57% i:-<'~““:3 3'“; ”§\\I§s:”'f6:1““°"."m S .\ » if}, <.2 32:: U :3“? fin-“1's I‘V- - A '.‘.,- Q ., I «13? ‘ ‘ .. ‘ 3, .l ’t . , -59: r -: -*- _ .':._ Mm 15,. if”? ,9; , 4: cl: “ .. - 0—! Fwy-fin . «fir-a .8 99...“ ' 4 J“.' J «:9 3 . ‘ #L‘ ‘ . " . 3, lgflmw‘ “‘3; mpg; . it: 4" 2")!” hi: .‘ " ‘}¥,<.;E‘at‘1“ ,, . 1..“ (I; «1:! . _ “‘133h'1‘.‘,',gm :N- ' ‘. {)‘M‘J 1 (‘3' “a: . 1:33:32; 1' ”:53, . .4 1Q, “1%; -' y 3. {1, "I“: ‘4‘ . ‘ ( ‘ . '2‘ “32:“, V m. 3, 3 "" 3%}: ‘Q ‘9‘: 3 193,53, " . ' -‘ (#2. . h.“ ‘5’ MES: 34?; w J A)!“ . 3.3:" $595: 05“??? 13.52“: "[5 fi' ‘ ~ - t “ , ~N .. ’ 4“???“ ”‘4': . _ A" “$551“ ,5“: 7 Mfg? 1'75 q,” ‘ 4.4 d. “‘1 yZ‘FEq-FN ‘ I" a}! (Air?! 4; :r;:;:’ hfi/J'itgt‘x r“ “:a:n::: :14: :m 4"?th '. 45" “f;- .é*:{?f«~?i{€¢z€'§ m‘ 1. A, $.41!“ ' I'v- ,w N.- ‘ ‘ 40.. a" - ~:-.~r' 4:“ s w... . ‘ :.- . At,- , -3" ”p . pan)?! V '5 . us m e 7% s 0 b - -— ——— " lllllllllfl’l”i‘”Ill‘IIifTI'IT ‘ USE-ARV 3 1293 00609 7400 Michigan State ”Riversity Ju-u This is to certify that the thesis entitled THE FINANCIAL SYSTEM SUPPORTING RICE MARKETING IN ECUADOR presented by HUGO H . RAMOS has been accepted towards fulfillment of the requirements for Ph.D. degree in W31 Economics 7%M/z 72‘?sz Major professor Date October 131 1989 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or baton dots duo. DATE DUE DATE DUE DATE DUE MSU Is An Affirmative ActiorVEquol OpportunIty Institution -__. _—_~—~_____ THE FINANCIAL SYSTEM SUPPORTING RICE MARKETING IN ECUADOR BY Hugo H. Ramos A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Economics 1989 €14.55 Q04 74 ABSTRACT THE FINANCIAL SYSTEM SUPPORTING RICE MARKETING IN ECUADOR BY Hugo H. Ramos This study examines the financial system supporting rice marketing in Ecuador. The main findings are based on field research conducted during 1988 and direct observation of rice market participants behavior during credit and price policy changes. ‘While the research focused on the physical rice flow system, the emphasis was directed to the financial mechanism inherent in rice marketing activities. The rice marketing system has been stagnant and the rice milling industry has become obsolete. The only noticeable change in the system is an increasing concentration of market power in the hands of a few national wholesalers. This concentration has its roots in the accessibility this group has to bank loanable funds. Meanwhile, access to credit has been limited for small rice market participants. The scarce subsidized interest rate loans provided by the BNF, CFN, and private banks has been rationed away from certain rice market participants such as traders and small borrowers in general. Increasing transactions costs has been a common measure to discourage small borrowers from demanding subsidized loans. Inflation induced liquidity problems have been found in long-term investments. Although debt financing is attractive when interest rates are below the inflation rate, firms undertaking long-term investments confront severe liquidity problems during the first payment periods. The study recom- mends a number of‘ways to help firms to mitigate this problem. ACKNOWLEDGEMENTS The completion of the program and this thesis has been the richest learning process I have ever experienced. The teachings of my major professors, Drs. Harold Riley and Lindon Robison, continued during the research and dissertation writing process. I deeply appreciate and thank them for their permanent support and advice and their patience to endure the interminable revisions of drafts. I want to thank Dr. Michael Weber for his valuable direction during the planning stage of the research. I am in debt with Drs. Riley, Robison and Weber for their commitment to help me go through the process of finding financial support for this task. I owe my gratitude to many people who contributed to the completion of the program and this thesis. I want to mention Dr. Joseph Goodwin of USAID, who encouraged me to continue my training: to Drs. Duty Greene, of Sigma One Corporation, and Morris Whitaker, from Utah State University, for their support while accomplishing the field research; to my fellows at Michigan State University, especially to Thomas Jayne and David Tschirley, for their stimulating friendship; to Mrs. Eleanor Noonan for her moral support and interest in my career . iv Without the financial support of the Institute of Agricultural Strategies (IDEA) and USAID/Quito the completion of this program would have not been possible. I want to thank especially to Mr. John O'Donnell, from USAID, and Eng. Neptali Bonifaz, the Executive Director of IDEA, for their encourage- ment and confidence in my responsibility. Finally, I want to express my profound gratitude to my family, Angela, Veronica, Hugo and Dennis for their under- standing and patience while I was heavily involved in my doctoral program. I wish to have many years to devote to them to compensate the forever lost opportunity to be close to all of them when they needed me the most. TABLE OF CONTENTS LI ST OF TABLES O O O O O O O O O O O O O O O O O O O 0 LIST OF FIGURES O O O O O O O O O O O O O O O O O O O 0 LI S T o F ACRONYMS O O O O O O O O O O O O O O O O O O 0 CHAPTER ONE INTRODUCTION . . . . . . . . . An overview of the economic and political dimensions of rice marketing in Ecuador . . . . . . . . . . Product and financial flows in the rice marketing system . . . . . . . . . . . . . . . . . . . . . . l. The rice subsector in Ecuador . . . . . . . . 2. The role of credit in Ecuador's rice marketing system . . . . . . . . . . . . . . . . . . . 3. Premises about the financial system supporting rice marketing investments . . . . . . . . . 4. Hypotheses of the study . . . . . . . . . . . Study purpose and objectives . . . . . . . . . . . Research procedure . . . . . . . . . . . . . . . . a. Field reconnaissance . . . . . . . . . . b. A field survey of the financial mechanisms facilitating rice marketing . . . . . . c. In depth survey of rice milling enter- prises . . . . . . . . . . . . . . . . . d. Limitations of the field survey . . . . vi xiv xvii xix 10 11 e. Descriptive and diagnostic analysis of the financial flows facilitating ‘marketing activities . . . . . . . . . . . . . . . f. Profitability and liquidity analysis of rice milling enterprises . . . . . . . E. Organization of the study . . . . . . . . . . . . CHAPTER TWO METHODS FOR ANALYZING FINANCIAL MECHANISMS SUPPORTING THE RICE MARKETING SYSTEM . . . A. The subsector analysis approach . . . . . . . . . l. The concept of subsector analysis applied to the study of the financial mechanisms sup- porting rice marketing activities . . . . . . a. The conceptual subsector analysis frame- work . . . . . . . . . . . . . . . . . . b. The applied subsector analysis frame- work . . . . . . . . . . . . . . . . . . 2. Functions of the rice subsector and their relation with financial mechanisms . . . . . 3. Vertical coordination mechanisms in the financial system supporting the rice subsector a. Vertical integration . . . . . . . . . . b. Forward contracting . . . . . . . . . . vii 12 13 13 15 16 16 16 21 24 26 27 29 c. Spot markets . . . . . . . . . . . . . . d. Informal trading arrangements . . . . . 4. Participants in Ecuador's financial system and rice subsector . . . . . . . . . . . . . . . Methods used to assess the profitability of finan- cing marketing activities and to study the liquidity problems of long-term investments . . . . . . . . 1. Design of prototype rice mills . . . . . . . 2. The profitability of rice storage . . . . . . 3. The profitability of prototype mills . . . . 4. Long-term capital investment problem . . . . Performance criteria used to evaluate the func- tioning of the financial system supporting the rice subsector . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . CHAPTER THREE THE FUNCTIONAL ORGANIZATION OF THE RICE SUBSECTOR The functional organization of rice production . . 1. Geographical distribution of rice production 2. Credit needs of rice producers . . . . . . . The functional organization of the rice marketing system . . . . . . . . . . . . . . . . . . . . . 1. Rice flows, and rice marketing participants . a. Paddy rice marketing system . . . . . . viii 30 3O 31 34 34 36 37 4O 42 45 46 46 46 47 50 50 2. b. Milled rice marketing system . . . . . . Vertical coordination in the rice subsector and identification of its financial needs. . . . The structure of the rice milling industry . . . . 1. THE CREDIT SYSTEM FACILITATING RICE MARKETING ACTIVITIES Functional organization of the rice milling industry 0 O O O O O O O O O O O O O O O O 0 Identification of functions and credit needs a. Credit needs for rice processing . . . . b. Credit needs for rice wholesaling . . . Summary . . . . . . . . . . . . . . . . . . . CHAPTER FOUR Functional organization of the financial system . 1. 2. The BNF as the official credit administrator Private banks participation in the rice subsector . . . . . . . . . . . . . . . . . . Trading and financial institutions for rice marketing . . . . . . . . . . . . . . . . . a. The issuance and use of warehouse re- ceipts . . . . . . . . . . . . . . . . . b . Critical issues concerning the use of warehouse receipts . . . . . . . . . . . Non-institutional sources of credit . . . . . ix 53 56 61 61 69 69 72 73 75 75 75 83 85 85 87 89 Official allocation of funds to the rice sub- sector . . . . . . . . . . . . . . . . . . . . . . 92 Impacts of subsidized credit . . . . . . . . . . . 94 1. Transaction costs and credit rationing . . . 94 2. Effects of subsidized interest rates . . . . 96 summary 0 O O O O O O O O O O O O O O O O O O 9 8 CHAPTER FIVE THE USE OF CAPITAL TO FINANCE RICE MARKETING AND STORAGE O O O O O O O O O O O O O O O O 0 O O O O O O O O 99 Working capital needs for marketing and storing rice . . . . . . . . . . . . . . . . . . . . . . . 100 1. Financing millers working capital . . . . . . 100_ 2. Financing wholesalers' working capital . . . 102 Profitability of financing rice storage . . . . . 103 1. Market participants in rice storage . . . . . 103 2. Seasonal movement of rice prices . . . . . . 109 3. Profitability of financing rice storage . . 117 ENAC's participation in the storage of rice . . . 125 1. Allocation of financial resources . . . . . . 125 2. Profitability of financing official stocks of rice . . . . . . . . . . . . . . . . . . . . 127 summary 0 O O O O O O O O O O O O O O O O O O 129 CHAPTER SIX PROFITABILITY OF LONG-TERM INVESTMENTS IN PRIVATE STORAGE FACILITIES AND DRYING AND MILLING MACHINERY O O O O O O O O O O O O O O O O O O O O O O O O O O O O 130 A. Identification of long term investments oppor- tunities . . . . . . . . . . . . . . . . . . . . 130 1. Piecemeal machinery replacement and additions of equipment . . . . . . . . . . . . . . . . 130 2. Investments in new rice milling plants . . . 132 B. Profitability of financing long-term investments in rice processing and storage . . . . . . . . . . . 134 1. Current capacity utilization of prototype rice mills . . . . . . . . . . . . . . . . . . . . 134 2. Profitability of financing piecemeal replace- ments and additions of equipment and ma- chinery . . . . . . . . . . . . . . . . . . . 139 3. Profitability of financing the construction of new plants . . . . . . . . . . . . . . . . . 143 4. Comparison of two financing strategies . . . 146 smary O O O O O O O O O O O O O O O O O O O 147 xi CHAPTER SEVEN THE IMPACT OF INFLATION ON THE LIQUIDITY OF LONG TERM INVESTMENTS . . . . . . Definition of the problem . . . . . . . . . . . . 1. Theoretical background . . . . . . . . . . . 2. The theoretical model in the context of rice mills investments . . . . . . . . . . . . . . Application of the conceptual model to the rice milling industry of Ecuador . . . . . . . . . . . Credit conditions to reduce or eliminate the liquidity problem of long-term investments partially financed with loans. . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . CHAPTER EIGHT CONCLUSIONS AND RECOMMENDATIONS . . . . Summary of the principal findings of the research Concluding remarks about financing operating capital . . . . . . . . . . . . . . . . . . . . . 1. Working capital financing . . . . . . . . . . 2. Operating capital to finance rice storage . . 3. Financing long-term investments in rice marketing activities . . . . . . . . . . . . xii 148 148 148 151 152 158 162 164 165 170 170 172 173 C. Related demand and supply considerations about subsidized credit . . . . . . . . . . . . . . . . 174 D. Recommendations . . . . . . . . . . . . . . . . . 175 1. Private banks participation in financing marketing, storage, and long-term investments in the rice subsector . . . . . . . . . . . . 176 2. Innovations in the credit delivery system . . 178 3. Guiding' the construction. of cost-effective mills, their number, and locations . . . . . 180 4. Defining the role of ENAC in the rice marketing system . . . . . . . . . . . . . . . . . . . 181 5. Negotiation of moderated payments schedule in loans aimed to finance long-term investments 183 E. Some suggestions for future research . . . . . . . 183 APPENDIX 0 O O O O O O O O O O O O O O O O O O O O O O 18 5 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . 194 xiii Table Table Table Table Table Table Table 1.1 3.4 LIST OF TABLES Ecuador's Population in Census Years 1962-1982 and Projection for 1988 (Thousands of Peo- ple). . . . . . . . . . . . . . . . . . . . . Average percentage of rice harvested, for the Winter and Summer cycles, estimated for the period 1985-88. . . . . . . . . . . . . . . . Land utilization in rice production by size of farms and number of independent farmers during the 1988 production year. . . . . . . . . . . Land utilization in rice production by size of farms and number of cooperatives during 1988 production year. . . . . . . . . . . . . . . Drying, storage, and milling capacities of mills in metric tons and as percentages of national totals, in 1988. . . . . . . . . . . Flows and consumption of Winter harvested paddy rice and the need for storage facilities, based on 1985 Winter harvest data. . . . . . . . . BNF's loans to finance the production and marketing of paddy rice during the period 1970-88, reported in nominal and real sucres, with May 1978-April 1979 = 100. . . . . . . . xiv 36 48 50 64 66 76 Table Table Table Table Table Table Table Table 5.2 Percentage participation of financial institutions in the seasonal rice production credit as an average of the years 1986-88, in nominal terms. . . . Sources of credit for the 1988 rice Winter crop. . . . . . . . . . . . . . . . . . . . . Financial costs of wholesalers' "adelantos" for prototype rice mills during 1988 opera- tions for a three months loan. . . . . . . . Stocks of rice at the end of June and October, and the Winter and Summer production for the 1982-88 period (Metric Tons). . . . . . . . . Seasonal wholesale price index per cwt of milled rice in Quito estimated for the period 1971-1988. . . . . . . . . . . . . . . . . . Annual rates of return for rice purchased in June and stored during part or the entire July-October period using market interest rates to estimate storage costs. . . . . . . . . . Annual rates of return for rice purchased in October and stored during part or the entire November-April period using market interest rates to estimate storage costs. . . . . . . Average official and.market prices for 200 lbs sacks of paddy rice, ENAC purchases (milled rice equivalent), and BNF's credit to ENAC to finance rice trade and storage, for the period 82 83 101 104 109 119 120 Table Table Table Table 1972-88. . . . . . . . . . . . . . . . . . . Maximum capacity and actual capacity utiliza- tion of mills measured in hours of work and cwt units of milled rice per year for the year 1988. . . . . . . . . . . . . . . . . . . . . Average total cost per cwt of milled rice estimated at current capacity utilization, at break-even level, and at maximum capacity for large, medium and small prototype mills. . . Net cash flows from estimates associated with investments in rice mills, and of annual loan installments assuming current commercial loans conditions. . . . . . . . . . . . . . . . . . Net cash flows estimates associated with investments in rice mills and of annual loans installments under the assumption of sub- sidized BNF loan. . . . . . . . . . . . . . . xvi 126' 135 136 153 157 Figure Figure Figure Figure Figure Figure Figure Figure Figure 3.1 4.1 LIST OF FIGURES Schematic view of the rice subsector framework of analysis. . . . . . . . . . Market channels and flows of paddy rice. . . . . . . . . . . . . . . . . . Market channels and flows of milled rice. . . . . . . . . . . . . . . . . . BNF's production and trade credit, in real terms (May 1978-April 1979=100), for the period 1970-88. . . . . . . . . End of June rice stocks and Winter production, in MT, all converted to milled rice for the period 1982-88. . . End of October rice stocks and Summer production, in MT for the period 1982-88. Seasonal wholesale price index for cwt of milled rice in Quito for the period 1971-88. . . . . . . . . . . . . . . . . Percentage changes in real wholesale prices between April and October in Quito for the period 1971-88. . . . . . Percentage changes in real wholesale prices between September and June in Quito for the period 1971-88. . . . . . xvii 23 52 54 77 107 108 110 114 115 Figure Figure Figure Figure Annualized rates of return to storage during one or more months of the Winter 1983, 1985 and 1988 periods. . . . . . . Annualized rates of return to storage during one or more months of the Summer 1983, 1985 and 1988 periods. . . Average cost curves for large, medium, and small prototype rice mills in Ecuador. . . . . . . . . . . . . . . . . Schedules of loan repayments(A) and net cash returns (R), for invest- ments in large rice mills in Ecuador, assuming 100 percent debt financing. . xviii 123 124 138 155 LIST OF ACRONYMS ALGRACESA Compafiia Almacenadora de Granos, Sociedad.An6nima - ALMACOPIO ALMAGRO ANIA BCE BNF BPA CFN COFIEC EMSEMILLAS the Grain Storage Company, a semi-private firm. Compafiia de Almacenamiento y Acopio - the Assembly and Storage Enterprise. The majority of its capital belongs to BCE and BNF. Almagro - a grain elevator private company. Asociacion.Nacional de Industriales Arroceros - the National Association of Rice Millers. Banco Central del Ecuador - the Central Bank of Ecuador. Banco Nacional de Fomento - the National Development Bank. Bolsa de Productos Agropecuarios - the Agricultural Products Exchange Board Corporation. Corporacién Financiera Nacional - the National Finance Corporation, a state owned enterprise. Corporacion Financiera Ecuatoriana - the Ecuadorian Finance Corporation, a private finance company. Empresa Nacional de Semillas - the National Seed Enterprise. A firm under MAG's administration. xix ENAC ENPROVIT FENACOMI FERTISA IDB IDEA IBRD PNA PROARROZ Empresa Nacional de Almacenamiento y Comer- cializacion - the National Grain Marketing and Storage Enterprise. A firm under MAG's adminis- tration. Enpresa de Productos ‘Vitales - ‘the Basic Food Distribution Enterprise. A firm under MAG's adminis- tration. Federacién Nacional de Comerciantes Minoristas - the National Federation of Small Traders and Retailers. Fabrica de Fertilizantes, Sociedad Anénima - the Fertilizer' Mixer Company. The majority of its capital belongs to BNF. Inter-American Development Bank Instituto de Estrategias Agropecuarias - the Institute of Agricultural Strategies. International Bank for Reconstruction and Develop- ment (the World Bank). Ministerio de Agricultura y Ganaderia - Ministry of Agriculture and Livestock. Programa Nacional del Arroz y Control de Piladoras - National Rice Program, a technical division of MAG. Productores de Arroz - the Rice Producers Associa- tion. XX CHAPTER ONE INTRODUCTION A. An overview of the economic and political dimensions of rice marketing in Ecuador Food demand in Ecuador has been influenced by rapid population expansion and growing urbanization. From 1962 to 1988, the population in Ecuador more than doubled, growing at an average annual rate of approximately 3 percent. Now, 54 percent of Ecuador's population lives in urban areas (See Table 1.1). Quito, the capital of Ecuador, and.Guayaquil, the major port city on the Pacific coast, are the urban centers where population increases have been the greatest. Another important demographic change has occurred in the Sierra and the Costal regions. In 1962 the Sierra population accounted for 51 percent of the total population and.the Costa 48 percent. In 1988 the Sierra accounted for only 46 percent while the Costal region accounted for 50 percent. These changes in population and the location of people in Ecuador have caused changes in the food marketing system in general and the rice marketing system in particular. Larger amounts of rice are now channeled to ever larger concentrations of people, requiring better marketing facili- ties, including transportation and storage, and requiring at the same time larger amounts of financial resources to fund rice marketing efforts. Table 1.1 Ecuador's Population in Census Years 1962—1982 and Projection for 1988 (Thou- sands of People). POPULATION POPULATION CENSUS YEARS PROJECTION DISTRIBUTION 1962 X 1974 X 1982 X 1988 X ------------ Urban and Rural Population ------°----- URBAN 1612 36 2699 41 3968 49 5529 54 Quito 355 8 600 9 866 11 1186 12 Guayaquil 511 11 823 13 I199 15 1635 16 Other Towns 746 17 1276 20 1903 24 2708 27 RURAL 2864 64 3823 59 4092 51 4674 46 TOTAL 4476 100 6522 100 8061 100 10204 100 -------------------- By Regions ----------------- SIERRA 2271 51 3147 48 3802 47 4695 46 COSTA 2127 48 3179 49 3947 49 5056 50 ORIENTE 78 1 196 3 312 4 453 4 Source: Whitaker and Alzamora, 1988c Another change affecting rice marketing in Ecuador has been an increasing per capita consumption of rice. Navarrete (1979) reports that annual consumption of rice was 9 kg per capita in 1970 and 23 kg in 1978. The National Rice Program (PNA), a technical division of the Ministry of Agriculture (MAG), estimates»that annual consumption for 1988 was.approxi- mately 6 million quintals (cwt) of rice, equivalent to about 273,000 metric tons (MT). A recent study by Stewart and CueSta (1988, p 5) estimates an annual per capita rice consumption of 27.5 kg for the same year. The process of urbanization and the increase in per 3 capita consumption help to explain why urban industrialists and organized unions have been lobbying for lower food prices. Meanwhile, vested interests on the production side, represent- ing a strong political force, have been lobbying for higher support. prices and lower input costs. ‘Politicians are sensitized to these political forces and conflicting inter- ests. A recurrent mechanism utilized by policy makers to face this conflict has been to offer subsidized credit to finance production. and. to set subsidies for ‘the importation of agricultural inputs, while setting and attempting to enforce official consumer prices. B. Product and financial flows in the rice marketing system 1. The rice subsector in Ecuador The rice subsector is conceived in this study to be a sequence of vertically coordinated stages from the supply of financial resources to acquire farm inputs to the final consumption of the product. The flow of rice from one stage to the next usually implies: 1) a physical exchange of rice for a financial instrument, such as cash, check, or a promis- sory note, associated with 2) a transfer of property rights over the product: and 3) a change in the value of the product. The marketing process for rice requires operating capital to facilitate the flow of rice from one stage to the next. 4 Moreover, additional capital is needed to finance rice processing operations such as cleaning and drying, storage, milling, and physical transportation to final consumers. Channeling rice through the marketing system can not take place without the use of financial resources. Each stage of rice processing, including such functions as processing, transportation, and storage, cannot be undertaken without the use of financial resources such as cash or cash substitutes. An efficient and smooth flow of rice requires physical facilities and equipment to fulfill these marketing functions. Among these are trucks to assemble and distribute the grain, drying and processing equipment to make the rice suitable for human consumption, and silos and warehouses to store the product during off—harvest periods, since production is seasonal. All these physical facilities need to be operating adequately for the marketing system to maintain stability between supply and demand forces. 2. The role of credit in Ecuador's rice marketing system The capital needed to finance rice marketing assets are supplied as either equity or debt capital. In this study, the focus of the analysis is on debt capital funds which can be made available either through non-official financial interme- diaries or through official financial institutions. Officially allocated credit often has been rationed away 5 from some important marketing activities, such as distribution of the product, storage, and investments in facilities. This allocation has reflected a bias against financing marketing activities and. has, been. based. on. political factors and misconceptions about the role of assemblers, millers, whole- salers and retailers. In the past legal and institutional restrictions have limited access to administratively allocated credit. Some of these restrictions have been imposed by allocating credit disproportionately to priority sectors and target benefici- aries. Other restrictions have taken the form of subsidies in the interest rate and.easy loan terms. ZNon-official finan- cial resources have been also constrained by overestimating the risk and underestimating the profitability of agricultural marketing activities. These constraints in the financial system have been more binding in the rice trade and storage segments of the marketing system. It was especially so during 1988 when official credit was rationed away from rice trade and storage. Bottlenecks in the flow of rice through the marketing stages are somehow created by financial constraints. These bottlenecks occur because marketing participants lack incen- tives for undertaking marketing investments. The incentives they lack are attractive rates of return on investments. 6 3. Premisesi about. the financial system. supporting' rice marketing investments Policy makers have often expressed their strong anti- intermediary biases. During political campaigns, government officials and policy makers have blamed middlemen and traders for increases in food prices. The frequently stated solution to reduce consumer prices of food items, such as rice, has been the elimination of the "long chain of intermediaries". This bias against this group of participants has influenced the official allocation of financial resources. Financial policy makers have expressed, in a number of inter-institutional meetings, their perception and fear that commercial credit to finance short-term marketing activities, such as distribution and storage, contributes to price increase. They reason that these activities are not produc- tive, i.e. do not add real value to the economy. Yet the financial policy' adopted increases the :money supply' and inflation. Inflation.is a pervasive, detrimental financial phenome- non constraining in many ways the use of financial resources to invest in market facilitating assets. It limits invest- ments in durable assets because loan repayments generate cash deficits. The cash flow deficits created by inflation cause liquidity problems for the firm, especially during early periods of an investment's life. Inflation and its attendant 7 liquidity problem have influenced private lenders to reduce the loan term making it even more difficult for private entrepreneurs to undertake capital investments. Finally, credit policy makers have asserted that there is evidence that the installed capacity to dry, mill, and store rice is sufficient for domestic production. This assertion, irregardless of its truthfulness, has resulted in a limited allocation of financial resources to these marketing activities. 4. Hypotheses of the study The main hypothesis that this research tests is that there are artificially imposed capital market constraints that limit access to credit that would otherwise be available to finance rice marketing activities. These suboptimal con- straints supposedly affect both rice millers and traders. They also affect consumers since rice is made available to them at higher prices than would be possible if the financial system could function without these artificial constraints. These artificial constraints we examine are both institu- tional and economic in nature. Some take the form of credit subsidies and preferential treatment for rice production, creating problems in both the supply and demand sides of the financial system. Other restrictions are the high costs of processing loans. By increasing transaction costs lending 8 institutions ration credit away from small borrowers. C. Study purpose and objectives The main purpose of this study is to identify those constraints that limit the timely and competitive access to credit for rice marketing activities and to recommend finan- cial mechanisms and credit policies that will contribute to the development of a more efficient, progressive and equitable rice marketing system. The specific objectives of the study are: (1) To provide a concise diagnosis of the functional organization of the rice subsector and the credit system facilitating the functioning of the rice marketing system. (2) To identify the major constraints limiting access to operating capital needed.to finance the trading and storage of rice. (3) To investigate the profitability of long-term investments in rice mills, storage facilities and equipment. (4) To assess the impacts of inflation on enterprise liquidity and the economic feasibility of long-term 9 investments when capital assets are debt financed and interest rates incorporate part or all the inflation rate premium. (5) To recommend changes in credit policies and related administrative loan delivery procedures to facilitate the economic flow of rice through the marketing process. D. Research procedure a. Field reconnaissance A rapid field reconnaissance was conducted during January of 1988 to identify the main stages in the rice marketing system, the flows of rice through the main market channels, and the key participants in the system. b. A field survey of the financial mechanisms facilitating rice marketing The first task was to collect relevant information about the problem stated and to get acquainted with related efforts being conducted in the country and elsewhere. This activity included the collection of relevant secondary data from public and.jprivate. organizations, especially from. the PNA, the Marketing Division of MAG, and the National Development Bank (BNF), as well as from the Rice Producers Association 10 (PROARROZ) , and the National Association of Rice Entrepre- neurs (ANIA). While collecting the secondary data, informal interviews were conducted with key market participants. The interviews were based on a set of questions about the rice marketing system and its associated financial mechanisms. The group of selected participants included rice producers, millers, wholesalers, and credit officials from BNF, National Finance Corporation (CFN), Ecuadorian Finance Corporation (COFIEC), Corporacion Bolsa de Productos Agropecuarios (BPA), Banco del Pacifico, and Central Bank. c. In depth survey of rice milling enterprises A pre-selection of 70 first and second class mills was made, based mainly on their size, location, and years in business, using the PNA files of registered mills. Technical personnel from the PNA helped in this pre-selection. Most of the preselected mills were visited but not all millers were interviewed, either because the plants were shut down or owners/managers were not at the plants. After this task was completed, selection of a better sample was possible. The criteria used to select a representative sample of 22 mills was: i) size, organized under two basic categories: sheller and. huller mills, which corresponds closely to the 11 PNA classification: ii) plant facilities for receiving, drying, storing, and milling rice; iii) operating level: and, iv) financial structure and operations. By working closely with the selected millers, it was possible to collect and organize most of the data required to construct prototype annual operating budgets and cash flow statements. These were then used to study the profitability of long term investments in equipment and storage facilities. They were also used to study the impact of inflation on the liquidity’ of rice enterprises ‘when investments are. debt financed. d. Limitations of the field survey The research. effort at ‘the ‘wholesalers level faced serious difficulties because of the political environment surrounding rice trade and storage at the time of the field work. After the 1988 winter harvest period (May and June), milled rice prices began to rise due to the presence of higher prices.at the northern (Colombia) and southern (Peru) borders. Unofficial export trade increased, reducing the domestic supply of rice and increasing milled rice prices. The upward price movement continued until the 1988 summer harvest (October). During this time the milled rice price rose 12 rapidly from about 6,000 sucres in October to more than 10,000 sucres by the end of November. In an attempt to control both unofficial trade and rising prices, the government began to seize stocks of rice from wholesalers in the main cities. Under these circumstances most rice wholesalers were unwilling to provide information for this study. e. Descriptive and diagnostic analysis of the financial flows facilitating marketing activities After collecting and organizing the secondary and primary information gathered during the field research period, an effort was made to describe and analyze the financial system supporting rice marketing activities. The observations and preliminary findings of this research work completed prior to December of 1988 were presented and discussed in a seminar held in Guayaquil on December 15, 1988. The participants in this seminar' were rice producers, millers, and 'traders, representatives of ENAC and the PNA, and Drs. Riley, Robison and Tschirley from Michigan State University. The preparation for this seminar and the discussions of the preliminary results of the research with the seminar participants resulted in a number of innovative and thoughtful recommendations to deal with the problem of credit for rice marketing activities. Some of these are included at the end of this report. HI fa fil vii 0v: SyS cri Cha exi. PM W Icti 13 f. Profitability and liquidity analysis of rice milling enterprises The final stage of the research analyzed the profitabili- ty of rice storage as well as the profitability of long term investments in rice marketing facilities. These investments were mainly for drying and milling equipment and storage facilities. This analysis included a study of the impact of inflation on the liquidity position of rice milling enter- prises when durable assets are loan financed and the interest rate incorporates an inflation rate premium. E. Organization of the study The conceptual approach and methods for analyzing the financial system associated with the rice marketing process will be described in Chapter Two. Chapter Three contains an overall description and diagnosis of the rice marketing system. It focuses on those stages where financial funds are critically needed to improve the physical flow of grain. Chapter Four presents a description and diagnosis of the existing credit system associated with the rice marketing process. It emphasizes the financial arrangements and operational mechanisms needed to support the rice marketing activities. Chapter Five investigates the use of credit to finance operating capital to physically assemble, store and 14 distribute the grain. Chapter Six evaluates the profitability of long-term investments in marketing facilities when they are financed by bank loans. Chapter Seven analyzes the impacts of inflation on 'the liquidity’ position of rice ‘milling enterprises when capital assets are financed by long—term loans carrying interest rates which incorporate all or'a large proportion of the inflation rate. The final chapter presents the main conclusions of the research. It also includes important recommendations to reduce or eliminate the con- straints that limit access to the financial resources on a competitive basis that are required to finance investments which would improve the performance of the rice marketing system. CHAPTER TWO METHODS FOR ANALYZING FINANCIAL MECHANISMS SUPPORTING THE RICE MARKETING SYSTEM Introduction This Chapter is organized into three main parts. Each part is divided into sections. The first section of part A defines the conceptual subsector analysis framework and its application to the financial system supporting rice marketing activities. The second section defines the main functions of a rice marketing system. The third section describes the vertical coordination mechanisms used in the financial system facilitating rice market activities. The fourth section identifies the main participants in the financial and rice marketing systems. Part B presents the research methods, emphasizing the analytical tools used in this study. The first section of part B presents the procedure used to build prototype rice mills. The second section shows the method utilized to evaluate the profitability of rice storage, concentrating the analysis of this function at the wholesale level. The third section presents the method utilized to assess the overall profitability of mill operations, both as they are currently structured and with additional investments in drying and milling equipment and storage facilities. Part C will discusses some issues related to financial market 15 16 performance measures and methods to evaluate the overall functioning of the system. A. The subsector analysis approach 1. The concept of subsector analysis applied to the study of the financial mechanisms supporting rice marketing activities a. The conceptual subsector analysis framework The concept of "subsector" appeared originally in Shaffer (1973), who defines it as "the vertical set of activities in the production and distribution of a closely related set of commodities". The development of this concept follows the works of Mighell and.Jones (1963) and Shaffer (1968 and 1973), French (1974), Marion (1986), and Riley (1989). The approach gained recognition and wide application through the numerous publications produced by the North Central Regional Research project 117 (NC 117). The usefulness of the subsector analysis stems from its pragmatism and applicability. It can be used effectively for comparative static analysis by facilitating the understanding of the impacts of policy changes on the main functions and participants of the subsector. It can be used to prepare diagnostic and analytic studies of commodity systems, 17 providing a sound basis for designing normative policy changes. Moreover, since the concept includes the substantive characteristic of interdependence and interactions among the components of the subsector, the approach can lead to an increased understanding of the forces of change and evolution in an economic system. The tools used in subsector analysis are much the same as those used in other economic and agri-business studies. These tools are applied considering the economic system as an integrated set of components adapting itself to changing conditions. Some of the financial analysis tools that can be used in the subsector approach are the net present value model. This tool was developed to evaluate the profitability of long-term investments. The cash flow technique can also be used in the subsector approach to identify cash flow deficits and liquidity problems. Finally, financial ratios that compare the performance of a specific firm with standards estimated for similar enterprises is another financial tool that can be used in the subsector approach. The subsector is looked upon simultaneously as a struc- ture consisting of stages and.as a movement or flow; From.the structure point of view, the key elements of a subsector are the stages where important economic activities take place. From the flow perspective, the main elements of a subsector are i) the fUnctions, ii) the coordination mechanisms, and iii) the participants. 18 i) Functions: These refer basically to the marketing utilities known as form, space, time, and possession. Bressler and King (1978, p 165) define form utility as the value added to the primary product by transformation of its characteristics. This value, in a competitive economic environment, should equal the costs of that transformation, including a normal profit. These costs are associated with the size and scope of the plants used for transforming the products. Important considerations are advanced.with respect to the size and characteristics of fixed assets (See for example Williamson, 1986) . With respect to the space utility, Tomek and Robinson (1972, p 144) argue that, in a competitive situation, the difference in the price of a homogeneous product between two markets should equal the cost of trans- ferring the product because arbitrage is the force that would equilibrate the price difference to the transfer cost. The implicit market integration of this idea has been tested empirically by a number of authors (see for example Ravallion, 1986, and Delgado, 1986). In a competitive scenario, the price difference between two points in time should equal the costs of storage, both, operating and financial, including depreciation and a normal profit (Tomek and Robinson, 1972: Bouis, 1983: Goetz and Weber, 1986). Finally, the possession utility concerns property rights and transaction costs included in the exchange. All these functions are productive economic activities which add value to the products. 19 ii) Coordinating mechanisms: Basically two forms of coordination are critical: vertical integration and external organization. In the case of vertical integration, the coordination is internal within a single business enterprise and the exchanges are decided by a unique body of decision makers (Williamson, 1985). The external organization takes various forms, such as spot markets, future markets, and forward contracting (Marion, 1986). iii) Participants: These are the economic agents deciding upon functions and coordination mechanisms. They are repre- sented by individuals as well as by organizations and public institutions. All these components are immersed within a set of broader economic, social, cultural and political forces regulating their activities. An effective way to organize the analysis of these components is to look at the subsector under three interacting dimensions: environment, behavior, and perform- ance . 1) Environment: Includes the factor and product markets: the set of rules, property rights, prices, contracts and taxes: and the social, cultural and political forces guiding the functioning of the subsector. Institutions and transaction costs are the key elements within the environ- ment. Institutions are the rules governing exchange and the set of norms that people follow when trading. Transaction 20 costs are the expenses, monetary and non-monetary incurred in performing the exchange. ii) Behavior: Williamson (1985) considers three kinds of behavior: 1) opportunistic behavior defined as "self- interest seeking with guile"; 2) open and simple self-interest seeking behavior: and, 3) obedience, which means following established norms in a mechanistic manner. Opportunistic behavior sets in motion social forces leading towards what Platt (1973) calls "social traps". These take the form of individuals pursuing their own self interest but damaging the group, or groups of individuals in the process. These individuals may still benefit in the short run but may impair and damage the source of the benefits in the medium and long- term. iii) Performance: Is the set of results and impacts of the actions of the participants. These results are important and more so when they feed back into the behavior of the participants and stimulate the undertaking of changes which can improve the functioning of the system. According to Schmid (1984), performance is a way to look at the distribu- tion.of’costs and.benefits: that is, a way to investigate "who gets what". 21 b. The applied subsector analysis framework Applied to the objectives of this research, the subsector analysis framework can be defined as a comprehensive and integrative approach to the study of the rice marketing system. The system is treated essentially as a sequence of vertically coordinated stages between which two basic types of flows occur: a physical flow of product and a flow of financial instruments. Key participants are responsible for coordinating these flows at each and every stage of the system. These participants have their peculiar ways of acting in the system, according to their perceptions and the chal- lenges that changing conditions impose on them. The function- al organization of the system and the behavior of the par- ticipants are influenced by a set of rules, policies, and institutions set forth to regulate the evolution of the system. Some of these policies are more closely related to a particular commodity system, while others are broader in scope. An example of the first is the credit and financial policies. .An example of the second is the set of sectoral and macro-economic policies. The functioning of the rice subsec- tor, as described here, generates performance results, which, in turn, have impacts and a strong influence on the evolution of the system. The subsector framework can be illustrated as in Figure 2.1, which shows the rice subsector as a sequence of stages 22 through which two basic flows take place: the physical flow of grains and the flow of financial instruments. This market process is influenced.by the credit and financial policies and institutions, as well as by the sectoral policies and institu- tions. The entire subsector is incorporated within a more general economic, social, and cultural environment. The focus of this study encompasses the transactions taking place from the farm gate forward, through the pro- cessing stage, and up to the wholesaling stage. The study emphasizes the credit and financial policies and institutions. The sectoral policies and institutions will be investigated only tangentially, to the extent they are related to the financia1.mechanisms facilitating rice transac- tions. As applied in this research, the subsector analysis framework has many similarities with the "agribusiness commodity system analysis" approach developed at Harvard University and applied in other Universities and overseas field research (see for example Goldberg, 1974). From the structure point of view, these are some of the main issues that need to be considered. The exchange of rice for financial instruments is prevalent in the rice subsector. The flow of rice and financial means through the marketing system should be smooth and balanced, reflecting a stable equilibrium through time between supply and demand. This smooth and.balanced flOW'Of rice needs to be free of technical constraints as well as financial restrictions. 23 TECHNICAL FARM INPUTS I PRODUCERS PROCESSORS /’ ”””””” \\ / 1’ ” \“‘\ I \ \ I x / \ / \ / \ / \ \ I \ I...--—— ------.>. '1’ AGRICULTURAL \\ III CRED I I AND \I‘ I I I SECTOR }—->[ WHOLESALERS “—1 FINANCIAL I l ‘ ,' ‘ POL I c I ES / \ POLICIES I 1‘ \ \ I \ \ I \ I / \. I : \\ ‘/ \ /' / ________ | ‘~_____,,’ RETAILERS A CONSUMERS —> FIG. of WWI! ---> Flmnclal Hons Figure 2.1 Schematic vieW'of the rice subsector framework of analysis. 24 The receiving and drying capacities for paddy rice at the processor level should be sufficient to process all the rice produced. Rice has to be dried after harvest. If it is to be immediately milled, the moisture content has to be lowered to about 13 percent. If it is to be stored, the moisture content has to be reduced to levels recommended for the time the grain will be stored. Unhulled rice left without drying for more than 24 hours after harvest begins to heat up. This in turn.causes burn spots and.discolorations of the.grains and yields a lower amount of milled rice. The flOW’Of paddy into the processing stage could.be con- strained both by technical factors and financial restric- tions. These constraints could be ameliorated with a suffi- cient injection of financial resources, and changes in the credit delivery system. This innovation would allow timely and less costly access to credit for all investors and entrepreneurs. 2. Functions of the rice subsector and their relation with financial mechanisms All four economic functions (time, space, form, and possession) are implemented in the rice subsector. Moreover, all require a flow of financial resources to undertake them. First, there is the issue of ownership or possession. Even though rice can be located at some specific place (farm, mill, 25 or warehouse), it is important to investigate who owns that grain. It is common to find farmers whose rice has been sold or committed to some buyer before harvest or millers who buy, dry and store rice in behalf of a wholesaler. Second, rice is delivered to consumers spread throughout the country, from the Costa area where it is produced.to towns in the Oriente. To be effectively performed, this physical distribution requires large amounts of financial funds. The important aspect to investigate is whether the price differ- ences over spaces equal the transfer costs. This price difference will be tested in this study, using the major consumer centers of Quito, Guayaquil and Ambato as market test cases. The simple correlation of historical prices among markets will be explored first, and.then the correlation.among the first differences of monthly prices in those markets will be examined. Third, the production of rice is seasonal while its consumption is relatively constant within a year. To match the supply of rice with its demand, the usual mechanism used in Ecuador is to store the grain during cropping periods, acquiring rice at harvest and depleting it gradually until the next harvest. Ensuring a stable supply of rice is critical in Ecuador given the economic and political importance of rice in the diet of the Ecuadorian population. It requires an efficient financial system and an environment free of uncer- tainty for this storage function to be performed effectively, 26 so that the demand can be satisfied by a stable supply of rice. The price difference between two points in time should be enough to cover the storage costs. Finally, rice needs to be transformed from paddy to milled rice. This function is critical because it requires financial funds to facilitate the millers purchase of paddy, and to finance the facilities required: receiving, drying, milling, and storing, to mention the most important ones. This study emphasizes this financial function and looks at the profitability and liquidity of long-term investments in the rice milling industry. 3. Vertical coordination mechanisms in the financial system supporting the rice subsector Vertical coordination is defined by Mighell and Jones (Marion, 1986) as "the general term that includes all ways of harmonizing the vertical stages of production and marketing". Hojjati and Staatz (1987) define coordination as "all ways of equilibrating supply and demand in adjacent stages of com- modity systems taking into account how the interest of various actors get counted". Vertical coordination is not a static concept but rather a dynamic mechanism that achieves the objectives identified by the marketing system. However, the simple flow of goods and financial resources between stages does not define dynamic 27 vertical coordination. Supply and demand at every stage may clear at a certain price and every time the exchange takes place. But the critical question is whether or not consumers are satisfied. What if the market clears at given prices but producers accumulate unnecessary stocks?, or consumers would like goods of better quality?. Dynamic coordination should mean adjusting the supply and demand to changing preferences. It should also change incentives when needed to encourage modernization of the production and marketing systems. It should also exploit profitable opportunities for development and look for alternative ways to satisfy in a more effective way the individuals' demands for employment, food and health. A diagnosis of the functional operations of the financial mechanisms supporting the rice subsector, with emphasis on those stages from the farm gate through wholesaling, can be achieved by analyzing at the manner in which the flows are vertically coordinated. a. Vertical integration Vertical integration means that two or more stages of the marketing system are under the same administrative unit. Rice production and milling, being two different stages in the marketing process, administered by two different decision units can be integrated into one single administrative unit. This may take place when a rice grower decides to integrate 28 his/her farm business with a rice processing plant. The flow of rice and money between separated stages is basically determined by the market, i.e. decisions are made by respon- ding to market signals and bilateral negotiations, which include transaction costs. When the stages are integrated under a unique decision unit, the flow of products and financial funds are independent of market forces and negotia- tions, which reduces or even eliminates transaction costs. A.rice farmer who is also a miller, moves paddy rice from farm to mill as an internal decision.and without the inconveniences of setting a price or inspecting the quality of the grain. Vertical integration offers the advantage of securing a certain level of product to the mill, at least by an amount equal to own production quantity, which may allow the firm to expand operations by purchasing more rice with the proceeds from the sale of its own milled rice. Using this method of organizing the flow of rice between production and processing, the physical transfer of grain may not be limited by financial constraints but by unbalanced processing facilities. Although the flow of rice from farm to mill may not require a flow of funds, it may be difficult to proceed if the plant does not have the facilities to receive the grain, or to dry it at the rate required, or to mill it at the pace set forth by the market. Thus, efficient vertical integration requires that neither technical nor financial constraints block the flow of rice. 29 In the rice subsector of Ecuador there are a number of cases of vertical integration. These occur mainly between production and processing and between wholesaling and retail- ing. Most of the medium and large rice mills in Ecuador are vertically integrated with rice production: the large majority of urban rice wholesalers have a retail outlet. b. Forward contracting Under this form of vertical coordination, the exchange of rice for money is not simultaneous. This mechanism has to consider the time, form and space utilities included in the exchange, plus the risk cost and normal returns to this activity, and the consequent increase in the value of rice. Money is exchanged with an agreement to deliver rice after a certain period of time, and to a certain place. This mecha- nism is often used to coordinate exchange between.millers and producers ("fomento") , and between wholesalers and millers ("anticipos"). If the financial flow'is not sufficient to facilitate the rice flow, it becomes crowded in some point of the marketing system causing the returns to be reduced and the incentives to invest to be hampered. The entire system is damaged and the rice gets to consumer at higher prices. 30 c. Spot markets This form of vertical coordination consists of a simul- taneous exchange of rice for a liquid financial instrument. For example, a check or money in exchange for rice. An efficient spot market equilibrates demand and supply and provides enough financial resources to be exchanged for the rice. If the flow of financial means is not enough to support the physical exchange for rice, the rate of return for some participants will be lower than expected, and future flows of rice will not balance supply and demand, making the system more costly. d. Informal trading arrangements This form of vertical coordination consists of exchanges arranged by telephone in which the product is described and payment conditions set. Several necessary conditions are required for this mechanism to perform efficiently: 1) information which must be reliable, timely, and as complete as possible to facilitate the formation of rational expecta- tions: ii) grades and standards must be widely acceptable if it is to reduce transaction costs and uncertainty; iii) a flexible and timely financial system. Deficiencies in these conditions increase uncertainty and adverse selection and moral hazard may occur, which in turn increases transaction 31 costs and prices. It may also cause market failures if only a few participants, such as few wholesalers, have access to this type of market and the information gathered becomes asymmetrical. 4. Participants in Ecuador's financial system and rice subsector The key participants providing financial resources for rice marketing activities are the National Development Bank (BNF), the National Finance Corporation (CFN), private banks, rice millers and wholesalers, as well as local moneylenders, participants in the informal financial system Ecuador's rice subsector. The main participants in the rice subsector are producers, millers, wholesalers, retailers, and consumers. Some aspects of their behavior will be addressed in a subse- quent chapter. However, some relationships between rice market participants and the financial system are described in this section. Millers who have easier access to credit and whose access is timely, gain advantages over those who have to overcome financial and institutional barriers such as high transaction costs and the fulfillment of a long list of requirements. This segmentation of borrowers at the milling stage could result in lower farm prices for paddy, since the number of effective buyers is lower than would be otherwise, driving up 32 the price of milled rice. Millers using informal sources of funds to complete their financial needs, such as wholesalers advances, will have higher financial costs. These costs will be passed on to farmers or forward to consumers. The rate at which.paddy has to be milled depends not only on the market demand for milled rice, which is more or less constant throughout the year, but also on the returns to the storage of paddy. The storage of paddy at millers facilities represents financial capital tied up in rice inventories. ZBut this financial capital has an Opportunity cost. It also requires that returns pay for depreciation of facilities, treatment of the rice, losses, etc. The expected difference between the price at purchase time and the price at release time should cover all of these costs (including cleaning and drying) plus a comparable return to any other use of the capital, under similar conditions of risk. If there are constraints to the financing of this critical activity, then rice will be milled and shipped to wholesalers, who will have to store milled rice. This possibility will generate a need for financial resources at the wholesaling stage to purchase the grain and to store it, plus it will crowd the available storage facilities at this stage while leaving idle those of the millers (not a cost-effective way of using existing facilities of millers). The transfer of the storage function from millers to wholesalers does not eliminate the need for financial resources to finance the flow of rice and its 33 storage costs. It only implies a reassignment of facilities. In summary, the flow of rice from the processors to the wholesalers should respond to economic factors such as price expectations, supply, demand, and returns, but should not be unnecessarily accelerated by financial pressures and limita- tions. The rice wholesaling stage shows a wide spectrum of participating groups, differentiated by their scale of operations and geographic concentration. The first level of wholesalers consists of a relatively small number of participants operating at a national level (i.e., distributing milled rice to all consumption centers of Ecuador and even outside its borders, to Colombia and Peru), who are identified in this study as "national wholesalers"; the second level is represented.by a larger number of'partici- pants operating mainly within provincial borders, identified as ”provincial wholesalers"; the third level consists of an even larger number of participants operating basically within the major urban centers, identified as "urban wholesalers". The question to investigate is whether or not.the flow of rice through all these strata of wholesalers is carried out under competitive conditions. A market that fails to allow entry of participants and restricts participants to a small group of wholesalers, permits profits higher than those comparable to other activities. This situation necessarily means higher prices to consumers and lower prices for producers. Unequal 34 access to financial resources contributes to the existence of concentrated markets. On the contrary, a reduction in the limitations to access to financial funds will allow new participants to enter and compete at the wholesaling stage, especially at the national wholesaling level. B. Methods used to assess the profitability of financing marketing activities and to study the liquidity problems of long-term investments 1. Design of prototype rice mills A field reconnaissance was conducted during January, April and May of 1988. These field visits allowed the author of this thesis to observe rice production and processing, and to interview a number of producers, middlemen and particularly millers. These contacts were helpful in selecting a sample of 17 first class and 32 second class millers, located within Guayas and Los Rios provinces. The main criterion for the selection of millers was the location of the mills within the production area. In this task, direct participation of PNA's field technicians was invaluable. Other factors were consi- dered as well with the objective of selecting a representative sample of all first and second class mills, such as: i) size: measured basically by the rate of rice milled per hour, i.e., cwt per hour: and ii) additional facilities: including 35 receiving facilities and capacity, drying capacity and method, storage facilities and type, and office space and utilities supply. Other selection criteria that were explored were the way millers financed their investments and operating capital, the coordinating mechanisms used by millers, the rate of capacity utilization, and the operating experience of mills. These factors proved not to be important in the selection of mills because the majority of mills presented similar patterns of investment financing, vertical coordination, operating experience, and capacity utilization. The pre-selected mills were visited during April and May and a smaller sample of mills was selected from among them: one large mill, with a milling capacity of 120 cwt per hour (one of the 15 mills with a milling capacity of more then 50 cwt per hour); three medium size mills, with milling capaci- ties ranging between 20 and 25 cwt per hour: and five small mills, with milling capacity between ten and 15 cwt per hour. Frequent visits were made to the selected mills to obtain more detailed information about their operations and the financing mechanisms they used, and to design prototype mills for the small, medium, and large sizes. For all three mills, two basic situations were established. i) The Winter and Summer harvest are 57.5 and 42.5 percent of the annual harvest, respectively. These are the averages for the period 1969- 88. ii) The same annual profile of purchases was assumed. The average annual profile of harvest is as follows: 36 Table 2.1 Average percentage of rice harvested, for the Winter and Summer cycles, estimated for the period 1985-88. Apr May Jun Jul Aug Sep Oct Nov Dec ..................... x -----------.------------- winter (1001) 8 32 45 15 Sumner (100%) 6 24 34 21 15 Source: PNA's Statistics 2. The profitability of rice storage Using a series of monthly wholesale prices (sucres per cwt) in Quito for the period 1971-88, a seasonal price index were calculated (Goetz and Weber, 1986) as well as the trend, cycle, and random component of the series. Based on these seasonal indexes, and knowing from field research the peak Winter and Summer harvest months, the evaluation of profitabi- lity of storage assumed two purchasing months, June and October, and two storage periods, June-September and October- April. Profits from storage were calculated for every month of the two storage periods. The profits were estimated on the price difference between the release and the purchase months (September-June and April-October, for the Summer and Winter cycles, respectively). The storage costs, including the opportunity cost of capital and operating cost (treatment, 37 depreciation charges, and losses)were discounted from this difference. The net difference was divided by the purchase price to find the percent margin (Mears, 1982). That is: L £35113.3:91L-.§i _____ ‘36) where: m = percentage profit (loss) to storage PH], = release price, in month j Pam = acquisition price, in months i (May—June and October-November) S. = Storage cost accumulated up to month j Since it was not possible to estimate operating costs, 5% represents only the opportunity cost of capital, measured by the market interest rate prevalent in each year, during the period 1981-88. Sensitivity analysis*was performed.to see how the percentage profit changed by varying the opportunity cost, assuming that the opportunity cost is represented by the inflation rate, and the inflation rate plus and minus 10 percent (Tschirley, 1988). 3. The profitability of prototype mills The basic financial model used to assess the long-run profitability of mills is the net present value, defined by: 38 n 2 R, t=l va = -------- - Io (1+r)t where: Rt = real and risk adjusted net cash income in period t Io = initial investment r = real and risk-adjusted discount rate t = each of the years of the project, t = 1, ., n n = the useful life of the project. Two important considerations are worth discussing regarding the use of this simple model: i) the assumptions about.the discount rate to be used, and the degree of certain- ty of the cash flows: and, ii) the "with and without" approach needed to evaluate additional investments to the existing mills. i) About the discount rate: It is generally defined as the opportunity cost of an investment project. If the investment were entirely financed by a bank loan, the discount rate should be the commercial loan rate: if the investment were financed by a mix of loan and own resources, then the discount rate could be a weighted average cost of capital. The discount rate must reflect the basic principles of 39 present value models as described by Robison and Brake (1984) . One of these principles is consistency: a net cash flow series in nominal (real) terms must.be discounted by a nominal (real) rate. When the commercial interest rate used as the discount rate is below inflation, the resulting real discount rate is negative: using a negative real discount rate in the NPV'model would increase (not discount) the real net cash income series. One way to cope with this problem is to use the certainty equivalent cash flow; as discussed in Brealey and Myers (1981, Ch 9). Following the consistency principle, a certain cash flow'must be discounted at a real, risk-free discount rate r,, while an uncertain one, at a real, risk-adjusted discount rate r. Brealey and Myers demonstrate that REQt = at Rt where: REQt = certainty equivalent net cash flow in period t Rt = net cash flow in period t and (1 + r,)' a = ----------- 40 Using this equivalence, the NPV model can be written as: where r}, the real, risk-free rate can be easily known. Here also a sensitivity analysis will be conducted, changing basically the discount rate to reflect a number of scenarios: a discount rate that incorporates fully the inflation rate; one that is simply the commercial loan rate; and one that is the commercial rate plus 5 and 10 percent. ii) About the "with and without" approach: This idea is presented here simply to differentiate this approach from the wrong one summarized in the words "before and after". The "with and without" approach is closely related with the incremental budget technique, that will be used in this study to prepare the financial statements for the prototype mills. 4. Long-term capital investment problem The hypothesis regarding this topic is that capital investments in periods of high inflation causes cash flow deficits for the firm. As the inflation increases the nominal rate of interest increases as well. This reduces the net cash 41 flow of the firm under ceteris paribus condition. The debt service can become large enough as to make the net cash flow negative, which in turn causes liquidity problems to the firm. An explanation of this phenomenon is found in Robison and Brake (1980). Their argument follows. Assume that net cash receipts R grows annually at i, for an infinite period of time. The present value V, discounted at a nominal interest rate r3,'will be given by: R(1+i) which can be rewritten in the form rnV - iV = R(1+i) In the above expression, the right hand side term is the net cash inflow and the left hand side is the difference between interest expenses and capital gains. Assume that V is entirely debt financed and that the lenders require payment of interest rgV. The net cash inflow R(1+i) must equal the interest payment rnV in a competitive capital market. With i>0, (rgv-iV) equals R(1+i). Therefore the net cash inflow is less than rnV by iV, the magnitude of the cash deficit. In later periods, of course, the cash flow deficit decreases. The methods used to investigate this topic are again the NPV model and the cash flow statement. The use of a nominal discount rate is required to make apparent the impact of 42 inflation in the behavior of the net cash flow of the firm. This study will be complemented by presenting financial credit arrangements to manage this problem, such as the use of models developed to moderate repayments of loans. C. Performance criteria used to evaluate the functioning of the financial system supporting the rice subsector The assessment of the financial system' performance that supports the evolution of the rice subsector requires an agreement on specific criteria and the establishment of impact measures. This study is concerned with the efficiency of the organization of the credit system to facilitate rice marketing activities. The study poses the questions of whether the economic. situation. and. the financial policies allow' the undertaking of basic marketing functions by the private sector?. How are the benefits and costs of the system distributed among participants, or "who gets what"?. Do the economic and financial policies stimulate long-term invest- ments that would modernize the rice milling enterprises?. Is the financial system adequate to allow modifications in its credit delivery mechanisms that would adjust repayments to the projects' return schemes?. These basic questions are asso- ciated.with such criteria as cost-effectiveness in performing marketing functions, equity in the distribution of benefits, 43 flexibility of the system to adopt new technologies, and progressiveness in terms of modernizing the rice processing. The performance assessment of the financial mechanisms facilitating the rice marketing system is based on a relevant list of objectives that the rice subsector must fulfill. The supply of rice has to satisfy domestic consumption, that is, the market has to match supply and demand through the form, space, and time functions, in a cost-effective manner. To achieve this ultimate objective, the financial mechanisms that support the rice marketing system must be fluid and free of economic and institutional restrictions. The flow of finan- cial resources to support the rice subsector, must be suffi- cient to cover the value of the product at every point in its market process, including the value added due to its changed form, space, and time utilities. This performance criteria applied in this study means that the returns from using financial resources to facilitate form, space, and. time transactions in the rice subsector must be enough to cover costs and provide incentives for further investments in rice marketing activities. Any constraint in this financial flow, either in amounts, timing, or accessibility, would increase the costs of financing and make rice transactions more costly, ultimately resulting in higher consumer prices. The financial system should facilitate improved efficien- cy and increased technology adoption. Efficiency, as measured by the highest level of output per unit of input, can be 44 achieved by facilitating the acquisition of better technical equipment and facilities, as well as by facilitating higher levels of plant capacity utilization. Progressiveness, measured by the rate of technology adoption, can be reinforced by allowing the financial system to operate with less restric- tions in terms of allocations, and access. If financial resources required to support the rice marketing system are artificially restricted, there may be groups of participants who may be hurt by the constraints. But also there may be groups who may benefit. By reducing or eliminating artificial barriers to financial resources there may be a more even distribution of benefits. As a result, larger groups of participants, including small farmers, medium size millers, small wholesalers, and especially consumers, will benefit from the improved financial system while larger rice markets and informal capital market participants might see their returns diminished. Technological change usually is coupled with higher capital-labor ratios. In the rice subsector and especially the rice milling industry, this seems to be the case. The objective. of the financial system ‘to contribute to the modernization and increased cost-effectiveness of the rice milling industry seems in conflict with a broader and funda- mental objective of the economy, that of increasing the employment rate and people's income. This conflict represents a classical trade-off between efficiency and equity and often 45 is a matter of political decisions and an issue that must be assessed within a broader sectoral and macro-economic perspec- tive. A diagnosis of the unintended effects of subsidized credit is utilized as another criteria to assess the perform- ance of the financial mechanism supporting the rice subsector. Summary The subsector analysis approach is the basic conceptual framework utilized in this study. The study looks at the financial system facilitating the functioning of the rice subsector as an integral and interdependent set of stages, functions, and coordinating mechanisms. The basic analytical tools utilized in this study are the same as those developed within economic and business disciplines. Briefly, these tools are the net present value methods. This method allows us to evaluate the profitability of marketing activities and capital investments in the rice subsector. The cash flow technique used with nominal net present value models allows us to investigate liquidity problems caused by inflation when long-term investments are undertaken. CHAPTER THREE THE FUNCTIONAL ORGANIZATION OF THE RICE SUBSECTOR Introduction This chapter contains a diagnostic analysis of the rice marketing system, including the identification of the physical flows of paddy and milled rice, and the financial flows that finance them. Although the emphasis is on the physical movement of rice, the stages, functions, and transactions where financing is needed are identified. Particular conside- ration is given to the rice milling and wholesaling activities and consequently to participants, millers and wholesalers, who perform them and their backward and forward financial link- ages. A. The functional organization of rice production 1. Geographical distribution of rice production Rice is produced almost entirely in Guayas and Los Rios: provinces of the Costa region; only'a small proportion.of rice is produced in other tropical zones of Ecuador. On average, these two provinces have produced approximately 90 percent of the domestic supply of rice during the last five years (1984- 1988): Guayas province, on average, has supplied 43 percent 46 47 of the annual volume produced domestically, and Los Rios, 45 percent. However, there exists inter-seasonal variation in production between these two provinces. While the production of rice in Guayas province has remained fairly equal during the Winter and Summer cycles, in Los Rios province Summer production ‘usually' drops to about. 50 percent. of ‘Winter production. The average amount of land utilized in Guayas is 121,000 hectares during the Winter, and 120,000 hectares during the Summer» .Averages for Los Rios are 136,000 hectares during Winter and only 58,000 hectares during Summer. The main reason for the dramatic change observed in Los Rios is the reliance of rice production on rainfall cropping and the lack of irrigation and water management and control facili- ties. This situation seems to be closely associated with restrictions on loans for land improvement works, especially for small farmers not.associated.to some type of organization. 2. Credit needs of rice producers Small farmers with less than five hectares, the majority of whom are organized into production cooperatives, constitute the largest group of rice producers” There are few large rice farmers, some of whom farm more than 1,000 hectares. During the 1988 Winter production cycle, about 89 percent of individual farmers used approximately 31 percent of the rice land in farms of up to 20 hectares. Only 2 percent of 48 farmers used 32 percent of rice land in farms greater than 100 hectares. During the 1988 Summer cycle, 68 percent of farmers used about 16 percent of the rice land in farms of up to 20 hectares. Only 5 percent of farmers used 41 percent of land in farms greater than 100 hectares (Table 3.1). Table 3.1 Land utilization in rice production by size of farms and number of independent farmers during the 1988 production year. SIZE OF WINTER No. OF :SIMER No. OF RICE FARMS 1988 X FARMERS X : 1988 X FARMERS X (HAS) (NAS) : (HAS) Less than 5 9136 17 3654 77 : 1141 4 456 41 From 5 to 20 7212 14 577 12 3732 12 299 27 From 20 to 50 8623 16 246 5 7207 24 206 19 From 50 to 100 10813 20 144 3 : 5953 19 79 7 More than 100 17193 32 110 2 12583 41 60 5 TOTALS 52977 100 4731 100 : 30616 100 1100 100 Source: National Rice Program: Annual Statistics. Adding the usage of rice land during Winter and Summer cycles of 1988, the figures show a very skewed distribution of land. About 85 percent of farmers used 25 percent of rice land in farms of up to 20 has, while only 3 percent of farmers used 36 percent of land in farms greater than 100 has. The distribution found for 1988 appears to be representative of other years as well, as the Annual Statistics of the National Rice Program suggest. 49 During the Summer, the small farmers shifted land from rice to either maize or soybean production because these craps perform better than rice in the dry season. This shift reduced the land area utilized by small farmers in rice production by about 70 percent below their Winter season use of land. Although large producers also shifted from rice to maize or soybeans for the Summer cycle, their reduction of land area used in rice production amounted to only a 27 percent reduction below the Winter season. These shifts suggest that small farmers have less access to irrigation facilities compared to large farmers. Cooperative organization of mostly small farmers pos- sessing less than 5 hectares is extensive in rice production. Small farmers organize into cooperatives to gain better access to credit and to obtain lower prices for farm inputs sold by parastatals (BNF, FERTISA, a fertilizer supplier, and EMSEMILLAS, a seed distributor). The president and manager of each cooperative are responsible for'bank loans.but all the land owned by the cooperative members is collateralized and will not be free until all loans have been repaid. The advantage of the cooperative organization is that it lowers transaction costs for obtaining credit and it gives the opportunity to gain discounts on their purchases of farm inputs. According to PNA's estimations, the membership of the cooperatives typically ranges between 20 and 40 farmers. PNA's technicians argue that members of these cooperatives 50 individually do not own more than 20 hectares. Teble 3.2 shows the distribution of land among cooperatives during the 1988 production year. Table 3.2 Land utilization in rice production by size of farms and number of cooperatives during 1988 production year. SIZE OF HINTER No. OF : SIMER No. OF TOTAL coop. FARMS 1988 CIDPS : 1988 CWPS 1988 (HAS) (ans) : (44s) (44s) Loss than 50 2200 as : 2976 119 5178 Fran so to 100 4402 59 : 5162 69 9564 From 100 to 150 5361 43 : 3636 29 8997 Fran 150 to 200 4224 24 : 4119 24 8343 More than 200 17796 66 = 7560 25 25358 TOWALS 33985 280 E 23455 266 57440 Source: National Rice Program: Annual Statistics. B. The functional organization of the rice marketing system 1. Rice flows, and rice marketing participants a. Paddy rice marketing system There are three major marketing channels for paddy rice: 1) the National Grain Marketing and Storage Enterprise (ENAC) , ii) grain elevators, both private, such as ALMAGRO, and semi- private, such as ALGRACESA (Grain Storage Company), and iii) private mills. The flows of grain, indicated as 51 percentages in Figure 3.1, have been estimated as an average over a number of years. For ENAC the number of years con- sidered is 10, about 5 for grain elevators, and 5 for mills‘. The flow from producers to millers is handled by a group of rural assemblers or middlemen. They usually have receiving outlets in the towns located in the production areas. They transport paddy rice in their own or contracted trucks from the farms to their facilities. The scale of operations of this group of participants is relatively large but they trade with many small farmers who sell small amounts of rice even at peak harvest time. All paddy rice eventually arrives at private mills, except for a negligible percentage that ENAC mills, and about 1 percent that is milled at ALGRACESA's facilities. Also the on-farm consumed rice, which is less than 1 percent of the national production, flows to private mills. The flow of paddy to ENAC has been about 12 percent of domestic production, while the flow of paddy to grain eleva- tors has been about.8 percent, on average» The bulk.of grain, about 80 percent of domestic production, flows to private mills. However, private mills process about 98 percent of all the paddy rice produced, buying rice from farmers and under contracts with ENAC and some private grain elevators. 1 Figure 3.1 is similar to the one prepared by Grupo de Consultores Asociados and presented in its report "Estudio del Sistema de Mercadeo de Granos en la Region Costera del Ecuador", Vol III, Mercadeol de .Arroz, published. by the Institute of Agricultural Strategies, IDEA, July 1987. 52 PRODUCERS I I I 12% 8% 58% 22% A I RURAL ASSEMBLERS I 1 I ENAC ------ > ELEVATORS ----> MILLS 1.----_-__> .---—n—v————a——- 12% 8% 80% ._> Flow of paddy ..... > Contracts for Mi I I Ing Figure 3.1. Market channels and flows of paddy rice. Source: Grupo de Consultores Asociados, "Mercadeo de Arroz", Vol III, IDEA, July 1987. 53 b. Milled rice marketing system The milled rice marketing system is more complex than that for paddy rice. This is because there are more stages and participants. IFigure 3.2 shows theselchannels, flows, and participants. The percentages of rice flowing between stages and participants represents estimations based on field observations and comparisons with former studiesz. The rice flows from ENAC have three main destinations: national wholesalers, retailers, and consumers. The flow originated at the grain elevator level ends up at the national wholesalers level. ENAC is supposed to sell to retailers under some rationing scheme based on quotas determined on the basis of their operating capital. Other important buyers of ENAC stocks are larger wholesalers and government employees, through inter-institutional arrange- ments. The main flow of rice (65 percent of marketed milled rice) goes from millers to national wholesalers, either directly or through brokers who perform their intermediation role by receiving money advances (usually 50 percent) from national wholesalers, paying this amount to millers, and getting credit (usually the other 50 percent) frmm millers until the time that they deliver the product to wholesalers. 2 Basically the study by Grupo de Consultores Aso- ciados, 1987. 54 ‘—— ENAC I ELEVATORS MILLS I I I I I —I 1% 4% 7% 8% 25% 25% 18% 12% v [BROKERS I, 65% <1 I NATIONAL l WHOLESALERS I I 15% 30% 20% I ~—————J 38% 50% URBAN WHOLESALERS ' I 46% 4% /T\ \ /”\ (J h» 82% <1——7;——- I 5% \I I 7% SURERMKTS RETAILERS ENRROVIT I I L I CONSUMERS Figure 3.2 Market channels and flows of milled rice. 55 National wholesalers are relatively few in number and are concentrated mainly in Guayaquil, Quito, and Ambato. They distribute rice at the national level and are involved in border trade with Colombian and Peruvian wholesalers. They form a relatively closed group but are far from being a cartel. Nor do they act in collusion. The main barrier to becoming a national wholesaler is the capital required. Each one handles very large amounts of capital, which are difficult to raise. They can remain in business because of their close personal ties to commercial bank officials. This relationship helps them obtain the funds needed to sustain their high volume of operations. Another important characteristic of this group is their substantial financial and physical capacity to store grain when price expectations are favor- able. The provincial wholesalers are more numerous than the national wholesalers and their operations are usually carried out largely within the boundaries of a province. The distinc- tion between provincial and national wholesalers, as well as between provincial and urban wholesalers is often blurred. The main difference between provincial and national whole- salers is the volume of operations. The difference between provincial and urban wholesalers is that the latter handle a broader group of products, i.e., they wholesale a number of basic commodities and groceries as well as rice. The urban wholesalers handle about 50 percent of the 56 marketed milled rice although they form a large and broadly dispersed group. Despite this importance, they store very little rice, keeping in storage only a working stock. Retailers deliver about 82 percent of the milled rice to consumers. There are a large number of participants and most of them are organized under the National Federation of Marketing Retailers (FENACOMI) , which claims to have more than 600 thousand members. ENPROVIT, the retailing parastatal, has a rice market share of only 7 percent. Originally this parastatal was created to serve as the food distribution mechanism to target populations. Now it is mainly a food retailing network for urban based medium income families. About 6 percent of the marketed rice is delivered by super- markets, mainly to medium and high income families of Quito, Guayaquil, Ambato and Cuenca, the major cities of Ecuador. 2. Vertical coordination in the rice subsector and identi- fication of its financial needs. Rice milling is often vertically integrated with farm production. The reasons for this integration include: i) The need to take advantage of subsidized production credit from the BNF or a commercial bank. These lending institutions have access to the rediscount facility managed by the Central Bank of Ecuador (BCE) through the Financial Funds Mechanism. ii) The need to secure available drying facilities for the grain 57 produced when the harvest exceeds existing facilities. iii) The need to conserve and raise operating capital by milling its own production and selling milled rice. iv) The mitigate the risk associated with investment in specialized assets. Vertical coordination between farm production and milling also takes the form of forward contracts, called "fomento". Through this mechanism a ndller advances money and/or farm inputs toIa farmeerith.the agreement that he/she will deliver their paddy to the miller. The "fomento" is mostly used to finance ‘harvesting expenses. But if the production is expected to be poor because planted area is less than expected or because bad weather is expected to reduce yields, then the "fomento" can be used to finance other production activities as well. Millers "foment" to assure the provision of rice at harvest time, especially when competition for paddy rice is strong among millers and ENAC. Some millers openly charge interest to farmers. Others recover their capital plus interest by down-grading the quality of the offered paddy rice and asking the farmers to add some extra pounds over the 200 lbs a sack of paddy rice usually weighs. This practice, however, has been the basis of the widely held view that millers who "foment" are exploiting farmers. 'The fact.is that this mechanism has existed for a long period of time and its maintenance requires a close relationship, often a kinship, between millers and farmers. Some farmers have tried to substitute for "fomento" 58 financing by a formal credit arrangements with the BNF. But eventualLy most have returned to the traditional "fomento" form of financing, convinced that this is faster, more timely, and even less costly than BNF loans especially when transac- tion costs are counted as part of the cost of obtaining a loan. Millers and producers usually agree on prices at harvest time based on what the market is offering (although farmers are free to search for better prices). But rarely do they not honor their commitment to deliver their paddy rice to the "fomenting" miller. A similar vertical coordination mechanism operates between national wholesalers and millers under the name of "anticipos". Wholesalers agree to advance capital to millers to finance their purchases of paddy during harvest time with the agreement that millers will deliver milled rice upon the wholesalers' demand. In general, wholesalers benefit from this forward contract in at least two ways. First, they fix the 'price at. harvest time (rarely offering' premiums if delivery takes place after one month). Second, they benefit from "free storage" since millers have to assume the costs of this activity. The advantage of this agreement for millers is twofold. One, they partially finance their operating capital needs during times of financial stress“ They also can turn the funds over more than 2 times before they deliver milled rice. Two, they are sure of their market since part of the "anticipo" agreement is to deliver not only the 59 financed grain but some additional amounts. For this part of the, delivery, millers can 'use current. market. prices to compensate for those fixed at harvest time. Here also a long term relationship is required and transient wholesalers do not have access to this kind of agreement. This characteristic establishes clearly two distinct markets for millers: a contract market and a spot market. Transient wholesalers can only access spot markets while those with a longstanding relationship with millers can access the contract market. The rigidity of prices set at harvest time emerge as the outcome of some sort of insurance arrangement between wholesalers and millers by which risk averse millers are induced not only to increase their level of seasonal operations but also to choose larger initial investments in fixed assets. In addition to the "fomento" and "anticipo" mechanisms, which are examples of contract markets, another mechanism of vertical coordination consisting of the spot market. The spot market is the simultaneous exchange of goods for’money (or any other acceptable financial means). This mechanism becomes important because of ENAC's intervention. To fulfill the objective of price stabilization, the government, through the Economic Council, sets a minimum purchase price for paddy and maximum sale price for milled rice. ENAC has been the principal implementing agency of this policy. Its implementa- tion, however, has usually required the allocation of large 60 amounts of financial resources, to both sustain the minimum price and to build buffer stocks to be released when prices are above the maximum price. ENPROVIT has contributed to the defense of the maximum price consumers pay by retailing rice, especially in urban centers. Two characteristics of this pricing mechanism have been relevant. First, the minimum prices have usually been fixed at high levels (an ex-post evaluation indicates that they have usually been above market prices during the period 1972-88). Meanwhile maximum retail prices have been fixed at low levels. In fact, the opportu- nities for private sector intervention could have been eliminated had these prices been rigorously enforced. Second, both prices have been kept effective during a whole season (often, they have been set for a whole year). Due to these characteristics of the jpricing system, ENAC has had. to intervene in the rice market and to assume much of the storage function. These activities have required large amounts of financial resources from the government's budget to cover ENAC's operating losses and to subsidize the sale of milled rice. These activities have distorted the functioning of the rice marketing system, especially when ENAC has stored a large proportion of available rice in the Winter months (see Chapter Five). In addition to these distortions, ENAC adds a great deal of uncertainty to the rice market. The lack of transpa- rency about ENAC's plans concerning the monitoring and managing of the stocks of rice. They raise questions such as 61 when the rice stocks will be released and at what price?. The lack of concrete signals generates uncertainty among market participants that translate into higher operational costs and eventually higher costs to consumers. The trading by description of the product through telephone communications is a coordination mechanism that is becoming more important. An informal grades and standards system has been generated through the use of this mechanism. Among rice traders there is a clear understanding of what white ("flor"), broken ("quebrado"), and yellow ("crema") rice means, as well as what should be understood when they refer to long and medium grain. To these standards, the moisture content is often added by naming the well dried rice as old ("viejo"), and the less dried rice, as new ("nuevo"). C. The structure of the rice milling industry 1. Functional organization of the rice milling industry Supreme Decree No 1593, dated October 22, 1971, estab- lished the National Rice Program and transferred the rights and obligations of the former Superintendency of Mills. The Program's constitutive law, in its article 6, classifies rice mills into three categories: 1) first class mills: those having a milling capacity of more than 20 cwt per hour (cwt/h); 2) second class: those having a milling capacity of 62 more than 8 cwt/h and up to 20 cwt/h: and, 3) third class: those with a milling capacity of up to 8 cwt/h3. Direct observation permitted the identification of most first class mills as "sheller mills". These consist of a mechanical cleaner and dryer, disc sheller, compartment-type separator and vertical cone polisher. There are also 15 mills with a milling capacity of more than 50 cwt/hour that can be classified as modern mills. These include truck scales, receiving hoppers, mechanized cleaners and dryers, rubber roll huskers, electronic separators, polishers with glazers, a full system of conveyors and elevators, and a packaging system. Second class mills have more traditional equipment, such as a simple huller, manual receiving and weighing systems, a mixture of mechanical and solar drying systems, and a single machine that combines the huller and polisher into one unit. Machinery and equipment suppliers agree that "sheller mills" can yield about 65 percent of milled rice, 3 percent of broken, 9 percent of bran, and 23 percent of hulls. “Muller mills" can yield about 63 percent of milled rice, 5 percent of broken, 8 percent of bran, and 24 percent of hulls, on average. The milling recovery rates vary according to size of grain. Yields are higher for short kernels than for long ones, and for dried grains than for more humid grains. 3 In practice, the PNA has increased the upper limit for third category mills up to 10 cwt/h. 63 Third class mills are neglected in this study because they have processed, on average, only 2 percent of all rice produced during 1982-87. These mills are family type enter- prises with a capacity to mill the amount of rice families need for their own consumption. In October 1988 the PNA reported the existence of 1,235 mills in total. There were 226 (18%) of first class mills, 844 (69%) of second class, and 165 (13%) of third class. Among the first class mills there were 3 large mills, each with a milling capacity over 100 cwt/h, totaling a milling capacity of 600 cwt/h and a drying capacity of 1,420 cwt/h. There were also 13 mills with capacities ranging between 50 and 100 cwt/h, totaling 765 cwt/h milling capacity and 1,613 cwt/h.drying'capacityu First class mills have a total milling capacity of 278 MT per hour (MT/h), equivalent to approxima- tely 36 percent of the total milling capacity. The first class mills have milled 59 percent of the total rice produced during the 1982-87 period. Second class mills have a total milling capacity of 458 mt/h, equivalent to about 59 percent of the total milling capacity. They have milled about 40 percent of the rice produced during the 1982-87 period (Table 3.3). 64 Table 3.3 Drying, storage, and milling capacities of mills in metric tons and as per- centages of national totals, in 1988. MILL ZNUMBER DRYING MILLING STORAGE AVERAGE CLASS (*) ----------------- MOT. ------------------ FIRST 226 458 278 134844 113067 SECOND 844 361 458 59835 76572 THIRD 165 10 35 1489 3311 TOTAL 1235 830 771 196168 192950 ------------- % OF TOTAL ----------—-----—-- FIRST 18 55 36 69 59 SECOND 68 44 59 31 40 THIRD 13 l 5 1 2 TOTAL 100 100 100 100 100 (*) It is the annual average of rice milled during each of the years 1982-87. Source: National Rice Program: Annual Statistics. First category mills have larger drying capacities‘ than milling’ capacities and. 69 jpercent. of the total storage capacity. Second class mills have more milling capacity than drying capacity and only 31 percent of the storage capacity. A larger drying than milling capacity is needed, however, ‘ Drying capacity is based on the time paddy rice, with an average of 22 percent moisture content, can be dried down to 13 percent moisture. Using solar heat, the rice needs usually two full days to lose 9 percent of moisture (field research). The drying rate estimated and reported by the PNA for this system demands a careful use. 65 since grain has to be dried approximately at the same rate as harvesting, while milling can be paced to market (consump- tion) needs. First class mills have better facilities to serve rice processing needs during harvest time than second class mills. About 30 percent of the first class mills total drying capacity corresponds to a solar system, based on the use of concrete slabs, called "tendales". The solar drying system exceeds 50 percent of the drying capacity among second class mills. Discounting these proportions of solar drying facilities, because they are unreliable and because they often are used to dry maize, soybeans and coffee, the total mechan- ized drying capacity is approximately 500 MT/h. Considering that 1) about 53 percent of the'Winter production is harvested during the peak month of June; 2) the peak harvest for June reached 198,600 MT in 1985; 3) the harvesters operate only 26 days per month during harvest time: and, 4) the harvesters work at most 12 hours per day, the hourly rate of harvest results in approximately' 636 MT/h, 27 percent over ‘the estimated mechanical drying capacity of mills. Only’the largest first.class mills (about 15) have modern metal silos for storing paddy. Their total storage capacity is 48,000 MT, of which 22,300 MT belong to the three largest mills. The rest of the first and second class mills have simpler warehouses, made of brick and wood, without protection against birds and rodents, and facilities to aerate the grain to control its relative humidity. 66 In connection with these findings and considering that 1) the largest Winter harvest reached 374,600 MT in 1985: 2) the harvest profile is 3 percent in April, 21 percent in May, 53 percent in June, and 23 percent in Julys: 3) the monthly consumption is 25,000 MT of paddy rice equivalent to milled rice, the need for storage space would.be equal to the figures shown in Table 3.4: Table 3.4 Flows and consumption.of‘Winter harvested paddy rice and the need for storage facilities, based on 1985 Winter harvest data. APRIL MAY JUNE JULY ----------------- 000 ur-------—------------------------- (a) Harvest 11.2 78.6 198.6 86.2 (b) eonsmption 25.0 25.0 25.0 25.0 (a-b) Difference 43.8 53.6 173.6 61.2 cunulative -13.8 39.8 213.4 276.6 SEurce: National Rice Program: Annual Statistics. By the end of July, the needed storage capacity amounts to 274,600 MT. The existing storage capacity of mills totals 192,900 MT. About 75 percent of this capacity could be considered adequate, according to technical personnel of the PNA, resulting in.a deficit of about 130,000 MT. 'This deficit is supplied by ENAC's storage capacity for both paddy and 5 "Plan de Comercializacion y Financiamiento de Productos Agricolas para el Afio 1989 con Enfasis en el Arroz y Maiz Duro", Comisién Especial, Abril 5, 1989. 67 milled rice (86,600 MT), and the facilities of private and semi-public grain elevators (82,500 MT)‘. The location of rice mills within the rice production area is relatively dense. The harvested area during 1988 Winter was about 90,000 hectares, and during 1988 Summer, 55,000 hectares. Taking the Winter rice area, the location of mills shows an average ratio of one mill for every 70 hectares, or one mill for every 350 MT of paddy, considering a yield of 5 MT/ha. Assuming 60 ten hour working days during the Winter harvest period, this average means that each mill would have to dry rice at a rate of less than 2 cwt/h. There are several reasons for the establishment of a large number of small mills in the rice producing area of Ecuador. The first reason is the poor transport facilities, especially before the petroleum boom of the mid-seventies. This situation favored the location of mills close to farms. A second reason is the exclusive allocation of administered funds to finance production and processing of agricultural products as a vertically integrated operation. This condition limits the access to credit by non-farmers entrepreneurs willing to invest in rice mill equipment and machinery. A third reason is the scarcity of funds at BNF and CFN to finance long-term investments. And finally, another reason 6 These additional facilities must share their space with other crops, such as maize and soybeans. The storage facilities of animal feed manufacturers are excluded, as well as those of rice wholesalers. 68 is the relatively high profitability of these small mills, given their low volume of investments in drying and milling equipment. Overall, the rice milling industry is characterized by the predominance of small scale, relatively obsolete machin- ery. About 75 percent of mills were built more than 15 years ago, and about 85 percent more than 10 years ago. Only a few of the more modern mills (no more than 50) have been built during the last 5 years. The recent trend has been to build larger mills adding modern equipment not used before such as truck scales, large hoppers, and continuous drying machinery. There is evidence of sub-utilization of installed milling capacity. Some possible reasons for sub-utilization might be the limited access to funds to finance working capital, the accumulated depreciation of most mills that has already covered the book value of machinery, and the small average cost reduction due to increases in the operating levels. The usual operating volume of most mills corresponds to the flat part of the average cost curve, reducing the incentives to aggressively increase operations. Most of the milling machinery needs replacement and modernization. Restrictions on such an endeavor currently are the inflation and liquidity problems, which cause delays in replacing some major parts of the old plants, and the con- straints on the accessibility to financial resources to finance long-term investments. 69 2. Identification of functions and credit needs a. Credit needs for rice processing There are two important functions where credit may improve the functioning of the rice marketing system at this stage: i) purchase of paddy rice to be dried, stored and milled: ii) investments in equipment, machinery and storage facilities either to add to the existing ones, or to replace them partially or totally. i. Purchase and storage of rice The study of the rice milling industry has shown the use of informal mechanisms to finance part of the purchases of paddy rice. One is the "fomento" mechanism and the other is the "anticipo". Vertically integrated millers usually take advantage of the BNF's production credit to partially finance their "fomento" activities. This practice reduces the availability of loanable funds which can otherwise be ad- dressed to a larger group of rice producers. Although this procedure does not necessarily show a lack of financial resources, it does indicate the need to revise the credit delivery system of the BNF if a larger number of small farmers are to be the direct beneficiaries of this facility. This indication does not lead necessarily to the conclusion that 7O millers financial intermediation needs to be eliminated, but that this credit mechanism should be made more transparent (see recommendations in Chapter Eight). The general consensus among millers is that the "ade- lanto" mechanism is a recourse of last resort, utilized basically because of their limited access to formal credit, either from the BNF or from commercial banks. The study of prototype mills does not lead to an unambiguous conclusion concerning the costs and benefits of this financial mechanism for millers. There are some indications that it is expensive for millers, more so in periods of high inflation and when delivery of milled rice is somehow delayed. But on the other hand, it permits millers to increase their operations, and therefore, their profits. Obviously, it has an advantage as a multiplier. Additional amounts of credit to finance these activities, especially storage, and more accessibility to it, would reduce or even eliminate the need for millers to rely on "adelanto" financing. ii. Investments in equipment, machinery, and storage facilities The lack of adequate drying equipment and storage facilities, and the obsolescence of the majority of milling machinery is apparent. One important and urgent investment is the shift from a solar drying system to a mechanized 71 drying system. The advantages of modern drying equipment in terms of cost, the ease of installation and operation, the relatively low cost of additional civil works, and the ease of controlling temperature, relative humidity, and cleanness are evident. The conversion rates from paddy to milled rice are higher when paddy has been dried by mechanical means than when paddy has been dried with solar heat. The losses are less and the milling process yields less broken kernels. The probability of increases in the price of diesel and electrical energy, however, are additional considerations that may serve as a deterrent to investments in mechanical drying equipment. The need for additional mechanical drying capacity will increase as production expands. The possibility of exporting rice requires to improve the rice processing technology to obtain better quality of rice (see Whitaker and Alzamora, 1988b, p 24). There is a large deficit of silo storage space in the mills and of adequate warehouse space. The majority of existing warehouses are simple constructions of brick and wood, and metal roofing. They are built at the ground level, without adequate protection from water, rodents, and small reptiles. The roof also lacks adequate protection, and birds and insects can easily get into the warehouse. Warehouses do not have facilities to aerate and to control the relative humidity of grains. 72 b. Credit needs for rice wholesaling The first stage of wholesaling is performed by a small group of participants with access to large amounts of capital made available by commercial banks. Entry into this group of market participants is limited by the capital needed to compete effectively by advancing money to millers to secure a large volume of milled rice. Provincial and urban whole- salers would like to gradually form part of the national wholesalers group and.to be able to buy larger amounts of rice directly from millers. Less concentration of credit at the national wholesaling stage could have the effect of gradually eliminating one or two steps within the wholesaling channel to the benefit of both consumers and producers. Surely a larger allocation of trade credit to millers would avoid the recourse to the "anticipo" financing, and millers might have more freedom to search for better markets. This change would reduce the power national wholesalers seem to have over an important part of the rice flow. Yet, credit is still necessary to finance the wholesaling function. The assurance of a market for milled rice that the "adelanto" (contract market) seems to provide could be replaced by new market institutions, such. as forward, contract. markets and. more intense use of the Agricultural Commodities Exchange (Bolsa de Productos Agropecuarios) where warehouse receipts are traded. 73 Credit also is needed to finance grain handling equip- ment, transportation facilities, and warehouse remodeling works. The process of loading and unloading sacks of rice to and from trucks, as well as the process of filling and evacuating a warehouse is completed using manual labor. Grain handling at ‘warehouses can. be improved by' adding small equipment to sack rice into bags of less than 100 lbs weight. Credit is needed to finance trucks which will be used by urban wholesalers to directly access mills and to deliver the product to retail outlets. Wholesalers have expressed their need to have adequate warehouses, although urban wholesalers recognize that these warehouses will be multipurpose, that is, to store rice, sugar, and other commodities and groceries. Summary Rice production is an activity involving many small farmers and a few large, mechanized agricultural enterprises. The majority of independent small farmers do not have access to formal credit and they have to use funds supplied by millers through the "fomento" mechanism. Small farmers appear to gain access to formal credit when they join cooperatives while at the same time benefiting from lower costs for their farm inputs and equipment. The milling industry is facing serious problems. Millers seem to lack enough formal financing for their operating 74 capital and have to rely on the "anticipo" mechanism, a forward contract agreed upon with national wholesalers. It is apparent that the majority of mills are old and urgently need remodelling. They also need to significantly improve their storage facilities. Yet, it appears that there are not enough incentives for millers to undertake major investments. So most of them.are surviving with fixed assets that have been completely depreciated. Better access to credit at the national wholesaling level might help reduce the relative concentration of economic power among national wholesalers. credit access might also make possible more direct linkages between millers and urban wholesalers, thus reducing distribution costs. Many wholesalers have expressed the need for financing the purchase of grain handling equipment as well as transpor- tation facilities and warehouses remodeling. CHAPTER FOUR THE CREDIT SYSTEM FACILITATING RICE MARKETING ACTIVITIES Introduction The functional organization diagnosis of the rice subsector has identified the activities where financial resources are needed. This chapter will investigate how credit is allocated and who has access to it. In this chapter, some financial constraints hindering a more efficient rice marketing system operation will be identified. One such constraint is subsidized.credita These issues will be studied in detail in this chapter. A. Functional organization of the financial system 1. The BNF as the official credit administrator The BNF has supplied credit to finance the main rice subsector activities. Credit has been provided to finance production inputs such as seeds and chemicals, farm equipment and machinery, irrigation equipment and land leveling. Credit has also been provided to finance paddy rice trading, with the goal of helping producers delay the sale of paddy rice post harvest dates. Finally, credit has been provided to finance long-term investments in drying and milling equipment. 75 76 The BNF also extends credit for financing a broader category of milling equipment under its agro-industrial credit program. This makes it nearly impossible to quantify the amount of the official credit for long-term capital investment. The same data problem occurs with the credit extended to finance fixed farm assets, included in the farm machinery and land improve- ments credit categories. Table 4.1 BNF's loans to finance the production and marketing of paddy rice during the period 1970-88, reported in nominal and real sucres, with May 1978-April 1979 = 100. YEAR PRODUCTION TRADE PRICE PRODUCTION TRADE CREDIT CREDIT INDEX CREDIT CREDIT --S/000 nominal-- ---S/000 real--- 1970 68821 40155 38.5 178756 104299 1971 57874 7225 40.6 142547 17796 1972 53875 7488 43.7 123284 17135 1973 225501 8874 49.0 460206 18110 1974 613422 23534 60.1 1020669 39158 1975 743764 45411 68.7 1082626 66100 1976 850300 17957 75.7 1123250 23721 1977 572337 138437 85.5 669400 161915 1978 547345 29217 96.7 566024 30214 1979 745519 22417 105.9 703984 21168 1980 863265 28379 118.3 729725 23989 1981 915851 31251 135.9 673915 22996 1982 948743 27463 158.1 600090 17371 1983 1969286 137847 234.6 839423 58758 1984 3133442 556710 307.8 1018012 180867 1985 5916193 608755 394.0 1501572 154506 1986 6622702 969828 484.7 1366351 200088 1987 6454682 1064797 627.7 1028307 169635 1988 9645780 2652340 983.6 980655 269655 Source: Credit: BNF: Annual Reports Price Index: Banco Central del Ecuador: Memorias 77 1.5 1.54 .- 141 ‘ 1 3 1 1.2- m 1.1—1 5 " mg 0.91 32 08— " gig, 0.7 ~ “ s 7 = .1 06- .. g 0 S — ' 041 u 0.3— 0.2 u 01 = -‘ OR I I I I I I I I T I I I I I I I I 70 71 72 73 74 75 75 77 78 79 80 81 82 83 84 85 85 87 88 CALENDAR YEARS D SEASONAL CREDIT + TRADE CREDIT Figure 4.1 BNF's production and trade credit, in real terms (May 1978-April 1979 = 100), for the period 1970-88. Source: National Rice Program: Annual Statistics. 78 The amount of production credit has decreased during the last 4 years. Meanwhile the amount of credit to finance rice trading has increased, both in nominal and constant sucres. The amount of credit to finance rice trading includes the credit for ENAC that the Central Bank allocates and passes through the BNF. Most of this credit has been used by ENAC to liquidate overdue loans, leaving a small proportion of new operating funds in each year. Three subperiods are important to notice in Figure 4,1. The first subperiod includes the 1974-76 years, characterized for an accelerated growth of credit financed mainly with oil export revenues. The second subperiod includes the 1983-85 years, characterized. also 'by Ianother increase of credit accompanied by high support prices (Stewart and Cuesta, 1988) . The third subperiod is the 1986-88, characterized. by a dramatic reduction of real credit as a consequence of high inflation rates, especially during 1988. Two observations are worth noticing in relation to the BNF's trade credit. First is the relatively low amounts of funds allocated for this purpose. Second is the recent trend to increase this allocation. This trend is apparent during 1984-88, despite the increasing inflation rate during this period. The beneficiaries of these trade loans in the rice subsector have been mainly farmers and, in a lesser degree, large millers. The average term for production loans is 6 months. The funds are disbursed by parts, as major 79 investments are undertaken, that is, after the farmer has financed his/her expenditures. This procedure causes cash flow deficits at harvest time, and farmers have to search for informal financing to cover their expenses. Immediately after harvest, farmers have the obligation to repay the loans because the collateral (the grain) has been disposed of, even though they may still have some time before the end of the term. During the last 4 to 5 years, rice farmers have been negotiating the swap of loans from the production to the trade category. This gives them another 3 months of time after harvest to repay their debts. This practice explains why trade loans have been increasing. BNF's participation in financing long-term investments in rice milling equipment and machinery peaked during 1974- 78 period. These years correspond to the petroleum boom period. It was also a period of aggressive borrowing from the Inter-American Development Bank (IDB) and World Bank (IBRD), through their global agricultural credit programs. However, since 1985 the BNF has not approved loans to finance drying and milling equipment, and storage facilities for millers. The corresponding IDB and IBRD credit programs regulations allow long-term investment financing in milling equipment and storage facilities only at farm level. The experience of these programs indicates a negligible demand for funds by farmers for long-term investments related to rice processing. One of the reasons to limit financing of long-term 80 investments in the rice processing business is the belief that there is enough drying, milling and storage capacity in the country' to process rice. 'This conjecture neglects the importance of the adequacy of the existing processing facili- ties, which are not acceptable. From the credit demand side, some reasons that explain the lack of effective demand for long-term investment funds are the following. First, the low profitability of these investments in the farm, given the low premiums for a drier and cleaner grain. Second, the lack of managerial capacity of most farmers and the lack of confidence impeding farmers to hire external management for either their farm.or their rice processing enterprise. .And third, the lack of technology dissemination throughout rice growers. The interest rate charged by BNF for both production and trade loans has been kept below the inflation rate during the 1970-88 period. During 1988 this rate was fixed by the Monetary Board at 23 percent, while the inflation rate was never below 45 percent. In April 1989, (El Comercio, April 21, 1989) the Monetary Board agreed to raise the interest rate to 32 percent. But the inflation rate had increased to over 90 percent, making even more negative the real interest rate. The negative effects of "subsidized" interest rates have been extensively studied and will not be analyzed in this study (See for example Adams, 1984; Von Pischke and Adams, 1980). Finally, we need to consider BNF's relationship to ENAC. Since 1974 ENAC has been getting large amounts of credit from 81 the Central Bank through the BNF. The loans allowed ENAC to intervene in the trading of grains, including rice. These actions, however, reduced considerably the amount of funds that could otherwise be allocated to private traders. The price system imposed on ENAC by the government.has not allowed it to profit from its intervention. On the contrary, it has been consistently loosing significant amounts of funds and increasing its overdue accounts at BNFu By December 1988, the overdue portfolio of ENAC exceeded 18 billion sucres. This may represent the costs of ENAC's intervention or the cost of the subsidy to consumers. But the government has not been willing to accept it as a subsidy payment and the deficit (losses) has been attributed to poor and perhaps corrupt administrators. The Annual Statistics of the PNA show that, on average, the BNF's share of rice production financing has been 58 percent for'Winter crop, and 50 percent for Summer, during the period 1986-88 (Table 4.2). 82 Table 4.2 Percentage participation of financial institutions in the seasonal ricejproduc- tion credit as an average of the years 1986-88, in nominal terms. Cycle BNF Cannerci el “Fonento” Own Total Banks Fmds X x X X X Winter 58 9 8 25 100 Smaller 50 12 12 26 100 Source: National Rice Program: Annual Statistics. During the 1981-88 period, the BNF's production credit has financed approximately 70 percent of the total area cultivated. The BNF's Statistics do not show the composition of its beneficiaries. However, it has been discovered that BNF has provided more financing for large farmers than for small and medium ones (Table 4.3). 83 Table 4.3 Sources of credit for the 1988 rice Winter crop. Fern BNF Caunerci at Other Si ze Banks Sources (has) 2 X X 0 - 50 45 30 25 50 - 100 52 28 20 100 - plus 70 12 18 ‘ Represents “fomentofi and farmers' own re- sources. Source: "Hoy" newspaper, December 10, 1988. 2. Private banks participation in the rice subsector The extent of private banks participation in rice production is limited to the legal obligations imposed on them. The Monetary Board regulations require that private banks have at least 20 percent of agricultural sector loans in their portfolios. The agricultural loans can be redis- counted at the Central Bank, under the Financial Funds Mecha- nism. Commercial banks fulfill this requirement with a few large loans and, demanding large amounts of security' as collateral. They seek investments considered highly profit- able with low risk. Additional reasons for limiting loans to few and large borrowers is to reduce transaction costs and to reduce their need for trained technicians to closely control and supervise investments (Ramos, 1984). The interest rate spread.on the use of the Central Bank's Financial Funds Mechanism is considered by private banks 84 officials insufficient to stimulate an extensive participation of commercial banks in agricultural loans, especially because the rate does not allow them to cover default risk. With regard to commercial banks financing of rice production and processing, the few loans granted by them are devoted to large rice farmers. The loans devoted to invest- ments in equipment and machinery for rice processing are negotiated for less than 2 years. These short-term loans, however, can be rolled over up to 3 times. At each time the interest rate is determined and the value of the collateral is revalued. The major participation of private commercial banks in the rice subsector is found at the wholesaling level. Here private banks provide the operating capital needed to facili- tate trading and private storage of rice. The interest rate charged for these loans is the market rate, which has been below the inflation rate but higher than the "subsidized" rates charged by the BNF and the Central Bank's Financial Funds Mechanism. During 1988 this rate fluctuated from an average of 45 percent in January, to a high 48 percent by mid year, and reached an average of 52 percent by the end of 1988. Meanwhile the inflation rate reached 85.7 percent by December 1988 and averaged about 56.7 percent (E1 Comercio, Martes Econémico No. 10, February, 1989). The main beneficiaries of these funds have been largely national wholesalers, and to a lesser extent, provincial and urban wholesalers. 85 3. Trading and financial institutions for rice marketing a. The issuance and use of warehouse receipts Warehouse receipts are legal trading instruments issued by public and private grain warehouses operating under the National Storage System (Sistema Nacional de Almacenamiento), of which the main participant is ENAC. These receipts were used primarily to collateralize commercial loans but also to facilitate grain exchange. IBefore the creation of the Agricultural Products Exchange Board (BPA) in January 3 of 1986, the main issuers and users of these instruments were ALGRACESA and ALMAGRO, a division of Banco del Pacifico. The creation of the EPA in 1986 was a government initia- tive. The intervention of government in its origin and functioning gave rise to bitter opposition from interested sectors, mainly from producers associations and from the bureaucracy of ENAC, leading the BPA.almost.to its extinction. The creation of the BPA was not accompanied by an effective educational process to teach potential users about the Exchange Board objectives and operations. The primary participants in the grain marketing system expected that this institution would solve their problems of transportation and trading of their products. The role of the BPA as a facili- tating institution for price discovery and market transparency was not'widely’understood, Nor were the uses of the so called 86 certificates of deposit well understood. Their purpose was to facilitate trading, reduce transaction costs, and serve as collateral for bank loans. During the first year of operations the main product traded in the Bolsa was rice. This product alone included about 56 percent of the Bolsa's annual trading operations and represented about 16 percent of the 1986 rice production. From January to November of 1987 rice trading in the BPA reduced to only 11 percent of the 1987 rice production (Analisis Semanal No. 47, December 9, 1987). After' a period of relatively intensive use of the warehouse receipts (1986 and 1987), their use was still not completely understood and accepted by the majority of farmers and traders. Small rice farmers accepted the warehouse receipts, especially during 1986, because ENAC was obligated to issue them to avoid using new loans from the government. In addition to this, the market was saturated by the bumper crop of 1985 and market prices were below official prices. Large rice farmers accepted with less resistance the use of these warehouse receipts, perceiving the advantages of negotiating with these instruments and of using them to collateralize their commercial loans. 87 b. Critical issues concerning the use of warehouse receipts It is important to briefly discuss some aspects of the role of ENAC, BNF and the BPA to understand the reasons why warehouse receipts were not accepted. i) Role of ENAC: Policy makers saw in the warehouse receipts an expedient means to avoid funding ENAC with new money. All of ENAC's 1986, 1987 and most of 1988 purchases were made by issuing indiscriminately warehouse receipts. These receipts were supposed to be redeemed at the BNF. But a number of problems surrounded this mechanism. Farmers, especially small ones, were not prepared to receive an unknown document instead of cash for their product. This situation created resistance and apprehension. Then ENAC's paper work required to issue a receipt became painful and time consuming, increasing insecurity and transaction costs to the already cumbersome exchange. Also, there was lack of freedom to negotiate these receipts at banks or even at the Bolsa, since they were earmarked for redemption only by the BNF. Finally, these receipts contained a discount for storage costs, giving them the appearance of depreciable papers that had to be rapidly passed on to another participant. This perception was exacerbated by the rigidity of official prices, which were not allowed to be adjusted for storage costs. 88 ii) Role of BNF: BNF officials saw this mechanism as an expedient way to recover their loans, and consequently supported its intensive use. The redemption process took this form. The certificates of deposits were received by the BNF and then sent to the BPA to be sold. Only after these warehouse receipts were transferred to a buyer (almost entirely the same ENAC) and cancelled, would the BNF proceed to liquidate them. The BNF would discount the value of the loan, the interests accumulated up to this date (not the date the farmer handed in the receipt) and the storage costs, and would return the balance to the farmer (the discounted storage costs were later credited to ENAC's accounts). It is easy to see why farmers considered this mechanism cumbersome and exploitative. iii) Role of the BPA: In addition to the cumbersome procedures described before, inside trading of some brokers of the BPA increased the doubts that market participants had about the system. Rice trading promoted by some brokers were never registered in the BPA and the exchange took place without any contact between either the seller or the buyer with the BPA. 89 4. Non-institutional sources of credit Too often informal financial arrangements have been used in the rice subsector such as the "fomento" and the ”ade- lanto". The "fomento" mechanism consists of millers financing farmers harvesting expenses. The "adelanto" consists of wholesalers advancing money to millers as part of an agreement for future delivery of milled rice. The study of these two informal credit mechanism emphasizes their economic rationale and financial relevance. From an economic theory perspective, these financial arrangements can be associated with the markets for paddy and milled rice. These markets include the spot market and the contract market. Rice producers can choose to wait until harvest to sell their production or to contract paddy rice delivery for a future date, agreeing upon prices and lead time. Rice millers can also sell milled rice in the spot market, at the ongoing price, or they can contract with wholesalers for future delivery of milled rice, agreeing also upon prices and lead time. The analysis that follows refers more specifically to the "adelanto" or contracts between millers and wholesalers. In the spot market, prices fluctuate relative to the strength of the demand and supply forces. In the contract market, both prices and lead time are subject to negotiation and can vary in some inverse relationship. A miller can agree 90 to fix the price at the time of the contract agreement, but he/she can negotiate the extent of the lead time to create the opportunity to turnover the "adelanto" several times and thus compensate any foregone benefit caused by freezing the price. The longer the lead time, the larger is the number of times the "adelanto" can turnover. Or else, a miller can negotiate higher prices (premiums) for shorter lead times if wholesalers are turning over their capital faster to accommodate supply to a growing demand. The contract guarantees the wholesaler will have a preferred place in the queue when demand is high. The contract, on the other hand, guarantees the miller that the wholesaler will continue to purchase milled rice at the fixed price even though the*wholesaler could find lower'prices on the spot market, when demand is low. If the wholesaler switches to the spot market.because of better prices, and lets the miller extend the lead time beyond the end of harvest date allowing him/her to incur on unexpectedly higher storage cost, then the wholesaler may not be able to return to the contract market as a loyal customer and may loose his/her preferred position in the queue when demand is high. Millers use the wholesaler's "adelanto" especially during the Winter harvest period, that is, during May and June, extending the lead time up to 2 months, to turnover this capital several times. Beyond this term (usually 2 months, ending early July), when the opportunities to turnover the "adelanto" are reduced considerably. Millers want to deliver 91 the contracted milled rice to avoid incurring storage costs. If they want to store some rice beyond harvest period because of expectations of higher prices, they would like to own the grain to benefit from the returns to storage. Millers are assumed to be risk averse. They would prefer to invest small amounts of their own capital and larger amounts of the lenders funds in the business of buying paddy, storing it, milling it, and then selling it as milled rice. If millers were to perform their businesses with their own capital only, the volume of operations would be lower than otherwise because of risk considerations. If millers were able to finance their businesses entirely with debt, the volume of operations would be larger because any profits they get would represent an infinite rate of return, ‘The advantage of the "adelanto" is to induce millers to behave as if they were risk neutral, stimulating them to buy, mill and sell larger amounts of rice. This behavior' can explain some aspect of the rice marketing system. First, the efforts of millers to gain a contract with ENAC to serve as its assembly center, buying paddy with ENAC's resources and delivering milled rice after an extended lead time. Millers can make good profits even after'paying official prices for paddy rice when market.prices are lower, since the funds they use are external to their finances and they can extend the lead time for several months, delivering milled rice to ENAC only when storage cost becomes 92 larger than benefits. Second, a rice subsector economy of risk averse millers would allocate smaller capital amounts to finance trading than a rice subsector economy of risk neutral participants. The supply curve for the risk averse subsector would lie to the left of the a supply curve for the risk neutral subsector. Therefore, for a given demand, prices in the risk averse subsector will be higher than in the risk neutral subsector. Millers are induced to behave as if they were risk neutral when they have the opportunity to reduce investment of their own capital and increase the use of lenders' funds. Millers can know with certainty both the cost of the borrowed capital and the purchase and the sale price of the products they handle, estimating with great certainty the profits they can make. B. Official allocation of funds to the rice subsector Official credit has been targeted to three main destina- tions, toIsupport rice production, to finance trading of rice, and to finance long-term investments in equipment and machin- erYe i) Credit allocated to increase rice production: Credit and production are assumed to have a positive relation- ship. However, since credit is fungible, it may be diverted to finance activities offering a higher rate of return, or to 93 substitute funds otherwise used to finance the purchase of consumption goods. This possibility has attracted powerful borrowers who have obtained large amounts of credit, leaving lenders short of loanable funds to satisfy the.demand.of small borrowers. ii) Credit aimed at financing marketing activities: Biases against intermediation and private storage have limited allocation of official funds to finance trade and, more rigorously, storage of rice. Perceptions of a linkage between short-term credit and inflationary pressures have been expressed by credit policy officials, neglecting to consider that trade, processing and storage (space, form, and time utilities) are economic activities that add value to the product. iii) Credit aimed at financing long-term investments in the rice marketing subsector: Currently, funds to finance investments in Idrying' and. milling equipment. and storage facilities are scarce. These investments can be officially financed only when they are placed in the farms and borrowers are farmers. This limitation derives from the assumption that storage and processing facilities are sufficient to handle the flow of grain. Although the assumption seems to be realistic, it needs to be revised considering the current status of the industry and some performance dimensions such as efficiency, 94 cost-effectiveness, progress and.modernization of the plants, and also taking into account adjunct services such as trans- portation, which have improved during the last two decades. Another cause for limiting the allocation of funds to finance these long-term investments stems from the impact of inflation on the interest rate and the term of the loans. Under a financial policy which sets controls over the interest rates, fixing them below the inflation rate, lenders ration the loanable funds away from long-term investments, in general, and, particularly, away from investments in acti- vities which the political environment makes more uncertain and riskier. Profitability’ of 'the investments. and. the impact. of inflation on the liquidity of these investments, could be another cause for limiting the allocation of financial resources to these long-term investments. C. Impacts of subsidized credit 1. Transaction costs and credit rationing It is critical to include in this analysis the concept and implications of transaction costs in the credit delivery systems of the financial intermediation. From the lender's viewpoint these costs include the inspection expenses and credit supervision, court and legal expenses incurred to 95 recover overdue loans, and the cost equivalent to the default risk, if it is not yet included in the financial (interest) cost. From the borrower's perspective, transaction costs include all the expenses incurred in getting the loan, that is, the trip(s) to the bank, the opportunity cost of the time spent in the negotiation, the expenses to fulfill all the requisites demanded by the lender, and some additional expenses such as food and lodging. A real negative interest rate (represented by the vertical axis in a two-dimensional graph) calls for a demand for loans (represented by the horizontal axis) asymptotically approaching the horizontal axis and located outside the supply curve. When this situation exists banks have to ration their loanable funds among borrowers. If in addition to this, the interest rate is administratively fixed and regulated, rationing must be accomplished using mechanisms that can affect transactions cost. These mechanisms are negotiations over the time span of the repayment schedules, the amount of the loan, the proportion of borrower's equity required as counterpart, the value of the collateral, and the list of requisites to be fulfilled. All these factors affect the magnitude of transaction costs. For example, the author's field survey revealed that informal sources of credit offered much lower transactions costs to small borrowers and over small amounts of loans than formal lending institutions, especially when confidence and trust have been cultivated 96 between the borrower and the lender. 2. Effects of subsidized interest rates Subsidized credit generates a chain of important effects that influence the functioning of the entire sector and economic system. To organize the discussion of these effects, they will be looked at from the supply of and demand for financial resources. i) From the supply side: Low interest rates limit savings and resource mobilization. An extensive body of literature concerning this issue has been published, led especially by the applied research conducted over a number of years and less developed countries by faculty and researchers of the Ohio State University (See for example Von Pischke, Adams and Donald, 1983: Adams, Graham and Von Pischke, 1984). Low interest rates affect negatively the amount and purchasing power of the lender's capital. The deterioration of capital occurs when the inflation rate is superior to the interest rate: lenders capital loses purchasing power, even when the default rate is zero; the capital may fade away if write-offs have to be made over and above the reductions of purchasing power. Given these two detrimental effects, the lender's financial activities, particularly when the lender is an 97 official bank like the BNF, are sustained by two sources of funds. First, government budget allocations, which increase fiscal outlays and, most likely, the fiscal deficit. The increase of fiscal deficits bring about an increase of inflation. If this increase of inflation is not fully absorbed by the interest rates, the magnitude of the distor- tion increases, causing larger and larger erosions of the lender's capital. Second, loans from external donors, which increase the costs of the debt services, demanding larger allocations of fiscal funds to honor them, which in turn increase the fiscal deficit, setting in motion inflationary pressures. ii) From the demand side: Real negative interest rates cause, theoretically, an infinite demand. Subsidized interest rates do not improve the profitability of investments. Loaned funds have to be converted into productive inputs and their prices are likely to increase when cheap loans are made available in larger amounts than under real positive interest rates. The cost of production does not necessarily decrease by financing investments with low interest rates. However, capital gains are likely’toIbe obtained.by fixing low interest rates, which are capitalized in fixed assets (See for example Lee and Rask, 1976: Robison and Brake, 1980). The value of these assets increase to reflect capital gains, making their owners richer: the larger'the.amount.of’subsidized.credit.they 98 can get, the richer they are. Thus subsidized credit distorts the distribution of income, exacerbating the existing inequi- ty. Summary This chapter has identified the main suppliers of financial resources and has also identified the main credit mechanisms. Structural and institutional restrictions have been identified both in the formal and informal capital markets. Particularly those that support the rice marketing system. Few incentives exist for private banks to finance rice marketing activities. Primarily because controlled interest rates and risk perceptions give them earning below market rates of return. Biases against middlemen and.miscon- ceptions about the benefits of short term credit for trading and storage are the major causes for an insufficient alloca- tion of formal credit to finance these activities. The lack of credit to finance long term investments in rice drying and milling equipment, and storage facilities is due the uncritical assumptions about overall capacity for rice processing, the detrimental effects of inflation on interest rates and term of loans, and the negative impacts on profita- bility and liquidity of these types of endeavors. CHAPTER FIVE THE USE OF CAPITAL TO FINANCE RICE MARKETING AND STORAGE Introduction This Chapter's main objective is to assess the profitabi- lity of financing rice trading and storagewwith formal credit. The chapter will focus on millers and wholesalers, including the role of ENAC, especially in storing rice. It will be shown that, despite low margins between the official minimum and maximum prices, private sector storage exists when price enforcement is not rigorous. The first two subsections of section A will discuss the profitability of financing rice trade at milling and wholesaling stages, respectively. The first subsection of section B will discuss the profitability of storing rice, identifying the key participants in this function of marketing, and the quantity and seasonality of stocks. The second subsection will develop a brief analysis of the seasonality of prices. This will help identify the pattern of rice purchases, storage and stocks release. The Chapter's third section discusses the profitability of storing both paddy and milled rice. Finally, the Chapter ends with a discussion of the role of ENAC in the storage function. 99 100 A. Working capital needs for marketing and storing rice 1. Financing millers working capital The author's field survey revealed that about 20 percent of millers' financial resources used to trade paddy rice during the Winter'harvest.of 1988 was financed by wholesalers, through the "adelanto" mechanism. The hypothesis tested in this subsection is that the substitution of formal sources of credit for "adelanto" financing would improve the financial and economic outcomes of mill operations. To test this hypothesis, the preparation of annual operating budgets and financial statements for the three prototype mills construc- ted in this study were prepared (See Appendix A). Several assumptions were used to analyze the profitabi- lity of rice trading at the mill level. First, the sources of finance for the prototype rice mills are assumed to be the BNF and. commercial banks (40 percent), the ‘wholesalers' "adelantos" (20 percent), and the millers' own resources (40 percent). Second, the commercial loans used to finance rice trading from banks assumed to be obtained at 90 days terms. Third, the "adelantos" are supposed to be paid back one third every month after the May and June disbursements, and one half every month after the October disbursement. Finally, the commercial loans are assumed to be negotiated at an annual interest rate of 44 percent (the average rate during 1988). 101 The financial cost for the wholesalers' "adelantos" was estimated as the difference between the money received by millers and the value of the rice delivered to wholesalers. Table 5.1 Financial costs of wholesalers' "adelan- tos" for prototype rice mills during 1988 operations for a three months loan. X Cost of Cost per cut Size of nills "Adelantos" of “Adelantos” Large 11.32 X S/ 43 82 Medun 12.99 X S/ SS 65 Small 13.33 X S/ 44 87 Source: Annual operating budgets for prototype mills (Appendix). The commercial bank interest for a 3-month loan was equi- valent to 11 percent during 1988. It is apparent that the financial costs of using the "adelantos" are higher than those carried on by bank loans and that they are more expensive for small mills than for medium and large ones. Dividing the interest expenses by the number of cwt of milled rice pro- cessed by each mill, the unit (cwt) financial cost is larger for the medium size mill, and smaller for the large mill. It is apparent also that a substitution of bank loans for the "adelantos" improves the financial outcome for rice mills. Using the annual operating budgets prepared for each prototype mill (Appendix), a number’ of changes in the percentage participation of the "adelantos", changes in the pattern of 102 paddy rice purchases, and changes in the pattern of repayment of these "adelantos" were explored. Considering a positive (greater than zero) participation of the "adelantos", it was observed that the relationship between the "adelantos" and the profits was inverse. With respect to the pattern of monthly (or weekly) purchases and repayments, it was difficult to establish a definitive conclusion. The actual results of the ways millers purchase paddy rice and deliver milled rice (repayments of the "adelantos") would depend on a number of factors, most of which are subject to miller's management capacity to perceive the trend of paddy and milled rice prices, the timing of paddy rice purchases, the timing of the repayments (deliveries of milled rice) to wholesalers, the pace at which rice is milled, and to select the appropriate turnover ratio. 2. Financing wholesalers' working capital Interviews with national wholesalers revealed their excellent accessibility to financial resources from private banks, both through normal commercial loans and contracted overdrafts. This situation has been the result of a number 0 factors: the amount of fixed assets national wholesalers have, the high daily average balance of their deposit ac- counts, the high turnover ratio of this balance, and their economic solvency, demonstrated during a relatively long 103 period of time. Provincial and urban wholesalers have less access to large commercial loans, because of their lower average balances in their bank accounts and lower value of the collateral they can offer when banks require additional guarantees for their loans. B. Profitability of financing rice storage 1. Market participants in rice storage The main participants in the storage of rice are the rice millers, the wholesalers, and the parastatals, including ENAC, ALMACOPIO, and ENPROVIT. Millers store both their own paddy rice and that actually owned by national wholesalers. National, provincial and urban wholesalers store milled rice in their own warehouses: national wholesalers store rice during periods ranging between 2 and 4 months: provincial wholesalers, during periods not longer than 2 months : and urban wholesalers, during very short periods, no more than two weeks. ENAC stores both paddy and milled rice. It has been difficult to get reliable information about the stocks of rice held by millers and wholesalers. The PNA records the rice inventories held by all participants in the rice market, especially millers, inspecting directly the mills and sampling the private warehouses in the Sierra and the 104 Costa, twice a month. PNA personnel also collect information from ENAC and private grain elevators, twice a month. However, the most reliable data are that provided by ENAC. The rest of the data are believed to be biased downwards because millers have to pay a tax of 80 cents for each cwt milled, and because wholesalers are afraid of being accused of "speculators" if they are found holding large inventories of rice. Table 5.2 Stocks of rice at the end of June and October, and the Winter and Summer pro- duction for the 1982-88 period (Metric Tons). YEARS PRIVATE ENAC PRwUC' PRIVATE ENAC PRwUC- SECTCR STOCKS TION SECTOR STOCKS TION STOCKS STOCKS 1982 43702 32006 140570 39850 38923 69609 1983 7541 4285 53978 6633 3752 95121 1984 22756 0 138807 24180 0 100683 1985 51409 7596 114557 50575 13178 91987 1986 75307 33159 208264 43575 69784 106971 Source: National Rice Program: Annual Statistics. For the Winter cycle, the stocks are reported at the end of June, when stocks are building and grow to their highest level. For the Summer cycle, the rice stocks are reported at the end of October, the peak.harvest month of this cycle, when stocks are built to supply rice during the Winter months of 105 January to mid-April. The data on production corresponds to the whole Winter and Summer harvests. Observing the Winter harvest of 1988, we can see that production reached 150,010 MT of milled rice; the stocks reported to have been built by the end of June reached only 62,890 MT. This suggests that 87,120 MT were consumed during the harvest months (May and June), which could not have happened. Comparing in this way the figures on production and stocks, it is clear that reports on stocks are biased downwards. During periods of relative oversupply, as the 1986—87, ENAC's stocks were higher than those of the private sector, while during periods of normal and relative short supplies, the private sector's stocks were higher than those of ENAC. The reason for this behavior is that prices are lower during periods of abundance than during periods of scarcity and, therefore, the monthly price differences are insufficient to cover storage costs. Analyzing the data on rice stocks on a monthly basis, it is found that millers stored paddy rice during the Winter months of January-April in relatively small amounts. The explanation concerns the adequacy of the millers' storage facilities (warehouses) to keep rice in good condition during these months of high humidity and related infestations of insects and fungus. Paddy rice during these months was mostly stored in the silos of large millers, grain elevators, and ENAC. Relatively more rice than the market needs to satisfy 106 current consumption is milled during the months of November and December and it is transferred to the Sierra warehouses, where storage is assumed to be less costly than in the Costa. 107 210 200 — 190 —I 180 — 170 J ,C‘ V 150 - § 140 - B 130 - 87:? 120 99 110 O. O "‘ 93 100 — :1- 90 ~ éu 80 — 1— 70 “ "’ 60 .- LU 5 so — '3 40 - 30 -I 20 'I a 10 0 7, ,.A j I I I 1982 1983 1985 1988 wINTER CYCLES IZZ] pawns seer. IS] ENAC Figure 5.1 End of June rice stocks and Winter produc- tion, in MT, all converted to milled rice for the period 1982-88. Source: Table 5.2 108 a gig E r E g girl/IA Egg 2 gig/gig L z .7/ /_////L_ 000000000000 11111111111 nmpc nmaocku nezu zo_eusooan oz< myqum amepuo a 8 9 1 7 8 9 1 6 8 9 1 S B 9 1 4 B 0.. 1 3 B O.— 1 2 8 g 1 DRODlXZI ION an SMR CYCLES [:3] we PRIVATE SECT. End of October rice stocks and Summer produc- tion, in MT for the period 1982-88. Figure 5 . 2 Table 5.2 Source: 109 2. Seasonal movement of rice prices Using a monthly series of wholesale prices in Quito for milled rice (sucres per cwt), for the period 1971-88, the seasonal index was calculated (Goetz and Weber, 1986). The lowest index value of 95.7 corresponds to June, and the highest index of 104.4 corresponds to April. Table 5.3 Seasonal wholesale price index per cwt of milled rice in Quito estimated for the period 1971-1988. MONTHS GRAND STANDARD GSI+/- GSI+/- SEASONAL ERROR OF ONE TWO INDEX THE MEAN STD ERR STD ERR JAN 100.59 1.23 99.36 98.13 FEB 103.79 1.88 101.91 100.03 MAR 102.51 1.86 100.65 98.79 APR 104.44 2.41 102.03 99.62 MAY 101.07 1.52 99.56 98.04 JUN 95.71 1.71 97.42 99.13 JUL 97.42 1.09 98.51 99.60 AUG 98.73 5.60 104.34 109.94 SEP 99.09 2.01 101.10 103.11 OCT 97.67 1.69 99.36 101.05 NOV 99.02 1.17 100.19 101.36 DEC 99.96 1.02 100.98 102.01 Source: Marketing Division of MAG. The price range does not seem wide enough to conclude that there is a strong seasonal pattern in rice prices. The standard errors of the mean seasonal indexes are relatively small, except for April, September and especially August. 110 104 103 UJ O 2 7 101 LU 9 E 100 LU .J I}? 99 LU .1 8 i 98 97 96 95 I I I I j fl I I I I ENE FEB MAR ABR MAY .JUN JUL AGO SEP OCT NOV DIC WNTHS D SEAS. INDEX Figure 5.3 Seasonal wholesale price index for cwt of milled rice in Quito for the period 1971-88. Source: Table 5.3 111 The mean standard errors variation can be explained by the Winter harvest. It starts around April, depending on whether the rainfall period is advanced, in which case harvest starts in early April, or it is delayed, in which case harvest starts at the end of April. These changes necessarily affect the seasonality of rice prices. A similar phenomenon can explain the large production fluctuation during August, a month that often marks the beginning of the Summer harvest. The strength of the seasonality indexes was measured by adding and subtracting 1 and 2 standard errors to the grand seasonality indexes. It was found that the June index as still below the annual average when 2 standard errors were added and that the April index fell below the average when 2 standard errors were deducted. October's index, the second trough of the series, rose above the average when 2 standard errors were added. The seasonality index for this month does not seem to be remarkably strong. Using the 1985-88 series of wholesale prices, the seasonality pattern did not change. The April index continued to be the highest and those of June and October continued to be the lowest. However, the value of the indexes did change and April's index became 107.6: June's index changed to 96.2: and October's index changed to 94.3. All these indexes maintained their position in refer- ence to the annual average when 2 standards errors were added (subtracted), and all showed smaller mean standard errors. Again, August's index showed an even larger mean standard 112 error, equal to 10.4, indicating the high variability of production (and prices) during this month, which depends largely on weather conditions and the emerging practice of cropping rice 3 times in 14 months. One of the harvest periods of this three-crops sequence takes place every other year in August. To understand better the variability of wholesale prices, percentage changes in real prices were analyzed, taking the differences between the prices in June and October, and between the prices in April and October of every year, during the period 1971-88. Apart from seasonal variations, real wholesaLe prices show severe fluctuations, due largely to changes in the level of production caused by natural disasters, and by factors related to international trade, credit and pricing policies. The sharp declines in rice production (and in other products as well) during 1977-78 and 1982-83 periods were caused by a severe drought and a disastrous flood, respectively. These falls in production were not supplemented by timely imports of rice. In fact, imported rice arrived at the end of 1978 and beginning of 1979, when domestic production had already recovered thanks to a reversal of the drought, increasing domestic supply (imported stocks plus production) well above domestic demand and driving down rice prices until 1980. Imports to offset the poor Summer harvest of 1982 arrived at the same time as the 1983 Winter harvest was being collected, 113 again overcrowding domestic supply and depressing rice prices. The low prices received by farmers during 1983 led them to plant crops such as maize and soybeans instead of rice, causing low production in 1984, which drove prices up. These phenomenon and the untimely counteracting measures generated a dramatic fluctuations in prices during these periods. 114 SD 40 A 304 20— 10-4 96 CHANGE OF PRICES: APRIL-OCTOBER \\/\J I/ AM V '10‘I '20 I r I I I I I I I I r I I I I T 71 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 YEARS 13 PERCENTAGE CHMGE Figure 5.4 Percentage changes in real wholesale prices between April and October in Quito for the period 1971-88. Source: Marketing Division of MAG. 115 40 35 A 30 a NJ 2 D "3 25 - O. LU "’ 20 d a UJ 9 15 .. E 5 10 — LU g s U & 0 II I -5 _ W ‘10 I I I I I I I I I I I I I I I I I 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 YEARS D RERCENTAGE CHANGE Figure 5.5 Percentage changes in real wholesale prices between September and June in Quito for the period 1971-88. Source: Marketing Division of MAG. 116 During 1984-85 rice was being unofficially imported from Peru at prices well below those prevailing in Ecuador (El Comercio, August 25, 1985), driving domestic prices down. During the last quarter of 1988 (and early 1989) rice was being unofficially exported to Colombia, where rice prices were about 50 percent above Ecuadorian prices (El Comercio, October 17, 1988). This unofficial trade moved domestic rice prices up. Another factor affecting rice price fluctuations was the administratively allocated credit. From 386 million sucres of loans to finance rice production in 1973, the government decided to increase to 1,033 million sucres for rice produc- tion in 1974, increasing this amount to 1,284 million sucres in 1975 (BNF's Annual Reports). These increases of loans for rice production brought about above normal harvests, causing prices to move downward from 1974 to 1976. During 1985 and 1986, the government adopted a policy of incentives to expand rice production, fixing support prices well above world prices (Stewart and Cuesta, 1988). This policy was effective in increasing rice production. Rice production during 1985 became the largest obtained in Ecuador. Large stocks were accumulated during 1985 and 1986; finally ENAC had to export rice assuming substantial losses because the export price was below the purchase price. The pattern of the June-September rice price changes shows a tendency towards greater instability, that is, towards 117 larger fluctuations. This is more evident for the period after 1978. One of the reasons for this increasing instabi- lity is the ever larger influence of the border prices (Colombia and Peru). Another reasons for the greater in- stability observed during the June-September period are the low participation of ENAC in setting reference prices and the higher expectations about price behavior during the Winter months. 3. Profitability of financing rice storage This analysis uses the wholesale prices in Quito for the 1981-88 period. Using the author's research data, a similar analysis is presented for the millers' storage function during 1988. To use the method of estimation of storage margins presented in Chapter Two, the following assumptions were formulated: i) the cost of storage SJ. was assumed to be equal to the opportunity costs of capital, measured as the market interest rate. A sensitivity exploration was conducted using the savings interest rate. It is recognized that interest is not the only storage cost: however, data about operations costs and depreciation charges were difficult to obtain: ii) it is assumed that wholesalers would buy in June and October, when rice prices are seasonally low, and will store during the following months, until the next harvest. Tables 5,4 and 5,5 118 show these estimations, using the market interest rates to compute storage costs (monthly interest rate, times purchase priceSIper cwt.of:milled rice, times the number of months rice is stored), and annualizing the monthly returns to storage. An average annualized return for each month of storage was calculated, as well as the mean standard error, the mean minus one standard error (to include about 83 percent of the observations, within this lower limit and.the right end.of the normal curve, assuming a normal distribution of the rates of return). 119 Table 5.4 Annual rates of return for rice purchased in June and stored during part or the entire July-October period using market interest rates to estimate storage costs. BUY 001 sAv. MKT 510 : ANNUAL RATES 0r RETURN IN PRICE 1 i 0081 : STORING RICE UNIIL (1) <2) (3) : Nov DEC ENE FEB MAR APR 001 81 667 0.01 0.01 8 :-0.08 -0.11 -0.13 -0.03 -0.02 -0.03 001 82 717 0.01 0.01 9 : 0.90 1.33 2.03 0.34 0.24 0.17 001 83 1950 0.01 0.02 31 :-0.61 -0.61 -0.59 -o.48 -0.31 -0.23 001 84 2000 0.02 0.02 38 :-0.47 0.19 0.40 0.24 0.42 0.40 001 85 2268 0.02 0.02 46 : 0.96 0.84 0.54 0.24 0.25 0.17 001 86 2225 0.02 0.02 53 :-0.47 -0.49 0.10 0.32 0.27 0.18 001 87 2785 0.02 0.03 73 :-0.08 -0.00 0.45 0.67 0.56 0.81 001 88 5705 0.02 0.03 172 : AVRG. ANNUAL RET. : 0.02 0.16 0.35 0.16 0.18 0.21 STANDARD ERROR : 0.25 0.26 0.27 0.12 0.10 0.12 NEAN - s: :-0.23 -0.10 0.08 0.04 0.08 0.09 PROBABILITY or LOSS : 0.63 0.50 0.29 0.29 0.29 0.29 (1) Savings interest rate (2) Market interest rate (3) Monthly storage costs Sources: Prices: Marketing Division of MAG. Interest: Central Bank: Annual Reports. 120 Table 5.5 Annual rates of return for rice purcha- sed in October and stored during part or the entire November-April period using market interest rates to estimate storage costs. BUY JUNE SAV. NKT STO : ANNUAL RATES OF RETURN IN PRICE 1 1 COST : STORING RICE UNTIL : JUL AUG SEP OCT : 0.97 0.97 0.80 0.59 : 0.33 0.05 -0.01 0.03 JUN 83 1311 0.01 0.02 21 : 1.63 1.30 1.98 1.27 : 0.12 0.12 0.55 0.73 . : 0.21 -0.13 0.22 0.37 JUN 86 2288 0.02 0.02 54 :-0.75 -0.43 -0.40 -0.37 JUN 87 2200 0.02 0.03 57 : 2.14 1.16 0.68 0.48 JUN 88 4240 0.02 0.03 128 : 0.29 0.43 0.29 0.68 AVRG. ANNUAL RET. : 0.46 0.31 0.43 0.41 s1AN0AR0 ERROR : 0.33 0.23 0.25 0.17 NEAN - se : 0.14 0.09 0.18 0.24 PROBABILITY or LOSS : 0.13 0.25 0.25 0.25 (1) Savings interest rate (2) Market interest rate (3) Monthly storage costs Sources: Prices: Marketing Division of MAG. Interest: Central Bank: Annual Reports. 121 The average return to storage for the Winter period fluctuates dramatically moving from a low of 2 percent for the first month of storage, to a high 35 percent for the third month, and diminishing for the next months. This suggests that wholesalers would be better off purchasing rice in October and storing it until January, instead of selling it either before or after the third month. This situation can be explained by the fixity of official prices, which does not allow for storage cost recuperation after a certain period of time. Once the market price reaches the ceiling official price, wholesalers are not stimulated to hold stocks of rice. Yet, if they do hold inventories after this point, their returns diminish or even vanish. The returns to storage obtained during the Summer months are in general higher than those of the Winter period. It may be because ENAC's stocks liquidation is moderate or even nil during this cycle, and official prices are not rigorously enforced, allowing market prices to vary in a way that permit recuperation of storage costs and an attractive profit. Despite this characteristic of the annual returns, and assuming that.operations costs plus depreciation charges constitute a low amount, the storage at the wholesale level seems to be have been profitable. This statement may not hold in reference to periods of high inflation, such as the Winter and Summer cycles of 1988 and Winter of 1989, when the inflation rate exceeded 55 percent and reached 90 percent at the end of 1988, while the market 122 interest rate was held below 52 percent. One mean standard error was reduced from the average rates of return. These rates were still positive, except for those for the first two months of storage during the Winter period. 123 ANNUAL PERCENTAGE RETURN fl. n—___F——-FFfl—flEr-_~—~______4}____——————-—&3 U I I I DIC ENE FEB MAR ABR 1 §~—s m MONTHS OF STORAGE D 1983 + 1985 O 1988 Figure 5.6 Annualized rates of return to storage during one or more months of the Winter 1983, 1985 and 1988 periods. Source: Table 5.4 Note: Complete data for 1988 were not available. 124 1 444444444 1 liiLllJl I I l ANNUAL PERCENTAGE RETURN 11A] ODODDDDOD NJOJMUAUIU'JUCDLD—h—IMUJIUIUIQDLDM I I I JUL AUG SEP OCT MONTHS OF STORAGE D 1983 + 1985 o 1988 Figure 5.7 Annualized rates of return to storage during one or more months of the Summer 1983, 1985 and 1988 periods. Source: Table 5.5 125 Observing the behavior of the rates of return to storage during years of shortages in supply and years of overproduc- tion, it is apparent that returns were consistently higher during periods of shortages than during periods of oversupply. During 1985, the monthly differences of wholesale prices were not enough to stimulate private storage of rice, and this function had to be largely fulfilled by ENAC. The opposite situation occurred during the periods of supply shortages (1983 and 1988, for example). C. ENAC's participation in the storage of rice 1. Allocation of financial resources The foundation. of IENAC's participation in 'the rice marketing subsector has been its commitment to sustain government support prices and reduce price instability. ENAC has attempted to sustain minimum producer prices by purchasing rice through the assembly centers it has opened during the Winter harvest periods. ENAC has made efforts to reduce price variability'by keeping buffer stocks of paddy and milled rice, and by importing and exporting rice‘when ENACIofficials consi- dered that domestic supplies were not adequate to match domestic demand. Although official prices have been above domestic market.prices during 14 out of 17 years of the period 1972-88, ENAC has not purchased significant amounts of rice, 126 basically because of the long procedures to receive the grain, to issue the payment (either a check or a warehouse receipt that had to be cleared through the BNF), and because of the rigorous grading of the grain. In other words, because of the higher transaction costs that selling to ENAC demands. Table 5.6 Average official and market prices for 200 lbs sacks of paddy rice, ENAC pur- chases (milled rice equivalent), and BNF's credit to ENAC to finance rice trade and storage, for the period 1972- 88. OFFICIAL MARKET X ENAC BNF YEARS PRICES PRICES OFI/NKT PURCHS CREDIT S/ SACK S/ SACK PRICES X ENAC S/ 000 1972 152 105 145 1973 187 140 134 1974 272 175 155 3.5 459673 1975 345 225 153 9.5 1145953 1976 345 235 147 1.9 261721 1977 335 245 137 7.1 41500 1978 335 280 120 5.6 369000 1979 390 325 120 2.8 416847 1980 435 355 123 15.3 900000 1981 571 470 121 9.6 1150000 1982 571 490 117 11.5 900000 1983 870 900 97 0.0 1450000 1984 1460 1375 106 0.0 0 1985 2175 2000 109 5.1 4021250 1986 2175 2200 99 22.1 8780000 1987 2395 2400 100 32.4 0 1988 6000 6200 97 0 Source: Economic Council's Decrees Marketing Division of MAG BNF's Annual Reports Comparing the percentage ratio between the official and market prices with both the percentage purchases of ENAC and the trade credit granted by the BNF to ENAC, it is not 127 possible to find a definite pattern. This lack of consistency in ENAC's participation might have been caused by the use of quantities and not of prices as triggers for its interven— tion. ‘The decisions to purchase rice, to import and.to export have been taken based on information about quantities and not on price signals, which are supposed to be more accurate to indicate shortages and oversupplies (Cornelius Hugo, D. Tschirley, and H. Ramos, 1989). The amounts of BNF's credit for ENAC have been, on average, 25 times higher than those that BNF allocates to private sector to finance rice trade and storage. However, millers and wholesalers have been able to move larger amounts of rice by revolving the funds at higher velocity than ENAC. ENAC buys rice and basically stores it, foregoing the benefits of turning the capital over a number of times. 2. Profitability of financing official stocks of rice ENAC intervention in the rice marketing system has been characterized by at least two critical factors: 1) official prices have been established at higher levels than market prices, which has obligated ENAC to request considerable amounts of funds to attempt to sustain them: ii) official prices have not included either regional or time utilities. Official prices have been fixed for all the regions in the country and often for an entire year. 128 These factors have caused ENAC to suffer considerable losses every year it has intervened, making it impossible for ENAC to pay its credits to BNF. Since 1980, ENAC has had to request either that the government pays the debt to BNF or that Central Bank renews the debt for a longer term (more than 180 days). By 1986 the overdue loans of ENAC were about 5 billion sucres, plus interest charges of more than 1 billion (BNF' Annual Report). In 1987 the government requested the Central Bank to acquire government bonds, totaling 5 billion sucres, and to use the proceeds to cancel ENAC's debt with BNF, which, in turn, has not.been able to complete the payment to Central Bank (Supreme Decree No.2971 of June 8, 1987). This operation was the last.of 5 negotiated in previous years, totaling more than 9 billion sucres. This amount could be equivalent to the losses accumulated by ENAC during the 1980- 88 period, of which a significant part is the value of the "subsidies”, accounted for the low margins between buying and selling prices, which.have:not covered.operational and.storage costs. This situation not only increased the fiscal deficit but also deteriorated the BNF's capital, since the credit allo- cated by the Monetary Board to ENAC has been automatically debited by the Central Bank from the BNF accounts at due date, without considering whether ENAC has paid or not to BNF. The secondary effects of these events are suffered by BNF bor- rowers, who are left with less loanable funds. 129 Summary The use of formal financial resources to finance trading of rice has usually been profitable for millers and whole— salers. The use of the wholesalers "adelanto" by millers has been a recourse of last resort. If ndllers had access to formal commercial credit, they would have not used these funds. The storage of rice by wholesalers has usually yielded positive rates of return. These returns have been inversely related to the level of rice supply, being relatively high in periods of shortages and relatively low in periods of over- supply. The allocation of official loans to finance rice trade through ENAC has been substantially higher than the allocation of trade loans to private millers and wholesalers. However, because official prices do not include premiums to cover inter-regional transfer and storage costs, ENAC has been loosing money, having requested government support to cancel its debt with BNF. The amount of the subsidies have con- tributed to a deterioration of BNF's operating capital, hampering other borrowers as well. CHAPTER SIX PROFITABILITY OF LONG-TERM INVESTMENTS IN PRIVATE STORAGE FACILITIES AND DRYING AND MILLING MACHINERY Introduction This chapter assesses the profitability of financing long-term investments in storage facilities drying and milling machinery, and equipment. This chapter identifies investment opportunities in these items, and their costs and benefits. It measures the profitability of piecemeal financing of machinery' and. equipment. replacements and. compares it 'to financing the construction of completely new plants. The last section of this chapter presents a cost-effectiveness and economies of size analysis of investments in new plants. A. Identification of long term investments opportunities 1. Piecemeal machinery replacement and additions of equip- ment The records of permits submitted to the PNA either to modify the existing plants or to construct new ones show that more and more millers are substituting a mechanical drying system for the solar drying method. Most of the equipment replacements undertaken during the last 3 to 5 years have 130 131 consisted of small size drying equipment. A few more modern drying systems (the batch or continuous flow systems) have been assembled in no more than six mills during the last two years. The installation of new drying equipment in the majority of mills have taken place gradually. The installation often begins by first buying a portable dryer (motor and burner) and building the needed drying tunnels. These drying tunnels consist of two floor closed rooms, divided by a metal screen. The upper level is the space to pile the rice. The lower level is an empty space whereby the hot air is blown inside the room. The mechanical for the solar drying substitution process rarely has been a one time decision. It is usually completed during at least two harvest cycles. The reason for both ‘the piecemeal substitution and the selection of a relatively small size equipment derives from the lack of long- term loanable funds in the BNF and private banks to finance this type of investments. Millers, however, have invested their own resources in small size, simple drying equipment, in a piecemeal manner, both to increase operations and to slowly substitute more modern systems for the traditional processes. Replacement of major parts of milling equipment have had a lower priority. Millers are constantly replacing the parts of their mills that wear out the most, such as sieves, sifters, screens, and more frequently, rubber rolls. 132 The supply of these parts has not been limited, but their prices have generally increased because of changes in the import taxes and the exchange rate. There is still a need for major improvements in storage facilities. Technicians of the PNA estimate that of the 195,000 MT storage space no more than 30,000 MT is silo capacity. The conditions of the remaining warehouse space often provide inadequate protection from ground humidity, rodents, birds and insects. Millers recognize these inadequa- cies and are willing to undertake investments in remodelling and improving their warehouses. Some of them have expressed a desire to acquire silos and small machinery to move rice within the plants, such as payloaders and carts. For them, the major limitation of these investments seems to be the lack of long-term funds, both at the BNF and at commercial banks. 2. Investments in new rice milling plants Only a few new rice milling plants have been built during the last three years (no more than 7 mills, according to PNA information). All of these plants have been classified as first category mills because their milling capacity exceeds 40 cwt/h. All of them are vertically integrated with large rice farms. Entrepreneurs have financed the construction of these plants with their own resources and commercial bank loans. These loans have been negotiated in such a way that 133 they can be rolled over every six months after payment of interest and usually 20 to 25 percent of the outstanding principal amount. This practice of rolling over short term loans has allowed lending institutions to adjust the interest rate to approximately the prevailing inflation rate. The size trend in the new plants constructed has been towards larger and more modern facilities, including automated receiving equipment (grading, weighing and unloading), batch or continuous drying equipment, and automated sacking equip- ment. There are some concerns among technicians of the PNA and the BNF about the size, location and number of the new plants. In fact, the PNA's Engineering Department is re- questing millers who plan to install a new plant with a milling capacity of 50 cwt/h or more, to include in the construction an automated receiving facility. Such a condi- tion is currently being discussed with BNF credit officials to coordinate some regulating procedures aimed to make the rice milling industry more adequate to process the current and potentially larger rice production. 134 B. Profitability of financing long-term investments in rice processing and storage 1. Current capacity utilization of prototype rice mills For'a number of reasons, rice mills are not.busy all year around. One reason is the time needed to repair and maintain the equipment and machinery. Another reason is the pattern of rice processing which parallels the harvest. Although milling can be undertaken at a pace different from that of drying, millers usually mill rice at almost the same pace as they dry it. This allows then to turnover the operating capital the greatest number of times possible. Once the peak harvest period ends, millers either reduce the rate of milling or shutdown the milling phase. They resume milling when they have accumulated a volume of rice that can justify turning on the mill. However, drying is a permanent activity, as long as millers purchase paddy rice. To estimate the maximum capacity of rice milling, the hourly rate (cwt/h) has been multiplied by 8 hours of work per day. Annually, this means that 2,304 hours per year (8 hours of work per day, times 24 days a month, times 12 months) are available. These hours per year could, of course, be used in 96, 24 hours working days requiring three shifts. Based on these data, Table 6.1 shows the actual capacity utilization of mills. I 135 Table 6.1 Maximum capacity and actual capacity utilization of mills measured in.hours of work and cwt units of milled rice per year for the year 1988. HILL MILLING : MAXIM CAPACITY : ACTUAL CAPACITY : X SIZE CAPACITY : UTILIZATION :UTlLl- cwt/h : hrs/yr cut/yr : hrs/yr cwt/h :ZATION LARGE 120 : 2304 276480 : 1676 201104 : 73 NEDILII 25 : 2304 57600 : 1627 40664 : 71 SIMLL 12 : 2304 27648 : 2141 25691 : 93 Source: Field Survey: Annual . Operating Budgets for prototype rice mills. Small, second class mills operate at about 90 percent of their maximum capacity, while the large and medium size mills operate at about 70 percent. Using maximum capacities estimated by the equipment suppliers as a reference, the actual capacity utilization is even lower since the number of hours per year is assumed to be at least 50 percent over the figures used in Table 6.1. One other reason for a reduced capacity utilization is the lack of financial resources. Mbst millers interviewed complained that funds were scarce during harvest periods. They all agreed that the practice of turning over the rela- tively small working capital as many times as possible helps them to increase their operating volume to levels where milling could yield positive returns. But they recognize that this practice puts a great deal of stress on their milling equipment and labor. Moreover, it requires that they 136 sacrifice the potential benefits of storing rice until its market price increases. Table 6.2 compares the average total cost for'these mills at their current operating level, at their break-even point, and at their maximum capacity, as estimated previously. Table 6.2 Average total cost per cwt of milled rice estimated at current capacity utiliza— tion, at.break-even level, and.at.maximum capacity for large, medium and small prototype mills. MILL BRK-EVEN:AVG.COST :AVG.COST :AVG.COST SIZE cwt :AT ACTUAL: AT : AT UNITS :UTILIZAT.:BRK-EVEN : MAX.CAP LARGE 150980 : 1400 : 1698 : 1177 MEDIUM 59614 : 1478 : 1188 : 1210 SMALL 39516 : 1491 : 1165 : 1428 Source: Author's Field Survey These estimations show that unit cost per cwt of milled rice is lower for the larger mill than for the medium and small mill, both at their current level of operations and at their maximum capacity. The unit costs estimated at the break-even level differ from the expected pattern because the break-even point for the large mill is found at a lower level than its current and maximum capacities, while for the other two mills, it is found at levels above their current and maximum capacities. These findings give an indication that there may be 137 economies of scale in the rice milling industry in Ecuador. In fact Figure 6.1 illustrates the existence of economies of scale. The curves in Figure 6.1 are short run average cost curves. The range of mill sizes is too limited to attempt to estimate a long run average cost curve. It is apparent that the short run average cost curves for the small and medium mills turn up after certain volume (cwt units), while the curve for the large mill continues descending after a volume of about 350,000 cwt of milled rice; it turns up only after about 600,000 cwt units per year. Because these results are based on a mathematical model, it is not safe to assume that in actuality these curves show the mills cost behavior. The average costs may turn up faster and perhaps at lower oper- ating volumes. Cost elasticities measure the ratio between marginal and average costs. These have been used to inves- tigate economies of scale. Wailes and Holder (1987) , however, suggest the inverse of the cost elasticity is.aibetter'measure of economies of scale, yet not a strict and definitive measure. To estimate both the cost elasticity and its inverse, the long-run average cost curve is needed. But it has not been possible to estimate this curve given the few point estimates that three mill sizes can offer. 138 AVERAGE COST ( Thousands) + 1.5- (3)3 R8 + a +0 1301‘:] + ++W IIIIIIIIIIIIIIIIIIIIIIIIrIIIIT—IIIIIIIIITj—T cwt UNITS OF MILLED RICE C] SMALL MILL + MEDILM MILL 0 LARGE MILL Figure 6.1 Average cost curves for large, medium, and small prototype rice mills in Ecuador. Source: Field Survey7 7 The average cost curves were calculated using a double log quadratic function. Large mill: L(AC) = 15.9262 - 2.7983 L(CWT) + 0.2195 [L(cw'r)]2 (0.2251) (0.0224) R2 = 0.9888 Medium mill: L(AC) = 15.9940 - 4.1842 L(CWT) + 0.4885 [L(cw'r)]2 (1.2193) (0.1522) R2 = 0.5629 Small mill: L(AC) = 17.7295 - 5.9883 L(CWT) + 0.8397 [L(cw'r)]2 2 (1.9877) (0.2794) R = 0.3352 AC=average cost: CAP= cwt/h: CWT=cwt units of operation. NOTE: The X-axis scale goes from 10 thousands cwt units to 350 thousands. 139 2. Profitability of financing piecemeal replacements and additions of equipment and machinery Annual operating budgets have been prepared for the three types of rice mills. One type of budget represents mills without additional investments, i.e. as they are currently operating. Another type of mill includes replacements and additional investments. The difference between the "with" and the "without" budgets results in. the incremental annual operating budget, which captures the additional costs and benefits of the new investments. For the large mill, these new investments consist of replacements of major components of the drying system, valued at 32 million sucres, and the replacements of major components of the milling machinery, valued at 71 million sucres (au- thor's field survey), for a total of 103 million sucres. The net present value of the incremental investment, at 15 years (the estimated useful life of the new equipment), and at a real discount rate of 2 percent, is 411 million sucres. It has been considered that 60 percent of the value of these investments is financed by a private bank loan, negotiated at 3 years term.and 52 percent interest rate (as it actually hap- pened in one case). The first 3 years' net receipts were negative, although the net cash flows were positive because of the depreciation charges (see the Appendix). 140 To get a positive NPV, the mill would have had to increase its operation by at least 50 percent. The average total costs per cwt would have been reduced from 1,400 to 1,298 sucres. Looking at the incremental situation, the unit cost of the additional cwt units of milled rice is only 1,048. The NPV without these new investments is 232 million sucres, for an initial capital outlay of 310.5 millions, while the incremental NPV is 411 for an initial investment of 103 millions. The contribution of the additional investments is substantial, but it depends on the increase in the operating level: the units must mill approximately 300.000 cwt units (13,636 MT) of milled rice per year, 100,000 over the current output. The additional investments planned for the medium size mill are an additional mechanical drying system, valued at 5.13 million sucres, and warehouse improvements, estimated to cost 2.5 million sucres, totaling 7.63 million sucres. These investments would be 60 percent financed by the BNF assuming that the miller is also a rice producer and can get access to BNF's funds, for 7 years at an interest rate of 32 percent. The NPV of the incremental investment is 22.8 million sucres, assuming 15 years as the useful life of the equipment and a real discount rate of 2 percent. To increase NPV, an increase in the operating level of at least 20 percent is required. However, an increase of 30 percent is expected, that is, an annual volume of approximately 52,860 cwt units of milled rice 141 (2,400 MT), which is still less than the original estimated maximum annual capacity. The total average cost with these new equipment and improvements is 1,275 sucres per cwt of milled rice or about 15 percent less than the original average cost. The NPV for the initial investment of 75.6 millions estimated to last 15 years is a negative 0.5 million sucres. The discount rate used is the same as for the earlier cases. It is apparent.that.the incremental investments would increase the value of the firm, increasing even more the NPV of the entire project. The additional investments.planned for the small mill are a new drying system, valued at 6.48 million sucres, and a major improvement of the warehouse. The cost of these investments is estimated at 4.5 million sucres“ The total new investment would be 10.98 million sucres. The NPV of these incremental investments is 8.8 million sucres, assuming also that 60 percent is financed by BNF to be repaid over 7 years term at an interest rate of 32 percent. In this case it is projected to increase capacity utilization by at least 30 percent. If this goal is achieved, the total average cost.per cwt would decrease to 1,437 sucres from its original level of 1,491 sucres. Despite these outcomes, the NPV including the incremental investments will still be a negative 2.8 million sucres. To substantially improve the outcome of this mill, its operating level would have to double, making the mill work at almost full capacity. 142 The net receipts of the medium and small mills are negative, although the net cash incomes are positive because they include depreciation charges (Appendix A). This finding suggests that medium and especially small mills are surviving by neglecting to make depreciation allocations. The incremental investments may help medium and small mills to improve their financial position and profitability. But there are at least 3 problems with such undertakings. One is the impossibility for mills intending to make these long- term investments to borrow from BNF. If they obtain 60 percent financing with private bank loans, at 52 percent interest and 3 years term, both the NPV and the net cash income for the first 3 years are negative. Two, if these investments are made, mills will still need additional working capital to increase their output to be able to pay off this capital outlay. Given an expected increase in the demand for funds, it is reasonable to expect that.many mills will be left without adequate financing. Finally, cost-effectiveness and economies of scale are critical issues to consider. Large mills show a relatively large unused capacity that could be utilized before their average cost curves begin to turn up, and millers recognize that the more volume they process the lower is the unit cost. Consequently, large millers will be stimulated to buy ever larger amounts of paddy rice, com- pelling medium and especially small mills to eventually exit the industry. Large millers have easier access to credit, 143 both to finance investments and working capital. This gives them an additional advantage over the other millers. Who benefits from this potential trend toward fewer, larger and.more modern rice mills? Assuming competition among large mills, rice producers are expected to benefit from this trend, since the need for larger supplies of paddy rice will stimulate aggressive procurement tactics by millers. Con- sumers are also expected to benefit from lower prices and better quality rice. Lower prices will be possible due to lower per unit processing cost (cost-effectiveness) , and better quality because larger, more modern machinery has a higher conversion rate with lower percentage of broken kernels. The owners and workers of medium size mills and the majority of owners and workers of the small mills will be adversely affected by this change toward larger mills. Exit costs are expected to be large because of the low liquidity (high specificity) of the assets, and the lack of a well developed market for used machinery. 3 . Profitability of financing the construction of new plants To assess the profitability of building new mills, the same prototype operating budgets prepared to evaluate piece- meal investments are used. Some of the basic assumptions are changed in this case. Now the interest rate is assumed to be 52 percent, which was the commercial loan rate during the 144 first months of 1988 and consistent with the date at which mills assets were valued. The term of the loan is assumed to be 3 years, the average term for commercial loans aimed at financing investments (not consumption). The real discount rate is assumed to be 2 percent. It is also assumed that the useful life of the investment is 15 years. Building a complete new mill with the same characteris- tics as the large prototype mill, including the additional investments, and assuming 50 percent increase in the operating level and 60 percent debt financed, appears to be a profitable investment since the NPV is a positive 450 million sucres. Assuming 100 percent debt financing, the NPV is 135 million sucres: while with 0 percent debt, the NPV is 945 million sucres. The NPV remains positive under a number of assump- tions about levels of debt, terms of the loan, and discount rate, up to about 7 percent. The NPV'becomes negative at real discount rates above 8 percent. The same evaluation was conducted with the medium and small mills. Under a reasonable range of discount rates, interest rates, terms of debt, the NPV is positive. However, as the proportion of debt financing increases, the NPV decreases but remain positive. But net cash income is negative while loan payments are being made. This is the situation when debt financing is over 85 percent of the investment for both the medium and small size mill. The net cash incomes do not become negative even with 100 percent debt 145 financing, when the interest rate of the loan reduces to 32 percent and the term of the loans increases to 7 years (see the Appendix). Throughout all this discussion the need to increase the level of operations has been a prevalent condition for the investments to be profitable. This increase would have to derive from these sources: one, an expansion of domestic production: two, a reduction of operations by some mills, likely the small ones; three, a combination of the previous 2. An expansion of production is already taking place and is expected to accelerate in the near future (Whitaker and Alzamora, 1988a). Yet, this is a long-term process while the investments analyzed in this chapter are short-term decisions, demanding an immediate increase in supplies of paddy rice. In the short-run, increasing the operations of the large mills will have to be achieved by reducing the amount of paddy rice handled by the small mills. This suggests many more small, inefficient, and. obsolete 'mills will exit the industry, leaving room for the installation of more modern, larger, and progressive mills. The credit related consequence of this trend is the need for a larger volume of loanable funds, and, more importantly, the need for an increase in the maximum amount of loans granted to a single firm. 146 4. Comparison of two financing strategies This comparison refers especially to credit considera- tions. It is apparent that new plants require far more funds than partial investments to add or replace equipment to an existing mill. Given the current restrictive financial and monetary situation in Ecuador, it is now difficult to obtain large loans. This situation is exacerbated in the case of investments in the rice milling industry. Therefore, from a financial perspective, it is more likely that more partial investments will be undertaken than the construction of new plants. This preference towards partial investments in rice mill enterprises is reinforced by the inflationary situation of the country. Inflation is currently causing two negative impacts: i) delays in the decision to replace fixed assets. Millers are stimulated to maintain their plants for the longest time possible before considering whether to replace some parts or to build a complete new plant (Nelson, 1976, 930); ii) liquidity problems when firms do not.have enough cash.to cover loan installments. This problem is exacerbated when firms roll over short-term credit at ever higher interest rates. A piecemeal investment can reduce the illiquidity effect (saused by inflation since a positive large cash income from the whole plant can offset the negative cash flow due to the incremental investment. Investing in the construction of a 147 complete new plant will not have the advantage of some older assets covering the cash deficits of the new ones. Another important aspect relates to the goodwill of existing plants. By maintaining the old plants and replacing gradually the obsolete equipment and.machinery, and improving some other facilities, millers are able to maintain their clientele and attract other clients. Summary The analysis presented in this chapter has identified a number of potentially profitable investments in the rice milling industry. This analysis shows that piecemeal replace- ment of drying and milling equipment, as well as investment in storage facilities, are generally profitable investments, under the current economic and financial environment. The construction of new mills are also an attractive investment from the financial point of view. However, this decision now faces serious financing limitations, including limitations of official and private sources of funds. Ecuador's current inflation adversely affects investments in fixed assets in two ways. First, it compels entrepreneurs to delay the replacement of depreciated assets. And second, it causes cash flow deficits when they are undertaken with a Significant proportion of debt. The next chapter will examine this latter effect. CHAPTER SEVEN THE IMPACT OF INFLATION ON THE LIQUIDITY OF LONG TERM INVESTMENTS Introduction It. has been. argued (Robison. and Brake, 1980) that inflation generates cash flow problems for long-term invest- ments in depreciable and non-depreciable assets, when they are debt financed. 'This argument.has been proposed.as an hypothe- sis in this study; The liquidity problem will be defined and, if it exists, identified and illustrated using data from the rice milling enterprise of Ecuador. Assuming the hypothesis is true in the rice milling firms, an attempt will be made to demonstrate that the presence of cash flow deficits prevents long-term investments. A. Definition of the problem 1. Theoretical background Robison and Brake (1980) state that: ”Using present value techniques, we show that even accurately anticipated inflation creates liquidity or cash flow problems for farm firms as capital gains increase in relation to cash returns: 148 149 moreover, the higher the rate of inflation, the more severe the liquidity or cash flow problem of the firm." The definition of the liquidity problem follows this reasoning. Let V be equal the value of an asset. The net cash income that this asset is able to generate is R which grows annually at rate 1. Moreover, let r be the real, inflation-free discount rate. Assuming an infinite time horizon and a constant net cash income, the value of the asset will be equal to: Now, assume that the net cash income grows annually at a rate i (the inflation rate), affecting also the discount rate. The present value of the asset will be: (1+r)(1+i) (1+r)" (1+i)" (1+r)" (1+i)" where V. is the salvage value of the asset. It is easy to see that the terms (1+i)t can be cancelled out. Assuming the series is infinite and that V = V the 3’ Value of the asset is again 150 But, by solving uncritically the model in this way, an important issue is missed. Consider net cash flow in the first period. It equals R(1+i). Replacing R for its equiva- lent value rV, the net cash flow is also equal to (r + ri) V With 100 percent financing, the interest costs equal (r + i + ri) V The difference between the receipts of the first year, which is (r+ri)V, and the interest cost is difference (r + ri) V - (r + i + ri) V from where, difference = - iV It should be apparent that the first year's cash flow is 1101: sufficient to pay the interest cost of the asset. It must 151 also be clear that this difference is the capital gains inflation adds to the asset's value. 2. The theoretical model in the context of rice mills investments Two basic assumptions are made in this study that allow us to discuss the liquidity problem defined in the previous section. First, the useful life of the investments is estimated to be 15 years. And second, the salvage value of the assets is assumed to be zero, to simplify the calcula- tions. This assumption becomes more realistic when it is recognized that there is not a well functioning market for used rice milling machinery in Ecuador. Another assumption concerns the relationship between the nominal interest rate, the real rate of interest, and the inflation rate. The real interest rate has been defined as equal to the nominal rate of interest minus the expected inflation. The assumption in this study is that the inflation rate is constant, and thus the expected inflation is known. With this assumption, the denominator of the present value model can be substituted in this way: (1+rn)=(1+r) (1+1) where r" is the nominal discount rate equals to the product of 152 one plus the real rate times one plus the inflation rate. Then the net present value model can be written as If V were 100 debt financed, the constant annual install- ment of the loan, at t years term and r3 nominal (inflation adjusted) interest rate would be: The last two equations will be applied to analyze the liquidity situation of the prototype mills constructed in this study, assuming they are projects to be undertaking. Several values for the relevant parameters will be used as a sen- sitivity approach to test the hypothesis proposed. B. Application of the conceptual model to the rice milling industry of Ecuador As a first approximation, it is assumed that the esti- mated investments for the 3 type of mills are undertaken under iflris set of conditions: 100 percent loan financed (shown also 153 60 percent debt financed), at a lending interest rate of 52 percent (commercial loan rate), and at 5 years term (t of loan). The project is assumed to have a useful life of 15 years. The rate of inflation g is assumed to be 85 percent, equal to the one observed by the end of 1988. Since the assets are 100 percent debt financed, the discount rate i is the lending rate, that is, 52 percent. When the debt financed is less than 100 percent, the discount rate is the weighted average between the commercial lending rate and the oppor- tunity costs of capital, estimated at 60 percent. Table 7.1 Net cash flows from estimates associated with investments in rice mills, and of annual loan installments assuming current commercial loans conditions. MILLS V R A : R A LARGE 413.50 137.67 245.25 : 137.67 147.15 MEDIUM 83.20 42.90 49.35 : 42.90 29.61 SMALL 58.50 28.27 34.70 : 28.24 20.82 V Asset's value, equal to the loan amount R Net cash income A Loan installment Source: Operating budgets for prototype mills. It is apparent that the estimated net cash income (R) is far from being enough to cover the installment of the loan (A), under the assumption of 100 debt financing. Using the 154 calculated figures for the large mill, the behavior of R and A for the assumed life of the investments is presented in Figure 7.1: 155 7D 50-1 50-1 40—I 3D-< NATURAL LCIS m R AM) A 2D—I 10- I I I I I j j I I I I I I I I 1 2 3 4 S 6 7 8 9 1O 11 12 13 14 15 PROJECT LIFE AND TEFN 0F LOAN (YEARS) D LCIS CF R + L% OF A Figure 7.1 Schedules of loan repayments(A) and net cash returns (R), for invest- ments in large rice mills in Ecua- dor, assuming 100 percent debt financing. Note: The values of R and A are in natural logs. Source: Table 7.1 156 Reducing the proportion of debt financing to 60 percent and keeping the other assumptions unchanged, the net cash return for the large mill is not sufficient to pay the loan installment, while the returns for the medium and small mills are enough to pay the amortization of the loan. If profits before debt service outlays were considered to make these comparisons instead of net cash returns, which include depreciation charges, the deficit problem would be dramatic: all mills show profits before loan payment that are insuffi- cient to cover the amortization.of their loans, even when debt financing is as low as 25 percent; with 100 percent debt financing, mills will have negative net profits for at least four years, perceiving positive net gains only after the period of amortization. The model was tested also assuming that entrepreneurs can access credit from the BNF, at 32 percent interest and an exceptionally 12 year term. Both 100 percent and 60 percent financing were assumed. 157 Table 7.2 Net cash flows estimates associated with investments in rice mills and of annual loans installments under the assumption of subsidized BNF loan. MILLS V R A : R A LARGE 413.50 137.66 137.22 : 137.66 82.33 MEDIUM 83.20 42.90 27.61 : 42.91 16.57 SMALL 58.50 28.23 19.41 : 28.22 11.65 V Asset's value, equal to the loan amount R Net cash income A Loan installment Source: Annual operating budgets for prototype mills. It is apparent that mills are better off when softer conditions are negotiated for their loans; net cash returns from investments are sufficient to pay the loan installments, even assuming 100 percent debt financing. This analysis indicates that inflation generates cash deficit problems in long term investments in rice mills in Ecuador, when loans are negotiated at relatively short terms and prevalent commercial interest rates, assuming that credits finance over 75 percent of investments. The liquidity problem encountered does not necessarily prevent investments to be profitable in the long run. In fact, even under the extreme conditions of financing (100 percent), interest (52 percent), and term (5 years), all mills appear to be profitable within 15 years of operations. However, if mills are not able to externally finance their cash deficits, even when they are 158 temporary, then their profitability is seriously hampered, and even fades away if problem is persistent. The sensitivity analysis performed in this section suggests that softer conditions of loans to finance durable assets are able to modify the outcome of the analysis and to moderate the impacts of inflation. Extending the term of the loans and reducing' the interest. rate can. eliminate the liquidity problem while maintaining the profitability of the investment project. C. Credit conditions to reduce or eliminate the liquidity problem of long-term investments partially financed.with loans. An evident solution to reduce or even eliminate the liquidity problem is to schedule the loan installments to be congruent with the pattern of net returns. Such schedules of amortization fall into the moderated payment loans category. A number of methods to moderate the loan repayment schedule are discussed, such as: 1) the skip payment loan, which consists of including a grace period in the contracted term: during this period the payments of principal are skipped, and often the payments of interest; 2) the buy down loan, by which lower installments are established for the first years of the term, estimated using lower interest rates, while higher payments are demanded during later years, offsetting the 159 initial "subsidies"; 3) the constant principal payment loan, allowing the negotiation of an especial schedule of payments for the interests; 4) the compensated balance loans, by which the borrower agrees to maintain a certain amount of resources (balance) immobile in his/her account, earning or not a determined interest rate. Under this set of methods to moderate the payments, the graduated payment model becomes important. The lender agrees to receive payments that grow as the term of the loan reduces, in parallel with the pattern of net returns. The graduated payment model is a modified version of the general amortization one that allows payments to gradually increase from an initial one, at a rate similar to the growth rate affecting net returns, or any other rate agreed upon by lender and borrower. The graduated payment model can have at least 2 presentations: i) A constantly increasing schedule of payments, developed with the following rationale: Defining the first loan payment as A”, the following payments will be given by: At = A1 (1+9)t where g is the growth rate (not necessarily equal to the inflation rate 1), and t = 1,2, ...n-1, being n the term of the loan" The sum of these payments will be equal to the loan amount, L, plus interest, I, that is: 160 L+I=2At = (At/g) [(1+g)"-1] t=1 The problem is to find A, such that the present value of the payments will be equal to the amount of loan L. This equation is: where rh in this case also stands for the nominal interest rate. It a simple task to solve this equation for the value of A, once the loan amount is known as well as g, rn and n, the term of the loan. ii) An increasing schedule of payments during part of the loan's term, and then a constant schedule of payments: The present value of such an schedule, which must be equal to the amount of the loan, will be: A, (1+g)" A, (l-l-g)" (1+rn)" - (1+rn)" where k is the number of years that payments increase. 161 The first term is the present value of the increasing stream of payments, up to year k, and the second term is the constant stream of payments that starts in year k+1 and ends in year n, the loan's term. It is again an easy task to solve this equation for'Aw once the other parameters are know. Working with this model, it is not difficult to modify some terms to include other variations, such as a decreasing schedule of payments after an initial period of increasing payments, in such manner as to make the stream of payments to appear'a.bell shape: or to split annual payments into semester installments (or quarter, or monthly payments). All these models show a precise equality between the loan amount and the present value of the payments stream: there- fore, neither option means any lost for the lender. Conse- quently, from a financial point of view, lenders should not reject accommodating their loans to these models, which may better represent the pattern of the project's net returns. In fact, the undertaking of these alternative payment sched- ules may reduce the default rate that banks, especially the BNF, are confronting in their long-term portfolio. The cash deficits that borrowers face when making long-term investments with debt financing have been the cause of bankruptcy of firms in many cases. One of the objections banks may argue against is the fact that the interest rate remains fixed for a certain period of time while inflation is rising. Currently, lenders are 162 granting loans for a maximum term of 6 months (most often, the loan term is only 3 months), which allows them to revise the interest rate every 6 months. These short-term loans can be rolled over by paying a portion of the principal plus interest (Author's field survey). This mechanism, however, does not alleviate the liquidity problem, especially when interest rates are high, as previously demonstrated (that net returns are insufficient to cover interest payments). It only allows borrowers to tacitly count on relatively long-term financing, but not to solve their cash deficits. In the next chapter some suggestions to cope with this problem will be discussed. Summary The hypothesis that inflation causes cash deficits in the financial position of firms undertaking long-term investments with debt financing was proposed for empirical investigation. The theoretical foundation of the liquidity problem was discussed before empirically testing the hypothesis. Using the constructed annual operating budgets for prototype mills, the effects of inflation in the cash flow of these enterprises were analyzed. The inflationary effects werefactored into the net cash returns, making them grow annually at a rate 9, and into the discount rate, making it approximately equal to the real opportunity cost of capital (the real commercial lending rate, when financing was assumed to be 100 percent 163 debt) plus this growth rate 9. For simplicity, this rate was defined as the nominal discount rate 1 (= r + i + ri). Under the assumption of 100 percent debt financing, the investments in rice mills showed liquidity problems. The net cash returns proved to be insufficient to cover the payments of the loans, contracted at current conditions: 52 percent interest and 5 years term. These liquidity problems were eliminated when the loan conditions were modified to lower the interest rate to 32 percent, to extend the term to 12 years, and to finance less than 100 percent of the invest- ments. Finally, some alternative schedules of payments were presented as mechanisms developed to alleviate the liquidity problem, These alternative models would attempt to adjust the payments stream to the pattern of net cash returns of the enterprises. Financially, these models precisely equilibrate the amount of the loan with the net present value of the payments stream: therefore, lenders should not be concerned about the profitability of these loan negotiations. CHAPTER EIGHT CONCLUSIONS AND RECOMMENDATIONS Introduction The central purpose of this study was to diagnose how rice marketing activities in Ecuador are financed. Attention was focused on the activities that occur as rice moves from farmers through wholesalers. The nature of the constraints that limit access to financial funds by a broader group of participants was identified. The five operational objectives “haweere: 1) To provide a concise diagnosis of the functional crganization of the rice subsector and the credit system facilitating the functioning of the marketing system. 2) To identify the major constraints limiting access to operating capital needed to finance the trading and the storage of rice. 3) To investigate the profitability of long-term 153—drrvestments in rice mills, storage facilities and equipment. 4) To asses the impacts of inflation on enterprise liquidity and the economic feasibility of long-term invest- ments when capital assets are debt financed and interest rates :iL#Iucorporate part or all the inflation rate. 5) To recommend changes in the credit policies and related administrative procedures to deliver loans so as to 164 165 facilitate the economic flow of rice through the marketing system. Three basic hypothesis were proposed for research. The f1 Ist hypothesis was that there are artificial constraints that adversely affect the availability of credit to finance rice trading, storage and investments in drying and milling equipment and storage facilities. The second hypothesis was that official financial resources are being allocated away from investments in equipment, machinery and storage facili- ties associated with rice processing. Moreover, that private If financial institutions are reluctant to grant loans for long- term investments. Finally, the third hypothesis was that inflation generates cash deficits when investments are financed with debt. Only when capital is sold are capital gains caused by inflation converted to cash, solving any 1 iquidity problem associated with its purchase. A -. Summary of the principal findings of the research Approximately 50 percent of the rice production credit is provided by BNF. The participation of private banks is relatively small. They fulfill legal requirements about QOlllposition of their portfolio by lending to large farmers, gSince transaction costs are lower than lending to small farmers and the default risk is perceived to be lower. Private banks do not have enough trained field personnel 166 to supervise the utilization of credit and to follow up on a large number of farm level investment projects. Broader access to subsidized production credit through private banks would require a larger cadre of field technicians and higher administrative costs. The alternative to hiring their own field experts would be to contract outside technical person- nel which would not only be expensive but would take the supervision of loans out of the direct control of the banks. So private banks meet their legal requirement to allocate part of their loanable funds to the agricultural sector by lending to a few, large borrowers. Financing land improvements is possible under officially managed lines of credit. However, there are restrictions for small farmers to access these financial resources. Rationing subsidized interest rate loans is achieved by raising the 3.— evel of transaction costs. Even though the BNF credit regulation permits loans to farmers without legal ownership O 15 land, in practice, BNF and private banks participating in the Financial Funds Mechanism do require land ownership t itles. This title as well as other documents requested by the banks as requisites for loans are costly to obtain. PIt‘oducers cooperative organizations have proved to be helpful in ameliorating these restrictions as they affect small farmers. The study has indicated that part of the production Qil'i‘edit needed, especially by small farmers, are channeled 167 through millers under the "fomento" mechanism. Although most of the miller loans finance only harvesting expenses, some also finance the acquisition of agro-chemical inputs. At the wholesaling level, access to credit seems rela- tively concentrated in the hands of a small group of large, national wholesalers. This group of participants handles about 65 percent of the total milled rice marketed in Ecuador. The author's field survey reveals that this group of national wholesalers numbers less than 20 entrepreneurs, while the group of provincial wholesalers is estimated to be about 3 times larger, and the group of urban wholesalers is con- Siderably larger. The modernization of the banking system through the automation of deposit account management obligated many participants to leave the wholesaling business. Previ- ously, the purchase of rice using check payment instead of cash gave wholesalers at least 2 weeks of costless credit, the ‘2 ime required for their checks to clear. Meanwhile the wholesalers could liquidate their stocks and return to millers for another stock of rice. At present, almost all banks have computerized national accounting systems so that the clearing Q 3'5 checks is almost instantaneous. The use of checks has now become equivalent to the use of cash. The capacity of existing rice mills seems to be more than a~