virtually dismanteled during the early 1970s, in the early implementation phase of the agrarian reform, and the Government did not assign a high priority to the generation of new production technologies through public or private agricultural research institutions. BAP's long-standing policy was that it was strictly a financing institution and that it was neither authorized nor adequately staffed to provide technical and/or managerial assistance or advice to sub-borrowers. Farm visits by BAP staff were limited to verifying that subloan funds had been spent as planned; while such visits were supposed to take place twice a year for each subloan, in reality they appear to have been less regular. This absence of adequate technical and managerial assistance to sub-borrowers may have seriously reduced the productive impact of the investment sublending program. Producer Returns 4.07 In the absence of an on-going monitoring system to track the per- formance of sub-borrowers or a sample thereof, the BAP was impeded from assessing project impact and the returns to its sublending program. More- over, the nature of the major project component--annual crops subloans, which were largely for machinery purchase, unrelated to a meaningful on-farm devel- opment program--effectively impedes estimating reliable rates of return. - 30 - Finally, the exceptionally high rate of inflation during the project exe- cution period, which was far from adequately reflected in sublending terms and conditions, further hampers reliable estimates of producers' returns anid seciously distorts the "burden" of subloan debt service. Nevertheless for the purpose of the PCR mission, BAP carried out a sample survey of a limited number of livestock sub-borrowers in several areas (paragraphs 4.03 to 4.05). The rates of return (before debt service, in 1979 prices) estiylAted ori the hasLs o these sample surveys, and a projection of the livestock enterprises to full development, are as follows:_1/ Item Appraisal BAP Survey Dairy 21% 16% (Arequipa) 32% (Tacna) Beef production 17% 9% These rates of return are increased substantially when allowance is made for debt service, which was not effectively readjusted for inflation. The more favorable return in Tacna derives mainly from the relatively lower investment per ha there (mostly owing to less machinery purchase and construction than Arequipa), accompanied by a relatively larger expansion of milk production (82% in Tacna compared to 44% in Arequipa). 4.08 Rate of return estimates are available only for dairy and beef sub- loans, which accounted for about 31% of total project stihlerding. The bulk of project sublending--for farm mechanization--was largely unrelated to spe- cific on-farm development programs and output expansion. 2/ To some extent this machinery investment represented scheduled, or even deferred, replace- ment which otherwise might not have taken place; therefore, it prevented possible declines in output. 3/ The data needed to permit a rate of return calculation for such replacement investments are not available. 4.09 In these circumstances it appears reasonable to expect that the project's economic rate of return has been appreciably below the 26% estimated at appraisal, although probably not below the country's estimated opportunity cost of capital of 12%. With the soft sublending terms and conditions--as accelerating inflation rapidly reduced real debt service burdens--sub-borrowers did not need to pursue sound management techniques or adopt improved technologies which would result in output increments in order to service subloans. The uses to which farmers put the subloans appear to be in line with the economic signals prevailing at the time. They benefitted from a substantial transfer of wealth unmatched by pari passu increase in financial liabilities. I/ Incremental herd values are excluded, as they were not included in the Appraisal Report estimates. For more detail see report prepared by 0. Cordon (Consultant). 2/ To determine the need for farm tractors, BAP applied a standard co- efficient (62.5 HP per 100 ha). 3/ See, however, the comments in fungibility in paragraph 5.11 and 5.12. - 31 - V. INSTITUTIONAL DEVELOPMENT BAP Staffing and Training 5.01 BAP was committed by the Loan Agreerent (Section 4.02) to hire, in consultation with the Bank, for the duration of the Project a qualified and experienced credit specialist to assist in improving the agency's appraisal procedures for medium and long-term investment and in the coordination of its departments and branch and field offices. 1/ The hiring of this specialist was a condition of Loan effectiveness. The first slate of candidates for the position was not acceptable to the Bank. Subsequently, the BAP proposed as a candidate--and the Bank accepted--a former Manager of the BAP (para 3.20). 5.02 In addition to hiring the Credit Advisor, BAP established a "World Bank Project Office" in Lima, staffed by four professionals (three agricul- turalists and one veterinarian). The main tasks of this group were (a) to train BAP field staff in project preparation and evaluation and (b) to back- stop field staff via periodic visits to branch offices to assist in clearing backlogs of subloan requests. Although four previous projects had been carried out, it appeared that BAP staff were not sufficiently familiar with the Bank style of incremental output and cash flow analysis as a basis for preparing and evaluating subloan proposals. This unfamiliarity resulted in some delays, especially in the early project execution period. Although BAP's "World Bank Project Office" sought to alleviate this problem, its success appears to have limited in this respect, since throughout project supervision, Bank staff commented on the apparently inadequate training of BAP field staff in subloan preparation and evaluation. The BAP staff comprising the Project Office was subsequently dispersed at project completion, and the Unit Chief was put in charge of the IDB Project Office in BAP. 5.03 BAP's staff continued to increase during project execution, resulting in rising administrative costs relative to its average portfolio: 8/31/72 8/31/79 % Change Professional staff 337 524 + 55 Administrative and Support staff 1722 2,489 + 45 Total 2,059 3,013 + 46 Of the total increase in personnel, about one-fifth (187) was in the profes- sional ranks, which grew at a slightly higher rate than the administrative and support staff. The expansion of the professional staff at the Main Office was relatively modest (32% or 27 positions); most of the increase took place in the branch and field offices (Table 19). 1/ To original (August 1968) appraisal report recommended BAP hire an accounting consultant and an expatriate livestock technician. - 32 - 5.04 This staff expansion accompanied an increase in BAP's lending; its outstanding loan portfolio (own resources) rose from S/. 6,173 million in March 31, 1972 to S/. 28,648.5 mmllion on December 31, 1978. However, adjusted by the exchange rate, the figure actually fell from US$159.5 million to US$150.8 million equivalent. Moreover, BAP's short-term lending rose more rapidly than its long-term lending from about 70% of its portfolio on March 31, 1972 to about 80% of its portfolio at the end of 1978. While the number of outstanding loans rose from 70,374 at the end of March 1972, to 118,371 on December 31, 1978, much of this increase took place in short-term lending, for which staff demands are relatively less than for preparing, evaluating and supervising investment subloans. Financial Performance 5.05 BAP's financial position deteriorated at the start of the project, as its losses rose from S/. 23 million in 1973 to S/. 45 and S/. 46 million in 1974 and 1975, respectively. The situation improved (nominally) in 1976 and 1977 when losses decreased to S/. 2.7 million and S/. 4.6 million, respectively. Finally, in 1978 and 1979, BAP reported profits of S/. 130.6 million and S/. 562 million, respectively. BAP's capital and reserves rose from S/. 2,347 million at December 31, 1973, to S/. 14,662 at December 31, 1979. However, when de- flated by the sol/dollar exchange rate, these amounts represent a decline from US$60.6 million equivalent in 1973 to US$58.6 million equivalent in 1979. This decline has occurred despite substantial capital contributions by the Government over this period; in the three years 1976-1978 some US$91 million of BAP's capital was lost owing to inflation/devaluation. 5.06 Receipts and expenditures as a percentage of the annual average port- folio during 1976-78 were as follows: 1976 1977 1978 Receipts 9.7 13.7 16.6 Financial 9.2 13.2 16.0 Other 0.5 0.5 0.6 Expenditures 9.8 13.8 16.2 Financial 1.2 2.8 4.4 Administrative 6.9 7.6 8.9 Provisions 1.7 3.4 2.9 Total (0.1) (0.1) 0.4 The recent increases in BAP's financial receipts, arising from higher interest rates on BAP loans, have barely more than kept pace with rising costs. The increasing financial costs in large part reflect the higher rediscount rate for BAP at the Central Bank, which is the major source of BAP resources for short-term lending. The relative increase in adminis- trative costs raises concerns about BAP staff efficiency. 1/ 1/ The provision for bad debts was abnormally low in 1976 and even the amounts for 1977 and 1978 are most likely insufficient to compensate for the effect of inflation. - 33 - 5.07 Despite the considerable nominal increase in BAP's lending during the period of project execution, the real flow of credit resoi.urceS to the sector appears to have actually declined. In a very summarized form, the BAP balance sheets at the end of 1974 and at the end of 1978, in both nominal terms and deflated by the CPI (Consumer Price Index, 1974 = 100) were as follows: Assets Liabilities and Equity 1978 1978 1974 Nominal Deflated 1974 Nominal Deflated --------S/. M---------- ---------S/. M ---------- Available Checking and Funds 913.6 1,675.8 465.5 Savings Deposits 1,448.9 1,853.0 514.7 Loans Borrowings 11,220.8 30,442.9 8,456.4 Short- term 6,989.5 24,637.1 6,843.6 Capital and Long- Reserves 2,658.1 10,284.5 2,856.8 term 6,620.8 11,168.5 3,102.4 Other 803.9 5,099.1 1,416.0 Total 15,327.8 42,580.4 11,827.9 Total 15,327.8 42,580.4 11,827.9 In real terms not even the short-term lending portfolio grew; the long-term portfolio fell by more than 50%. Resources mobilized through deposits fell by two-thirds. Capital and reserves increased by only about 7%, despite the very substantial Government contributions to the BAP's capital during this period; merely in the three years 1976-1978 the Government's contribution to BAP was the equivalent (in 1974 soles) of about S/. 5,620 million--twice the BAP's capital and reserves at the end of 1978. The lending of these resources to BAP clientele at highly subsidized interest rates represents a substantial transfer of wealth to a relatively limited group of beneficiaries, resources which will never be recovered. It is very misleading to label such a resource transfer program "credit". 5.08 During project execution the Bank repeatedly expressed to the BAP its concern over the losses being incurred by BAP and their impact on BAP's financial viability. It was only when significant interest rate adjustments were introduced in 1978 that the BAP began to report profits. Nevertheless, owing to inadequate adjustments in interest rates and failure to apply account- ing procedures which compensate for inflation, such profits are only nominal and BAP's true financial position will continue to be precarious. Fortunately, for BAP, the Government is responsible for servicing the Bank Loan, thus distributing the debt burden over the general economy. 5.09 Subloan Recoveries. It was anticipated in the Appraisal Report, and set out in the Loan Contract (Schedule 4), that subloans would be for periods of not less than 5 years. Maximum grace periods could vary from 2 to 7 years, and maximum subloan term could range from 8 to 15 years, depending - 34 - on the type of production activity being financed. As it turned out, subloan terms imposed by BAP were appreciably stiffer than foreseen. According to the project cash flow in the Appraisal Report, it was anticipated that by year 6 (1979) about 12% of subloan principal would have been repaid; in reality, 41% had been repaid (Table 20) and, at the current rate, full repayment should be obtained by 1982, rather than the Appraisal Report's projection of 1991. Owing to inflation, however, when measured in soles of 1974 purchasing power, the amount recovered through 1979 represented only about 12% of the amount lent; adding interest receipts increases the figure to about 20%. In view of the anticipated inflation up to 1982/3, it appears unlikely that subloan debt service, in real terms, will represent much more than 25%-30% of the amounts lent. 5.10 Subloan Arrears. Relatively few subloans have wholly expired under the project. Principal and interest collections in 1979 are summarized below and presented in more detail in Table 21: Collected Overdue Total (1979) (12/31/79) Principal S/. Million 544.8 56.7 601.5 90.6 9.4 100.0 Interest S/. Million 531.1 85.4 616.5 86.1 13.9 100.0 At the end of 1979 arrears of principal repayment on expired subloans amounted to about S/. 1.1 million, of which about half (S/. 0.6 million) were less than 90 days overdue at the end of 1979 (Table 21). Arrears in repayment of prin- cipal on outstanding subloans at the end of 1979 (S/. 55.6 million) represented about 10% of the amounts due in 1979; about one-third of these arrears were less than 90 days overdue. Interest payments overdue amounted to S/. 85.4 million at the end of 1979, representing about 16% of the interest payments actually collected during 1979; again about one-third of these arrears were less than 90 days overdue. The status of the project portfolio is presented in Table 22. Excluding those subloans granted extensions of repayment periods, at the end of 1979, 78% of the project portfolio was being recovered with no problems. This is the average for BAP's total investment lending portfolio, of which the Bank project accounted for about 15% at the end of 1979. 1/ 1/ Thus, by the criterion of subloan recovery, it does not appear that BAP made a conscious effort to limit project sublending to only the "better" sub-borrowers. - 35 - Fungibility of Project Funds 5.11 Subloan supervision carried out by BAP was intended to assure that subloan finance would not be diverted to uses other than those specified in the farm investment plans. BAP appears to have conscientiously carried out this task and diversion was not a problem under the project. However, there is a likelihood that subloan funds did substicute for other resources which sub-borrowers put to other uses. Such substitution was specially likely among the AREs, which have tended to increase member returns, e.g., increasing wage payments to members, rather than channelling enterprise earnings into reinvestment. The notably negative real interest rates which persisted throughout the project period also provided a strong incentive to maximize debt financing, rather than encourage c:lf-flnancing. Finally, at the present time, term deposits in the BAP--and elsewhere--pay a higher rate of interest than the cost of borrowing at BAP, which also encourages use of borrowed funds in lieu of self-financing. 5.12 At the sectoral level, it may well be that some substitution took place, as the Government would most likely have provided resources for term lending to the newly-formed politically sensitive AREs regardless of the availability of funding under a World Bank loan. On the other hand, during the first few years of Loan implementation, the Government assigned a rela- tively low priority to attending to the credit and technical assistance needs of the private farming sector. The requirement under this project that half of the lending be to private farmers very likely made resources available to them that otherwise would not have been available in the prevailing political environment. Reporting and Audits 5.13 A system of quarterly reports was established under the project. Compliance on the part of BAP was reasonably good in this respect, although reports were occasionally slow in arriving to the Bank. The reporting system was solely concerned with the evolution of subloan approvals and disbursements. There was no planning process by which lending programs were established in accordance with targets in terms of output objectives, etc. There was not built into the project a means by which sub-borrower performance--even on a sample basis--was tracked and assessed. While the BAP sent to the Bank a relatively comprehensive final project evaluation report in 1979, in the absence of any on-going monitoring system, this report presented relatively few quantitiative measures of project impact. 1/ 5.14 Annual audit reports of BAP were sent to the Bank in accordance with Section 5.03 of the Loan Contract, although not always within the four month period set out in that Section. The Bank was especially interested in reviewing through these reports BAP's performance in improving its situation of loan arrears and providing for operating costs. 1/ The original (August 1968) appraisal report recommended that BAP establish a record system whereby the results of the credit project could be sampled and analyzed. - 36 - VI. BANK PERFORMANCE 6.01 Owing to economic and political factors beyond its control, the Bank was unable to complete processing of the project originally appraised in 1967. The project which was subsequently appraised and carried out omitted several aspects of the originally appraised project which could have been important: (a) There was no financing to support technical assistance to BAP; I/ and (b) The original project anticipated possible index-lending of subloan principal. 6.02 Moreover, several aspects of the project could have been strengthened: (a) Provision for financing the training and/or equipping of BAP staff; (b) Arrangements for the provision of adequate technical services to sub-borrowers and provision for strengthening the research-demonstration-extension system so as to encourage the use of credit being accompanied by the adoption of improved technologies and inputs. 2/ (d) Subloan preparation and evaluation procedures; and (e) Provision for a reporting system which would permit a quantitative assessment of sub-borrower performance and project impact. 3/ I/ The original project contemplated technical assistance in accounting procedures and a livestock technician. At the time of the reappraisal the political environment had changed and the Government was in general not receptive to expatriate technical assistance. Moreover, there was provision of technical assistance to BAP under ongoing IDB loans. 2/ On the other hand, the highly distorted economic signals transmitted to producers--especially the very subsidized terms and conditions of sublending--might still have swamped and vitiated the possible impact of such technical services. 3/ Establishment of such a system was foreseen in the 1968 Appraisal Report. - 37 - 6.03 Despite the Bank-wide changes in organizational structure in 1972, and the reorganization of the LCP Agricultural Projects Divisions in 1975, there was reasonable continuity of supervision. Supervision reports reason- ably reflected the progress of project execution and the problems supervIsion missions observed. Perhaps surprisingly, the project was never designated a "Problem Project" for the SVP review. Hindsight suggests that the Bank should have more vigotoucciy pursued compliance with the covenant relating to interest rate adjustments and encouraged BAP to accelerate staff training and provide adequate technical assistance to sub-borrowers. Assessment of Bank/ country relations played an important role in determining the stance taken by Bank management in this respect. However, Bank leverage under the project was perhaps more than appreciated; of total BAP medium- and long-term lending during the project execution period (Table 23), that carried out under this project was about 41%. As the project was nearing completion, the Bank and the Government began conversations regarding a possible repeater project. Unfortunately, over the period from late 1976 to early 1980 such a project became repeatedly mired as country economic problems intensified and a series of project issues arose regarding tractor procurement and interest rate policies -- issues that should have been more straight-forwardly faced up to earlier, in the course of the execution of the Fifth Project. VII. CONCLUSIONS AND LESSONS LEARNED 7.01 The sectoral policy environment at the time the project was imple- mented was a difficult one. The sector was experiencing notable structural changes and the economic signals transmitted to both private and ARE producers were often confusing and inconsistent. With the advantage of hindsight, we can observe that the stage was being set for the severe economic crisis which arrived in 1976. In view of the institutional environment prevailing in the early 1970s, the Bank may have had relatively little freedom of action in terms of the agricultural credit policy changes which it could hope to have implemented in conjunction with the Loan. Nevertheles3, policy issues might have been given greater emphasis at the time the project was appraised. The Loan was viewed as an important step in renewing Bank contacts with a Government which had undertaken wide-ranging structural reforms. External pressures had for sometime worked against Bank operations in Peru. With this Loan, the Bank sought to demonstrate its willingness to accommodate diverse policy orientations and work with such "reformist" Governments. Had the Bank more vigorously sought to achieve improved agricultural credit policies under this project, the possibility of an agricultural credit operation may have evaporated. As this was number five of a series of projects through which the Bank had sought to promote sound sectoral credit policies and strengthen the financial viability and autonomy of BAP, Bank expectations in this respect were high. However, in view of BAP's very limited autonomy on such policy issues as interest rates, the Bank could achieve relatively little by leaning very heavily on BAP in the absence of a firm Government commitment in this regard. In view of the limited production impact resulting from the credit program, the failure to achieve improved sectoral credit policies, and the negligible institutional development, it might have been better not to make a Loan until a more favorable policy environment were in evidence. - 38 - 7.02 The sectoral lending policies that the Government imposed on BAP-- and any other financial institution which might lend to the sector--resulted in establishing BAP as the only financial institution lending for agriculture. Thus BAP had to take over lending operations which private banks had earlier carried out -- mostly for export crops now cultivated by AREs. BAP became virtually the only short-term lending institution for the sector, diverting staff and resources from its development lending program. Such considerations support the importance for the Bank to analyze such credit programs within the context of rural financial systems, the stress put on the structure and performance of financial institutions, the importance of strengthening farm technical assistance services and the need to assure that input, output, and financial markets are permitted to transmit appropriate signals to farm producers. I/ On the basis of experiences such as those encountered in this project, the evidence suggests that the channelling of Bank loans through a sole--usually Government-owned--lending agency, rather than encouraging wider participation of both public and private lending institutions, is often not the most effective approach for improving sectoral credit policies and strengthening rural financial markets. Future agricultural credit projects in Peru should incorporate this experience. 1/ Cf. AGR-Interim Guidance Note No. 3, "Sector Lending Approaches to Agricultural Credit Projects". - 39 Table 1 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Cumulative Loan Disburs,.rents IBRD Fiscal Year Appraisal and Quarter Estimate Actual - - - - US$ ('000) - - - - 1974 12/31/73 200 3/31/74 400 6/30/74 600 1975 9/30/74 800 12/31/74 1,800 900 3/31/75 3,300 2,600 6/30/75 5,000 3,400 1976 9/30/75 7,000 4,600 12/31/75 9,000 5,000 3/31/76 11,000 7,600 6/30/76 13,000 10,100 1977 9/30/76 16,000 11,700 12/31/76 18,000 14,100 3/31/77 20,000 15,900 6/30/77 22,000 17,700 1978 9/30/77 24,000 19,500 12/31/77 25,000 20,800 3/31/78 - 22,000 6/30/78 - 23,700 1979 9/30/78 - 24,800 12/31/78 - 25,000 March 25, 1980 - 40 - Table 2 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Subloan Approvals US$1/ Year No. '000 Soles Total Average 1974 307 108,821.3 2,811,921 9,159 1975 907 442,432.6 11,033,233 12,165 1976 1,111 738,223.8 13,253,570 11,929 1977 1,328 805,434.2 9,111,246 6,861 1978 967 923,646.3 5,943,670 6,147 Total 4,620 3,018,558.2 42,153,640 9,124 1/ Conversion calculated by using the following exchange rates, which are the weighted averages derived from quarterly Bank loan disbursement data: 1974 S/.38.7/US$ 1975 S/.40.1/US$ 1976 S/.55.7/US$ 1977 S/.88.4/US$ 1978 S/.155.4/US$ March 25, 1980 - 41 - Table 3 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Sublending by Category Appraisal Actual S/.M % S/.M % I. Dairy and Livestock (a) Private Farmers 277.6 21 831.9 28 (b) AREs 277.6 21 114.3 4 II. Annual Crops and Alfalfa (a) Private Farmers 160.8 12 439.0 15 (b) AREs 160.8 12 1,161.9 38 III. Permanent Crops (a) Private Farmers 77.6 6 215.0 7 (b) AREs 77.6 6 121.6 4 IV. Agroindustries, etc. (a) Private Farmers 129.6 10 11.1 1/ (b) AREs 129.6 10 123.7 4 Total 1,291.2 100 3,018.5 100 (s) Private Farmers 645.6 50 1,497.0 50 (b) AREs 645.6 50 1,521.5 50 1/ Less than 0.5 April 10, 1980 - 42 - Table 4 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Original and Final Loan Allocation by Category and Type of Sub-borrower Category Original Final US$'000 1. Beef and Dairy Production 10,800 7,982 (a) Private farmers 5,400 7,003 (b) AREs 5,400 979 2. Annual Crops and Alfalfa 6,200 13,316 (a) Private farmers 3,100 2,089 (b) AREs 3,100 11,226 3. Permanent Crops 3,000 2,143 (a) Private Farmers 1,500 1,173 (b) AREs 1,500 970 4. Agroindustry, Poultry, Swine and Sugar Cooperatives 5,000 1,559 (a) Private farmers 2,500 121 (b) AREs 2,500 1,438 TOTAL 25,000 25,000 (a) Private farmers 12,500 10,386 (b) AREs 12,500 14,614 April 11, 1980 - 43 - Table 5 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Number and Amounts of Subloans by Purpose No. of Subloans Amount S/. '000 1. Dairy 2,043 627,611 (a) Private farmers 2,026 580,780 (b) AREs 17 46,831 2. Beef cattle 568 293,590 (a) Private farmers 554 246,123 (b) AREs 14 47,467 3. Sheep 7 44,851 (a) Private farmers 3 5,027 (b) AREs 4 39,824 4. Alpacas 6 8,850 (a) Private farmers 3 577 (b) AREs 3 8,273 5. Poultry, etc.1/ 9 54,064 (a) Private farmers 4 10,574 (b) AREs 5 43,490 6. Annual crops 1,290 1,627,112 (a) Private farmers 879 439,889 (b) AREs 411 1,187,223 7. Permanent crops 571 304,816 (a) Private farmers 510 189,529 (b) AREs 61 115,287 8. Alfalfa and other pastures 122 31,789 (a) Private farmers 115 26,211 (b) AREs 7 5,578 9. Agroindustry 4 25,875 (a) Private farmers --- (b) AREs 4 25,875 TOTAL 4,620 3,018,558 (a) Private farmers 4,094 1,498,710 (b) AREs 526 1,518,848 1/ Includes one loan for raising rabbits (S/.745,000). March 28, 1970 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Geographical Distribution of Subloans AREs Private Farmers Total Region No. % S/.'000 % No. % S/.'000 % No. % S/.'000 % .I Coast 411 78 1,292,476 85 2,240 55 891,116 60 2,651 57 2,183,592 72 Sierra 75 14 127,994 8 1,186 29 355,215 24 1,261 27 483,209 16 Jungle 40 8 101,059 7 668 16 250,698 16 708 16 351,757 12 Total 526 100 1,521,529 100 4,094 100 1,497,029 100 4,620 100 3,018,558 100 March 25, 1980 FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Subloans by Category and BAP Branch I II III IV - Total Grand Total Private Private Private Private Private Branch Farmers AREs Farmers AREs Farmers AREs Farmers AREs Farmers AREs ------------------ S/.'000 ------ ------------------- No. S/.1000 No. S/.1000 No. S/.*000 Arequipa 455,218 4,943 120,739 15,954 19,823 - 513 - 1,668 596,293 22 20,897 1,690 617,190 Ayacucho 360 - 1,140 6,077 - - - - 2 1,500 7 6,077 9 7,577 CaBete 512 1,000 28,800 152,766 8,088 6,680 - - 43 37,400 55 160,446 98 197,846 Cajamarca 16,793 1,456 402 9,046 - - - - 40 17,195 9 10,502 49 27,697 duzco 40,018 18,035 31,899 54,643 2,412 - - 1,782 83 74,329 43 74,460 126 148,789 Chiclayo 5,605 1,000 47,910 80,443 1,472 - - 8,451 168 54,987 22 89,894 190 144,881 Huacho 1,240 16,395 17,348 216,500 52,667 51,973 - 53,975 128 71,255 93 338,843 221 410,098 Huancayo 15,301 1,319 6,175 4,981 - - - - 24 21,476 7 6,300 31 27,776 Ica 2,888 7,097 36,062 218,237 24,646 47,859 - 1,246 67 63,596 122 274,439 189 338,635 Iquitos 74,804 955 14,902 1,328 42,483 - 999 - 411 133,188 5 2,283 416 135,471 Jaen 7,949 47 13,589 14,013 623 - - - 93 22,161 7 14,060 100 36,221 La Merced 33,630 - 16,288 24,986 3,797 - - 12,000 109 53,715 4 36,986 113 90,701 Lima 3,487 1,002 1,020 34,959 3,820 - 7,157 - 19 15,484 13 35,961 32 51,445 Piura 8,326 14,634 19,545 149,281 - - - - 146 27,871 43 163,915 189 191,786 Puno 37,471 46,426 8,322 12,873 - - - - 94 45,793 14 59,299 108 105,092 Tacna 124,743 - 41,524 8,729 20,256 15,100 2,418 - 710 188,941 4 23,829 714 212,770 Tingo Maria 1,640 - 1,382 7,066 1,144 - - 3,125 13 4,166 8 10,191 21 14,357 Trujillo 1,918 - 31,999 150,035 33,762 - - 43,112 276 67,679 48 193,147 324 260,826 TOTAL 831,903 114,309 439,046 1,161,917 214,993 121,612 11,087 123,691 4,094 1,497,029 526 1,521,529 4,620 3,018,558 March 27, 1980 -46 - Table 8 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Size Distribution of Subloans by Type of Sub-borrower AREs Private Farmers Total Size Range (S/.) No. S/.'000 No. S/.'000 No. S/.'000 Up to 100,000 5 298 1,200 69,787 1,205 70,085 100,001 - 200,000 9 1,355 862 128,236 871 129,591 200,001 - 500,000 38 14,019 1,186 403,947 1,224 417,966 500,001 - 1,000,000 93 72,025 579 410,895 672 482,920 1,000,001 - 2,000,000 149 213,342 193 244,293 342 457,635 2,000,001 - 5,000,000 152 504,747 67 193,457 219 698,204 5,000,001 - 10,000,000 61 411,883 7 46,414 68 458,297 10,000,001 - 15,000,000 10 122,977 - - 10 122,977 15,000,001 - 20,000,000 5 88,180 - - 5 88,180 20,000,001 - 30,000,000 4 92,703 - - 4 92,703 More than 30,000,000 - - - - - - Total 526 1,521,529 4,094 1,497,029 4,620 3,018,558 April 11, 1980 - 47 Table 9 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Number and Average Size of Subloans Appraisal Estimate Actual No. of Average Subloan No. of Average Subloan Type of Subloan Subloan S/. 000 US$I Subloan S/. 000 US$2 Dairy 1,500 124.0 3,204 2,043 307.2 4,156 Beef/Sheep/Alpaea 400 920.8 23,793 581 597.7 8,086 Pigs/Poultry/ 10 10,404.8 268,858 8 6,664.9 90,163 Annual Crops and Alfalfa 50 6,436.0 166,305 1,412 1,174.9 15,894 Perennial Crops 600 259.2 6,698 571 533.8 7,222 Agroindustry 30 5,184.0 133,953 4 6,468.8 87,510 "OTAL 2,590 498,535.9 12,882 4,620 653,367.5 8,839 1/ Appraisal exchange rate: US$ = S/. 38.70. f/ Average exchange rate for Bank Loan disbursement: US$ = S/. 73.92. 3/ Only poultry subloans were made under the project. April 11, 1980 - 48 - Table 10 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Investments Financed by Subloans Appraisal Actual % (%) Model and Item: 1. Dairy (a) Pasture Development 21 7 (b) Buildings 16 12 (c) Machinery 7 - (d) Fence and Water Supply 10 0.4 (e) Breeding Stock 46 79 (f) Other - 2 100 100 2. Beef (a) Pasture Development 23 5 (b) Buildings 8 33 (c) Machinery 28 - (d) Fence and Water Supply 9 5 (e) Breeding Stock 32 53 (f) Other 5 100 100 3. Annual Crops (a) Drainage 7 0.4 (b) Irrigation 8 3 (c) Land Levelling 18 6 (d) Machinery 67 90 (e) Other - 1 100 100 April 11, 1980 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Subloans for Dairy Enterprises Type of Sub-borrower and No. of Subloans AREs: 17 Private Farmers: 2,026 Total: 2,043 No. of No. of No. of Purpose head Amount % head Amount % head Amount % (S/.'000) (S/.'000) (S/.'000) Animal purchase 902 31,242 67 11,035 464,214 80 11,937 495,456 79 4 Pasture establishment 3,333 7 37,703 6 41,036 7 Purchased feed 160 0.3 14,596 3 14,756 2 Fencing 481 1 1,949 0.3 2,430 0.4 Other buildings 11,615 25 62,318 11 73,933 12 TOTAL 902 46,831 100 11,035 580,780 100 11,937 627,611 100 (D April 3, 1980 F PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Subloans for Beef Enterprises Type of Sub-borrower and No. of Subloans AREs: 14 Private Farmers: 554 Total: 568 No. of No. of No. of Purpose head Amount % head Amount head Amount (S/.'000) (S/.'000) (S/.'000) Animal purchase 2,005 22,338 47 7,433 134,464 55 9,438 156,802 53 C) Pasture establishment 440 2 12,722 5 13,162 5 1 Purchased feed 10,617 22 1,790 0.5 12,407 5 Fencing 1,672 3 12,797 5 14,469 5 Other buildings 12,400 26 84,350 34 96,750 33 TOTAL 2,005 47,647 100 7,433 246,123 100 9,438 293,590 100 April 3, 1980 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Subloans for Sheep Enterprisesi' Type of Sub-borrower and No. of Subloans AREs: 3 Private Farmers: 3 Total: 6 No. of No. of No. of Purpose head Ariount % head Amount head Amount % (S/.'000) (S/.'000) (S/.'000) Un Animal purchase 5,808 25,162 88 3,020 3,699 74 8,828 28,861 86 Pasture establishment 3,485 12 80 1 3,565 10 Fencing - - 1,248 25 1,248 4 Other buildings - - 0.3 50 0.3 TOTAL 5,808 28,697 100 3,020 5,027 100 8,828 33,724 100 1/ Excludes one subloan of S/.11,127,000 to a coastal breeding/fattening enterprise. April 3, 1980 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Subloans for Annual Crop Enterprises Type of Sub-borrower and No. of Subloans AREs: 411 Private Farmers: 879 Total : 1,290 No. No. No. tractors Amount % tractors Amount % tractors Amount % (S/.'000) (S/.'000) (S/.'000) Purpose Tractors 587 613,500 52 177 167,477 38 764 780,977 48 Vehicles 150,023 13 41,553 9 191,576 12 Motors 56,343 5 38,082 9 94,425 6 Implements and other machinery 309,272 26 88,779 20 298,051 24 Storage facilities 1,025 0.2 1,025 0.1 Silos 546 0.1 942 0.2 1,488 3 Irrigation 25,044 2 26,780 6 51,824 1 River protection 19,326 1 312 0.1 19,638 1 Drainage 1,035 0.1 3,660 1 4,713 0.4 Land levelling roads, etc. 12,116 0.9 71,279 17 83,395 6 TOTAL 587 1,187,223 100 177 439,889 100 764 1,627,112 100 April 3, 1980 - 53 - Table 15 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Subloans for Permanent Crops AREs: 61 Private Farmers: 510 Total: 571 Crops ha Amount ha Amount ha Amount (S/.'000) (S/.'000) (S/.'000) Vineyards 257 25,009 243 40,867 500 65,876 Apples 211 30,644 146 28,995 357 59,639 Asparagus 40 2,273 220 15,109 260 17,382 Oranges 125 24,905 126 20,135 251 45,040 Olives 75 6,900 170 11,683 245 18,583 Peppers - - 128 41,687 128 41,687 Pears 78 10,999 34 4,771 112 15,770 Mandarin oranges 25 8,189 30 5,878 55 14,067 Peaches 25 2,754 26 9,454 51 12,208 Pecans 13 2,175 23 3,516 36 5,691 Avocado 10 1,439 15 2,219 25 3,658 Lemon - - 13 2,095 13 2,095 Maracuya 12 1,632 12 1,632 Mango 3 600 3 600 Plums 3 67 3 67 Other fruits 8 821 8 821 TOTAL 859 115,287 1,200 189,529 2,059 304,816 April 3, 1980 -54 - PERU Table 16 FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Appraisal Report Yield, Cost and Price Assumptions Cost/Hectare (SI.) Total Invaestment Annual Fare Gate Price Enterprises Yi-ld/Hectare . Year 1-3 Costs Year (SI.) Plantation Establishment Bananas 20.0 m' 34,750 20,440 2-12 2,000/at Mango 11.3 m/ 21,525 17,535 10-25 5,OO/et Mechanization and Reclamatio-3 Cotton (unginned) 2.1 mt!/ ) 16,053 1 1,1301.t Sorghum 3.4 nt ) 6,220 1 2,00/m.t 16,032 Corn (hybrid) 3.9 nt ) 9,200 1 3,450/mt Potat. 13.5 mt )) 18,845 1 2,10O/mt Pasture Establishment for Beef Production Annual after (Stocking Rate 1.5 AU/ha)/ 178 kg liveweight Year 8 Pasture establishment incluiding fertilizer and weed control 2,180 1,021 6 Fencing 390 Fencig 39037. 7/kg fun breeding helfers Rench structures including water supply 1,223 Machinery 2,626 Livestock purchases 3 090 Pasture E.tablihment for Milk Production Annual after (Stocking Rate 1.5 AUlha)5/ 3,660 liter per cow/year Year 6 5,500 liter per ha/year Pasture establitshment including 5/liter fertiltzer and used control 1,968 Pe ii z r e84 1,572 :Y 1,575 for 4 days old male calves 20,500 for breeding heifers Ranch structures including water supply 1,565 23/kg for culled cosm Machinery 656 25/kg f or 100 kg finished pigs Livestock purchases 4 334 9.. 20/kg for 200/250 kg culled asw and boare Pigs Breeding/Fattening 315 Breeding Sous Unit Annual after 462,000 kg tiveweight Year 3 per farm Parmnstcucturn 4,341,496 5,426.326 6/ 25/kg for 100 kg finished pigs Machinery and equipment 1,163,250 20/kg for 200/250 kg called so"e and boars Livestock purchases 2,755,560 Feed and medicines 2,415,189 Labor and maintenance 1 148 003 1/ Yields: first year 307., second year 707., third to twelfth years as shown. 2/ Vo yield until fifth year. Yields gradually increase from fifth to tenth years and then level off as shown. 3/ Under annual coste, depreciation on tractor and implements is not included. / Yields increase to rate shown in fourth year. 5/ Alt Animal 1 year and older, not including calves. 2/ At full development building up gradually. Source: Report 55e-PE. February 4, 1980 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Summary Technical Parameters, Dairy Subloans Sample 1/ Before Project With Project (Year 4) Itam Arequipa Tacna Arequipa Tacna (36) (36) (36) (36) Calving Rate (%) 78.4 80.0 75.6 80.0 Calf Mortality (%) 8.0 7.0 7.0 6.0 Adult Mortality 3.0 3.0 2.0 2.0 Culling Rate (male calves, %) 100 100 100 Cows 11.3 7.9 17.5 12.1 Breeding 13.2 9.3 21.9 14.5 Bulls 0.4 0.5 0.8 0.5 Breeders/Bull 33.0 18.6 27.4 29.0 Cows in milk: No. 8.7 6.3 13.7 9.8 Cows in milk: % 77 79 78 81 Milk Prodiction/cow lactation (1 ) 3,777 2,287 3,476 2,676 Milk Production/ha (1 ) 3,692 1,600 5,351 3,497 Milk Production/Cow/Day (1 ) 12.6 7.6 11.6 8.9 Total Milk Production 32,860 14,408 47,621 26,225 Stock Rate (AU/ha) 1.8 1.1 2.7 2.3 Average Subloan (S/. 000, 1979 prices) 2,096.1 1,202.4 IRR: Without Incremental Herd 16% 35% With Incremental Herd (Year 12) 19% 32% 1/ For more detail see report by 0. Cordon (Consultant). April 11, 1980 Table 18 - 56 - PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Summary Technical Parameters, Beef Production Subloans Sample 1/ Item Without Project With Project (Year 4) Calving Rate (%) 67 57 Calf Mortality (%) 6 5 Adult Mortality (%) 3 3 Cows 45 68 Breeders 15 27 Breeders/Bull 29 30 Stocking Rate (AU/ha) 0.9 1.2 Beef Production (kg/ha) 92 58 Average Subloan (S/. 000, 1979 prices) 1,445.1 IRR: Without Incremental Herd 9% With Incremental Herd (Year 12) 13% 1/ 31 farms in Pucallpa, Oxapampa and Tarapoto. For more detail see report by 0. Cordon (Consultant). May 28, 1980 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report BAP Staff as of Augut 31. 1979 ----------------------------------------MATH--- H--MANBRNHE ---------C----------------------------------------------- Main Office Lima Canete Huacho Tingo Marla Aeecipe Tacoa C arca ChiJlaao nam Ceco Pira uancayo Ayacooho La Mered Iquits Puno TrIjillo Ica Total Profe-i-onal Staff Agriculturaliate 61 11 9 23 17 18 8 7 19 15 36 28 20 10 13 35 31 19 25 405 Veterinarians 3 1 1 2 7 Layer, 16 2 2 4 3 5 1 1 2 2 4 5 3 1 3 4 5 8 71 Accountanto 25 1 1 1 1 1 1 2 2 35 Economiet. 4 4 Othe-a 2 2 Subtotal 111 13 11 28 21 24 9 9 21 17 42 33 23 11 14 39 39 26 33 524 Support Staff Administrative 329 36 42 99 74 106 35 34 88 53 124 112 63 36 68 147 94 93 115 1,747 Technical 46 1 1 6 6 9 2 2 5 8 13 4 3 5 12 25 5 8 5 166 Drivers 13 5 3 7 8 15 4 3 12 7 7 9 8 5 5 7 14 12 12 156 Manual Laborers 96 7 9 17 14 17 3 5 16 8 33 17 11 13 13 80 24 17 19 420 Subtotal 484 49 55 129 102 147 44 44 121 66 177 142 85 59 98 259 137 130 151 2,489 TOTAL 595 62 66 157 123 171 53 53 142 93 219 175 108 70 112 298 176 156 164 3,013 No. of Agenciea 1 3 5 4 1 1 3 3 5 5 1 2 2 12 7 7 3 65 No. of In.pectorates 2 4 1 1 5 4 4 1 1 3 26 - 58 - Table 20 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Project Subloan Recoveries, 1974-1979 Year Principal Interest Total - - - - - - s/. '000 - - - - - - - - - - - 1974 80 15 95 1975 15,460 7,083 22,543 1976 64,514 40,093 104,607 1977 192,174 119,039 311,213 1978 430,162 289,593 719,755 1979 544,822 531,101 1,075,923 TOTAL 1,247,212 986,924 2,234,136 March 28, 1980 - 59 - Table 21 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Project Subloan Arrears, December 31, 1979 Up to 91 to 181 to Over 90 days 180 days 360 days 360 days Total - - - - - - - - S/. '000,000 - - - - - - - - - Repayments of Principal 1. Payments in Arrears on Active Subloans 19.4 8.5 7.7 20.0 55.6 2. Payments in Arrears on Expired Subloans 0.6 0.4 -4- 0.1 1.1 Subtotal 20.0 8.9 7.7 20.1 56.7 Interest Payments 1. Payments in Arrears on Active Subloans 29.7 23.4 13.2 18.7 85.0 2. Payments in Arrears on Expired Subloans 0.2 0.1 -0-- 0.1 0.4 Subtotal 29.9 23.4 13.2 18.8 85.4 TOTAL 49.7 32.4 20.9 38.9 142.1 March 28, 1980 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report Status of Project Portfolio. December 31, 1979 Total Portfolio Under Collection Capital Payments Overdue Portfolio Normal Total Recovery Rescheduled Overdue Total Active Subloans Expired Subloans Up to 91 - Over up to 91 - Over Total 90 days 360 days 360 days Total 90 days 360 days 360 days S/. Million 1,487.9 1,422.8 1,158.4 62.4 202.0 56.7 55.6 19.4 16.2 20.0 1.1 0.6 0.4 0.1 % 100 95.6 77.9 4.2 13.6 3.8 0.1 May 15, 1980 - 61 - Table 23 PERU FIFTH AGRICULTURAL CREDIT PROJECT Completion Report BAP Medium and Long Term Loans, 1974-1978 Private Year AREs Farmers Total ----------------(S/. million)------------------ 1974 914.8 108.6 1,023.4 1975 991.7 170.6 1,162.3 1976 873.3 277.9 1,151.2 1977 595.8 551.3 1,147.1 1978 1,527.1 1,422.4 2,949.5 Total 4,902.7 2,530.8 7,433.5  C~ é.~ s,77- 733 T, A MSE U T HA E c u A D O R 0 T1. Z El Alt. 47 71 F. / H,, 9< 9P c.-\ hTa 72< CI IB.~C .....E 9 ~ .7 -" CTI bres9- -e curm Cbicc- 5p-. P L-ETE EOD R< - ASPHAL DIT ROAD5 RAl lWAYS- +++RAILlWAY FERRY -ý PORTS 9Z -> * AIRPORTS s e c- NR,VERS- - ---INTERNATIONAL BOUNDARIES -, ALTITUDE5 IN METERS - - LA 0 - 500_ - o so0o-2000 200D 4000- .E OVER 4000 -- K'LOMMER C HILlE