I Congresso da SBI-Agro


Lack of Monitoring Agricultural Cooperatives in Brazil: Evidence and Prospects for Improvement


Luis M. H. Requejo
603 West Springfield #B
Champaign, Illinois 61820
Phone (217) 355-8507



Analysis of the financial statements of a sample of agricultural cooperatives in Brazil indicates a widespread dispersion in measures of financial liquidity, solvency, profitability and efficiency. LOGIT regression models are used to predict stress in the form of sequential negative earnings in a sample of agricultural cooperatives in the Rio Grande do Sul state. Successive negative earnings are modeled using the return on assets, asset turnover, debt to equity ratio, members’ debt in the cooperative, and members’ equity in the cooperative. This study also provides evidence on the lack of financial monitoring in cooperatives in Brazil. One implication is that the recently created banks for cooperatives need to improve coordination and monitoring of the system in order to reduce the impact of future crisis in the sector. Loan restrictions and covenants can be used to align incentives for more sound financial practices and to insure that cooperative principles are not violated. However, the banks for cooperatives also need to be monitored. The federated structure of the banks for cooperatives, when properly devised, can generate a self-monitoring organization in which all participants monitor each other’s financial activities.



The recent crisis in the agricultural cooperative sector in Brazil undermined the confidence of both their public and private lenders. The early crisis of the 1990s culminated with the liquidation of large cooperatives and the enormous deterioration of the cooperative debt repayment capacity. The credibility of the overall agricultural cooperative system was severely damaged. Questions remain whether the lessons from the crisis will be used to restore the credibility of the sector and reduce cooperatives’ vulnerability to future downturns in the agricultural economy.

This paper investigates some organizational, financial and legal inefficiencies in the agricultural cooperative sector in Brazil. Existing inefficiencies may have worsened, instead of lessened, the 1990s agricultural cooperative crisis. It is believed that in order to reduce the impact of future crisis on this segment of the agricultural economy, improvement in cooperative monitoring and the creation of incentive mechanisms within the cooperative system are necessary. This paper links theoretical aspects and empirical evidence reported in the literature with an historical overview of various aspects of the agricultural cooperative sector in Brazil.

Particular attention is devoted to the relatively ineffective government monitoring in the past and weak monitoring of cooperative institutions in the present, and to flaws in some provisions of the cooperative law, which give rise to internal incentive problems that tend to magnify cooperatives financial distress during periods of downturns in the agricultural sector.

The large dispersion of cooperative financial indexes during the early 1990s reflects the crisis that affected the sector. The large variability of financial indexes is used to model sequential successive negative earnings for a sample of cooperatives in Rio Grande do Sul state for the years 1993-94 and 1994-95. LOGIT models are used to capture the predictive power of cooperatives’ measures of financial liquidity, profitability, efficiency, leverage, and members’ motivation towards their cooperative. Financial variables used in these models are greatly associated to the existing problems that persist in the cooperative sector.

With the advent of the creation of two Banks for Cooperatives in 1996, prospects for improvement in the coordination and monitoring of agricultural cooperatives exist. Improvement in monitoring can occur if term- loans offered to the credit starved cooperative sector are tied to loan restrictions and covenants imposed by the Banks. The federated structure of the Banks for Cooperatives can also facilitate the monitoring of the Banks by their cooperative borrowers creating incentive mechanisms not yet present in the agricultural cooperative sector.

The remainder of the article is organized in additional five sections. Section two discusses some evidence reported in the literature of self-monitoring and incentive mechanisms applied in rural financial institutions in both developed and developing countries, and agency related problems. Existing monitoring schemes are also briefly discussed based on successful experiences. Section three provides an overview of the agricultural cooperative history in Brazil, and discusses the relatively ineffective government monitoring coupled with the description of some organizational, financial , and legal inefficiencies that affect the cooperative system. Conditions are examined under which the cooperative sector can attain a self monitoring status without the presence of the government sector. Section four describes data sources and the financial structure of a sample of agricultural cooperatives in Rio Grande do Sul state. The methodology utilizes the information contained in the large dispersion of cooperatives’ measures of financial liquidity, solvency, profitability and efficiency to predict sequential negative earnings with a LOGIT model is also described in section four. The results and concluding remarks are presented and discussed in section five and six respectively.



Recent theoretical research presents some important evidence on successful group lending in developing countries. Particular attention is given to small size group lending in which peer or group monitoring substantially reduces the default on loans (Stiglitz, 1990; Varian 1990; and Besley and Coate, 1995). These programs have received considerable attention mostly because of the reported failure of agricultural development banks and other rural lenders to provide low-income producers with affordable credit (Desai and Mellor, 1993; and Von Pischke et al, 1983).

A key characteristic of such lending programs is the joint liability incurred by all members in the group. All members are treated as being in default if the member who receives the loan does not repay the loan. The idea of social collateral is also embedded in this scheme (Besley and Coate). If group social cohesion and ties within the community are strong, a powerful incentive device naturally arises to stimulate loan repayment and not upset group members. This scheme has been successfully used, particularly at the Grameen Bank of Bangladesh where loan default rates are approximately 2 percent of the total outstanding loans (Lurie, 1988).

The lending structure illustrated by peer monitoring in rural developing economies carries some features that reduce agency problems for rural lenders. Because it is costly to monitor borrowers in rural regions, the informational advantages of group lending increases the likelihood of loan repayment. When loan officers properly determine the size and characteristics of each group and loans, loans provided by the lender become an important and reliable source of financing in rural regions. With some adaptations, these programs can be applied to other types of rural lending institutions.

2.1. Credit Cooperatives as Rural Lenders

Credit cooperatives and lending groups are also reported to exhibit some incentives to improve loan repayment. Huppi and Feder (1990) stress the importance of credit cooperatives and lending groups in supplying loans to low income producers. While lending groups involve some form of joint liability and group homogeneity is a desirable attribute to reduce the risk of default, incentives in credit cooperatives are quite different. The existing embedded incentive mechanisms in credit cooperatives refers to equity contribution by members in the form of initial equity capital and capital contributions according to use of the cooperative. Credit cooperatives’ net income is either distributed to members in the form of dividends and/or patronage refunds via a revolving fund according to their patronage in the cooperative, or is retained to increase the organizations’ capital. The distribution of net income to members guarantees that the benefits go to members rather than possible intermediaries. As members have equity and deposits in their credit cooperatives, and also depend on them as a reliable source of credit, the overall cooperative arrangement appears to be a valid and promising scheme to reduce the existing costly monitoring in rural regions. In fact, governments, international development donors, banks and agencies have used credit cooperatives to develop rural economies. The other incentive mechanism is simply the existence of the credit cooperative itself . Members’ contributions at the moment a loan is originated is considered a fee for having the cooperative providing credit. Members would not have the same borrowing environment conditions without the cooperative, and the loan fees in this case are not expected to be returned.

The advantages of credit cooperative organizations depend on the ability to mobilize savings by means of creating strong ties with their members. Their organization and structure may allow members to perceive that they have a stake in the institution either via equity capital or deposits, and that they need the institution for a sound and timely source of credit. Credit cooperatives may benefit from the close relationship that exists between managers, staff, and members. This provides management with more accurate and enhanced information than other lenders. These characteristics can improve the repayment performance of potential delinquent members because of the uncomfortable situation resulted by the fact that their peers have also equity and deposits at stake in the institution.

Although the potential use of credit cooperatives to channel credit to the rural sector seems compelling, mixed results are reported (Huppi and Feder; Yongjohns, 1980; and Braverman and Guash ,1988). They refer to various problems that arise in credit cooperatives that may weaken their potential success. The problems of moral hazard and concentration of the loan portfolio create potential weaknesses. The moral hazard problem stems from managers’ hesitation to penalize delinquent borrowers, because managers may be bound to members by friendship, family ties or political affinity. The concentration of the loan portfolio refers to highly correlated income and agricultural activities of individuals that undertake the loans.

Other factors may also contribute to weaken credit cooperatives’ success as rural lenders. Huppi and Feder argue that because many governments formed cooperatives from the top-bottom (meaning that the cooperative is formed by the government instead of farmers), the interaction between cooperative managers and members can be minimal or non-existent. They cite the importance of planning and educating members to maintain members’ cohesiveness and their sense of ownership, as well motivating participation in the cooperative affairs. Excess government interference and the failure to promote educational programs is reported to cause cooperative malfunction and failures. Excess reliance on government resources also tends to generate dependency on government programs and discourage efficient management. Government presence entices indebted members to think the government will bail them out to alleviate their debt burden, making default attractive (Rosegrant and Siamwalla, 1988). When the level of education of both managers and members is weak, it is not uncommon to find cooperatives with weak and incompetent leadership, poorly defined managerial objectives, arbitrary decisions, and lack of sound and reliable accounting and controlling.

Local credit cooperative problems tend to be amplified when weak institutional arrangements exist and when no properly designed incentive mechanisms exist. Lack of control and monitoring coupled with no enforcement procedures and punitive measures also give rise to inefficiencies in the sector (Braverman and Guasch; and Bonus and Schmidt, 1990). In general, when these conditions are not properly addressed by the cooperative system, the delinquency rates of member borrowers are high.

2.2. Weak Government Monitoring and Flaws in the Cooperative Legislation

Various governments attempted to undertake the role of monitoring agents to strengthen cooperative institutions, in particular agricultural cooperatives. Evidence shows, however, that in many cases governments were unable to perform effective sustainable and timely monitoring (Adams and Vogel, 1990). The limited success of government monitoring is attributed to poor government coordination, problems of political patronage, corruption, weak enforcement procedures towards delinquent lenders, and inconsistency of government policies. In addition, when the government is present, the sense of ownership and joint responsibility of members is generally lost. Thus, rather than monitor cooperatives, governments could pursue policies to strengthen the cooperative sector by providing training to managers and members, helping cooperatives to devise incentive and monitoring schemes, and assisting the creation of mechanisms to enforce and punish members that do not comply with the terms of credit. Improvements in the cooperative system may occur if governments take the role of a catalyst agent to devise and implement incentive schemes.

The potential success of cooperative organizations can be also hampered by flaws in the legislation that govern these type of institutions. Laws governing cooperatives may differ from laws governing investor oriented firms in various countries. The reason lies in the characteristics of cooperative principles that apply to cooperative organizations which differ from investor oriented firms. Laws pertaining to cooperatives may require a minimum percentage of the cooperative net income to be allocated to indivisible equity funds to serve as a cushion against losses or to be invested in educational programs. However, legislation may not specify maximum ceilings which gives precedent for managers to allocate large part of net income into indivisible funds to "capitalize" the cooperative in detriment of members’ equity account. This practice may lead to typical inefficiencies reported in the literature due to the underestimation of the cost of cooperative capital which gives rise to overinvestment in capital assets and under usage of leverage (Tubbs, 1971; Royer, 1985; Fisher, 1989; and Nadav, 1993).

Flaws in the legislation can also discourage members’ involvement with the cooperative’s affairs. This behavior is noticed when there are no provisions to mandate cooperatives to redeem equity, pay patronage refunds or dividends, and adjust members capital according to the inflation levels. Cooperative auditing may not be required by the cooperative law, further aggravating the problems of mismanagement and perpetuating mistakes in their financial statements. The legislation can also be particularly troublesome when cooperatives file for liquidation. Cooperative liquidation can be time consuming, and its threat causes concern from its lenders giving rise to the reduction of credit availability. When private sources of credit to cooperatives are diminished, the government generally fills this gap, taking on the greater risks .

Agricultural cooperatives may also suffer from similar problems encountered in credit cooperatives. Agricultural cooperatives have similar demands for term- and short-term credit as investor oriented firms do. Similarly, if incentive schemes and lack of adequate monitoring do not exist, lending agents in the economy tend to restrain agricultural cooperatives from credit. In general, lack of understanding of agricultural cooperative structure, the inherent risk of agricultural cycles and the social attributes conferred to agricultural cooperatives are factors that restrain private banks from lending to the sector. In various countries, government owned agricultural banks became the major lenders to agricultural cooperatives.

2.3. Federated Cooperatives

Some forms of cooperative organizational structures may help to ameliorate the agency related problems and the lack of monitoring in the sector. For example, a system of two- or three-tier cooperatives structured in the form of a national or a regional federation can be effective to reduce the problems related to lack of a monitoring agent. These schemes are widely used in many countries and can largely benefit from economies of scale and risk reduction by the diversification of their operations. Federated cooperative system can also improve coordination, monitoring, and assist their locals with managerial training, auditing and member education (Guiguet and Campbell, 1993; and Cropp and Ingalsbe, 1989).

Although federated systems may help monitor and coordinate local operations, it is not clear whether federated structures can effectively impose enforcement mechanisms and restrictions on operations on their local cooperatives (or primary cooperatives). In general, enforcement tends to be weak, mostly because the federated cooperative is owned by their locals which elect representatives to the board of the federated cooperative. In addition, federated cooperatives can, and very often do generate business dependencies that can curtail the decision making freedom of their affiliates, or even intrude in their locals business decisions. Business dependencies coupled with locals’ equity in the federated cooperative can jeopardize the overall federated structure when financial distress occurs both at the federated and/or at the locals’ level. When locals are jointly liable for the federated cooperative, the entire system can collapse due to the pyramidal liability scheme when the federated cooperative fails. When locals are not jointly liable though, failure of the federated system can also cloud their operations. This fact indicates that although federated cooperatives may provide monitoring at the locals’ level, monitoring must also occur at the federated level as well. The question rests on who monitors the monitor and how to solve or mitigate agency problems for the overall cooperative system.

2.4. Examples of Monitoring and Incentive Schemes

To overcome some of the inherent problems to monitor cooperative organizations, some incentive schemes have already been successfully implemented in some countries. Bonus and Schmidt point out the importance of auditing associations within the centenary German Raiffesein system of credit cooperatives. The German system is a federated cooperative structure which comprises direct lending associations or primary banks (local credit cooperatives) that are distributed throughout the country and are owned by their member-borrowers. Primary banks, in turn, jointly hold a majority of the stock of the regional central banks. There were five central banks operating to serve primary banks in their regions in 1990. Each regional central bank jointly owns most of the stock capital in the Deutsche Genossenschaftsbank - DG Bank. The DG Bank due to its large size (seventh largest bank in Germany with total assets of DM$ 130 billion) provides funding to the regional central banks and also assist other cooperative institutions that have capital in the bank (home loan associations, insurance companies, and investment and leasing companies). The total consolidated balance sheet of the federated banking group was reported to be DM$ 562 billion in 1987.

The primary banks that directly serve their member borrowers must be a member of one regional cooperative auditing association. Cooperative auditing associations audit the primary banks. Primary banks, in turn, finance the auditing cooperatives. The auditing associations are controlled by representatives elected among the members of the Management and Supervisory Board of the primary banks. The auditing associations are members of the Federal Association of Cooperative Banks which serves as a lobbying agent to defend the interests of the cooperative system at the federal government level. If financial anomalies and exceptions are found by the auditing associations in the financial statements of primary banks, the auditing associations report to the Supervisory Board of the primary bank which in turn, notifies the general meeting of members, who may decide to remove the management board from office or restructure its staff or operations. Because other primary banks also provide financing to the auditing associations, compliance to the norms imposed by the auditing associations is of fundamental importance to maintain the system. In addition, it is the responsibility of auditing associations to fulfill advisory functions and operate training centers for the employees of the primary banks. The Supervisory Board, instead, is concerned to ensure that the primary bank’s overall policy conforms to the members and the overall business interests.

Incentive mechanisms permeates throughout the federated system in Germany and certainly have strengthened its credibility. During the years that followed the inception of the cooperative system, a member had to accept mutual and unlimited liability for their own cooperative. If the cooperative went bankruptcy, members had to provide extra funds. Given the indefinite liability of all members, appropriate members’ selection was important and members often monitored members’ activities. As the system evolved, the advent of auditing associations was an important factor to control and assess credit applications. The result was a dramatic decline in the transaction costs of the system. Later, the system implemented a limited liability scheme instead of the unlimited liability. Members would risk a loss of their shares at the primary bank and an extended liability that could not go beyond DM 1,000. The extended liability was required when the bank had not enough funds to cover its losses. Bonus and Schmidt report that the extended liabilities scheme has never been claimed because a jointly insurance fund was formed and operated by the primary banks. In a complex system involving various cooperative institutions, avoiding cooperative bankruptcy is of fundamental strategy to guarantee that no loss of confidence in cooperative banking group is feared.

Institutional monitoring and incentive mechanisms are also encountered in the Farm Credit System (FCS) in the U.S. The Farm Credit System is a nationwide farmer user-owner and controlled network of financial institutions that lend to agricultural related activities. Its structure is also based on a federated system of cooperatives. The FCS lending institutions lend to farmers, rural homeowners, agribusiness, agricultural cooperatives and rural utilities. In 1995, the FCS had approximately US$57 Billion in loans outstanding (Lee and Irwin, 1996). The institutions are organized into six Farm Credit Banks and two Banks for Cooperatives distributed by region throughout the nation. In 1996, there were 228 direct lending associations that are affiliated to their respective regional banks. The direct lending associations have equity in the regional banks that fund them. The eight regional banks, in turn, obtain most of the loanable funds through the Federal Farm Credit Banks Funding Corporation which is an agency within the System that sell farm credit bonds and discount notes in the financial markets. The banks are jointly liable for the FCS securities sold in the market, however the existing capital at the direct lending associations is not available for direct use for the joint and several liability. The joint and several liability scheme adopted by the FCS was an important device to force FCS banks to reduce and control agency costs for both investors in the Farm Credit securities, taxpayers and for the U.S. government (Barry, Brake and Banner; 1993).

The Farm Credit Administration (FCA) is an independent federal regulatory body that regulates, supervises and has enforcement authorities on the FCS institutions. It examines and ensures the safety and soundness of the System institutions. During the formation of the System, the FCA was established as an independent federal government agency. Because the System was created with government capital, the FCS was subject to provisions enacted from the federal government. Government officials were present at the FCA board as long as the government had capital in the System. Government interest in the soundness and safety of the FCS institutions was clear as the System was a major provider of credit to agriculture. As the FCS grew and evolved, the government ownership gradually changed to joint and finally private ownership. As the government loans were paid off and changes in the legislation occurred, the FCS became a privately owned cooperative system of institutions. The FCA became a supervisory body ruled by a 13-member board representing 12 districts and the Secretary of Agriculture.

With the advent of the 1980s agricultural crisis that dramatically affected the American agriculture, the federal government had to, again, injected capital in the System. Legislation in the mid-1980s strengthened the FCA’ supervision, regulatory, and enforcement authorities. The FCA could intervene and even remove FCS institutions’ managers and directors. The board of the FCA is comprised today of a three-member board appointed by the President. The agricultural crisis also encouraged the creation in 1988 of the Farm Credit Insurance Corporation in order to insure timely payment of interest and principal of FCS securities sold to the public.

The U.S. experience shows the importance in monitoring, regulating and controlling the federated FCS of cooperative institutions. Both the FCA as an independent agency, and the FCS banks have certainly contributed to the improvement in the overall monitoring within the System, its strengthening and reduction of the agency costs. Differently from the cooperative banking group in Germany, where monitoring is performed by auditing associations that belong to the system, in the U.S. the government has important presence via its participation in the FCA board. Because equities of the associations are all tied within institutions of the System, a generalized crisis in the sector can cause a domino effect with profound consequences for the federated system of cooperatives and for the borrowers that depend of their source of credit. In addition, depending on the way joint liability schemes are implemented, the likelihood of a break down of the overall system through a pyramidal scheme increases. The 1980s U.S. agricultural crisis reflects the importance of maintaining sound and safe operations within a federated system of credit cooperative banks and banks for cooperatives.

2.5. Bank Monitoring to Reduce Agency Problems

The importance of bank monitoring in capital markets has long been reported in the literature (Morgan, 1993). Bank monitoring can overcome some of the agency problems commonly found in capital markets. Agency problems arise when managers and shareholders can reduce the value of the lenders’ claims. Some of the agency related problems are the large dividend payments by firms which reduce their ability to pay their debt, excessive over borrowing which dilutes lenders’ claim, and acceptance of riskier investments which increases lenders’exposure to default. Firms may even refuse to accept lucrative investiments when such investiments tend to benefit only the firms’ lenders. In general, the firms not the lender will be hurt due to the existence of agency problems. Lenders may not lend to the firm restraining its ability to borrow, or may charge higher risk premiuns. Credit constraints and higher risk premiums lead firms to finance investment with their own funds which may give rise to liquidity problems when internal funds are not sufficient.

Agency problems can be mitigated when banks include restrictive covenants in their loans. Such covenants affect the firms’ behavior either by limiting or requiring certain actions on the part of the firm. Covenants are generally used in certain financial matters as dividend payments, debt issuance, net worth, and working capital. A lender may prohibit any additional issuance of debt or restrict the range in which certain financial ratios are acceptable. Violations of covenants imposed on working capital, for instance, are considered an initial signal of possible default by the firm. Covenants also control underinvestment or overinvestment in firms and may lose liquidity. Efficient covenant monitoring tends to encourage other investors to lend to the firm because close banking relationships are viewed favorably by other lenders (Fama, 1985).

2.6 Agency Problems and Monitoring in Cooperative Organizations

Agency problems also arise in agricultural cooperatives as well. The control of agency problems and reduction of agency costs seem to be necessary to guarantee the sustainability of a system of cooperatives, being federated, centralized or in the form of banks. At the local cooperative level, agency problems arise when members delegate cooperative managers to act upon their behalf. As cooperatives grow, cooperative managers may pursue different objectives than previously stated in the original "contractual agreement" with members. Different objectives exist because of the conflicts of interest that may arise between cooperative managers and their members, largely in part due to the cooperative principles. The conflicts created can generate internal motivational schemes that not always benefit cooperative members and/or lenders. The existing cooperative principles and legislation can impact the incentive structure and organizational behavior leading to several important problems as: 1) potential for suboptimal decisions (Fisher, 1989), 2) dilution of members’ capital and excess power attributed to managers due to the increasing reliance of retained earnings, 3) lack of motivational incentives due to inconsistent patronage refund payment policies and systematic equity redemption programs, and 4) Difficulty or lack of member monitoring of cooperative managers as cooperatives become large and present more sophisticated operations.

The monitoring of agricultural or credit cooperative systems is particularly important when membership and/ or management is poorly educated, the cooperative legislation presents flaws, and little or no enforcement and incentive mechanisms exist in the system. Monitoring can be performed both from government or delegated by cooperative institutions themselves or both. However, if government institutions are not strong and reliable, government monitoring may be inefficient. Lack of monitoring tends to threaten the overall cooperative credibility and economic sustainability during periods of economic downturns. As cooperatives become more complex to operate, members’ ability to monitor their managers is diminished. This is particularly evident when cooperatives are formed from the top-bottom and when the government is largely present. Incentive mechanisms and monitoring tend to improve managers decisions towards their member/owners and financial institutions that lend to the cooperative. When effective monitoring occurs it can play an important role in the coordination of cooperative activities and their overall strategies. Cooperative monitoring can also reduce the likelihood of cooperative liquidation which can be cumbersome and may transmit the negative image to lenders about the overall cooperative system. Even cooperatives that are soundly operated and managed may suffer credit cuts when cooperative liquidation occur.

When cooperatives are monitored, the inclusion of covenants and other restrictions are important steps to promote cooperatives to undertake more sound financial measures and adopt more strict cooperative principles. Enforcement of covenants can make the overall system more sound and safe. If monitoring is performed by a cooperative bank, the cooperative bank should be monitored as well. The self monitoring system can be performed by an outside agency or can be accomplished by the cooperative bank itself. Devices adapted from group lending, or system described earlier in Germany, US, and Bangladesh, can play important role to improve cooperative monitoring and coordination. The banks for cooperatives can also exert a beneficial role to reduce the agency problems in agricultural cooperatives. They possess advantages over other financial institutions because they are already part of a cooperative system, and have ties with their cooperative borrowers via equity and /or joint liability schemes. The violation of or non-compliance to covenants could lead to withdrawal of the loan commitment by the banks leading to enforcement actions. By not complying with covenants the cooperative could jeopardize other possible sources of credit. Covenants can reduce the cooperative financial fragility, particularly when flaws in the legislation occur or when government is inefficient in monitoring the sector. Improvement in the whole sector can occur when safer levels of financial standards are adopted reducing the likelihood of future distress.

This section describes various forms of incentive mechanisms used by rural lending institutions with emphasis on credit and agricultural cooperatives. Attention is given to the benefits generated by the monitoring of cooperative organizations. Lack of monitoring can increase the financial vulnerability and diminish the credibility of the overall cooperative system. However, effective monitoring within the system occurs when the monitor is also monitored, or the government efficiently performs the task of a monitoring agent. The next section presents an overview of the agricultural cooperative sector in Brazil linking the theoretical evidence presented in this section to the particular conditions of Brazilian agricultural cooperatives.



The present section focuses on historical and present characteristics of the Brazilian agricultural cooperative sector. Emphasis is given on some aspects related to their organization, legislation, financial structure and monitoring. The importance of the newly created Banks for Cooperatives to improve coordination and monitoring is also discussed.

According to the census of the Organization of Brazilian Cooperatives (OCB), in 1993, there were 1374 agricultural cooperatives in Brazil with approximately one million affiliated members. The socio-economic importance of agricultural cooperatives is particularly evident in the in southern states of the country (states of Paraná, Santa Catarina and Rio Grande do Sul). Agricultural cooperatives are responsible for more than 50% of the gross agricultural product marketed in these states and account for approximately 5-12% of the total state tax collection on the circulation of merchandise (the major source of state tax revenue for the states in Brazil). Moreover, a large number of members are small farmers who heavily rely on their cooperatives to supply farm inputs and to market their products. On average, 70% of cooperative members have less than 50 hectares of land in the southern states. Hence, the financial strength of agricultural cooperatives seems essential to adequately support farm activity and to promote and maintain agribusiness operations.

A significant number of the agricultural cooperatives were formed in Brazil during the late 1960s and early 1970s. The Federal government played an important role in stimulating the creation, organization and establishment of agricultural cooperatives. This measures came about during a period of rapid expansion of Brazilian agriculture. Agricultural cooperatives were viewed as important organizations to make up for the deficiency in infrastructure, storage and technical assistance to farmers, and were thought to provide a rapid response to the federal government’s stimulus to increase agricultural production. They were also considered strategic organizations to carry out government agricultural policies. As a result, a large number of cooperatives were formed from the top to the bottom without the coordination and necessary representation of their members’ common objectives. Federal government agencies became the supervisory bodies of agricultural cooperatives.

Various credit lines, many of them heavily subsidized, were opened during the 1960s and 1970s to offer investment incentives to cooperatives . The 1960s were marked by tax exemptions and tariff reductions for machinery and fertilizer imports, and the 1970s were characterized by credit incentives aimed mostly at expanding grain and oilseed crops. Federal and state banks were the major lenders to agricultural cooperatives and a federally owned National Bank for Cooperatives (BNCC) was formed in 1966 to provide credit specifically to the cooperative sector. Initially, 51% of its capital stock was owned by the federal government and 49% by the borrowing cooperatives. Cooperative infrastructure build-up was largely due to the loans provided by BNCC. Later on, BNCC began to lose its strength and importance to the cooperative sector. The government became its major shareholder, and by the end of the 1980s it had 97% of the bank's stock. Poor administration, political patronage and lack of specific policies forced the federal government to close the bank in 1991. However, agricultural cooperatives continued to be important agents to execute government agricultural policies.

3.1. Monitoring of Agricultural Cooperatives

Until 1988, the cooperative system fell under the oversight of the Ministry of Agriculture, which had powers to register, monitor and intervene in cooperatives. With the advent of the 1988 constitution, government control of cooperatives was substantially reduced. Under the new constitution, cooperatives received functional and administrative autonomy and began operating independently and outside of the government tutelage. As a result of the government withdrawal from the supervision of cooperatives, state cooperative organizations became the supervisory body for agricultural cooperatives.

3.2. The State Cooperative Organizations

The state cooperative organizations (OCEs) represent cooperatives' interests and are maintained and financed by their affiliated agricultural cooperatives at the state level. The OCEs offer training, technical assistance, lobbying, and advise for their affiliates in different areas as well as the monitoring of cooperatives' financial performance. They also offer educational and communication programs to encourage and motivate more active participation of members in the cooperative decisions. The OCEs, in turn are represented at the federal level by the Organization of Brazilian Cooperatives (OCB) which represents the cooperatives principles and doctrines and assists the cooperative system in its operational and juridical form. The OCB system is a federated system in which each OCE has one vote in the election of its board of directors. It also lobbies for cooperative interests at the federal level. Although the OCB system is represented by the respective OCEs, the state organizations are not the sole organizational forum to defend cooperative interests. Some OCEs are not well-structured and organized giving rise to other forms of representation of agricultural cooperatives like federations and centrals that are not necessarily linked to the OCB system.

3.3. Agricultural Cooperative Crisis

During the late 1980s and early 1990s the agricultural cooperative system in Brazil was drastically affected by a financial crisis that deeply weakened its overall credibility. Some cooperatives filed for liquidation and various experienced financial distress. The level of past due loans reached historical highs for Brazilian standards. Banco do Brasil, a government owned bank which is the largest lender to agriculture in the country, on average reported past 60 day loans plus nonaccruals, around 20% of the total outstanding loans to agricultural cooperatives for the years 1991-1994 (Araújo, 1996). The liquidation of two large agricultural cooperatives, once known as solid and well-administered agribusiness organizations, also contributed to increase concern in various segments of the economy regarding the soundness of the overall system.

Another important factor that significantly contributed to the financial distress of agricultural cooperatives is their borrowing constraints. Cooperatives face borrowing constraints particularly from private banks in the economy. Lending constraints that arise for agricultural cooperatives are due to mismanagement, legal issues, high riskiness of agricultural activities, heavy dependency on government agricultural policies, and low debt repayment capacity. Government owned banks (both state and federal banks) have traditionally partially filled the credit shortages in the cooperative sector. Such banks have been the major source of term and short credit for agricultural cooperatives. Although guarantees are required for such loans, no covenants and strong monitoring of cooperatives’ activities exist from their part. Moreover, changes in the Brazilian financial system have arisen in conjunction with the economic stabilization plan that culminated with the introduction of the new currency (the Real) in July, 1994. The Brazilian government intends to gradually privatize state banks to lessen further pressures on the federal budget (Pereira, 1994). Nonetheless, the reduction of credit lines brought about by the privatization and restructuring programs of state banks has increased the importance of the private banking as a source of long term credit to agricultural cooperatives. However, private banks’ supply of credit to cooperatives is still modest.

It would be impossible to identify a single factor that caused the recent cooperative crisis in Brazil. A combination of several factors serves to better explain the distress which affected the sector in different magnitudes. Periods of poor crops, high inflation, recessions, and the drastic reduction of the rural credit supply directly impacted the sector. Credit restrictions from both private, state and federal banks occurred mostly due to the cooperative fragile finances. The liquidation of cooperatives, known to be a cumbersome and lengthy process, also added uncertainty, particularly for cooperative lenders. If cooperatives were already constrained to term credit, the advent and magnitude of the 1990s crisis worsened their credit supply and undermined to a large extent the overall confidence in the cooperative system in Brazil.

3.4. Capital Structure of Agricultural Cooperatives

The reduction of the rural credit availability and limited term credit supply coupled with high real interest rates prevailing in the economy pushed many cooperatives to heavily rely on members’ equity to finance their operations. Although internal funds can make up for some of the costly external funds in the economy, such policies greatly underestimate the value of members’ equity and may lead to inefficiencies reported in the literature as over usage of capital assets and under usage of leverage. As large proportion of cooperatives net income is allocated in the form of indivisible funds and reserves which do not belong directly to members, the cost of equity is normally underestimated by cooperative managers which leads to an artificially low weighted average cost of capital. Investment projects that would otherwise be rejected are accepted by cooperative decision-makers. In addition, the low levels of leverage employed by cooperative results in the reduction in monitoring from financial institutions.

The excessive use of retained earnings coupled with lack of patronage refunds payments and systematic equity redemption programs tend to discourage members to operate with their cooperatives. Systematic equity redemption programs rarely occur in agricultural cooperatives in Brazil. Such programs assure the return of funds to members who invested in the cooperative, and provide a more equitable means to finance the organization according to use.

Cooperative legislation does not have provisions to require cooperatives adjust members’ capital according to inflation levels. In several cooperatives, members’ capital decreased in real terms, mostly during the 1980s when high inflation levels plagued the Brazilian economy. The proportion of members’ capital in agricultural cooperatives in Brazil is less than 10% on average of total liabilities. Furthermore, although provisions in the cooperative legislation require that a minimal proportion of 5% of net income be allocated to funds to promote members’ education and training, and 10% of net income be allocated to reserves used as a cushion for periods of economic downturns, the cooperative legislation does not impose a ceiling for these values. In fact, various cooperatives allocate a larger than required proportion of net income to indivisible funds and reserves that do not pertain to members. This artifice is often used and abused by managers and the board of directors to "capitalize" cooperatives. As a result, members’ capital dilution occurs and the sense of ownership and the interest in cooperatives’ affairs is dramatically reduced. New potential entrants can perceive these practices and be discouraged to become members. Lack of members’ cohesion and interest for the cooperative’s affairs can jeopardize cooperative operations and its financial stability, deepening its financial problems during agricultural crisis. Members continuing involvement with their cooperative is often viewed by lenders as a favorable attribute.

3.5. Liquidation of Agricultural Cooperatives

The recent liquidation of various agricultural cooperatives reduced banks credibility in the overall cooperative sector adding to the deterioration in the supply of credit. The troublesome liquidation process of cooperatives in Brazil is certainly an important factor that constraints cooperative borrowing. In the state of Rio Grande do Sul it is estimated that 15 agricultural cooperatives underwent liquidation during the late 1980s and early 1990s (OCERGS,1995). The states of Paraná and São Paulo also had cooperatives under liquidation. In fact, two large agricultural cooperatives (Cooperativa Agrícola de Cotia-CAC and Cooperativa Sul-Brasil) were liquidated. The CAC was the largest agricultural cooperative in Brazil with net sales of U$ 900 million in 1990. Due the size and magnitude of these cooperatives, the liquidation crisis severely weakened the banker's confidence towards Brazilian agricultural cooperatives as a whole.

The lengthy and troublesome liquidation process of agricultural cooperatives is certainly another major constraint on bank lending. Cooperatives are governed by the Cooperative Law No. 5764 amended in 1971. Because of the characteristics of cooperative organizations, cooperatives are ruled by their own law and are not subject to the same bankruptcy proceedings as corporations and commercial firms are. There are three basic distinct forms of liquidation for a cooperative:

(i) Self-liquidation. Voluntary liquidation occurs when the cooperative board verifies that the cooperative is unable to meet its short term obligations or cooperative members simply want the cooperative to shutdown. If financial recovering is possible, the self-liquidation is a valid alternative for a cooperative because the cooperative can still operate. The board of directors call an extraordinary meeting to submit the cooperative to liquidation and to elect a liquidator(s) and a fiscal council composed of three members. The liquidator's duty is to collect pending credit and to restructure debt and other claims against the cooperative. If restructuring is not a viable alternative, the liquidator starts to sell cooperative assets through auctions ruled by official auctioneers. Debt restructuring could involve extending debt maturity, lowering the interest rates on it or reducing the principal amount owed. The liquidator files a petition with the federal district court and registers the minutes of the extraordinary meeting at the commercial council. The liquidator also notifies creditors that the cooperative is under liquidation. Every six months the liquidator calls for meetings, and, if within one year financial recovery does not follow, the recovery process can be extended one more year. This means that the law confers cooperatives with at most two years to continue operating and pursuing financial rehabilitation. If rehabilitation does not occur during this two-year period, the liquidator starts the dissolution process to shut down the cooperative and the self-liquidation process changes to either a judicial or extra-judicial liquidation for which no deadlines exist.

The advantage of the self-liquidation process is that it allows cooperatives to renegotiate with creditors at a more favorable position. If the cooperative indeed goes to dissolution after the two year grace period, the judicial process may take too long and in general creditors may only receive a small portion of their claims. When renegotiation takes place, a large discount on creditors' claims generally occurs which may help the cooperative to continue operating. Normally, banks do not recover their loans when cooperatives are liquidated. The average number of years for a complete cooperative dissolution has been around 10 years in Rio Grande do Sul. There are cases for which liquidation took 20 years. Sometimes two or three liquidators may serve during the liquidation process.

(ii) Judicial liquidation. Judicial liquidation is less common than self-liquidation. In general, it takes place when patrons are not satisfied with cooperative policies or because of substandard management or even theft. Members perceive that the cooperative is insolvent and if no liquidation occurs they may be responsible for its debt. In this case a group of members goes to the judge and files a petition soliciting the liquidation. However, the number of cases of judicial liquidation is very small (1 or 2 cases in Rio Grande do Sul). Cooperative mismanagement has generally not resulted in judicial liquidation.

(iii) Extra-judicial liquidation. It is an involuntary liquidation process. Creditors are the ones that file a liquidation petition with the federal district court. The procedure that follows may be similar to the self-liquidation procedure, however, either the federal court or a group of unsecured creditors appoint a trustee to perform the functions of liquidator. At least 5-6 years are necessary for the extra-judicial liquidation to dissolve the cooperative.

In both judicial and extra-judicial liquidation, all claims against the cooperative must be processed and the cooperative eventually dissolved, whereas in the self-liquidation the two year grace period exists and the cooperative may still function if it recovers financially. When the cooperative is too far gone to be reorganized, the liquidator or trustee will proceed to pay creditors according to the following priority order:

  1. wages due to workers.
  2. labor debt with the federal government.
  3. federal taxes (social security, etc...).
  4. creditors who file claims within 90 days after the publication in the State Official Diary (the previous ones do not need to habilitate).
  5. state taxes (ICMS).
  6. municipal taxes.
  7. mortgage creditors.
  8. unsecured creditors.
  9. members.

If all assets are sold and the resources to pay creditors' claim are finished the liquidator uses equity and verifies if there is any equity pending (not incorporated into the equity account yet). Because in most cooperatives the capital is limited (limitada), patrons obligations go up to the limit of the their equity capital (or social capital). The liquidation is only resumed when all assets are sold.

The liquidation process is slow, troublesome and time consuming for all parties involved. There are many factors that can retard the process. The justice system per se is slow in Brazil and not all courts are used to the unique aspects attending to agricultural cooperatives. In addition, the petition has to pass through at least three different law courts and wait the judge's referral. Even though the liquidator is in charge of the process, decisions to sell cooperative assets need patrons' approval and therefore another meeting. This operation can take time and is particularly cumbersome when valuable assets are to be sold. The liquidator can call for as many meetings as he/she wishes. On the asset side, the liquidator has to collect pending credits (receivables). As a large part of the receivables belongs to members, knowing that the cooperative is under liquidation, members tend to postpone paying their debt with the cooperative.

The priority order of claims can also slow the overall liquidation process. During the cooperative reorganization, the liquidator may dismiss some employees. If the employee does not agree with the dismissal, he/she can go to court up to five years after the liquidation started and file suit against the cooperative. If the Justice Department considers the employee' claims, the liquidation process can be extended two to three more years because the worker is a preferential claimant. During the negotiation with banks, sometimes agreements can be reached so that the banks can receive part of what the cooperative owes even at a large discount. During the liquidation process, the federal government can also intervene and not allow the negotiation with banks to take place because, according to the priority order, social security benefits have priority over claims from banks.

In general, private banks are reluctant to provide term loans to cooperatives. If a liquidation process take place, banks can barely recover their loans. Banks always attempt to reach a settlement to avoid the liquidation process to take place. The cooperative liquidation process is certainly a constraint for private term borrowing, nonetheless short term borrowing is more frequent. For this reason, most of the term loans provided to cooperatives come from federal and state banks.

3.6. Account Receivables of Agricultural Cooperatives

The reduction in the availability of rural credit to farmers by the federal government forced various cooperatives to make up for the lack of sufficient agricultural government funds by often anticipating credit to their members. Increasing credit advances to members generated a rapid expansion of agricultural cooperatives’ account receivables. When large swings in net farm income occur, members’ debt payment capacity is reduced and in many times they can not payoff their debt with the cooperatives (or simply leave the cooperative debt to be paid lastly, if resources are still available). Sometimes members default their loan payments abusing the weak guarantees imposed by cooperatives. When members default or restructure their payments, cooperative debt payment capacity is reduced dramatically generating even more concern for lending institutions. Members’ debt with their agricultural cooperatives is, on average, about 35% of cooperatives’ account receivables for Rio Grande do Sul state.

3.7. The Newly Created Banks for Cooperatives

The magnitude of the agricultural cooperative crisis is reflected in the unprecedented number of cooperatives that filed for liquidation and the large number of cooperatives that presented negative earnings in Brazil. Although impossible to identify a single factor that caused the cooperative crisis, historical lack of monitoring and no incentive mechanisms have certainly worsened the crisis. Government agencies and banks, as well as the state cooperative organizations were and are unable to efficiently monitor cooperatives’ finances and activities. Government agencies became inefficient in addressing cooperative problems and were also victims of political patronage. The shut down of the state-owned Bank for Cooperatives in 1991 and other government agencies in charge for the supervision of the cooperative system reflects the problems faced by the public sector to achieve theses specific goals. Cooperative State Organizations, in turn, are also not effective at monitoring their affiliates. Their monitoring is often very limited and occurs without any enforcement power over their affiliates. Many of the cooperatives notably diverged from their orientation. Cooperative State Organizations are an advisory body only, meaning that they can not impose and enforce cooperatives’ adoption of their counsel. Adequate monitoring, restrictions on cooperative activities either via covenants or other form of incentive mechanisms could have helped reduce the magnitude of the cooperative crisis in the early 1990s. Changes in the agricultural cooperative representation in Brazil are called for and necessary if the sector wants to regain it credibility and be able to exploit its enormous potential. If small farming operations and cooperative form of organizations are to receive governmental attention to promote the development of rural regions in Brazil, policies aiming at stimulating agricultural cooperatives’ financial strength can definitely contribute to the improvement of the economic conditions of the farm segment and cooperative agribusinesses operations.

Responding to the demands of the agricultural cooperative sector, the Brazilian Central Bank chartered, in 1996, two Banks for Cooperatives in 1996 to operate within two distinct geographic regions. The long awaited charter of cooperative banks is an important step towards improvement in coordination and monitoring of agricultural cooperatives in Brazil. Although the Banks’ major objective is to supply term credit to agricultural cooperatives, they can also play an important role in monitoring the overall sector by imposing loan restrictions and covenants on their affiliated cooperative borrowers to encourage desired outcomes. Because cooperative borrowers have to contribute equity capital to the cooperative bank at the moment a loan is originated, cooperatives’ equity capital and future dividend payments are at stake when cooperatives default on the Banks for Cooperatives’ loans.

The federated cooperative structure of the Banks for Cooperatives may form a self monitoring arrangement with its cooperative borrowers. The Banks can not only monitor the financial health of its loan applicants, but can also be monitored by their agricultural cooperative borrowers. This reciprocal feature arises because member cooperatives with equity in the cooperative bank want to guarantee that the source of term loans provided by the Banks for Cooperatives will last, and their equity capital at the cooperative bank is protected (and can be redeemed in the future). As a result, the banks for cooperatives may have the incentive to better select cooperative borrowers than their former public lender counterparts.

The potential role for the Banks for Cooperatives in Brazil is enormous. Depending on the structure and organization, a self-monitoring scheme can arise, lessening the impacts of future agricultural downturns and reducing the level of non-performing loans of agricultural cooperatives. If instead, no form of delegated monitoring ensues, limited potential for improvement in the overall cooperative system may result and the Banks for Cooperatives may be short lived with little impact on the restoration of the agricultural cooperative credibility. Various forms of restrictions and covenants can be imposed and enforced by the newly created banks to manage the supply of term credit and also strengthen the overall agricultural cooperative system. The Banks for Cooperatives can also ensure that cooperative principles are not violated by cooperative managers. Proper legal arrangements and incentive mechanisms can be implemented by the banks and their affiliated borrowers so that all agents have some stake in performance and responsibility in the monitoring process. If these conditions are not met, the tendency could be for the credit-starved sector to continue to be credit constrained.

The next section describes the financial characteristics of a sample of agricultural cooperatives in Brazil and presents statistical models designed to capture the explanatory power of the large dispersion of cooperative financial indexes to predict sequential negative earnings. The significance of the variables selected for these models reflect, to a large extent, some of the inefficiency problems that exist within the cooperative system in Brazil.



4.1. Methodology

Statistical techniques based on credit scoring analysis are frequently used to assess the most important characteristics of borrowers. Credit scoring is useful to evaluate borrowers’ creditworthiness and assess their credit risks. The information generated by scoring models are useful to accept or reject loan applications, price loans, identify credit applicants that require special supervision and monitoring, and analyze lenders’ loan portfolios. Credit scoring models are used to maximize the economic return of lending institutions by anticipating future loan performances. Such models utilize financial and/or non-financial variables to determine certain characteristics pertaining to borrowers. Scoring models are widely used in many countries to assess credit risks related to agricultural loans incorporating the particular characteristics of borrowers of a country or region (Turvey 1990; Ruiz, 1997).

Because bankruptcy of borrowers is certainly the major concern of lenders, this study utilizes a credit scoring model to predict successive negative earnings for a sample of agricultural cooperatives in Brazil. Although agricultural cooperatives in Brazil have an enormous agribusiness potential, and several of them are successful organizations, the number of liquidating cooperative institutions and the troublesome liquidation process is of particular concern to cooperative lenders. A sample of agricultural cooperatives of Rio Grande do Sul state shows this evidence. On average, out of 62 cooperatives, 26 cooperatives presented negative earnings for the years 1992, 93, 94, and 95. The widespread dispersion of cooperative’s financial ratios also reflects the financial distress. Because persistent losses lead to the erosion of equity and eventually the failure of the firm, identification of financial variables that can explain losses is helpful not only for lenders but cooperative managers as well.

Although various statistical models are used in scoring models, most of the recent studies have used logit or probit models. The present study uses dichotomous logit models on cooperatives’ financial indexes in 1993 and 1994 to predict sequential earnings for 1994-95, and 1995-96 respectively. A sample of agricultural cooperatives in Rio Grande do Sul state is used. The logit models assume a logistic cumulative distribution function of the form where Pi is the probability of sequential negative earnings for observation i, F(Zi)=F(’Xi), is the vector of estimated coefficients, and Xi is a matrix of k exogenous variables. The probability of non-sequential negative earnings is given by (1-F(Zi )). The variable Zi* is the cut-off variable in which, given the explanatory variables (financial indexes) the probability of sequential negative earnings obtained by the model is considered low. If the cut-off rate is 50% the Y variable takes on two values


Zi >0.5* Yi =1 (2)

Zi <0.5* Yi =0


If the model provides a Zi=0.40, the logit model classifies the ith observation as non-sequential negative earnings. Different cut-off rates are estimated in this analysis.

4.2. Financial Characteristics of Agricultural Cooperatives and Data Sources

This study uses financial statements of a sample of agricultural cooperatives affiliated with the Federation of Wheat Producing Cooperatives from Rio Grande do Sul (Gonçalves, 1994). Although this federation is not an Organization of State Cooperative, its structure is very similar to some organizations of state cooperatives (OCEs) that operate across Brazil. The Federation of Wheat Producing Cooperatives was initially founded by wheat cooperatives but as cooperatives grew they engaged in other agribusiness activities. Their affiliated cooperatives are active participants in other crops and dairy business as well. Cooperatives affiliated with the Federation of Wheat Producing Cooperatives represent 75% of all agricultural cooperatives in Rio Grande do Sul state. Although this study limits itself to the cooperatives located in Rio Grande do Sul, the general financial conditions and organizational characteristics of such cooperatives are very similar to agricultural cooperatives locate in other Brazilian states.

The financial statements used in this analysis comprise the 1992, 93, 94, and 95 years. End of the year financial statement accounts are transformed into dollar values according to the prevailing exchange rate as of December 31st of every fiscal year. Currency conversion into Dollar values facilitates data manipulation during the period prior to July of 1994 where high inflation rates prevailed in the economy. Financial ratio calculations were then carried out and some selected ratios were used in the LOGIT model.

Sixty two cooperatives affiliated to FECOTRIGO were included in the analysis. Three cooperatives that either merged or entered into the liquidation process in 1994, and four cooperatives that exhibited negative equity for the same period were not considered in the analysis. Table 1 shows general balance sheet and income statement statistics of the cooperatives used in this study. The combined asset values show a 92% growth in the four year period and total liabilities represented 51% of the total assets in 1992 and 57% in 1995. Current liabilities figures show the large proportion of short term credit to finance operations and low levels of term credit to finance long term investments. Current liabilities represented 85% of the total liabilities in 1992 and 72% in 1995. The proportion of members’ equity (social capital) in the cooperatives is very small. Members’ equity as a proportion of total cooperative equity was 16% in 1992 and 32% in 1995. The registered increase in the level of members’ equity is more due to losses and depletion of equity than new injection of capital. In fact, the cooperatives sampled registered increased aggregated losses during the period.

Members debt with their cooperatives is also notable during the four years showing a proportion of receivables of 32% in 1992, 36% in 1994, and 26% in 1995. The combined cooperative earnings for the four consecutive years evidence the distress in the sector. Out of 62 cooperatives used in this study, 40% reported negative earnings in 1992, 43% in 93, 43% in 94, and 54% in 95.

Two logit models are employed in this study. Both models use financial variables at year t to predict sequential negative earnings for both years t+1 and t+2. The dependent variable E9394 in model one takes on a value of 1 when both years 1993 and 1994 have negative earnings, and 0 otherwise. Negative earnings in 1993 and positive earnings in 1994 confers the variable E9394 a zero value. All the explanatory variables in model one are 1992 figures. Similarly, model two uses 1993 explanatory variables to predict sequential negative earnings for both 1994 and 1995. The dependent variable for model two is E9495, and it takes on a value of 1 when both years 1994 and 1995 have negative earnings and 0 otherwise. The explanatory variables are the same for both models. The models utilize five exogenous variables to predict sequential negative earnings in 1993-94, and 1995-96.

Return on Assets (ROA): Is a measure of profitability that takes into account the effectiveness with which management employ cooperative assets to generate net income. It is the cooperative net income divided by the total assets. The coefficient sign for this variable is expected to be negative meaning that when the variable ROAt is negative it increases the likelihood of sequential negative earnings for both years t+1 and t+2.





Std Dev



Sum of all coops

Assets 1992






Assets 1993






Assets 1994






Assets 1995






Liabilities 1992






Liabilities 1993






Liabilities 1994






Liabilities 1995






Current liabilities 1992






Current liabilities 1993






Current liabilities 1994






Current liabilities 1995






Equity 1992






Equity 1993






Equity 1994






Equity 1995






Social capital 1992






Social capital 1993






Social capital 1994






Social capital 1994






Members debt with the coop in 1992






Members debt with the coop in 1993






Members debt with the coop in 1994






Members debt with the coop in 1995






Earnings 1992






Earnings 1993






Earnings 1994






Earnings 1995






Sales 1992






Sales 1993






Sales 1994






Sales 1995






Members 1994






Coops with neg. earnings in 1992  




Coops with neg. earnings in 1993  




Coops with neg. earnings in 1994  




Coops with neg. earnings in 1995  





Table 1 Selected Balance Sheet and Income Statement financial information of 62 Agricultural Cooperatives from Rio Grande do Sul (in US$ values).


Equity per Member (EQM): It is a proxy to measure members’ motivation towards their cooperatives. It is the total cooperative equity divided by the number of its members. The coefficient sign for this variable is expected to be positive. The reason is as follows. As the amount of equity per member increases, members are (or may be) aware that large part of the equity retained goes to indivisible funds and will not be returned to them (at least not in its full amount). In addition, few cooperatives have systematic equity redemption programs for the funds that are in the social capital account. These facts may lead to a reduction in the motivation to operate with the cooperative which tends to increase the likelihood of sequential negative earnings in the coming years.

Term Debt Equity (DEBTEQ): It is a financial leverage ratio. This ratio is given by term debt divided by term debt plus equity. It measures the relative proportions of lenders’ term claims to ownership claims and it is used as a measure of term debt exposure. An increase in term debt relative to equity capital would normally increase the firm probability of bankruptcy if the firm is beyond its optimal level of term debt. However, the sample of agricultural cooperatives comprised in this study show a very low level of term debt. Hence the expected coefficient sign for this variable is negative meaning that the more term debt is used the less is the probability of future sequential negative earnings.

Asset turnover (ASTUR): This ratio is a measure of resource management. It is an expression of the effectiveness with which assets generate sales. It is defined as the ratio of net sales to total assets. The coefficient sign is expected to be negative. As more sales are generated with the same assets, more efficient is the cooperative and less likely to present sequential negative earning in the coming years.

Members Debt with the Cooperative (MDEBT): It measures the dollar value of members’ indebtedness with their cooperatives. This figure is obtained from each cooperative’s account receivables. The coefficient sign is expected to be positive. The larger is the debt amount hold by members with their cooperatives the larger is the cooperative exposure to members’ default on their payments. As a result, cooperative’s debt payment capacity and investment can be reduced, increasing the likelihood of future sequential negative earnings.



The results of the two LOGIT models are presented in table 2. Table 2 shows the estimated equations and coefficient values on the explanatory variables. This table also reports the elasticity coefficients calculated at the means for each variable. A positive coefficient sign indicates that the probability of sequential negative earnings increases as the value of the variable increases. Alternatively, a negative sign indicates that the probability of obtaining sequential negative earnings decreases as the variable increases. General consistency was found for the signs of the parameters with the exception of the variable MDEBT for both models. A possible explanation for the dubious coefficient sign for the variable MDEBT is the attractiveness of cooperative credit to members. Although the probability of default on cooperative loan increases, it is possible that members’ operations with their cooperatives also increases due to covenants imposed by the cooperative, which lead members to increase business with their cooperatives. As operations with members increase, the probability of obtaining negative earnings is reduced. Thus, the selected variables do interact to reduce the probability of sequential earnings for both models, and present consistent results when used for different periods.



LOGIT MODEL 1 - Dependent variable E9394



Mode Chi-Square  





Parameter Estimates of LOGIT Model 1




































































* indicates significance at the 5% level; ** indicates significance at the 10% level; *** indicates significance at 5% level for the Chi-square test.

LOGIT MODEL 2 - Dependent variable E9495



Model Chi-Square  





Parameter Estimates of LOGIT Model 2




































































TABLE 2 - Estimated Coefficients and Elasticities of Model 1 and Model 2

* indicates significance at the 5% level; ** indicates significance at the 10% level; *** indicates significance at 5% level for the Chi-square test.


All variables were significant at 10% level with the exception of the variable MDEBT for model 1. At 5% significance level the variables EQM and DEBTEQ were not significant in model 1. In both models the constant terms are not significant at 10% level. Similar LOGIT regressions estimated without the constant term slightly reduce the predictive power of model 2. For this reason the constant is kept in both models. No significant correlation was found among the selected variables when the Spearman’s correlation test was used. Both models are significant at 5% levels given by the 2 significance test.

The sensitivity of the model is given by the elasticities evaluated at the means. Measures of financial leverage (DEBTEQ) and profitability (ROA) have the greatest impact on the probability of sequential negative earnings. The elasticity at the means evaluated for DEBTEQ92 indicates that for a 1% increase in the term debt/equity ratio decreases the probability of sequential negative earnings by 1.02%, and for a 1% increase in ROA92 decreases the probability of sequential negative earnings to 0.65%. A smaller elasticity was found for the variable ASTURN. For model two, an increase in 1% in the asset turnover ratio decreases the probability of sequential negative earnings for 1994-95 by 0.19%.


Classification Table for Model 1 - Dependent Variable is E9394



Percent Correct  




















Classification Table for Model 2 - Dependent variable is E9495



Percent Correct  





















TABLE 3 - Prediction-Success Tables for Model 1 and Model 2

Classification tables for prediction-success for both models are shown in table 3. The cut-off rate (Z*) used to compile these tables is 60% meaning that if the probability calculated for the ith observation is greater than 60% the model classifies Yi =1 (sequential negative earnings). If the probability calculated is less than 60% the model classifies Yi =0 (non-sequential negative earnings). Because of the relatively small number of cooperatives used in this study (62) no hold-out sample was kept to test the accuracy of the models for both 1993-94 and 1994-95 periods. The results indicate that both models have a good predictive power for non-sequential negative earnings. Model 1 accurately predicts 97% of the non-sequential negative earnings and 68% of sequential negative earnings. Model 2 predictive power is inferior than model 2. It predicts 89% of non-sequential negative earnings and 64% of sequential negative earnings.

Both models misclassify sequential negative earnings. Type I errors arise from classifying cooperatives with sequential negative earnings as being cooperatives with non-sequential negative earnings. A type I error of 5 cooperatives (31.5% of cooperatives with sequential negative earnings) and 6 cooperatives (35.2% with sequential negative earnings) were found for model 1 and model 2 respectively. Type II errors arise from classifying non-sequential negative earnings as sequential negative earnings. Type II errors of 1 cooperative (2%) and 5 cooperatives (11%) were found for model 1 and model 2 respectively.

4.5. Discussion of the findings of the LOGIT models

The results obtained from both LOGIT models clearly indicate the importance of cooperatives’ capital structure on the future of cooperative’s profitability. The significance and negative signs encountered for the variable DEBTEQ, and the elasticity evaluated at the means close to unit reflects the potential that term-debt can play to improve cooperatives’ profitability. The level of term liabilities is relatively low. Term liabilities comprised 15% of the total liabilities in 1992 and 28% in 1995. However, possibilities to increase the supply of term-loans to agricultural cooperatives are small when one considers their present financial situation. In addition, lack of credibility on the agricultural cooperative system is still a widespread feeling among cooperative lenders.

The significance of the variable EQM reflects ongoing problems in the way members’ capital is used to finance cooperative operations. In Brazilian cooperatives, higher levels of equity capital per member may discourage members from operating with their cooperatives due to the lack of specific programs to retire equity, to pay dividends, and to distribute earnings. In fact, most of the cooperative earnings are transferred to indivisible accounts and never returned to members. Inefficiencies are reported in the literature when cooperative managers underestimate the cost of members’ equity. The existing motivational characteristic inherent in this variable reflects the importance in the adoption of proper equity capital management policies that are more consistent with cooperative principles.

The significance of variables related to profitability and efficiency is not a surprise. In most credit scoring models these variables can largely explain the probability of bankruptcy of firms and loan default by agricultural borrowers. Strategies aimed at improving cooperative profitability and efficiency tend to reduce the probability of occurrence of sequential negative earnings.

The large proportion of members’ debt in the cooperative’s account receivables is evidence of cooperative financing their members. Managers have greatly relied on member financing to entice members to operate with their cooperatives. The negative sign for the variable MDEBT indicates that this strategy may work. However, agricultural cooperatives are not specialized in lending as banks are. In general, collateral requirements and guarantees imposed on member borrowers are weaker than banks, which creates the potential for members’ default.



The purpose of this paper was to review some organizational, financial and legal aspects of agricultural cooperatives in Brazil. Due to its economical importance, the financial strength of this segment of the agricultural economy is essential to support farm and agribusiness activities. However, the unprecedented financial crisis that affected the sector brought about loss of credibility and confidence of cooperative lenders.

Some theoretical and empirical evidence of the importance in monitoring cooperative type of organizations is discussed. Successful evidence indicates that when cooperative monitoring occurs, and incentive mechanisms are present, the overall cooperative system can better perform their goals. Monitoring of the cooperative sector can attenuate inefficiencies that arise in cooperative type of organizations, and tend to be more efficient when it is aligned with financial incentives.

Cooperative monitoring by the federal government was inefficient in Brazil. Proof of this is the complete shut down of the National Bank for Cooperatives in 1991. Although some of the State Cooperative Organizations, Federations and Centrals are carrying on some sort of monitoring since government withdrawal from cooperative tutelage, no mechanisms of enforcement exist. The lack of agricultural cooperative monitoring is certainly one of the major problems faced by the cooperative system in Brazil. The government, State Cooperative Organization, federations, state and federal banks have not been efficient monitors and enforcement agents.

Two logit models are designed to identify variables that can predict future sequential negative earnings for a sample of cooperatives of Rio Grande do Sul state. Variables identified were related to measures of profitability, efficiency, and capital structure. Successive negative earnings are modeled using the return on assets, asset turnover, debt to equity ratio, members’ debt in the cooperative, and members’ equity in the cooperative. Both models present good predictive power for both sequential and non-sequential negative earnings. The models utilize variables that are easily observed in cooperative’ financial statements and can be easily monitored.

With the recent chartering of two Banks for Cooperatives in Brazil, monitoring of the system may occur and improve their overall financial condition. Adequate monitoring could not only restore confidence in the sector but could also attract other lending institutions to finance cooperative term investments. However, strong covenants or other restrictions need to be imposed to force cooperatives to operate within certain financial limits, and adopt policies more in accordance to cooperative principles on behalf of its members and lenders. The federated structure of the Banks for Cooperatives can improve monitoring and coordination in the sector.

However, without the creation of incentive mechanisms in which a self monitoring scheme arises, improvements in monitoring and coordination can be short lived. Success is more likely to occur when incentive schemes are implemented so that the Banks for Cooperatives are also monitored. If Banks for Cooperatives are to start operating, it is believed that the whole cooperative system need to be responsible, involved and committed to its operational and financial integrity. The unique opportunity that has arisen for the cooperative sector with the chartering of Banks for Cooperatives can dramatically strengthen the system if incentive mechanisms are properly devised.



  • Adams, D. W. and Robert C. Vogel, 1990."Rural Financial Markets in Low-Income Countries", in Eicher, C. K. and John M. Staatz, eds, Agricultural Development in the Third World. Baltimore, MD, John Hopkins University Press.
  • Aldrich, J. H. and Forrest D. Nelson, 1984. Linear Probability Logit, and Probit Models, Sage Publications, Beverly Hills, California.
  • Araújo, Uilson Melo, 1996."Assimetria de Informação no Crédito Rural: Aspectos teóricos e um Modelo para a Classificação do Risco dos Créditos Concedidos a Cooperativas Agropecuárias", Tese de Doutorado, Escola Superior de Agricultural "Luiz de Queiroz", Universidade de São Paulo.
  • Barry, P. J., Brake, J. and Delmar K. Banner, 1993."Agency Relationships in the Farm Credit System: The Role of the Farm Credit Banks", Agribusiness, Vo 9, No. 3, pp. 233-245.
  • Besley, Timothy and Stephen Coate, 1995."Group Lending, Repayment Incentives and Social Collateral", Journal of Development Economics, 46, pp. 1-18.
  • Bonus, H. and G. Schmidt, 1990."The Cooperative Banking Group in the Federal Republic of Germany: Aspects of Institutional Change", Journal of Institutional and Theoretical Economics, 146, pp. 180-207.
  • Braverman, Avishay, and J. Luis Guasch. 1988. "Institutional Aspects of Credit Cooperatives"World Bank Agricultural Policy Working Papers, WPS 7, April, p 25.
  • Cropp, Robert and Gene Ingalsbe, 1989."Structure and Scope of Agricultural Cooperatives", in in Cobia, D.W.,ed., Cooperatives in Agriculture, New Jersey: Prentice Hall.
  • Desai, B. M, and John W. Mellor (Eds), 1993. Institutional Finance for Agricultural Development - An Analytical Survey of Critical Issues. Food Policy Review 1, International Food Policy Research Institute, Washington, D.C.
  • Huppi, Monica and Gershon Feder, 1990."The role of groups and credit cooperativs in rural lending ", World Bank Researcher Observer, 5 pp.187-204.
  • Fama, Eugene F. 1985."What’s Different about Banks?", Journal of Monetary Economics, January, 15, pp.29-39.
  • Fisher, Martin Lee."Financing Agricultural Cooperatives, Economic Issues and Alternatives", Ph. D. Dissertation, University of Minnesota, 1989.
  • Gonçalves, M. F. R."Análise de Desempenho das Cooperativas Filiadas-Relatório 92/93", Porto Alegre, FECOTRIGO, 1992, 93,94, and 95.
  • Guiguet, E. Depetris and Gerald Campbell, 1993." in Csaki, C. and Y. Kislev, ed., Agricultural Cooperativesin Transition. Bolder: Westiview Press.
  • Lee, W. F. And George D. Irwin,"Restructuring the Farm Credit System: A Progress Report" Agricultural Finance Review, Vol. 56, (1996): pp. 1-21.
  • Lurie, T., 1988."Increasing Income and Employment in Bangladesh", The Ford Foundation Letter, 19, pp.1-5.
  • Nadav, Carmel."Incentives and Equity Capital in Cooperatives", Ph.D.Dissertation, University of Minnesota, 1993.
  • Morgan, Donald P., 1993."Bank Monitoring Mitigates Agency Problems: New Evidence Using Financial Covenants in Bank Loan Commitments", Federal Reserve Bank of Kansas City Research Working Paper 93-16, December, p. 20.
  • Pereira, L.C.B., 1994." A economia e a Política do Plano Real". Revista de Economia Política, Outubro-Dezembro. 14, 4, p. 56,
  • Rosegrant, Mark and Ammar Siamwalla, 1988."Government Credit Programs: Justification, Benefits, and Costs", in John Mellor and Raisuddin Ahmed (eds), Agricultural Price Policy for Developing Countries, Baltimore, MD, John Hopkins University Press.
  • Royer, J. S."Strategies for Capitalizing Farmer Cooperatives", in Farmer Cooperatives for the Future, Lee F. Schrader and Willian D. Dobson eds., NCR-140 workshop, St. Louis MO; Department of Agricultural Economics Purdue University, August (1985).
  • Ruiz, L. M. G. S., 1997."A Credit Risk Assessment Model for Dairy Farm Borrowers in Mexico", Ph. D. Dissertation, University of Illinois.
  • Stiglitz, Joseph E.,1990."Peer Monitoring and Credit Markets", The World Bank Economic Review, 4, pp. 351-366.
  • Tubbs, Alan Roy, 1971."Capital Investments in Agricultural Marketing Cooperatives. Implications for Farm Firm and Cooperative Finance", Ph. D. Dissertation, Cornell University.
  • Turvey, C. G., and Reginald Brown,"Credit Scoring for a Federal Lending Institution: The case of Canada's Farm Credit Corporation", Agricultural Finance Review,Vol. 50 (1990): 47-57.
  • Varian, Hal R., 1990."Monitoring Agents with Other Agents", Journal of Institutional and Theoretical Economics, 146, pp. 153-174.
  • Von Pischke, J. D., Adams, D. W., Gordon, D. (Eds) 1983. Rural Financial Markets in Developing Countries: Their Use and Abuse, Baltimore, MD, John Hopkins University Press.
  • Youngjohns, B.J. 1980."Agricultural Co-operatives and Credit", in J. Howell, ed., Borrowers and Lenders, London: Overseas Development Institute.